Personal current account banking services in Northern market investigation

15 May 2007 © Competition Commission 2007

Website: www.competition-commission.org.uk Members of the Competition Commission who conducted this investigation

Christopher Clarke (Chair) Professor John Baillie Laura Carstensen Ian Jones Jeremy Seddon1

Chief Executive and Secretary of the Competition Commission

Martin Stanley

The Competition Commission has excluded from this report information which the inquiry group considers should be excluded having regard to the three considerations set out in section 244 of the Enterprise Act 2002 (specified information: considerations relevant to disclosure). The omissions are indicated by [].

1Jeremy Seddon died in November 2006. Personal current account banking services in Northern Ireland market investigation

Contents Page

Summary...... 4 that provide PCAs and their business models ...... 4 Characteristics of PCA customers ...... 5 Market definition and market shares...... 5 Branch networks, entry and expansion...... 6 conduct ...... 7 Customer conduct...... 9 Unilateral effects ...... 10 Coordinated effects...... 11 Financial performance ...... 11 Findings and features ...... 12 Remedies...... 14

1. Introduction...... 18 The reference...... 18 Conduct of the investigation...... 19 Structure of the report...... 19 2. Overview of the sector ...... 20 Northern Ireland background ...... 21 Banks that provide PCAs ...... 23 Clearers...... 25 BoI ...... 25 First Trust ...... 25 Northern...... 25 Ulster ...... 25 Non-clearers...... 26 Abbey ...... 26 A&L...... 26 ...... 27 Nationwide...... 27 and ...... 27 Co-operative ...... 28 HSBC...... 28 Characteristics of PCAs...... 28 PCA charging structures ...... 28 Traditional PCAs...... 30 Fee-free PCAs...... 31 Recent changes to PCAs ...... 31 Overview of financial performance...... 32 Customer characteristics ...... 33 Levels of switching ...... 35 Regulatory environment, the Code, payment systems and provision of bank notes ..... 38 Banking regulation ...... 38 The Code ...... 40 Payment systems...... 42 Provision of bank notes...... 43 Market trends...... 44 3. Market definition ...... 45

1 Product market definition ...... 45 Basic bank accounts, instant access savings accounts and credit union accounts.. 46 Packaged accounts and offset/current account mortgages ...... 49 Personal financial products ...... 49 Market segmentation...... 50 Conclusions on product market definition ...... 50 Geographic market definition...... 50 Chain of substitution...... 51 Local markets...... 51 Conclusions on geographic market definition ...... 51 Conclusions on market definition...... 52 4. Competition in the market...... 52 Market structure...... 52 Business models ...... 52 Market shares and concentration...... 54 Volume-based market shares and concentrations ...... 54 Value-based market shares and concentrations ...... 61 Conclusions on market shares and concentrations...... 63 Branch networks ...... 64 Entry and expansion ...... 66 Entry ...... 67 Expansion...... 72 Conclusions on entry and expansion...... 75 Bank conduct ...... 75 Charging structures and levels...... 75 Overview of charging structures...... 76 Assessment of charging structures and information provision ...... 78 Recent changes to charging structures...... 84 Trends in interest rates and charges...... 88 Discussion of levels of interest rates and charges ...... 90 Conclusions on charging structures and levels of charges and interest rates...... 92 Competition for customer groups ...... 93 Non-price competition ...... 94 Customer conduct...... 97 Overview ...... 97 Reasons for searching and switching ...... 97 Barriers to searching and switching ...... 99 Customer attitudes to unauthorized overdraft charges...... 102 Complexity and lack of clarity...... 106 Customer satisfaction...... 107 Conclusions on searching and switching ...... 108 Analysis of competition ...... 109 Unilateral effects ...... 109 Coordinated effects ...... 115 Conditions for coordination...... 115 Evidence of past parallel behaviour...... 117 Recent behaviour ...... 118 Conclusions on coordinated effects...... 118 Local competition ...... 119 Other indicators of the extent of competition ...... 120 Financial performance and profitability ...... 120 Financial performance...... 121 Operating costs ...... 124 Return on capital and profitability ...... 124 Efficiencies ...... 125 5. Findings and features...... 126

2 Findings ...... 126 Features...... 128 6. Remedies...... 129 Overview...... 129 Framework for decisions on remedies ...... 130 The need for remedial action ...... 131 Assessment of remedy options...... 131 Remedies that we have decided to pursue ...... 132 Remedy (a): Easy-to-understand terminology and descriptions of PCA services...... 132 Remedy (): Explanations of the levels of charges and interest rates and how and when they are applied ...... 136 Remedy (c): Information on statements ...... 142 Remedy (d): Summary and breakdown of charges and interest ...... 146 Remedy (e): Advance notice of charges and debit interest incurred...... 151 Remedy (f): Regular ‘rights reminder’...... 157 Remedy (g): Changes to the switching process...... 160 Remedies that we have decided not to pursue ...... 164 Remedy (h): Provision of personal indicative ‘quotes’ for PCAs ...... 164 Remedy (i): Provision of ‘typical customer’ indicative quotes for PCAs ...... 165 Remedy (j): Switching statistics...... 167 Remedy (k): Changes to the switching process: number portability...... 169 Direct regulation of PCA prices ...... 169 Additional remedies proposed by third parties ...... 170 Scope of remedies...... 170 Implementation of remedies...... 174 Monitoring and enforcement of remedies ...... 175 Effectiveness and proportionality...... 177 The least cost, least intrusive package of effective remedies ...... 177 Proportionality to the scale of the AEC ...... 178 Conclusions on effectiveness and proportionality ...... 181 Relevant customer benefits ...... 182 Framework for assessment of relevant customer benefits...... 182 Conclusion on relevant customer benefits ...... 182 7. Our decisions...... 183

Appendices 1.1 Terms of reference and conduct of the investigation 2.1 Banking Regulation 2.2 Banking Code 2.3 Payment systems 2.4 PCAs in Northern Ireland and comparisons with Great Britain 4.1 Market shares in personal banking 4.2 Process of opening and switching PCAs 4.3 Searching and switching 4.4 Trends in charges and interest rates 4.5 Competition for customer groups 4.6 Unauthorized overdraft charges 4.7 Authorized and unauthorized overdraft scenarios 4.8 Review of the banks’ financial analysis of new PCAs 4.9 Analysis of non-price competition 4.10 Financial performance 4.11 Cost allocation and return on capital 4.12 Efficiencies 4.13 Pre-notification: current requirements, practices and future requirements Glossary

3

Summary

1. On 26 May 2005 the Office of Fair Trading (OFT) referred the supply of Personal Current Account (PCA) banking services in Northern Ireland to the Competition Commission (CC) for investigation and report under section 131 of the Enterprise Act 2002 (the Act). The reference followed the OFT’s receipt of a supercomplaint on 15 November 2004 from Which? and the General Consumer Council for Northern Ireland (GCCNI) on the market for PCA banking services in Northern Ireland. The CC is required to publish its report by 25 May 2007.

2. A PCA is defined in our terms of reference as ‘… an account, marketed to individuals not businesses, which provides the facility to hold deposits, receive and make payments (cheques and debit cards) and use automated teller machine (ATM) facilities and to make regular payments (direct debit and standing orders)’.

3. The current account is the most widely-held personal banking product in Northern Ireland. About 80 per cent of households in Northern Ireland have access to a current account, compared with almost 90 per cent in the UK as a whole.

Banks that provide PCAs and their business models

4. PCA banking services are provided by many different organizations in Northern Ireland. We discuss separately, where relevant, the Northern Ireland clearing banks (the clearers)—Governor & Company of the (BoI), AIB Group (UK) plc (trading as First Trust Bank (First Trust)), Northern Bank Limited (Northern) and Limited (Ulster)—and those that are not clearers (the non-clearers). We refer to the clearers and non-clearers collectively as the banks.

5. The non-clearers can be divided into three subgroups: current and former building societies (including plc (Abbey), Alliance & Leicester plc (A&L), Halifax plc (Halifax), Nationwide (Nationwide) and ‘’ (Woolwich));2 banks based in Great Britain that have opened one or more branches in Northern Ireland since 2000 (Co-operative Bank plc (Co-operative Bank) and HSBC Bank plc (HSBC)); and banks that provide a remote service by Internet, telephone, post and/or text messaging (the remote providers); these include (a division of Abbey), (a division of HSBC), (IF, a division of Halifax) and (a division of Co-operative Financial Services Limited (CFS)).

6. The banks’ businesses are not designed solely for the provision of PCAs. The clearers and many of the non-clearers offer a variety of personal banking products through various distribution channels including the branch, via the Internet, or by telephone. In addition, the clearers’ business model is based on the requirements of both personal and business banking. Each of the clearers has a more extensive branch network in Northern Ireland than each of the non-clearers, and offers a wide range of both personal and business products. All the non-clearers offer standard products across the UK and do not tailor their PCA for the Northern Ireland market.

7. We note that, despite overall similarities of business model, there are some significant differences between the individual clearers. The clearer with the most branches (Northern) has more than twice the number of branches of the clearer with

2Woolwich is owned by Barclays Bank plc (Barclays). Barclays is rebranding all Woolwich branches as Barclays during 2007. Among other changes, Woolwich PCAs will in future be provided by Barclays.

4

the fewest (BoI). BoI is the only clearer to have an agreement with the Post Office to provide counter services. There are also differences in strategy and in IT and operating platforms. Both BoI and First Trust’s parent company, Allied Irish Bank plc (AIB), have their group headquarters in Dublin. Ulster is part of RBS, headquartered in Edinburgh, and Northern is part of Danske, headquartered in Copenhagen.

Characteristics of PCA customers

8. We looked at the characteristics of PCA customers and the levels of switching between banks. We noted that the Financial Services Authority (FSA), in its work to assess levels of financial capability, found serious concerns about current levels of financial capability throughout the UK. We also noted that UK customers are generally not particularly interested in PCAs.

9. We commissioned three surveys of Northern Ireland customers: a qualitative survey (Millward Brown survey, January 2006), a quantitative survey (ORC survey, January 2006), and a further quantitative survey looking particularly at overdraft charges and charging practices (the BMRB survey, December 2006) to understand customer views on PCAs in Northern Ireland. We looked at the reasons given by customers for the choice of bank. In general, when customers first choose their bank, they are particularly influenced by branch location, whether close to their home, place of work, or other frequently visited location. They are also more likely to choose a particular bank if other family members already bank there, if charges are low, and, to some extent, if credit interest rates are high. Our ORC survey found that only around one- third of customers use the branch to transact the majority of their PCA business. We found that relatively fewer younger people said that branch location was important to them in their choice of bank.

10. PCA customers normally stay with their chosen bank for a long time. The ORC survey showed that the median time since opening a main PCA for all customers was 13 years and that more than three-quarters of all PCA customers had been with their main bank for more than six years.

Market definition and market shares

11. We concluded that the product market should include all PCAs, including packaged accounts, but should not be drawn more widely to include other types of personal account such as basic bank accounts, instant access savings accounts, credit union accounts, offset/current account mortgages or other personal financial products. We also concluded that the geographic market was Northern Ireland.

12. We looked at a range of bases for calculating market shares and concentrations. MORI MFS (MORI) survey data had been collected over a number of years on a consistent basis. It reports rolling annual averages of four quarterly financial surveys of a changing sample of 1,500 Northern Ireland respondents. However, there were important differences of definition between the current account as defined by MORI, and our PCA definition. We also collected data for 2002 to 2005 from the eight largest banks to allow us to calculate alternative data on market shares by volume, as well as value-based market shares (based on average credit balances and PCA revenues).

13. The MORI volume-based market shares showed that, overall, the clearers have lost significant share over recent years. These losses appear to have largely been driven by the relatively poor performance of two of the four clearers, First Trust and Northern. Depending on the measure of market share used, we found that the

5

clearers’ market share in 2005 was between 69 and 77 per cent. According to all measures, the combined market share of the clearers has fallen since 2002.

14. Market shares based on a measure of value rather than volume give a higher market share to the clearers. This suggests that the effect on some of the clearers’ revenues of a reduction in market share by volume might have partially been mitigated by retaining more of their high-value customers and hence increasing their average credit balance per account at a faster rate than the non-clearers.

15. We looked at the level of concentration in the market in 2005 based on each of these measures of market shares. Although the market is more concentrated based on the value of credit balances or revenue than suggested by the MORI market shares based on number of main accounts or the data provided by the banks on the number of accounts, all measures show the market in 2005 to be concentrated.

Branch networks, entry and expansion

16. We considered the role of the banks’ branch networks. The ORC survey showed that a branch is very important to customers when opening an account. Some of the banks told us that customers could open PCAs remotely, but, due to money- laundering regulations, many banks still required the customer to go to the branch in person to validate their identity and address. Although many customers transact much of their day-to-day business away from the branch, other customers, particu- larly those in socio-economic groups D and E, still visit the branch fairly frequently. We also understand from our Millward Brown survey that customers regard it as important to have the option of visiting a branch to resolve any difficulties that might arise.

17. All banks now offer remote banking as an additional channel. Whilst few customers choose to use remote banking as the only mode of accessing their PCA, we were told that they expected to be able to access their accounts from a variety of channels including branches, the Internet, and by telephone.

18. The banks have differing views on the importance of the branch network to them. Some claimed that the branch network was key to their PCA businesses, notwithstanding the growth of direct channels such as telephone and Internet banking. Others saw the branch network as less important and saw the need for fewer branches than some of the clearers currently had. In addition, the banks consider the branches to be important marketing tools, raising customers’ brand awareness and providing an opportunity for cross-sales. We noted that the clearers have many branches in rural locations or small towns with a relatively limited catchment area, and that the costs of providing a wide branch network are significant.

19. We thought that a substantial branch network provided a benefit to a bank in terms of attracting and recruiting both switchers and new-to-banking customers, and in providing existing customers with access to branches. Branch networks also carry substantial costs. Banks need to consider both the costs and the benefits and decisions on opening or closing branches will depend on a variety of factors, including the bank’s strategic view and the effect that branches can have on the number and type of products sold, of which PCAs are only a part.

20. We found no evidence of entry to the market in Northern Ireland on any significant scale, and that expansion in the market took many years. While the non-clearers collectively have achieved a significant market share, they have been providing PCAs in Northern Ireland for nearly 20 years. We found that barriers to entry and

6

expansion included, in particular, a lack of willingness to switch among existing customers and a distrust of unfamiliar banks. In addition, we found that the non- clearers are not focused in particular on expansion in Northern Ireland. The evidence that we received has not led us to form an expectation that the non-clearers will increase their number of branches in Northern Ireland in the foreseeable future.

Bank conduct

21. There are two basic PCA charging structures in Northern Ireland: the historical charging structure (the ‘traditional’ PCA) and the ‘fee-free’ PCA. In addition to ancillary charges, customers may incur the following charges on one or other (or sometimes both) of these PCA structures:

(a) transaction charges: levied on a per-transaction basis;

(b) maintenance charges: fixed payments (monthly, quarterly, yearly or ad-hoc) arising from the operation of the PCA; and

(c) charges for setting up, amending or cancelling standing orders or direct debits.

22. Authorized and unauthorized overdraft charges may include both transaction and maintenance charges. In addition, customers will normally pay interest when operating their PCA in overdraft.

23. We defined traditional PCAs to be PCAs where transaction and maintenance charges are payable when the account is operated in authorized or unauthorized overdraft. Both transaction and maintenance charges are levied during a quarter if the account is overdrawn at any point during that quarter, whether or not the overdraft has been authorized. Additional charges are levied when a PCA customer either exceeds their agreed overdraft limit or goes overdrawn without agreement. Most traditional PCAs also levy charges for setting up, amending or cancelling standing orders or direct debits, whether the account is in credit or in debit.

24. Fee-free PCAs do not levy transaction and maintenance charges when the customer is in credit or in authorized overdraft. However, fee-free PCAs, like traditional PCAs, are subject to a range of ancillary charges. They are also subject to unauthorized overdraft charges that are often set at higher levels than those for traditional PCAs.

25. There are also a number of other charging structures. Several banks offer packaged accounts, where customers typically pay a fixed monthly or quarterly charge regardless of whether the account is in credit or debit, but do not pay any transaction charges. Packaged account customers are also likely to receive preferential terms on other characteristics of the account, may be exempt from some ancillary charges, and often receive other bundled benefits such as free travel or discounts on some purchases. Some accounts levy a fixed monthly charge for authorized overdrafts but do not levy transaction charges or apply maintenance charges in months other than those in which the customer was in authorized overdraft.

26. We found that there was an inherent complexity in PCA charging structures that arose in part because PCAs service a wide variety of needs. We looked at the PCA charging structures and practices to see whether they were unduly complex; that is to say more complex than might reasonably be necessitated solely by the wide variety of uses or services offered.

27. We noted the following aspects which might add to complexity for the customer:

7

(a) There are a large number of different charges that might be levied on a particular PCA.

(b) The total amount payable in a particular month or quarter can be made up of several different elements (eg interest payments, transaction and maintenance charges).

(c) Transaction charges for operating a traditional PCA in authorized overdraft are incurred retrospectively for any transactions made since the start of a quarter and continue to be incurred for the remainder of that quarter even if the customer returns to credit.

(d) In many cases customers are not advised of charges prior to their being debited from their account, making it difficult for them to take appropriate action to ensure that sufficient funds (or an appropriate overdraft facility) are available to pay for the charges, or to confirm that the charges are correct, prior to their being debited.

(e) Certain charges are advised and debited a long time after they are incurred, making it more difficult for customers to understand the links between their actions and the charges incurred.

(f) Charging practices, particularly in relation to unauthorized overdraft charges, can vary between banks. For example, some banks may levy a charge for every transaction; others may charge just once a day, or limit the number of charges in a day or in a specified period, through a charge cap. Some banks have buffer zones of a certain value or duration before charges are triggered.

(g) Banks also have individual practices on honouring or refusing payments which create or extend an unauthorized overdraft. Some banks calculate in advance an undisclosed unauthorized overdraft payment limit which determines the highest payment value that the bank is willing to honour. Other banks will assess whether or not to make payments on an ad-hoc basis; these assessments may be automated or manual.

(h) Banks have individual practices on refunding charges once they have been debited from the account.

28. In addition, we noted that a bank is unusual in both providing a PCA to the customer, and also having access to the customer’s account which is used to pay for the service, either by employing the customer’s own funds, or by providing an overdraft facility. This dual role might be a factor in the banks’ charging practices.

29. We sought to understand the charges incurred by customers when operating their account in authorized or unauthorized overdraft by constructing plausible customer scenarios. Many of the banks found it difficult to complete our scenarios exercise without additional time and guidance, in part because of the complexities that arose from flexibility in individual banks’ practices in applying their own charging structures.

30. We found that authorized overdraft charges for traditional PCAs, and unauthorized overdraft charges for all PCAs in particular, showed considerable, and undue, complexity. We believed that this complexity in charging structures and practices makes it more difficult for customers to understand and react to overdraft charges and build their assessment into making efficient choices of PCA should they seek to do so.

8

31. We also found that, in general and across the banks as a whole, customers are not provided with the necessary information to enable them to have a sufficient understanding of the charges and interest rates that might apply to their PCA. This was particularly true for information on unauthorized overdraft charges. We also thought that ancillary charges could be explained in terms that were more customer- friendly.

32. There have been a number of changes in PCA offers in Northern Ireland over the past two years, particularly those offered by the clearers. These new charging structures show that the clearers are, in general, moving towards a fee-free business model. However, we note that the undue complexity associated, in particular, with unauthorized overdraft charges, remains. Furthermore, almost 20 per cent of customers in Northern Ireland remain on traditional PCAs.

33. The financial analysis that we carried out on the clearers’ new PCAs suggested that most of the new charging structures were forecast to be broadly revenue neutral for each bank, albeit the banks’ estimates were subject to considerable uncertainty. The forecasts represent only limited reductions in charges on average to the customer. We recognized, however, that depending on individual customer usage, some customers would benefit whilst others would incur higher charges.

34. The trends in PCA charges and interest rates that we looked at suggested similarities in pricing among the clearers’ traditional PCAs that we did not think could be explained by similarities in costs nor by high levels of competition in the market. We did not find such similarities in the pricing of packaged accounts.

35. With regard to unauthorized overdraft charges, we found that some banks were able to maintain unauthorized overdraft charges or interest rates significantly higher than their competitors for extended periods. The customer scenarios suggested that when all the charges were aggregated together, total charges varied significantly by bank. In the unauthorized overdraft customer scenarios, total charges were generally higher for fee-free PCAs than for traditional PCAs. We found, therefore, that there was a lack of competitive constraint on banks’ unauthorized overdraft charges and debit interest rates. This was also supported by the evidence that we had on the way in which banks set unauthorized overdraft charges.

36. We also looked at non-price factors. Overall, we identified several indicators of a relative lack of competition in Northern Ireland, particularly among the clearers in relation, for example, to branch opening hours and functionality of Internet banking. Product innovation in Northern Ireland appears to have been relatively limited in the past, and the continued existence of some traditional PCAs might reinforce our finding of a lack of competition in the market.

Customer conduct

37. We looked at levels of switching in the Northern Ireland PCA market. We considered the banks’ and survey data on levels of switching, as well as switching rates for other products and relative rates of switching in Northern Ireland and Great Britain. Whilst evidence was mixed on the absolute levels of switching, given the customer attitudes and barriers to searching and switching, it was clear to us that there was a lack of responsiveness to changes in charges or interest rates in the market.

38. We found that a customer’s decision to switch was more often prompted by dissatisfaction with their bank than the recognition of a better offer elsewhere. If a customer was pushed into switching, the decision about the choice of bank may be based on factors that can easily be observed, or on information that can readily be

9

obtained, rather than on a rational decision based on the costs and benefits of switching. Similarly, for new-to-banking customers the difficulty in searching may result in these customers making wrong, or inefficient, choices.

39. We looked at the barriers to switching and searching. Although switching is free under the Banking Code (the Code), we found that barriers included anticipation of hassle and delay, as well as the risks and costs of switching. Customers believed that switching bank might damage their longer-term banking relationship, jeopardizing, in particular, future availability of credit. We also found that customer indifference, lack of interest in PCAs and ignorance of some charges were important reasons for the lack of switching. The financial incentives to switch were likely to be low for many customers, but, for a customer regularly operating their account in overdraft, and particularly in unauthorized overdraft, the annual financial incentive to switch could amount to several hundred pounds.

40. We concluded that the obstacles to searching and switching were significant for many customers. Whilst some of the issues with switching were perceived rather than real, we believed that these perceptions affected customer behaviour, leading to a high level of customer indifference. This resulted in a lack of searching and switching which we believed contributed to a lack of competition in the market.

41. We found that in the case of unauthorized overdraft charges, customers who had incurred such charges were more likely to switch bank. However, few customers make comparisons of these charges between banks, and customers tend not to think about such charges when choosing a new bank, largely because they do not believe that they will be affected by such charges. This could be because they intend to avoid unauthorized charges in the future. This suggested that customers object to paying such charges but banks would receive little competitive benefit in terms of attracting switchers from lowering these charges.

42. We found that there was a level of complexity in PCA charging structures and practices and a lack of clarity of information that would be likely to make it difficult for customers to search and switch. As competitive price offers were not easily observed or understood by customers, competitive pressure on banks was likely to be reduced, giving banks an incentive to raise prices, or reduce service or quality levels, compared with those that would exist in a well-functioning market.

43. We believed that, as a result in particular of the barriers to searching and switching that customers face arising from the undue complexity of charging structures and practices and lack of clarity of terminology, customers may be paying higher charges or levels of debit interest, or receiving lower levels of credit interest, lower levels of service, or there may be less innovation, than would be likely to occur in a well- functioning market.

Unilateral effects

44. We looked at whether competitive conditions in the market might give individual banks market power. We found that both the clearers and non-clearers have the ability to exercise unilateral market power in relation to their customers, arising from customers’ low propensity to switch. However, their incentives to do so might vary due, for example, to their different strategic focus and business models, their different customer bases, or their different operating scale or costs, such as the size of their branch network. We observed several indicators that suggested that the incentive to exercise unilateral market power was, and to some extent still is, greater for clearers than for non-clearers.

10

45. We looked at recent changes in the market. It appeared to us that, despite the recent changes in behaviour that represent a shift in competitive positioning by the clearers, there remained several characteristics of the market which meant that all banks, clearers and non-clearers alike, continue to have the ability to exercise unilateral market power. Whether and in what way they choose to exercise this market power will depend on the particular incentives facing them. However, where it is profitable to do so, we expect banks to take advantage of such market power, whether by imposing higher charges or levels of debit interest, paying lower levels of credit interest, and/or by offering lower levels of service or less innovation than would otherwise apply.

Coordinated effects

46. We also looked at the possibility of coordinated effects. We considered whether the conditions existed to facilitate tacit coordination between the banks. As set out in the CC guidance, where markets are sufficiently concentrated, the actions of individual firms can have identifiable effects on their competitors, such that firms recognize their interdependence and act accordingly. Thus, through mutual self-awareness of interdependence, banks might find that if each of them adopts a less competitive strategy, they could all enjoy higher prices or lower costs, leading to higher profits or greater scope to sustain inefficient operating practices than would otherwise have been the case (coordinated effects).

47. We believed that any potential coordinating group would be likely to consist of the clearers, given the different business models and incentives of the non-clearers. We concluded that overall the conditions for sustained coordination were not met, principally because the clearers did not have an incentive to coordinate due to the low propensity of customers to switch between banks. For coordination to be possible, it is necessary that all three conditions set out in the CC guidance are met.

48. There was evidence of some prices moving in parallel in the past, in particular in the period prior to the supercomplaint. However, recent changes in charging structures by the clearers suggest that any parallel pricing that might have taken place in the past has now broken down.

Financial performance

49. We reviewed the financial performance of the eight largest banks in Northern Ireland: the clearers and the four largest non-clearers—Abbey, A&L, Halifax and Nationwide. Together these provide over 95 per cent of PCAs in Northern Ireland.

50. A significant proportion of each bank’s operating processes and costs, including the operation of its branch network and centralized head office and group functions, is shared across its businesses and products. All banks were asked to allocate revenues and shared costs to their PCA businesses in Northern Ireland. Whilst the allocation of revenues presented the banks with some difficulties, these were not as significant as the difficulties the banks faced when allocating costs.

51. The banks earn income in three forms from providing PCAs:

(a) net interest income (NII) on credit balances (value of funds elsewhere in the bank less interest paid to customers);

(b) NII on debit balances (interest received from customers less cost of providing funds); and

11

(c) charges.

52. In 2005 the eight banks earned income in total of £167 million from providing PCAs, which represented an average increase of nearly 8 per cent a year since 2002. On average the banks earn more than half their total income from NII on credit balances, over 40 per cent from charges and 5 per cent from NII on debit balances. The relatively high proportion of total income represented by NII on credit balances reflects the significant benefit to the banks of paying low average credit interest rates. We estimated that the eight banks earned over half of all income from charges from unauthorized overdraft charges. Average credit balances per account were higher for the clearers as a group (2005: £1,796) than for the non-clearers (2005: £1,184).

53. We estimated that, on average, the total revenue earned directly from PCA customers (in the form of charges and net credit and debit interest) was £64 for the clearers and £41 for the non-clearers, but this masked some differences between the individual banks, and some non-clearers earned more than some clearers.

54. We were unable to conclude as to whether the banks were making returns in excess of their cost of capital for their PCA businesses, given the interdependency of PCAs and other banking products, and the practical difficulties in identifying appropriate and accurate bases for cost and capital allocations.

55. The banks were unable to provide us with answers to detailed questions on efficiency sufficient for us to draw conclusions from them. We noted that, whilst the banks may monitor the efficiency of their retail businesses as a whole, they do not separately monitor the efficiency of their PCA business in Northern Ireland on a routine basis. In particular there is no benchmarking of their efficiency either internally within the banks’ Northern Ireland PCA businesses or against other entities except at the highest level of bank cost/income ratios. However, we noted that the clearers are all seeking to improve the efficiency of their PCA businesses.

Findings and features

56. Our findings are as follows:

(a) There is a general lack of customer interest in PCAs and customers tend to view PCAs as ‘all the same’.

(b) Customer perception is that switching PCAs is much more difficult and risky than it is in practice. Some problems do nevertheless arise, despite the success of the Code in relation to switching.

(c) In the present state of competition in the market, the financial incentives to switch are unlikely to outweigh the perceived risks for most customers.

(d) Customer understanding of unauthorized overdraft charges appears to be low; some are not aware of when they are charged; and few know the levels of their banks’ charges or how they compare with competitors.

(e) Annual rates of switching in the PCA market in Northern Ireland as a whole are between 1.5 and 4 per cent. There is a lack of responsiveness to changes in charges or interest rates in the market.

(f) A customer’s decision to switch is more often prompted by dissatisfaction with their existing bank than the recognition of a better offer elsewhere. Dissatisfaction often arises from the bank levying unauthorized overdraft charges; poor service,

12

unhelpful staff, and bank errors are examples of other causes of dissatisfaction which may prompt switching.

(g) The single most important factor in the choice of new bank is access to a branch; personal recommendation and, to some extent, levels of interest and charges are also important. However, customers have, up to now, been unlikely to compare and assess such charges when choosing a new bank.

(h) There is an inherent complexity in PCA charging structures, in part because PCAs service a wide variety of needs. However, even recognizing this, PCA charging structures and practices are unduly complex, particularly for authorized overdraft charges associated with traditional PCAs and for unauthorized overdraft charges levied on all PCAs. The incidence and frequency of charging on PCAs is not clear. There are often exceptions or discretionary items which are not visible to customers and add to complexity.

(i) A bank is unusual in both providing a PCA service to the customer, and also having access to the customer’s account which is used to pay for the service.

(j) The banks, particularly the clearers, describe their charges using terminology that is unclear and, in many cases, inconsistent between banks. This is particularly true of unauthorized overdraft charges and, to some extent, ancillary charges. In general, customers are not provided with the necessary information to enable them to have a sufficient understanding of the charges and interest rates that might apply to their PCA.

(k) Charges and credit interest rates on traditional PCAs showed similarities in pricing that are unlikely to be explained by costs since, in most cases, banks do not know their costs of providing particular PCA services. Nor are they likely to be explained by high levels of competition in the market.

(l) Unauthorized overdraft charges and debit interest rates are unlikely to be explained by costs. Charges and debit interest rates are likely to be above the levels that would apply in a well-functioning market.

(m) Clearers’ customers pay more, on average, than non-clearers’ customers, to operate their PCA, as evidenced by the total revenues earned directly per account.

(n) There is a relative lack of competition, particularly among the clearers, on several non-price factors such as branch opening hours and the introduction of full- function Internet banking.

(o) Clearers, in particular, whilst monitoring their business at a broader level, and despite some recent improvements, have limited management and financial information specific to their PCA business in Northern Ireland. This includes limited information on numbers and destinations of switchers; incidence of particular charges; activity-based costings; costs of acquiring new customers; benchmarking of the efficiency of different products; and levels of cross-sales.

(p) Whilst we recognize the costs of an extensive branch network, clearers benefit from their networks, as well as their strong brands. They continue to attract a relatively high proportion of new-to-banking customers in relation to their share of flow of all new accounts. Non-clearers are not looking to expand their branch networks in Northern Ireland in the foreseeable future.

13

(q) There was evidence of some prices moving in parallel in the past. However, the conditions for sustained coordination in the market are not met, principally because the clearers do not have an incentive to coordinate due to the low propensity of customers to switch between banks.

(r) Any competition concerns are unlikely to have a local market focus.

57. In many cases these findings are indicators or outcomes of a lack of competition in the market. The banks as a whole continue to be able to impose higher charges or levels of debit interest, pay lower levels of credit interest, and/or offer lower levels of service and innovation, in particular to existing customers, than might be the case if switching were more prevalent. Although the findings are caused by market-wide features, the findings may differ among individual banks depending, for example, on the bank’s business model, ownership or strategy.

58. We found that the features within the meaning of the Act which prevent, restrict or distort competition in the PCA market in Northern Ireland are as follows:

(a) banks have unduly complex charging structures and practices;

(b) banks do not fully or sufficiently explain their charging structures and practices; and

(c) customers generally do not actively search for alternative PCAs or switch bank.

59. The first two of these features lead to difficulties in customers making properly informed choices. This applies both to new-to-banking customers in searching for an appropriate PCA, and to existing customers who might otherwise switch to a bank that might provide a PCA to meet their needs at lower cost. This is exacerbated by the third feature: the perception and, to some extent, experience of difficulties in switching, as well as by customer indifference.

60. We therefore found that, on the statutory questions that we have to decide pursuant to section 134(1) of the Act, there are features of the relevant market, either alone or in combination with each other, that prevent, restrict or distort competition in connection with the supply of PCAs in Northern Ireland, and hence that there is an adverse effect on competition (AEC) within the meaning of section 134(2).

Remedies

61. We concluded that, despite significant recent changes in the market, the features outlined still persisted and prevent, restrict or distort competition in the market and that, without taking remedial action, the market as a whole would remain uncompetitive and that there would be a detrimental effect on customers.

62. We decided that the following package of remedies would represent a compre- hensive, reasonable and practical solution to the AEC and any resulting detrimental effect on customers that we had identified:

(a) Remedy (a): Easy-to-understand terminology and descriptions of PCA services. Banks operating in Northern Ireland must satisfy the Banking Code Standards Board (BCSB) that all information provided to customers when choosing a PCA, when opening a PCA, on statements, and when pre-notified of charges and interest payments, is easy to understand. The banks must ensure that all such communications are:

14

(i) certified by an independent organization specializing in plain English; or

(ii) otherwise tested with customers and found to be easily understandable.

(b) Remedy (b): Explanations of the levels of charges and interest rates and how and when they are applied. Banks operating in Northern Ireland must ensure that their Northern Ireland customers receive explanations of the levels of charges and interest rates on their PCA, the circumstances in which the customer would incur these charges and/or receive interest, and the way in which, and when, these charges and interest rates would be applied. Explanations must be provided of at least the following charges and interest rates:

(i) credit interest rates applicable to relevant balance limits;

(ii) current account charges;

(iii) charges for standard account services (eg charges for setting up direct debit or standing orders);

(iv) authorized overdraft debit interest rates;

(v) authorized overdraft charges;

(vi) unauthorized overdraft debit interest rates; and

(vii) unauthorized overdraft charges (including maintenance charges, paid and unpaid items charges and transaction charges).

Banks operating in Northern Ireland must also make clear to customers the existence, but not the detail, of discretionary policies in the banks’ application of charges and interest in relation to unauthorized overdrafts.

The banks must make this information freely available to customers when choosing a PCA, when opening a PCA, on statements, and when pre-notified of charges and interest payments.

(c) Remedy (c): Information on statements. Banks must provide customers with key account information, as part of their PCA statements. Banks must provide explanations of the level and application of at least the charges and interest rates set out in remedy (b). The information should be provided to each PCA customer, in the same form as a customer receives their statement unless the customer elects otherwise. It may be provided on the front or the back of the statement, on a separate sheet on ‘statement paper’, or as part of an electronic communication, provided the information is clearly part of the statement.

(d) Remedy (d): Summary and breakdown of charges and interest. Banks operating in Northern Ireland must provide PCA customers once a year with an annual summary and breakdown of charges and interest. This summary should cover all the charges and interest payments relating to the customer’s PCA but need not include ancillary charges. This summary must include an explanation of the level and application of all those charges included on the statement (see remedy (b)). The summary should be sent to the customer in the same form (electronic or hard copy) as the customer routinely receives their statements, unless the customer elects otherwise.

(e) Remedy (e): Advance notice of charges and debit interest incurred. Banks must pre-notify customers of all authorized and unauthorized overdraft charges.

15

Pre-notification may be sent together with a statement or in advance of a statement and should be in the same form as the customer routinely receives their statements, unless the customer elects otherwise. However, the banks must not deduct a pre-notified charge from a customer’s PCA until at least 14 days after the date of the statement. If the pre-notification is not sent together with a statement, the bank must provide the key account information that is required under remedy (b) on the pre-notification communication. Banks must not allow charges and interest to accumulate to a PCA for more than one month before the customer is pre-notified of the bank’s intention to deduct the charge or interest from the PCA. The bank must send the customer a statement at the end of any month in which authorized or unauthorized overdraft charges are incurred.

(f) Remedy (f): Regular ‘rights reminder’. Banks must include wording on the annual summary and breakdown of charges and interest that makes it clear that the customer can close their PCA with that bank and seek to obtain PCA services from a different bank. It should also specify where the customer can find further information on how to do this. This communication should be accompanied by a generic leaflet or electronic communication which explains the switching process, the fact that it is easy to switch using the banks’ switching services and that customers are not charged for switching. This communication should be developed by the banks working together, possibly through the British Bankers’ Association (BBA). It may be in electronic form or hard copy but should be sent to customers in the same form as they receive their summary and breakdown of charges and interest, unless the customer elects otherwise.

(g) Remedy (g): Changes to the switching process. For each Northern Ireland customer who uses the switching service to open a new PCA, banks must offer an interest- and charge-free overdraft facility, the amount of which is commensurate with the expected transactions on the PCA, for a minimum of three months after the PCA is opened, provided the customer would be eligible for such a facility under the bank’s usual credit scoring policy. Where a customer is not eligible for such an overdraft or does not wish to have one, banks must guarantee to refund the customer any charges and interest which are incurred within a minimum period of three months after the PCA is opened as a result of a failure in the switching process. This is regardless of whether the charges and interest were incurred as a result of an error by the new bank. Banks operating in Northern Ireland must publicize these ‘hold harmless’ provisions in their switching literature.

63. We decided that it would be appropriate to establish a minimum threshold for the application of our package of remedies. We decided that this threshold should be based on the absolute number of a bank’s PCAs held by customers with a postal address in Northern Ireland. A bank that is within a corporate group with more than 10,000 PCAs held by customers with a postal address in Northern Ireland will be required to apply the remedies, unless that bank can demonstrate that it has fewer than 5,000 PCAs held by customers with a postal address in Northern Ireland.

64. We decided that these remedies should be implemented by means of an order. We decided to require the implementation of remedies (a) and (b) by 1 April 2008, and to require the implementation of the other remedies within our package by 1 October 2008 at the latest. This latter date should allow the banks scope to implement the changes required as a result of our remedies alongside those changes that are likely to be required as a result of the Code review and the implementation of the Consumer Credit Act 2006 (CCA 2006).

16

65. The OFT has a statutory role in relation to the monitoring and enforcement of remedies. However, in line with the principles of better regulation, we do not wish to impose any unnecessary additional regulatory structure on an already highly regulated industry. We concluded that the BCSB should monitor compliance with our remedies. The banks should fund additional resource for the BCSB as appropriate to enable it effectively to monitor compliance with the remedies package. The BCSB should provide the OFT with regular written reports to the OFT’s specification.

66. We concluded that our package of remedies was the least cost, least intrusive package of remedies that would be effective in addressing the AEC. We believed that the estimates provided by at least some of the banks as to the cost of implementing our package of remedies may represent the upper end of a range of possible costs, especially taking into account costs that are likely to be incurred in any case following the Code review and the implementation of the CCA 2006. Even taking these costs at face value, we found that over time, the benefits from the remedies could reasonably be expected to exceed the costs by a substantial margin. We therefore concluded that our package of remedies was proportionate to the scale of the AEC.

67. None of the parties said that there were relevant customer benefits deriving from any of the features of the market for the supply of PCA banking services in Northern Ireland that we had identified. We have not identified any relevant customer benefit on which the remedies, alone or in combination, would have an effect. We therefore did not consider relevant customer benefits further.

68. We also recommended that BACS review the switching process, with a view to identifying and addressing any outstanding impediments to switching direct credits, direct debits and standing orders.

17

1. Introduction

The reference

1.1 On 26 May 2005 the OFT referred the supply of PCA banking services in Northern Ireland to the CC for investigation and report under section 131 of the Act. The CC’s terms of reference are set out in Appendix 1.1. The CC is required to publish its report by 25 May 2007.

1.2 A series of events prompted the OFT’s reference to the CC:

(a) a supercomplaint from Which? and the GCCNI on the market for PCA banking services in Northern Ireland submitted on 15 November 2004 under section 11 of the Act;

(b) the OFT’s response to the supercomplaint and a consultation paper on making a reference to the CC published on 11 February 2005. The consultation period closed on 18 March 2005; and

(c) the OFT’s consideration of the supercomplaint and the responses to its consultation.

1.3 Section 134(1) of the Act requires the CC to decide whether ‘any feature, or combination of features, of each relevant market prevents, restricts or distorts competition in connection with the supply or acquisition of any goods or services in the UK or a part of the UK’. If the CC decides that there is such a feature, then there is said to be an ‘adverse effect on competition’ (see section 134(2)).

1.4 A ‘feature of the market’ means:

(a) the structure of the market concerned or any aspect of that structure;

(b) any conduct (whether or not in the market concerned) of one or more than one person who supplies or acquires goods or services in the market concerned; or

(c) any conduct relating to the market concerned of customers of any person who supplies or acquires goods or services.3

1.5 If the CC decides that there is an adverse effect on competition, it is required under section 134(4) of the Act to decide whether action should be taken by it, or whether it should recommend the taking of action by others, for the purpose of remedying, mitigating or preventing the adverse effect on competition concerned or any detrimental effect on customers4 so far as it has resulted from, or may be expected to result from, the adverse effect on competition; and, if so, what action should be taken and what is to be remedied, mitigated or prevented.

3Section 131(2) of the Act. 4A detrimental effect on customers is defined in section 134(5) of the Act as one taking the form of: (a) higher prices, lower quality or less choice of goods or services in any market in the UK (whether or not the market to which the feature or features concerned relate); or (b) less innovation in relation to such goods or services.

18

Conduct of the investigation

1.6 In conducting our investigation we have had regard, in particular, to the CC’s published guidelines on market investigations.5 We published our Issues Statement on 28 October 2005, based on evidence received during the first five months of the investigation.

1.7 On 28 April 2006 we published our Emerging Thinking and seven associated working papers, setting out our thinking on the central issues of the investigation. These reflected submissions from parties in response to the Issues Statement and to draft working papers, hearings with parties, meetings in on 16 February 2006 with community sector and representative groups and with elected representatives, and the results of the qualitative and quantitative research (Millward Brown and ORC surveys respectively) that we had commissioned.

1.8 On 12 October 2006 we published our provisional findings report, taking account of the evidence received during the course of the first 16 months of our investigation. We also published our Notice of possible remedies. On 6 March 2007 we published our provisional decisions on remedies, taking account of comments received in relation to our Notice of possible remedies.

1.9 We published a considerable number of papers on the CC website during the course of the investigation. In addition to our Issues Statement, Emerging Thinking and associated working papers, provisional findings, Notice of possible remedies and provisional decisions on remedies, these include non-sensitive versions of parties’ written submissions and summaries of key arguments and views. We also published the reports from the Millward Brown and ORC surveys, as well as the second quantitative survey that we commissioned, looking particularly at overdraft charges and charging practices (the BMRB survey). These surveys, commissioned by the CC during the course of the investigation, are described in Appendix 1.1. Further details of the conduct of the investigation are also set out in Appendix 1.1.

Structure of the report

1.10 This document, together with its appendices, constitutes our final report setting out the decisions on the questions we are required to answer under the Act together with the reasons for our decisions and sufficient information to facilitate a proper understanding of them. It takes account of the evidence received during the course of the investigation. It refers, where appropriate, to material published separately on the CC website (see paragraph 1.9).

1.11 The rest of this report is structured as follows:

(a) overview of the sector, including relevant economic background on Northern Ireland; an outline of the banks that provide PCAs in Northern Ireland; characteristics of PCAs; overview of financial performance; customer characteristics and levels of switching; overview of the regulatory environment, the Code, payment systems and provision of banknotes; and future market trends (see section 2);

(b) market definition (see section 3);

5CC3: Market Investigation References: Competition Guidelines, June 2003.

19

(c) market structure, including an outline of business models; market shares and concentrations; branch networks; and entry and expansion (see paragraphs 4.1 to 4.107);

(d) bank conduct, including PCA charging structures and levels; competition for customer groups; and non-price competition (see paragraphs 4.108 to 4.206);

(e) customer conduct, including reasons for, and barriers to, searching and switching (see paragraphs 4.207 to 4.260);

(f) analysis of competition, including a discussion of unilateral and coordinated effects, and local competition (see paragraphs 4.261 to 4.311);

(g) other indicators of the extent of competition, including financial performance and profitability and efficiencies (see paragraphs 4.312 to 4.339);

(h) findings and features (see section 5);

(i) remedies (see section 6); and

(j) our decisions (see section 7).

2. Overview of the sector

2.1 In this section we set the context for our market investigation:

(a) relevant economic background on Northern Ireland (see paragraphs 2.4 to 2.13);

(b) an outline of the different banks that provide PCAs in Northern Ireland (see paragraphs 2.14 to 2.47);

(c) characteristics of PCAs, including PCA charging structures (see paragraphs 2.48 to 2.64);

(d) overview of financial performance (see paragraphs 2.65 to 2.71);

(e) customer characteristics and levels of switching (see paragraphs 2.72 to 2.93);

(f) banking regulation, the Code, payments system and the provision of banknotes (see paragraphs 2.94 to 2.120); and

(g) future market trends (see paragraphs 2.121 to 2.125).

2.2 The European Commission launched three sector inquiries into financial services in June 2005 covering payment cards, core retail banking and business insurance respectively. The inquiry into core retail banking is the most relevant for our investigation. It published its final report, Sector Inquiry under Art 17 of Regulation 1/2003 on retail banking (Final Report), on 31 January 2007. This report identified various symptoms suggesting that competition may not function properly in certain areas of retail banking. It noted that ‘the decisions of retail banking customers are … constrained by information asymmetry and high switching costs’ (paragraph 33). In particular, the report noted that:

(a) in some member states, the conjunction of sustained high profitability, high market concentration and evidence of entry barriers raises concerns about banks’ ability to influence the level of prices for consumers and small firms;

20

(b) some credit registers, holding confidential data that lenders use to set loan rates, may be used to exclude new entrants to retail banking markets;

(c) some aspects of cooperation among banks, including savings and cooperative banks, can reduce competition and deter market entry;

(d) product tying, eg where a loan customer is forced to buy an extra insurance or current account, is widespread in most member states. This could reduce customer choice and increase banks’ power in the market place to influence prices; and

(e) obstacles to customer mobility in banking—notably the inconvenience of changing a current account—are high. The inquiry’s analysis suggests that banks’ profit margins are lower where customers are more mobile.6

2.3 A number of other investigations and reviews relevant to PCAs in Northern Ireland have also been undertaken during the period of this investigation, or are still in progress:

(a) The new CCA 2006, which updates the 1974 Consumer Credit Act (CCA 1974), seeks to improve and extend the framework of consumer credit regulation and enhance the rights of consumers in relation to unfair lending and minimum standards of information. The final provisions of CCA 2006 will be implemented in October 2008.

(b) The Code is reviewed and if necessary updated every two years. The latest independent review of the Code began in November 2006 and is currently under way. The revised Code will be published in March 2008.

(c) The OFT has recently announced that it is carrying out a market study into personal bank current account pricing, alongside a formal investigation into the fairness of charges for unauthorized overdrafts and returned items under the Unfair Terms in Consumer Contracts Regulations 1999 legislation.7

(d) We also note that public awareness of bank charges has increased during the course of this investigation, and that there has been a substantial increase in the number of customers challenging these charges.

Northern Ireland background

2.4 We looked at the economic background of Northern Ireland to understand the context of our investigation.

2.5 Northern Ireland has a population of around 1.7 million, representing just under 3 per cent of the UK population.8 In terms of geographic area, it is approximately 10 per cent of the size of , 17 per cent of the size of Scotland and two-thirds of the size of Wales. Northern Ireland’s population density is around 51 per cent of the UK average.9 The East of Northern Ireland is much more densely populated than the West: around 16 per cent of the population live in Belfast, and more than one-third of

6Press release IP/07/114 Brussels, 31 January 2007 ‘Competition: Commission sector inquiry finds major competition barriers in retail banking’. 7OFT press releases 54/07 and 67/07. 8Source: 2001 Census. For comparison, Wales had a population of 2.5 million. 9In 2001, the population density in Northern Ireland was 1.2 people per hectare.

21

the population live in the Greater Belfast area.10 No other conurbation has a population of more than 100,000, and 35 per cent of people live in rural areas.11

2.6 Within the UK, Northern Ireland has a relatively young population with proportionately more children and the second lowest proportion of pensioners of the UK regions. In 2004, 22.4 per cent of the total population of Northern Ireland was under 16 years of age (19.5 per cent in the UK as a whole), and 18.4 per cent were 60 or more years old (21.0 per cent in the UK as a whole).

2.7 The early to mid-1960s were seen as a period of economic growth in Northern Ireland. Although the rural economy still dominated, industry—including shipbuilding, light and heavy engineering, and textiles—accounted for more than 30 per cent of the economy. However, following the 1968 civil rights marches, there were high levels of civil unrest, and the economy entered a period of long-term decline. This led to low levels of investment and entrepreneurship and generally depressed conditions. The cash economy was strong compared with the rest of the UK.

2.8 In the mid-1980s, the Government encouraged the small and medium-sized enterprises sector in Northern Ireland as well as inward investment. The 1994 paramilitary ceasefires, which ultimately culminated in the signing of the Good Friday Agreement on 10 April 1998, presented new economic opportunities for Northern Ireland. A number of US and European businesses relocated to Northern Ireland, contributing to increased economic confidence, new jobs, higher incomes, and a stronger property market.

2.9 In the five years from 2000 to 2005, the economy in Northern Ireland grew at around 5.1 per cent a year, as measured by gross value added (GVA). This is broadly the same rate as the UK as a whole, which grew at around 5.4 per cent a year over this five year period. However, its GVA per head remains relatively low at 80.3 per cent of the UK-wide figure. In August 2006, the Department of Enterprise, Trade and Investment (DETI) said that the Northern Ireland economy continued to perform well, with employment at a record high and unemployment remaining low.12

2.10 Nevertheless, Northern Ireland has a relatively high rate of economic inactivity: its working age employment rate (69.5 per cent) is the lowest among the UK regions and was 5.1 percentage points below the UK average in Q4 2006. However, the unemployment rate more than halved over the decade ending in 2006 to 4.3 per cent, which is lower than the corresponding UK rate (5.6 per cent).13

2.11 Median gross annual pay for all jobs in Northern Ireland was £16,787 in 2006, around 14 per cent below the UK median of £19,496 and the lowest of the UK regions. Public sector median annual earnings in Northern Ireland were £19,742, 30 per cent above private sector median earnings (£15,235).14

2.12 Northern Ireland’s employment structure is different from that of the UK as a whole. It is heavily dependent on the public sector with around 31.4 per cent of those in work employed in the public sector (compared with 20.2 per cent across the UK as a

10Report of the Inter-Departmental Urban-Rural Definition Group, Northern Ireland Statistics and Research Agency (NISRA), February 2005. 11‘Rural areas’ are defined as towns and villages with less than 10,000 people or open country. 12http://www.statistics.gov.uk/pdfdir/gva1206.pdf. 13http://www.statistics.gov.uk/downloads/theme_labour/LMS_FR_HS/WebTable18sa.xls. 14http://www.detini.gov.uk/cgi-bin/downdoc?id=2528. http://www.statistics.gov.uk/downloads/theme_labour/ASHE_2006/tab3_7a.xls.

22

whole). In 2004/05, total identifiable public expenditure as a percentage of GVA was around 59.6 per cent compared with a UK average of 38.3 per cent.15

2.13 Financial intermediation, of which banking is a part, is under-represented in the Northern Ireland economy compared with that of the UK as a whole. In April to June 2006, 2.7 per cent of employee jobs in Northern Ireland were in financial inter- mediation,16 compared with 4.0 per cent in the UK as a whole.17 Northern Ireland increased its share of GVA for the financial services sector slightly from 1.27 to 1.29 per cent of the UK figure between 2001 and 2003. Investment Belfast told us that there had been significant local growth in financial services back-office operations with a number of customer service centres being established.

Banks that provide PCAs

2.14 PCA banking services are provided by many different organizations in Northern Ireland. For the purposes of this investigation we discuss separately, where relevant, the Northern Ireland clearing banks (the clearers)—BoI, AIB Group (UK) plc (trading as First Trust), Northern and Ulster—and those that are not clearers (the non- clearers).18 The non-clearers can be divided into three sub-groups:

(a) current and former building societies (including Abbey, A&L, Halifax, Nationwide and Woolwich);19

(b) banks based in Great Britain that have opened one or more branches in Northern Ireland since 2000 (Co-operative Bank and HSBC); and

(c) banks that provide a remote service by Internet, telephone, post and/or text messaging (the remote providers); these include cahoot (a division of Abbey), first direct (a division of HSBC), IF (a division of Halifax) and smile (a division of CFS).

We refer to the clearers and non-clearers collectively as the banks.20

2.15 We found it helpful, at times, to distinguish between the clearers and the non- clearers. As discussed further in paragraphs 4.3 to 4.13, the clearers have some distinct characteristics, such as their wide range of products (both personal and business) which has always included a PCA, and their generally more extensive branch networks, which, taken together, differentiate them from the non-clearers. Table 3 shows that, overall, the non-clearers earn less revenue per account than the clearers. We recognize, however, that on other occasions, this differentiation may not

15http://www.detini.gov.uk/cgi-bin/downdoc?id=2653. http://www.statistics.gov.uk/elmr/03_07/downloads/Table2_04.xls. http://www.hm-treasury.gov.uk/media/377/3B/cm6811_08_Chap_7.pdf. http://www.statistics.gov.uk/pdfdir/gva1206.pdf. 16http://www.detini.gov.uk/cgi-bin/downdoc?id=2438 and http://www.nomisweb.co.uk/default.asp?Session_GUID={6D97FD0C- B655-4720-90EF-F991BFFC9587}. 17http://www.statistics.gov.uk/StatBase/Product.asp?vlnk=7359&Pos=2&ColRank=1&Rank=192. 18Paragraphs 2.15 and 4.3 to 4.13 discuss the characteristics of the clearers and the non-clearers. We use the terms for the purposes of categorization, rather than to suggest that the essential difference between the two groups is whether they are members of the Belfast Bankers Clearing System. 19Woolwich is owned by Barclays. Barclays is rebranding all Woolwich branches as Barclays during 2007. Among other changes, Woolwich PCAs will in future be provided by Barclays. For simplicity we refer to Woolwich through this report although the rebranding to Barclays is underway. 20This is not intended to be a comprehensive list of all banks and building societies which offer PCAs in Northern Ireland. We also note that, although we refer to them collectively as ‘banks’, certain organizations, including Nationwide, are building societies with mutual status. References to the banks refer to all clearers and non-clearers, whether or not they are named. Further details of the banks’ parent companies are set out in Table 1, paragraphs 2.21 to 2.47 and the glossary.

23

be relevant to our assessment of the market, and have, in any case, distinguished where relevant between individual banks within one or other of the two groups of clearers and non-clearers.

2.16 We do not consider in detail other banks which may provide a service to a small number of customers resident in Northern Ireland, but do not actively market their PCAs there (eg Lloyds TSB Group plc (Lloyds TSB) and National plc (NatWest), a subsidiary of Royal Group plc (RBS)).

2.17 We also refer, where relevant, to four large banks based in Great Britain (Barclays, HSBC, Lloyds TSB and RBS/NatWest) as a point of comparison for the banks in Northern Ireland. We do not, however, regard the banks based in Great Britain as a competitive benchmark, since we have not carried out an investigation into the PCA market in Great Britain.

2.18 Table 1 sets out some key details of the banks at the level of their respective groups and in Northern Ireland. Further details are set out in paragraphs 2.21 to 2.47.

TABLE 1 Key information on banks

PCAs in NI No of full No of Total group (December branches in branches in Country assets 2006) NI GB Ultimate parent of parent £bn

Ulster 87 RBS UK 871.4a First Trust 56 Allied Irish Bank plc RoI 113.2b Northern 95 A/S Denmark 251.3c BoI 45 BoI RoI 127.1d Halifax 15 (+ 30) IF/St James’s - HBOS plc UK 591.0a Place Abbey 21 Cahoot/Cater - Spain 547.3e   Central Hispano SA Allen Nationwide 14 (+12) Nationwide UK 129.6f A&L 14 A&L UK 68.6a Barclays 996.8a Barclays UK (Woolwich) 9 HSBC 5 HSBC Holdings plc UK 1,033.8g first direct - Co-op 1 Co-operative smile - Financial Services UK 38.6e Limited

Source: Datastream, company accounts, responses to questionnaires, company websites.

Notes: 1. Latest financial information, including unaudited interim and preliminary results. a as at 31 December 2006. b as at 31 December 2006 translated from euros at 1.4. c as at 31 December 2006 translated from Danish krona at 10.9. d as at 30 September 2006 translated from euros at 1.4. e as at 31 December 2005. f as at 30 September 2006. g as at 31 December 2006 translated from US dollars at 1.8. 2. PCA numbers refer to all accounts, including inactive and secondary accounts. 3. BoI branches in Great Britain are business branches. [] 4. [] 5. Numbers in brackets are the number of agencies (Northern Ireland only). 6. RBS has a large branch network in Great Britain under the RBS and NatWest brands.

2.19 In the following paragraphs we set out some key details of the structure of the banks in Northern Ireland. Further details on business models, market shares and branch networks are included in section 4 on market structure. We discuss PCA charging structures in more detail in paragraphs 2.50 to 2.64 and paragraphs 4.109 to 4.186.

24

Clearers

2.20 All the clearers provide a full range of financial products and services to personal and business banking customers in Northern Ireland. All have offices in Belfast, but are part of larger banks based outside Northern Ireland.

BoI

2.21 BoI provides personal banking services throughout the island of Ireland. It has its group headquarters in Dublin, where it opened its first branch in 1783. It does not currently offer PCAs in Great Britain.

2.22 BoI has had a joint venture with the Post Office since 2003, in which BoI- manufactured products (excluding PCAs) are delivered through the Post Office branch network throughout the UK. In addition, in September 2004, BoI entered into an agreement with the Post Office to allow all BoI UK customers to service their PCA through any Post Office in the UK.21 In 2005 BoI sold its UK & West branch network and associated deposit base to Britannia Building Society for £150 million.

First Trust

2.23 AIB, originally named Limited, was incorporated in the in September 1966 as a result of the amalgamation of three long-established banks: the Munster and Leinster Bank Limited (established 1885), the Provincial Bank of Ireland Limited (established 1825) and the Royal Bank of Ireland Limited (established 1836). AIB’s group headquarters are in Dublin.

2.24 In July 1991, AIB acquired TSB NI, a bank with 56 branches in Northern Ireland. The AIB and TSB NI businesses in Northern Ireland were integrated as ‘First Trust Bank’, with its head office and divisional processing centre in Belfast. In October 1996, AIB’s retail operations in Great Britain and Northern Ireland were integrated as AIB Group (UK) plc with two distinct trading names: ‘First Trust Bank’ in Northern Ireland and ‘Allied Irish Bank (GB)’ in Great Britain.

Northern

2.25 Northern was originally established in Belfast in 1809. Between 1962 and 1987 it was owned by Midland Bank. In 1987 Northern was acquired by NAB. On 1 March 2005, Danske acquired Northern and its Republic of Ireland sister company, National Irish Bank (National Irish), from NAB for £967 million.

2.26 Danske, based in Copenhagen, is the largest bank in Denmark and the second- largest financial services enterprise in Scandinavia. Danske’s main divisions include Danske Bank and BG Bank in Denmark, Danske Bank in Sweden, and Fokus Bank in Norway. The group serves 3.5 million retail customers across Scandinavia and the island of Ireland with a total of around 750 branches.

Ulster

2.27 Ulster Bank Group is a wholly-owned subsidiary of RBS. Ulster Bank Group offers retail, business and corporate banking services across the island of Ireland,

21Customers cannot open a BoI PCA in the Post Office, but can perform most other activities.

25

principally through Ulster Bank Ireland Limited and plc in the Republic of Ireland, and Ulster in Northern Ireland. Ulster was founded in 1836 and has its headquarters in Belfast. Ulster Bank Group was purchased by NatWest in 1917 and by RBS in 2000.

Non-clearers

2.28 We first describe the five non-clearers with the largest market shares of PCAs in Northern Ireland: Abbey, A&L, Halifax, Nationwide and Woolwich.22 We then describe more briefly the banks based in Great Britain that have launched PCAs in Northern Ireland since 2000 (Co-operative Bank and HSBC). The key remote providers are discussed under their parent companies.

2.29 In general, the non-clearers offer a broad range of products and services to personal customers across the UK, but their business banking offer is more restricted. The larger non-clearers have high market shares in personal loans and mortgages, consistent with their origins as building societies. The Building Societies Act 1986, which came into force in 1987, allowed building societies to increase the range of financial services offered, including PCAs. Several of the non-clearers’ PCAs can be accessed via the Post Office.23

Abbey

2.30 The Abbey National Building Society was formed in 1944 by the merger of two building societies. In 1988 the building society transferred its business to Abbey National plc, converted to a bank and was listed on the Stock Exchange.

2.31 On 12 November 2004, Abbey was acquired by Banco Santander Central Hispano SA (Banco Santander). Abbey remains a separate legal entity incorporated in England and Wales and authorized and regulated by the FSA. Banco Santander is Spain’s largest financial group. It has a presence in 40 countries, services more than 63 million customers through branch networks with a total of 10,000 branches. The group focuses on retail banking, which makes up around 85 per cent of its revenues, as well as providing other financial services.

2.32 Abbey also provides PCAs in Northern Ireland through cahoot (a division of Abbey) and Limited (a subsidiary of Abbey). cahoot is Abbey’s separately branded e-commerce retail banking and financial services provider. Cater Allen Limited offers banking services to high net worth individuals and small corporate clients by telephone and by post.

2.33 Abbey has had a retail branch presence in Northern Ireland since the 1940s. Its business in Northern Ireland is managed as part of the Northern Ireland and North- West of England region.

A&L

2.34 Leicester Permanent Benefit Society was formed in 1852. The Brighton and Sussex Equitable Building Society was formed in 1863, and later amalgamated with other

22In many cases we carried out our competitive assessment based on the five largest non-clearers listed. However, our financial analysis looked only at the clearers and the four largest non-clearers (Abbey, A&L, Halifax and Nationwide), covering more than 95 per cent of the market. 23This includes cahoot, A&L, Nationwide, Co-operative Bank and smile.

26

local building societies to form The Alliance Building Society. In 1985 these two societies merged to form Alliance & Leicester Building Society, with its headquarters in Leicester. In 1997 it converted to a public limited company and was floated on the .

2.35 Leicester Permanent opened its first branch in Northern Ireland in Belfast in 1963, and the majority of its branches in Northern Ireland opened in the late 1970s.

Halifax

2.36 The Halifax Permanent Benefit Building Society was established in 1852. In 1997, the building society converted to a public limited company and subsequently floated on the London Stock Exchange. HBOS plc (HBOS) was established in 2001 as a result of the merger between Bank of Scotland and Halifax plc. PCAs are part of the retail bank business which has its headquarters in Halifax. HBOS sells its PCAs mainly under the Halifax brand in Northern Ireland, England and Wales.

2.37 Halifax has had a presence in Northern Ireland for over 75 years. It has a mix of branches and agencies.

2.38 Halifax launched an Internet bank, IF, in 2000. In 2001, Halifax acquired a stake in St James’s Place Bank (St James’s Place), which serves high net worth private clients, a small number of whom are resident in Northern Ireland.

Nationwide

2.39 Nationwide is a mutual building society whose origins lie in Northampton (1848) and within the cooperative movement in London (1883). Following over a hundred mergers—most notably between the Nationwide and Anglia building societies in 1987—and the of many former building societies, Nationwide is now the largest building society in the world.24 It has its headquarters in Swindon.

2.40 Nationwide’s mutual status means that it is owned by and run for the benefit of its members. Nationwide returns profits in part through pricing benefits to its members. Given its mutual status, Nationwide has no equity and therefore its capital structure is different from that of the other banks.

2.41 Nationwide opened its first three branches in Northern Ireland in the late 1940s.

Woolwich and Barclays

2.42 Woolwich was formed in 1847 and converted to a bank in 1997. Barclays acquired Woolwich in 2000. On 28 June 2006, Barclays announced that it intended to convert Woolwich branches to Barclays branches in 2007.25 The first of these, in Belfast and Lisburn, were migrated to Barclays branches in March 2007, with the remaining branches in Northern Ireland expected to be converted by October 2007.26 Barclays told us that it intended to migrate all customers with Woolwich Current Accounts and Open Plan Current Accounts to the Barclays Bank Account, and customers with a

24On 12 September 2006, Nationwide and (Portman) announced that they planned to merge. The savers and borrowers of Portman approved the merger on 23 April 2007 and the proposed merger is expected to become effective on 28 August 2007, subject to confirmation by the FSA. The combined entity would be called Nationwide Building Society. Portman does not provide PCAs and has no branches in Northern Ireland. 25Barclays told us in September 2006 that a Barclays till had recently been introduced into the Woolwich branch in Belfast. 26Previously there had been no Barclays-branded branches in Northern Ireland.

27

Woolwich Open Plan Current Account linked to a mortgage would be migrated to a new product, Barclays Mortgage Current Account. Woolwich will continue to be the mortgage brand.

2.43 Barclays provides a full range of financial products and services to personal and business banking customers. It operates in more than 50 countries and employs over 120,000 people.

Co-operative Financial Services

2.44 CFS, part of the Co-operative Group, offers a range of financial products, including high street banking (the Co-operative Bank) and Internet banking (smile). CFS has its headquarters in .

2.45 The Co-operative Bank opened its branch in Northern Ireland in 2003; smile PCAs have been available to customers in Northern Ireland since smile’s launch in 1999.

HSBC

2.46 HSBC, headquartered in London, is one of the largest banking and financial services organizations in the world. It is listed on the London, Hong Kong, New York, Paris and Bermuda stock exchanges. HSBC opened its first branch in Northern Ireland in 2001.27

2.47 first direct is a division of HSBC and was launched in October 1989. It offers telephone, Internet and text messaging banking. It has no branches.

Characteristics of PCAs

2.48 A PCA is defined in our terms of reference as ‘… an account, marketed to individuals not businesses, which provides the facility to hold deposits, receive and make payments (cheques and debit cards) and use ATM facilities and to make regular payments (direct debit and standing orders)’.

2.49 Many accounts which meet this definition also have additional characteristics (such as a facility for short-term borrowing through overdrafts). We discuss product market definition further in section 3.

PCA charging structures

2.50 There are a large number of possible charges associated with operating a PCA, reflecting the many potential functions of a PCA. The number of charges listed in the banks’ fees and charges brochures will depend on the number of different ancillary services the bank chooses to detail, the number of accounts it offers, and whether the brochures cover both personal and business use.

2.51 In most cases, customers in Northern Ireland are not charged for operating a PCA when it is in credit, although all customers incur ancillary charges for irregular, one- off services provided on request, regardless of whether their account is in credit or

27HSBC had been present in Northern Ireland for more than 40 years through HSBC Asset Finance, part of its business banking activities. HSBC was formed through the acquisition of Midland Bank in 1992 by Hong Kong and Shanghai Banking Corporation (HongKong Bank). As noted in paragraph 2.25, Midland Bank owned Northern between 1962 and 1987. HSBC Asset Finance was formerly part of the Midland Bank group.

28

debit. Ancillary charges are incurred for the provision of certain services such as making CHAPS payments or providing bank drafts, and for additional administration such as stopping a payment or providing a replacement bank statement. Customers may receive interest when the account is in credit.

2.52 There are two basic PCA charging structures in Northern Ireland: the historical charging structure (the ‘traditional’ PCA) and the ‘fee-free’ PCA. These are described in more detail in paragraphs 2.58 to 2.62. In addition to ancillary charges, customers may incur the following charges on one or other (or sometimes both) of these PCA structures:

(a) transaction charges: levied on a per-transaction basis;

(b) maintenance charges: fixed payments (monthly, quarterly, yearly or ad-hoc) arising from the operation of the PCA; and

(c) charges for setting up, amending or cancelling standing orders or direct debits.28

Authorized and unauthorized overdraft charges may include both transaction and maintenance charges. In addition, customers will normally pay interest when operating their PCA in overdraft.

2.53 In addition to these two basic charging structures, there are a number of other structures. First Trust, Ulster, Northern, HSBC and the Co-operative Bank offer packaged accounts, where customers typically pay a monthly or quarterly fixed charge regardless of whether the account is in credit or debit, but do not pay any transaction charges. Halifax launched a packaged account in February 2007. Packaged account customers are also likely to receive preferential terms on other characteristics of the account (for example, on credit interest rates or authorized overdraft rates), may be exempt from some ancillary charges, and often receive other bundled benefits such as free travel insurance or discounts on some purchases. Some accounts, such as Northern’s ChequeMaster and A&L’s Current Account, levy a fixed monthly charge for authorized overdrafts but do not levy transaction charges or apply maintenance charges in months other than those in which the customer was in authorized overdraft.

2.54 The characteristics of each of the PCAs offered by the nine largest banks in Northern Ireland are listed in Appendix 2.4, including, for each PCA, the levels of interest and charges.29 These are grouped by charging structure. The proportion of ‘traditional’ PCAs in Northern Ireland has fallen following the migration of all traditional accounts to a fee-free charging structure by Ulster in November 2005 and BoI in November 2006, and the introduction of fee-free accounts by Northern (although Northern continued to operate traditional PCAs). Table 2 shows the distribution of PCAs by type of PCA in September 2006 and January 2007.30 We estimate that, as at September 2006, more than one-third of all PCAs in Northern Ireland, and almost half of the clearers’ PCAs, were still traditional PCAs.31 By January 2007 this had fallen to less than 20 per cent of all PCAs in Northern Ireland and less than 30 per

28These are sometimes referred to as ancillary charges by the banks. For the purposes of this report we treat these charges separately to ancillary charges as defined in this report. 29The nine largest banks are the clearers and the first group of non-clearers listed in paragraph 2.14(a). 30Data was collected on these two dates: September 2006 was shortly before we published our provisional findings, and January 2007 was the latest practicable date for obtaining market-wide data for inclusion in this report. 31Customers of the clearers not on a traditional PCA in September 2006 included all customers of Ulster, Northern’s customers on fee-free accounts, those with student accounts, other fee free accounts and other accounts including packaged accounts and accounts levying a fixed charge for authorized overdrafts such as Northern’s ChequeMaster.

29

cent of the clearers’ PCAs, largely due to BoI transferring its customers to fee-free PCAs in November 2006.

TABLE 2 Distribution of PCAs in Northern Ireland by type of account in 2006

per cent

September 2006 January 2007 Number of Share of all Number of Share of all Type of PCA PCAs PCAs in NI PCAs PCAs in NI

Fee-free 613,070 49.8 950,798 68.0 Traditional 446,621 36.3 264,034 18.9 Other, incl packaged 170,610 13.9* 182,898 13.1* Total 1,230,301 100.0 100.0

Source: CC analysis of information provided by the banks (updates to market questionnaire).

*More than 40 per cent of ‘Other’ are packaged accounts.

2.55 We concentrate in the following analysis on traditional and fee-free PCAs as these represent the great majority (more than 85 per cent) of PCAs. We describe the traditional PCA in paragraphs 2.58 to 2.60 and the fee-free PCA in paragraphs 2.61 and 2.62.

2.56 We note that there are some similarities in charging structures between the large banks based in Great Britain and the clearers (with the exception of Northern’s new account portfolio; see paragraphs 4.145 and 4.146). In particular, both groups tend to offer low levels of credit interest, and offer packaged accounts. However, traditional PCAs do not exist in Great Britain.32

2.57 We note that individual banks sometimes use different terminology to describe the same type of transaction, particularly in relation to unauthorized overdraft charges. This is discussed further in paragraphs 4.132 to 4.139. We have attempted throughout this report to use consistent terminology, defined in the glossary, to aid understanding.

Traditional PCAs

2.58 We define traditional PCAs to be PCAs where transaction charges and maintenance charges are payable when the account is operated in authorized or unauthorized overdraft. We therefore include First Trust’s traditional PCAs (under their current charging structures (see paragraph 2.63), as well as Northern’s older account (Current Account). Ulster and BoI offered traditional PCAs until November 2005 and November 2006 respectively. Transaction charges apply to a variety of regular transactions such as cheques, ATM withdrawals, debit card payments, and standing order and direct debit payments. Maintenance charges are fixed charges, generally levied on a quarterly basis. Both transaction and maintenance charges are levied during a quarter if the account is overdrawn at any point during that quarter whether or not the overdraft has been authorized. Most traditional PCAs also levy charges for setting up, amending or cancelling standing orders or direct debits, whether the

32The charging structures we discuss cover those currently in use in Northern Ireland. These are similar to those found in Great Britain. However, different charging structures are commonly found in other countries; for example annual subscription charges or debit card subscription charges are frequently applied (see for example the Oxera report ‘The price of banking: an international comparison, report prepared for the British Bankers’ Association, November 2006’).

30

account is in credit or in debit.33 There are no transaction or maintenance charges levied if the account is in credit for the entire quarter. Authorized overdraft limits and interest rates may be subject to negotiation.

2.59 Additional charges are levied when a PCA customer either exceeds their agreed overdraft limit or goes overdrawn without agreement. These unauthorized overdraft charges may include, for example: maintenance charges for breaching the agreed overdraft limit or notifying the customer of their unauthorized overdraft position; and transaction charges such as referral charges levied when the bank makes a payment which creates or extends an unauthorized overdraft or unpaid charges for returning cheques unpaid, or declining to pay a standing order or direct debit that, if paid, would create or extend an unauthorized overdraft. Customers also generally pay a considerably higher rate of interest on unauthorized overdrafts than they do on authorized overdrafts. In general, the charges for operating in unauthorized overdraft mode are lower on traditional accounts than on fee-free accounts. This is illustrated by the customer scenarios work that we did to estimate charges on a variety of accounts in authorized and unauthorized overdraft mode (see paragraphs 4.124 to 4.126 and Appendix 4.7).

2.60 Credit interest rates on traditional PCAs are typically very low (0.1 per cent) or zero.

Fee-free PCAs

2.61 The second basic charging structure is that of the fee-free PCA. It is described as ‘fee-free’ because transaction and maintenance charges are not levied when the customer is in credit or in authorized overdraft. However, fee-free PCAs, like traditional PCAs, are subject to a range of ancillary charges. Fee-free PCAs are also subject to unauthorized overdraft charges as described in paragraph 2.59. These are often set at higher levels than those for traditional PCAs.

2.62 Some banks choose to offer more attractive credit interest rates on certain fee-free PCAs than apply on traditional PCAs. If certain conditions are met, these can be as high as 6 per cent, although credit interest rates are often much lower on PCAs which do not meet such conditions (see paragraph 4.164).

Recent changes to PCAs

2.63 There have been a number of changes to PCAs in Northern Ireland in the past two years, particularly those offered by the clearers.34 We outline these changes by bank; these are discussed in more detail in paragraphs 4.140 to 4.161.

(a) BoI launched a fee-free PCA charging structure in November 2006; customers on traditional PCAs were migrated automatically to this new charging structure.

(b) Northern removed charges for setting up, amending or cancelling standing orders and direct debits from its traditional PCAs in February 2005.35 It launched Current Account Plus, a fee-free account which was available between February 2005 and April 2006 to new customers and to existing customers who applied to

33With the exception of Northern, where charges for setting up, amending or cancelling standing orders and direct debits were first removed from their student, graduate and retiree accounts in 2002/03 and were removed for all accounts from February 2005. 34During this period there have been no significant new offers by the non-clearers, and any such new offers that have been introduced apply across the UK. 35Northern removed these charges from its retiree, student and graduate accounts in 2002/03.

31

change their account; existing customers were not automatically transferred to Current Account Plus. In July 2005, Northern also removed certain charges, namely overdraft arrangement and renewal charges, on all its PCAs. Northern launched a new PCA product portfolio in April 2006, including fee-free and packaged accounts.

(c) Ulster launched a new, fee-free PCA charging structure in November 2005, applying to all PCA customers other than those with packaged accounts.36

(d) First Trust announced on 3 May 2007 that it intended to launch a new, fee-free PCA in June 2007. It would also offer free standing order and direct debit set up, amendment and cancellation on all PCAs.

2.64 We discuss issues associated with PCA charging structures, trends in charges, and recent changes in charging structure in more detail in paragraphs 4.109 to 4.186.

Overview of financial performance

2.65 We reviewed the financial performance of the eight largest banks in Northern Ireland: the four clearers—BoI, First Trust, Northern and Ulster—and the four largest non- clearers—Abbey, A&L, Halifax and Nationwide. Together these provide over 95 per cent of PCAs in Northern Ireland.37

2.66 We experienced considerable difficulties in obtaining reliable information to enable us to understand the financial performance of the banks’ PCA businesses. This was due, in part, to the fact that none of the banks report routinely on the profitability of their PCAs in Northern Ireland. The clearers do not routinely assess the profitability of their PCA business alone, looking instead at the profitability of their retail business in Northern Ireland as a whole including both PCAs and other products and services. The non-clearers generally assess the profitability of their PCAs, but only for the UK overall, or at least for regions larger than Northern Ireland. The information on their PCA business in Northern Ireland was provided for the specific purposes of our investigation, and required each bank to allocate their revenues and common costs. We therefore had to interpret the data broadly, drawing conclusions based on a range of multi-period, multi-bank data wherever possible.

2.67 Table 3 summarizes the financial position of the banks in 2005. Because two of the banks have March year ends, 2005 is the last complete year for which we have data. We requested certain 2006 information in order to understand how the changes to PCA charging structures affected the four largest banks. However, the 2006 information does not capture all changes (for example, BoI’s new PCA portfolio launched in November 2006). Furthermore, First Trust was unable to provide the requested information in the time frame. Further details are set out in paragraphs 4.312 to 4.339 and Appendix 4.10. Customers in Northern Ireland held 1.3 million PCAs in 2005, of which 1.0 million were active. The number of accounts has grown at an average of 3 per cent a year since 2002, with most of the growth coming from the non-clearers.

2.68 Credit balances provide a bank with funds that it uses elsewhere in the bank to earn income. Average credit balances in 2005 were £2.1 billion, which represents average

36We note, in addition, that First Trust’s packaged account levies a fixed charge rather than transaction charges, and BoI’s Clear account is fee-free only for the first year. 37We note that Nationwide is a building society with mutual ownership which passes benefits on to its members via pricing benefits.

32

annual growth of over 10 per cent since 2002. Average credit balances per account are higher for the clearers as a group (2005: £1,796) than for the non-clearers (2005: £1,184). Overdraft balances represent approximately 7 per cent of the value of credit balances.

2.69 The banks earn income in three forms from providing PCAs:

(a) NII on credit balances (value of funds elsewhere in bank less interest paid to customers);

(b) NII on debit (overdraft) balances (interest received from customers less cost of providing funds); and

(c) charges.

2.70 In 2005 the eight banks earned income in total of £167 million from providing PCAs, which represents an annual average increase since 2002 of nearly 8 per cent. On average the banks earn more than half their total income from NII on credit balances, over 40 per cent from charges and 5 per cent from NII on debit balances.

2.71 Our review of the financial performance of the banks over the period 2002 to 2005 illustrates the financial impact on the banks of the differences in charges and credit interest rates between the two charging models. During the period covered by our financial analysis the vast majority of the clearers’ PCA customers held traditional PCAs. As noted in paragraph 2.63, there have been a number of changes to PCAs in Northern Ireland. Our review suggests that the non-clearers earn less revenue directly from PCA customers per account (£41 in 2005) than the clearers (£64 in 2005), and pay higher average credit interest per account (1.02 per cent for the non- clearers compared with 0.10 per cent for the clearers) (see Table 3), although these averages mask some significant differences in the revenues per customer earned by individual banks. This is discussed further in paragraphs 4.312 to 4.339.

TABLE 3 Summary financial information, 2005 Clearers Non-clearers Total

Total number of accounts (‘000) 923 362 1,285 Total number of active accounts (’000) 762 267 1,029 Total revenue (net interest income and charges) (£m) 133 33 167 Average revenue per account (£) 145 92 130 Total average credit balances (£m) 1,657 429 2,085 Average credit balance per account (£) 1,796 1,184 1,623 Average revenue earned directly from customer per account (£) 64 41 58 Average credit interest paid on credit balances (%) 0.10 1.02 0.29

Source: CC analysis of response to Financial Questionnaire (including updates).

Note: All averages are weighted by the number of accounts. Average credit balances are calculated on a daily basis. Totals may differ due to rounding.

Customer characteristics

2.72 We looked at the characteristics of PCA customers and the levels of switching between banks.38 In order to set the context, we considered the work that had been

38We drew on various sources of evidence in order to gain an understanding of customer characteristics and behaviour. These included research commissioned by the banks, and the results of the MORI survey. In addition, we commissioned three surveys of Northern Ireland customers: a quantitative survey of 1,252 consumers in Northern Ireland, including a sample who were not PCA customers (the ORC survey); a qualitative survey of attitudes to banks and current accounts among 14 varied

33

carried out by the FSA to assess the level of financial capability in the UK.39 This work highlighted the fact that the economic and social environment in which people take financial decisions has been changing, and that this change is expected to continue. It also suggested that there are serious concerns about current levels of financial capability throughout the UK. Of particular relevance to our investigation, the survey found that people do not take adequate steps to choose products to meet their needs, and that 18- to 40-year-olds, and particularly 18- to 30-year-olds, are less capable than their elders to do so.

2.73 We also note that UK customers are generally not particularly interested in PCAs. Most consumers have a wide range of possible decisions to make on suppliers of a range of financial and non-financial services (eg mobile phones, utilities, insurance, mortgage etc), and switching bank is rarely considered to be a priority. The 2001 DTI report into switching suppliers says that ‘Most consumers are not motivated to even consider switching as their interest in this market [current accounts] is low’.40 In response to our Emerging Thinking, Which? told us that its research into financial brands suggested that the PCA was a ‘low visibility’ product; customers appear to pay little attention to it except when it goes wrong.

2.74 The current account is the most widely-held personal banking product in Northern Ireland. About 80 per cent of households in Northern Ireland have access to a current account, compared with almost 90 per cent in the UK as a whole.41 We were told that penetration rates in Northern Ireland increased from about 68 per cent in 1999. Penetration rates of PCAs in Northern Ireland are lowest for young people, pensioners, and lower socio-economic groups.

2.75 Evidence from the MORI survey for 2005 indicated some differences between the characteristics of different banks’ customers.42 The results suggested:

(a) The average age of a clearers’ customer is around 44, four years older than the non-clearers’. Clearers have a higher proportion of their customer base in the over-45 age groups, whereas the non-clearers have a higher proportion of their customer base among customers aged 24 to 44. The proportions of customers in the 18 to 24 age group are similar for clearers and non-clearers. Within the clearers, there are some significant differences: the average age of First Trust and of Northern customers in particular is above other banks’; BoI has the lowest average customer age of both the clearers and the non-clearers.

(b) Average customer income is similar for customers of the clearers and non- clearers.43 The most notable exceptions are Nationwide and A&L with above- average customer incomes, and Ulster with slightly below-average customer incomes.

(c) The distribution of customers by socio-economic group varies between banks. Northern, First Trust and Nationwide appear to have a higher proportion of customers from higher socio-economic groups than the average. Overall the distribution of customers between the clearers and non-clearers by socio-

discussion groups (the Millward Brown survey); and a quantitative survey looking particularly at overdraft charges and charging practices (the BMRB survey). These are described in more detail in Appendix 1.1. 39Financial Capability in the UK: Establishing a Baseline, FSA, March 2006. 40Switching Suppliers, Consumer Affairs Report, 2001—DTI (2001), Chapter 3, page 25. 41NISRA Family Resources Survey 2004/05 found that 83 per cent of all Northern Ireland households had access to a current account compared with 90 per cent in the UK as a whole. ‘Current account’ includes the basic bank account. 42These results are also consistent with the results of a consumer survey published by Mintel—Customer Retention, to Switch or Not to Switch, Special Report January 2006, pages 32–43. 43Although around 25 per cent of customers chose not to give their income.

34

economic group is very similar. The non-clearers have a smaller proportion of their customer base who are retired or unemployed.

(d) The market share of the non-clearers varies substantially between the regions of Northern Ireland, but is highest in and around Belfast. Ulster and Northern draw a proportionately higher share of their customers than average from rural areas while First Trust and Halifax are relatively strong in urban areas.

2.76 We looked at the reasons given by customers for their choice of bank. In general, when customers first choose their bank, they are particularly influenced by branch location, whether close to their home, place of work, or other frequently visited location. They are also more likely to choose a particular bank if other family members already bank there, if charges are low, and, to some extent, if credit interest rates are high. Our ORC survey found that only around one-third of customers use the branch to transact the majority of their PCA business. We found that relatively fewer younger people said that branch location was important to them in their choice of bank (14 per cent compared with 27 per cent overall), perhaps reflecting a move towards a greater use of remote banking (see paragraph 4.73 and Appendix 4.3).

2.77 PCA customers normally stay with their chosen bank for a long time. The ORC survey showed that the median time since opening a main PCA for all customers was 13 years and that more than three-quarters of all PCA customers had been with their main bank for more than six years.

Levels of switching

2.78 Customers can switch providers in one of two ways: by opening another PCA with a different bank and arranging for their direct debits, direct credits and standing orders to be transferred from the old to the new account (hidden switching), or by using a bank’s switching service. The evidence that we received suggested that around 40 per cent of switchers used a bank’s switching service (see paragraph 2.85). In the case of hidden switching, customers do not necessarily inform the original bank, or close down their existing PCA. Banks experience a high number of accounts becoming dormant, which they attribute to hidden switching.

2.79 If the switching service is used, the new bank takes responsibility for the switching process. The Code sets out the information that must be given to customers at each stage of the process, as well as outlining the roles and responsibilities of the old and new bank and the timescales which they must meet. Under the Code, customers are not charged for switching.

2.80 When opening a first account or switching to an account with a new bank, customers must be informed of the charges and interest rates that apply to their new PCA. They will then be treated in exactly the same way as the bank’s existing customers, being sent regular statements showing charges and interest incurred and interest earned. In addition, before certain charges are debited from their account, they are sent a pre-notification of the type and level of charges. Appendix 4.2 sets out the details of the switching process.

2.81 The banks had difficulty giving us overall switching rates because, in general, they only monitor the levels of switching occurring through their switching service, and cannot monitor hidden switching directly. In addition, the non-clearers generally collect data at a UK-wide level. The clearers told us that there were high levels of hidden switching that were not necessarily included in the survey results. Ulster reported total switching rates of over [] per cent a year after allowing for hidden

35

switching. This was arrived at by aggregating customers whose accounts became dormant for reasons other than switching to another bank (eg switched accounts within the same banking group (RBS), or where the customer died). A non-clearer ([]) told us that around 23 per cent of their PCAs were opened as second accounts, with the customer’s first PCA remaining open.

2.82 We therefore looked at survey evidence to give an indication of total levels of switching in Northern Ireland (ie including hidden switching). Both the MORI survey and the ORC survey asked customers whether they had changed their main bank, thus capturing both types of switching. Although we found that up to 25 per cent of customers have at least two accounts, the survey companies did not report any difficulties with respondents identifying their main PCA.44

2.83 The ORC survey estimated the total proportion of PCA customers who have switched suppliers to be 11 per cent over the past five years, equating to between 2 and 2.5 per cent each year.45 The MORI survey (2005) showed that 22 per cent of the total Northern Ireland population had switched bank at some time in the past. Annual rates of switching are calculated from this MORI data in Table 1 of Appendix 4.3. Switching rates have been between 1.5 and 2 per cent each year since 2002. A Mintel report from January 2006 found that 4.1 per cent of the survey sample in Northern Ireland had switched PCA in the previous 12 months, and a further 8.1 per cent in the period between one and three years previously.46

2.84 The BMRB survey looked at switching patterns among customers who had experienced unauthorized overdraft charges, and authorized overdraft charges on traditional PCAs. The samples are therefore not representative of all PCA holders. It was found that 8 per cent of customers who thought that they had paid unauthorized overdraft charges in the first half of 2006 had switched bank between January and October 2006, significantly more than the 2 per cent switching rate among those who knew they had not been charged. No significant difference was found in the level of switching between those who had or had not incurred authorized overdraft charges.

2.85 As set out in paragraph 2.82, the ORC estimate includes both switchers who used the banks’ switching service and those switchers who transfer their accounts themselves. Of the respondents classed as switchers by the ORC survey, only 38 per cent had used the switching service of their new bank while 60 per cent had made the arrangements themselves. This is in line with the estimates of the ratio of switching via the banks’ switching service to self-switching given to us by the banks, which estimates that approximately two-thirds of all switching is hidden.

2.86 In addition, we would expect the individual rates of switching experienced by the banks to differ from the overall market switching rate, to reflect changes in market shares over time. We attempted to obtain estimates of individual switching in and switching out (attrition) rates from the banks, allowing for account closures due to death, dormancy and bad debts. Only one of the banks was, however, able to supply this information (see paragraph 2.81). Further details are set out in Appendix 4.3. In the absence of other data, we consider the switching rates identified in surveys (see paragraphs 2.82 to 2.85) to be the best estimate of true overall switching rates.

44The MORI survey (2005 Q2) indicates that about 11 per cent of those with a PCA in Northern Ireland have at least two accounts; the ORC survey found that 25 per cent of respondents had another PCA with a different institution. 45The precise annual figure depends on whether account holders switched more than once during the five-year period. 46Customer Retention: to Switch or Not to Switch, Special Report January 2006.

36

2.87 Data from switching services indicates that overall, switchers from the clearers are more likely to move to the non-clearers, although recent patterns of switching over a short period have differed greatly between the clearers, which may indicate changes in the nature of the churn in the market.47 All the clearers have a lower share of switchers than their share of flow would suggest (see paragraph 4.30). However, 55 per cent of those who switched bank at some point switched to one of the clearers (based on 2005 MORI data). MORI data also shows that since the start of 2003 the clearers have picked up 35 to 45 per cent of all customers who switched their main accounts.48

2.88 By 2006 Q4 the clearers’ share of flow had reduced to 31 per cent of all switchers compared with 57 per cent for all accounts. This is considerably below their share of the stock of PCAs, and indicates that switching may be a significant factor in the decline of the clearers’ aggregate market share by volume.

2.89 We also looked at switching rates for other financial products and in other industries and considered whether this might inform our view of switching rates for PCAs. We recognized that there were various difficulties in comparing switching rates across industries, as different industries would have different incentives and costs of switching. Northern, for example, said that the incentives to switch PCAs were low since the majority of PCA customers paid no fees. BoI said that low switching was a characteristic of the market. PCAs were based on trust, loyalty, and a high degree of customer satisfaction. Nevertheless, we observe that switching rates for PCAs are low compared with the rates of switching seen in other industries across the UK (eg mobile phones, utilities, insurance, mortgages etc).

2.90 A DTI report49 researched the reasons for consumer reluctance to switch in a variety of utility and service industries in May 2000. This was updated by the National Consumer Council (NCC) in 2005. The results are shown in Figure 1. They show that, across the UK as a whole, the proportion of switchers for PCAs is lower than for any of the other industries or products surveyed (except savings accounts in 2005), and few consumers who had not switched had considered doing so. The DTI report said that the low switching rates compared with other industries demonstrated high levels of customer inertia and loyalty.

2.91 We also looked at comparative rates of switching PCAs with Great Britain. There is mixed evidence on switching rates for PCA customers in Northern Ireland compared with those in Great Britain. One non-clearer said that the NOP survey estimated that 20 per cent of all new current accounts in 2005 in Great Britain had been switched.50 This had fallen from 23 per cent of stock in a similar survey in July 2002. One clearer quoted NOP survey results of an annual rate of switching in Great Britain of 1.1 per cent in 1999, rising to 1.7 per cent in 2001 and falling to 1.4 per cent in 2003.

2.92 Current switching rates may be higher in Northern Ireland than in Great Britain. This view was supported by the experience of several of the non-clearers, who said that there was a greater propensity for customers to switch to them from the clearers in Northern Ireland than for customers to switch to them from the banks as a whole in other parts of the UK. Until fairly recently, rates of PCA switching were falling in Great Britain, possibly as public interest in, and media attention to, PCAs had been declining. We note, however, that more recently, there has been increased media

47[] 48This compares with 55 to 60 per cent of all account openings, since the clearers have a higher share of first-time accounts. 49Switching Suppliers, Consumer Affairs Report, 2001 (DTI 2001). 50NOP Financial Research Survey in Great Britain (July 2005).

37

attention on PCAs throughout the UK, particularly in relation to unauthorized overdraft charges. In any case, higher rates of switching in Northern Ireland than in Great Britain would not, in themselves, necessarily indicate a more competitive market in Northern Ireland. As set out in paragraph 2.17, we have not carried out an investigation of the market in Great Britain and have made no assumptions, nor have any view, as to whether Great Britain provides a competitive benchmark.

FIGURE 1

Proportion of customers who had switched or considered switching suppliers for different industries, 2000 and 2005

100%

90%

80% 40 46 48 47 52 52 56 70% 61 62 71 71 78 79 60% 82 neither (%) 50% 10 considered (%) 15 29 switched (%) 11 23 40% 20

13 30% 18 32

44 11 20% 18 37 37 31 15 28 30 15 26 11 10% 20 17 11 12 7 6 7 0% 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 Current Gas Electricity Fixed Mobile Home Savings Mortgage account telephony telephony insurance account

Source: NCC (2005).

2.93 We discuss customer behaviour, particularly in relation to switching, in more detail in section 4.

Regulatory environment, the Code, payment systems and provision of bank notes

Banking regulation

2.94 Banking is a heavily regulated activity. The Financial Services and Markets Act 2000 (FSMA) provides the regulatory framework for the financial services sector in the UK, including banking. Among other things, it sets the structure for bank supervision, which is also subject to extensive European Community (EC) legislation and inter- national agreements (see Appendix 2.1). The Basel Accord provides a framework of prudential regulation centred on minimum capital standards reflecting the level of risk to which each bank is exposed. It is translated in EU law through a series of EC directives. The European Commission is committed to developing a single market in financial services: a series of directives and regulations has been adopted which, for example, enable a bank licensed by any national regulator to offer services through- out the EU.

38

2.95 Responsibility for overall financial stability issues in the UK is shared among the FSA, the Bank of England (BoE) and HM Treasury (HMT).51 The FSA is responsible for the authorization and prudential supervision of banks, building societies and other financial institutions52 and for the conduct of business aspects of banks’ investment business. The BoE is responsible for the overall stability of the financial system. HMT has responsibility for the overall institutional structure of financial regulation and the legislation that governs it. Further details are set out in Appendix 2.1.

2.96 The provision of PCAs is a regulated activity. Subsidiaries of UK banks are authorized by the FSA to take deposits.53 They must comply, in particular, with all relevant FSA rules and obligations under the FSMA. The FSA undertakes their prudential supervision. Passported branches of an EEA bank providing PCAs in the UK will generally be authorized and regulated by the EEA state in which the bank’s head office is located, although certain of the FSA’s rules will still apply to them.54

2.97 All the banks operating in Northern Ireland are authorized and regulated by the FSA, with the exception of BoI, which is authorized and subject to prudential regulation by the Irish Financial Services Regulatory Authority (the Irish Financial Regulator).55 The prudential supervisory responsibilities for each bank’s parent company are set out in Table 4. All the Northern Ireland banks subscribe to the Code (see paragraph 2.100).

TABLE 4 Supervision of banks in Northern Ireland

Bank Ultimate parent company Prudential supervisor

BoI BoI The Irish Financial Regulator First Trust Allied Irish Banks plc The Irish Financial Regulator/FSA* Northern Danske Bank A/S Finanstilsynet/FSA* Ulster plc FSA Abbey Banco Santander Central Hispano SA Banco de España/ FSA* A&L A&L FSA Halifax HBOS plc FSA Nationwide Nationwide FSA

Source: CC analysis of information from banks and regulators

*The overseas financial regulator is the lead prudential supervisor for the group but the FSA has prudential responsibility for the UK registered subsidiary.

2.98 Lending to consumers, including overdrafts, is regulated under the CCA 1974, as amended by the CCA 2006 (see Appendix 2.1).56 The CCA 1974 requires most businesses that offer goods or services on credit or lend money to consumers to be licensed by the OFT. Among other things, the amendments made by the CCA 2006 will also enable borrowers (a) to challenge unfair conduct by lenders in court and to obtain redress; (b) to take complaints about lenders to the Financial Ombudsman Service; and (c) to receive more information about the state of their accounts, which should help them to identify problems at an early stage. We discuss the likely impact of CCA 2006 further in section 6.

51The FSA also works closely with other bodies such as the OFT, the BCSB and the Financial Ombudsman Service. 52Includes investment firms, insurance firms, insurance and mortgage companies and brokers, credit unions and friendly societies. 53As set out in FSMA (Regulated Activities) Order 2001. 54When a bank operates branches in another country without establishing a separate trading subsidiary in that country, pruden- tial regulation of its activities there is entirely the responsibility of the home state regulator. The necessary regulatory authoriz- ations are said to be ‘passported’. 55Certain FSA rules do, however, apply to BoI. 56The CCA 2006 is being implemented in phases; at present the final phase, bringing the post-contract transparency require- ments into force, is expected to be implemented in October 2008.

39

2.99 There are high regulatory barriers for a company entering banking in Northern Ireland for the first time. For an existing bank, much depends on where it already operates:

(a) A bank operating in Great Britain will already have satisfied UK regulatory requirements and will face no further regulatory barriers to operating in Northern Ireland.

(b) A bank operating in the Republic of Ireland or another EU or EEA country will already have been authorized and satisfied prudential regulation requirements in its home country. This would enable it to open a ‘branch’ business in Northern Ireland on a ‘passported’ basis (see first footnote to paragraph 2.96).

(c) For those banks based outside the EEA, the FSA will seek to apply its statutory minimum ‘threshold conditions’ to the bank’s operations as a whole, after taking account of supervisory work done by overseas regulators as appropriate.

The Code

2.100 The Code is a voluntary UK code of practice. It has been adopted by all the banks active in Northern Ireland. It sets minimum standards for how banks should treat their customers and for the information that they should provide to them.57 It covers a wide range of subjects including changes to interest rates and charges, terms and conditions, guidance on lending money to customers, how customers should be informed about changes to interest rates, charges and other terms and conditions on their accounts, and arrangements for switching accounts. It covers a range of personal financial products including PCAs.58

2.101 The Code is sponsored by the BBA, the Building Societies Association (BSA) and APACS.59 The Government and consumer bodies also have a role in its determination. It was established in 1991 and is now in its seventh edition (March 2005). The BCSB, an independent body, monitors the banks’ performance and enforces the Code. Its board is comprised of seven independent directors and three representative directors of banking industry stakeholders.

2.102 The Code sets out detailed requirements based on six key commitments that each bank must adopt:

• We will make sure that our advertising and promotional literature is clear and not misleading and that you are given clear information about our products and services.

• When you have chosen an account or service, we will give you clear information about how it works, the terms and conditions and the interest rates which apply to it.

• We will help you use your account or service by sending you regular statements (where appropriate) and we will keep you informed about changes to the interest rates, charges or terms and conditions.

57The Business Banking Code sets the standards for banks’ treatment of their business customers. 58The Code applies to PCAs; basic bank accounts; credit, debit and cheque guarantee cards; loans; overdrafts; savings and deposit accounts (including cash ISAs); payment services; and foreign exchange services. It does not cover investments, , general insurance or mortgages. 59The BBA is the principal trade association for banks operating in the UK. The BSA is the trade association for all the UK's building societies. APACS is the UK trade association for payments and for those institutions that deliver payment services to customers.

40

• We will deal quickly and sympathetically with things that go wrong and consider all cases of financial difficulty sympathetically and positively.

• We will treat all your personal information as private and confidential, and operate secure and reliable banking and payment systems.

• We will publicise this Code, have copies available and make sure that our staff are trained to put it into practice.

2.103 In 2001, in response to the Cruickshank Report (see paragraph 2.112), there was a review of the operation of the Code throughout the UK (the Julius Review). Many respondents to the Julius Review’s consultation exercise felt that the Code, while not perfect, was a good example of self-regulation. The Review concluded that the Code had generally been effective in raising standards but identified considerable concern among customers about switching between banks. These centred on the ‘hassle factor’ of changing payment instructions and the fear that things would go wrong. There was also evidence that customers had inadequate information about the Code, and that information on their accounts was not presented clearly enough to facilitate comparisons with alternative products. These concerns resulted in a series of recommendations which led to the introduction of portable credit histories and a ‘5- day start—5-week finish’ standard for switching accounts.60 All the banks in Northern Ireland signed up to the enhanced switching provisions in the Code.

2.104 Professor Elaine Kempson of Bristol University’s Personal Finance Research Centre reviewed the Code in 2002 and 2004. These reviews resulted in the Code including more detail on specific service levels. In particular, the 2002 review introduced a new standard for account switching. Following the 2004 review, changes designed to make basic bank accounts more readily available were adopted and the help to be given to customers in financial difficulty was clarified.

2.105 In addition, Professor Kempson considered the recommendation of the Julius Review that each customer be provided with a consolidated annual summary statement (CASS), showing the total charges and interest paid on each of the financial products held with a particular financial institution for each tax year. Every customer would receive a CASS from every financial institution with which they had a product, and all would be mailed at the same time in the year to aid comparison. However, Professor Kempson told us that she received convincing evidence from banks that the imple- mentation costs would have been high and difficult to justify within the cost benefit framework for changes to the Code. This, together with evidence from consumer research conducted by the BCSB, led her to reject the proposal in her 2004 review.

2.106 Summary boxes were introduced in March 2004 in all credit card newspaper and magazine advertisements, application forms and any other pre-contract marketing material, as part of an industry-wide initiative. These were designed to give consumers high-level information in a simple format.61 Professor Kempson recommended that research be carried out into the extension of summary boxes to products such as current accounts. The BBA has commissioned a consumer research survey which is currently under way. The survey aims to ascertain whether consumers would want to receive information in a summary box format and, if so, the

60Under this standard, banks must pass on details of a customer’s standing orders and direct debits within five working days, and must complete the transfer within five weeks. 61Such information might include APRs, monthly interest rates, the length of the interest-free period, how payments are allocated, the minimum payment, fees for having the card, charges that might be incurred, and default charges that might be incurred.

41

optimum content, format and presentation of summary boxes (pre- and post-sale of the PCA), including the frequency of their provision.

2.107 The latest independent review of the Code began in November 2006. Various inter- ested bodies, including the BCSB and the OFT have submitted recommendations to the independent reviewer, Mike Young. The revised Code will be published in March 2008.

2.108 Further details on the Code are set out in Appendix 2.2.

Payment systems

2.109 Payment systems consist of a set of instruments, banking procedures, and, typically, inter-bank funds transfer systems which facilitate the transfer of money. They are central to the operation of the banking system. The banks in Northern Ireland partici- pate in UK-wide payment systems, with the exception of the clearing of cheques and paper credits, for which Northern Ireland operates its own system.

2.110 The most important payment systems used by banks in Northern Ireland are:

(a) BACS (formerly Bankers’ Automated Clearing Services): provides a three-day service for clearing electronic payments such as direct debits, direct credits and inter-bank standing orders throughout the UK. A&L, Abbey, Halifax and Nationwide are settlement members of BACS; the clearers, with the exception of Northern, are indirect participants using agency arrangements (Agency Banks). Northern’s parent Danske is a full member of BACS and Northern participates through this group membership.

(b) Clearing House Automated Payment Scheme (CHAPS): provides a rapid service for clearing systematically important and time-dependent payments throughout the UK. Abbey is a settlement member of CHAPS; the other banks operating in Northern Ireland use agency arrangements to access CHAPS.

(c) Belfast Bankers Clearing system (BBC): provides a set of rules (the Belfast Clearing Rules), similar to those used for the Cheque and Credit Clearing Company Limited (CCCC) in Great Britain, governing the exchange, processing and settlement of internal paper payments in Northern Ireland. The Belfast Clearing Rules are administered by the Northern Ireland Bankers’ Association (NIBA). The clearers are settlement members of BBC, and the non-clearers access the BBC system using agency arrangements.

(d) Plastic card networks: cover transactions made using debit, credit and other ATM cards.

2.111 APACS is the UK trade association for payments and for those institutions that deliver payments to end-customers. It provides a forum for its members to discuss non-competition issues relating to money transmission and represents the payments industry to media and government departments.

2.112 In 2000, the Cruickshank Report concluded that there were profound competition problems and inefficiencies associated with UK payment systems. In 2003, the OFT carried out a review of payment systems, concentrating on open access, innovation and charging, which led to important governance changes for APACS and its associated payment schemes. Nevertheless, HMT considered that some, at least, of the problems identified in the Cruickshank Report remained and asked the OFT to set up the Payment Systems Task Force (Task Force).

42

2.113 The Task Force established a number of working groups to examine particular payment systems with broad representation from the banking industry and users. The BACS Payment Schemes Ltd (BPSL) Innovation Working Group’s report led to the industry agreeing to introduce faster payments for Internet and telephone banking, as well as an improved service for standing orders by the end of 2007. As a result, customers will be able to transfer funds via the new system in a matter of hours.

2.114 The Cheques Working Group report published in November 2006 led to the industry agreeing to introduce a standard set of maximum clearing times for cheques from November 2007.62 This will mean that customers will start to earn interest on money deposited via cheques two days after the cheques is deposited, that they will be able to withdraw money deposited via cheque four days after it is deposited, and that they will know that money paid to them by cheque cannot be taken out of their account without their agreement six days after the money is deposited, unless they are a knowing party to fraud.

2.115 The industry proposed a new, high-level body with responsibility for governance across all the payment systems. The Task Force and the OFT believed that this proposal could potentially resolve outstanding governance issues and remove the need for the Task Force to examine the remaining payment systems. On 14 November 2006 the Chancellor of the Exchequer formally gave his backing to the new governance model (the Payments Council (the Council)) for the UK’s payment systems industry and to the winding-up of the Task Force.

2.116 The Council focuses on three key objectives: promoting strategic vision across the payments industry, promoting increased transparency and innovation, and ensuring the integrity of all member schemes. The Council’s board is headed by an independent Chairman. There are also four independent directors. The new governance structure will help address the agenda for payments set out in the Cruickshank Report on competition in the UK banking industry. The new structure will be reviewed by the OFT after two years.

2.117 The European Commission has put forward proposals to make cross-border payments as easy, cheap and secure as payments within a Member State, in part, to address the legal barriers to the building of a Single Euro Payments Area (SEPA).63 The Payment Services Directive was agreed in December 2005 and is designed to enhance competition, increase market transparency for providers and users, and standardize rights and obligations of providers and users of payment services in the EU, with a strong emphasis on a high level of consumer protection. This Directive, intended to be implemented in 2010, will create a new EU-wide licensing regime for ‘Payment Institutions’ and introduce conduct of business rules for all payment service providers for payments under EUR 50,000.

2.118 Further details on payment systems are set out in Appendix 2.3.

Provision of bank notes

2.119 The clearers issue their own bank notes which must be covered by coins or BoE notes.64 As at April 2007, HMT was consulting on possible changes to the

62These arrangements would apply to the UK. 63http://ec.europa.eu/unitedkingdom/press/press_releases/2005/pr0520_en.htm. 64In terms of cover, the clearers must hold the equivalent of the value of their assets in coins or BoE notes.

43

arrangements under which the note-issuing banks (the clearers) in Northern Ireland and the note-issuing banks in Scotland issue their notes.65 At present, the calculation of note cover is based on the average value of notes in circulation and the average value of note-covering assets at the close of business each Saturday. HMT is proposing that all note-issuing banks should be required to hold note-covering assets at all times, rather than only once a week.

2.120 HMT estimates that the clearers as a whole receive a financial advantage of between £35 million and £40 million each year as a result of holding their own banknotes.66 The financial advantage accrues to the clearers because they do not have to meet the cost of holding note-covering assets for six days a week. Some clearers told us that this gave them an incentive to offer a more extensive ATM network than they might otherwise do and that clearers also benefited from providing the non-clearers with notes for their ATMs and tills. There may also be a brand marketing advantage to the clearers in having their own banknotes. We were told that the variety of notes in circulation creates additional costs, such as printing costs, security costs and sorting costs. GCCNI told us that it was often inconvenient and even costly for consumers to change local notes before travelling to other parts of the UK.

Market trends

2.121 We looked at the key trends in the market and across the UK as a whole over the past ten years and their likely movement in the future, to help to set the context of our market investigation.

2.122 Over the ten years to 2004, there has been a shift away from using banknotes and coins. UK non-cash transaction volumes have risen by about three-quarters. Within the total, there have been substantial shifts between the various types of payment. There has been a 170 per cent increase in plastic card payments, a 120 per cent increase in electronic payments, and the use of cheques declined by 39 per cent.

2.123 The rapid increase in the use of plastic cards and automated payments is expected to continue. Customers are increasingly choosing to use direct debits for regular payments: the number of transactions is expected to reach about 3 billion across the UK in 2011. Standing orders are predicted to grow steadily, but more slowly. Internet and telephone payments are also increasing. The use of cheques is currently declining by about 7 per cent a year in Great Britain and volumes are expected to decline further with continuing migration of payments to debit and credit cards and automated payment systems including the proposed faster BACS payments service. There were 2.5 billion ATM withdrawals in the UK in 2004: this represented 72 per cent of cash withdrawn from accounts, compared with 48 per cent in 1994.

2.124 We looked at the trends in the way in which people choose to access their accounts. Some had predicted that the Internet would transform banking, removing the need for a physical branch network. Mintel’s Current Accounts, Irish Series, May 2005, reported, however, that Internet banking has not replaced conventional banking ‘due to the very real preference for face-to-face banking and the somewhat delayed acceptance of the Internet among banking consumers’. Mintel’s Online Banking, Ireland, May 2004 report found that Internet penetration had increased considerably

65HMT: Banknote issue arrangements in Scotland and Northern Ireland, July 2005. 66HMT estimated the total financial benefit to the clearers in Northern Ireland and the note-issuing banks in Scotland to be £70 million a year.

44

but that the actual uptake of Internet banking lagged behind, with around one in ten consumers in the island of Ireland using Internet banking.

2.125 The MORI survey asked ‘How do you mostly transact your personal business with your main financial institution these days?’. Overall, 31 per cent of respondents stated that they mostly called at a branch and 64 per cent that they used an ATM. Smaller numbers of respondents mentioned Internet banking, telephone banking, telephoning the branch and talking to a relationship manager. Market research estimates from MORI, Mintel and Datamonitor suggest that 12 to 13 per cent of customers use Internet banking, and the clearers provided evidence for 2005 showing that 12 to 19 per cent of customers used Internet banking. However, the MORI survey estimates that only between 1 and 2 per cent of customers use Internet banking as their main transaction method.

3. Market definition

3.1 We considered the appropriate product and geographic market definitions, applying the methodology set out in our guidelines.67 We do not regard market definition as an end in itself, but rather as a framework within which to analyse the effects of market characteristics.

Product market definition

3.2 We took the terms of reference set out by the OFT as the starting point for our product market definition. We believed that a PCA must have all the characteristics identified in the terms of reference (see paragraph 2.48), ie that it is marketed to individuals not businesses, it provides the facility to hold deposits, receive and make payments (cheque and debit cards), use ATM facilities and make regular payments (direct debits and standing orders). Many accounts which meet this definition have additional characteristics (such as a facility for short-term borrowing through overdrafts). We considered in particular whether two types of account—packaged accounts and offset/current account mortgage accounts—might be in distinct markets to other PCAs. This possibility is considered in paragraphs 3.22 to 3.24.

3.3 We considered whether other products may also be in the same product market, by considering whether the potential for demand-side or supply-side substitution was sufficient to constrain pricing or other competitive factors. In order for such products to be in the same market, the extent of potential substitution must be strong and widespread enough to constrain competitive behaviour, ie it must change incentives on pricing.

3.4 We looked at several potential substitute products and considered them against the four functions of a PCA:

(a) to provide a facility to deposit and store money, with quick and easy access;

(b) to provide a facility to receive payments by cheque or electronic transfer;

67CC3: Market References: Competition Commission Guidelines, June 2003. We could not perform a quantitative SSNIP test because there is no single price for a PCA: the cost to a customer depends on PCA usage. However, we used the SSNIP test as an intellectual framework for assessing market definition.

45

(c) to provide a facility to make instant and/or regular payments without using cash, eg through cheques, switch payments, bank transfers, standing orders and direct debits; and

(d) to provide the means for short-term borrowing through an overdraft.

Basic bank accounts, instant access savings accounts and credit union accounts

3.5 We considered whether some financial products that did not have all the characteristics identified in the terms of reference might also be in the same economic market. We looked in particular at basic bank accounts,68 instant access savings accounts69 and credit union accounts.70

3.6 Some of the clearers argued that a product did not need to satisfy all the functions set out in paragraph 3.4 to be considered a PCA. BoI said that many customers used their accounts only for certain functions (for example, as a place to deposit and store money, to receive payments and to make payments). BoI argued that as there were more customers in Northern Ireland than Great Britain in social categories D and E, and a lower level of economic activity in Northern Ireland, a relatively large proportion of the population would require their account only to perform certain functions (for example, they may not require overdraft facilities, cheques, standing orders or direct debits). This was reflected in the fact that PCA penetration is about ten percentage points lower in Northern Ireland than in the rest of the UK.

3.7 We recognize that the proportion of the population that require their account only to perform certain functions may be larger in Northern Ireland than it is in Great Britain. However, we received no evidence to suggest that customers switch to products such as basic bank accounts in response to PCA price increases. The introduction of basic bank accounts after 2000 did not appear to reduce numbers of PCA customers, nor did we see evidence of PCAs being redesigned or repriced as a result of the introduction of basic bank accounts. Rather, basic bank accounts appear to have increased the proportion of the population with access to a bank account. Transfers from PCAs to basic bank accounts have been minimal; similarly very few customers have ‘upgraded’ from basic bank accounts to PCAs. [] estimated that [] per cent of its basic bank account customers had upgraded to a PCA since it introduced its basic bank account in April 2003; [] estimated that [] per cent of its basic bank account customers had similarly upgraded. Finally, we saw no evidence that the banks monitor products such as basic bank accounts or take account of their characteristics when designing their PCAs.

3.8 Northern has a large number of customers on its CashMaster account; this has an overdraft facility but no cheque or debit card function. Northern told us that CashMaster offered much the same functionality as a PCA and so believed that this could be a constraint on its PCAs. It also told us that it had seen customers migrating from CashMaster to Current Account Plus as much as from any other product. We excluded CashMaster from the product market because it lacked some of the func- tionality of a PCA (ie cheques and debit card) and was, we considered, more similar

68A basic bank account allows money to be paid, standing orders and direct debits to be set up, and money to be taken out using a cash card. It should not go overdrawn. 69An instant access savings account allows immediate withdrawals without incurring a penalty. 70Credit unions are cooperative financial organizations owned by their members and run for their benefit. Members save with their respective credit union and can borrow from it. It does not provide a facility to make payments.

46

to a basic bank account than to a PCA.71 We are satisfied, in any case, that our findings would not change whether or not CashMaster was included or excluded from the market, despite the large number of CashMaster accounts (see paragraph 4.20).

3.9 If prices rose, a PCA user might, in theory, consider switching to a combination of other products, including an instant access savings account or basic bank account, a credit card and a loan. This would in practice, however, be difficult for the customer to administer. We also note that customers with PCAs use these other products in addition to their PCAs, suggesting that they are complements to PCAs rather than substitutes. Ulster said that even if a customer viewed various financial products as complements, the customer could move balances, and possibly transactions, between, for example, their PCA, savings account and credit card, exerting competitive pressure on PCAs. However, if this were the case, we would expect to see the banks competing on credit interest rates between their PCAs and savings accounts. In general, however, we do not see credit interest rates on PCAs that are competitive with savings accounts. Until the introduction of Northern’s new accounts in April 2006, credit interest rates on the clearers’ non-packaged accounts had remained at or near zero over recent years.

3.10 We note that charges have increased over the last five years on traditional PCAs. Charges do not appear to have increased similarly on other financial products such as loans, credit union accounts and basic bank accounts. If these products were in the same product market, we might have expected significant numbers of customers to switch to these other products. However, we have seen no evidence of this, nor was there any evidence in the banks’ internal pricing papers that substitution to other products was considered by them, let alone viewed as a constraint when setting prices.

3.11 First Trust argued that we should look at whether, in aggregate, PCAs would be constrained by the threat of different customers switching to any of these alternative products in response to uncompetitive pricing or poor quality of service on PCAs. First Trust also pointed to various alternative products and innovations which it said provided a growing competitive threat such as offset mortgages and 360money’s pre- paid cash accounts.72

3.12 We looked at whether there was evidence that the cumulative impact of alternative products constrained pricing and competitive behaviours on PCAs. We saw no evidence in internal pricing papers provided by the banks that any other products were monitored or considered in the banks’ decision-making processes on PCAs. The banks provided no evidence of any competitive reaction to the introduction of other personal financial products. While some customers may have turned to alternative products, there is no indication that enough considered doing so to require any of the banks to alter their behaviour, and therefore there was no evidence of a competitive constraint being exerted on PCAs.

3.13 For these reasons, we found that other personal financial products could not readily be substituted for a PCA except for a small minority of customers with straightforward financial affairs. There is no evidence that there are enough such customers to constrain PCA pricing. We therefore conclude that basic bank accounts, instant

71Halifax’s Cardcash account does not automatically issue a chequebook, but one is available on request although there is no overdraft facility. Cardcash, however, has all the functionality set out in our terms of reference (see paragraph 3.2) and we therefore included it in our product market. 72360money is a trading name and a service provided by PrePay Technologies Ltd, a company authorized and regulated by the FSA.

47

access savings accounts and credit union accounts are not in the same economic market as PCAs from a demand-side perspective.

3.14 Supply-side substitution applies where a provider of a different service can quickly, easily and profitably, without significant investment, switch to providing PCAs in the event of a price increase. The usual period of reference is substitution within one year; longer time frames are generally taken to constitute new entry to the market.

3.15 An established financial institution, providing other personal finance products, such as a building society, or suppliers of savings accounts, loans, mortgages and credit cards, would be the most likely type of institution to move into the market quickly and easily. BoI and First Trust argued that credit unions had an established presence, customer base and distribution network in Northern Ireland and could easily move into the supply of PCAs.

3.16 We note that, in order to pick up a significant market share, such an institution would need a network of branches. Advertising and other forms of marketing would be required to establish its credibility, and, even with an appropriate branch network, the rate at which customers would switch to a new provider is likely to be low. Other delivery methods such as telephone and Internet banking have not been successful in establishing significant market share. We discuss these issues in more detail in the section on entry and expansion (see paragraphs 4.62 to 4.107).

3.17 There are few potential entrants that have an appropriate branch network and could easily move into the supply of PCAs. [] The Post Office provides an alternative branch network, although the remote providers—cahoot and smile—that use the Post Office for this purpose have not built substantial market share.

3.18 Credit unions would encounter a number of additional barriers to entry, including the need to obtain regulatory approval to be a licensed deposit taker and to join appropriate banking schemes for money transfer and clearing. Nine credit unions in Great Britain started trialling a scheme from November 2006 to offer bank accounts provided by Cooperative Bank. The accounts offer a Visa/ATM card, and the facility to pay direct debits and standing orders, but there is no overdraft or cheque book facility. However, this has taken three years of negotiation and planning, we understand that there is no current intention to move into PCAs, and it seemed to us unlikely that credit unions would be able to achieve a significant market share if they did launch a PCA within the timescales required for supply-side substitution (see paragraph 4.91 for discussion of credit unions as potential entrants to the market).

3.19 We did not think that entry by a remote provider would provide a supply-side constraint, in part because remote providers have already entered the market and have failed to capture a significant market share.

3.20 It was put to us that the growth of the market share of the non-clearers demonstrates potential supply-side substitution from the provision of other financial services, such as mortgages and savings products. However, we believe that this growth resulted from a one-off regulatory change (see paragraph 2.29), and it took many years for the non-clearers to achieve significant market shares. There have been no examples in the past of entrants to the Northern Ireland PCA market quickly achieving a

48

significant market share.73 This growth in market share by the non-clearers should therefore be regarded as new entry rather than supply-side substitution.

3.21 We therefore found that basic bank accounts, instant access savings accounts or credit union accounts should not be included in the market due to supply-side substitutability.

Packaged accounts and offset/current account mortgages

3.22 We also considered whether packaged accounts and offset/current account mortgages should be included in our product market.74

3.23 Packaged accounts constitute around 6 per cent of current accounts in Northern Ireland. Given the similarity of function, the ease with which customers can substitute between packaged accounts and regular PCAs, and evidence that customers do switch between the two, at least to some extent,75 we found that packaged accounts were in the same market as PCAs.

3.24 Offset/current account mortgages constitute a very small proportion of current accounts in Northern Ireland, and Northern is the only clearer to offer such a product. The customer of an offset/current account mortgage will be seeking an efficient way to operate their PCA and mortgage; as such, this type of account substitutes for both, and the functionality of the PCA is the same. Several banks told us that these products were targeted at mortgage customers rather than PCA customers, and selling such products was regulated in the same way as mortgages. Evidence we received suggested that creating this type of product was not straightforward: it was likely to require significant IT development, and there were legal, regulatory and operational impacts; therefore there was unlikely to be close supply-side substitution. We were given evidence that these products were mainly regarded by customers as a means of controlling mortgage costs. We thought it unlikely that customers would switch easily if charges rose on PCAs. We therefore found that these products were not in the same market as PCAs.

Personal financial products

3.25 We also considered whether there might be a broad market for all personal finance products rather than separate markets for individual products. The clearers tend to consider profitability at the retail banking level rather than by individual products, and banks try to sell multiple products to customers.

3.26 However, none of the banks suggested that there was a broad market. Whilst the banks see cross-sales as important, they achieve on average less than one other product sale per PCA customer, which would suggest that customers do not see personal financial products as a single market. Consequently, we do not believe that there is close supply-side substitutability between different financial products and we found that there was not a broad market for all personal financial products.

73The CC’s guideline CC3 on market investigation references (paragraph 3.19) says that the usual level to be judged as significant entry would be a 5 per cent share of the relevant market. 74In an offset/current account mortgage, credit balances in the PCA element of the account are used to reduce payments on, or speed up the rate of repayment of, the mortgage. 75For example, Ulster provided evidence that [] per cent of its U-First customers, and [] per cent of its U-First Gold customers, had transferred from another PCA held with Ulster as at 8 September 2005.

49

Market segmentation

3.27 We looked at whether the market was segmented, either according to delivery method or by type of customer. Although most banks have common pricing for their PCAs regardless of delivery mechanism, we found that there was some differentiation by delivery mechanism: for example, A&L offers more favourable terms on its Premier Direct account, some of Northern’s new suite of PCA products limit the number of times a customer can write a cheque or withdraw cash in a branch free of charge in a quarter, and there are several remote providers offering PCAs over the Internet and/or telephone.

3.28 However, we saw no evidence of a substantial and distinct segment of customers that favour remote banking through the Internet or over the telephone. The evidence that we received suggested that most customers expect to be able to operate their bank account in a variety of ways, with access through branches, ATMs, the Post Office, over the Internet, and by telephone. Nor did we see the clearers responding to innovation by the remote service providers in the design or pricing of their PCAs.

3.29 We found evidence of market segments for students, graduates, and high net worth individuals. We also found that banks have differed in how actively they target switchers (see paragraphs 4.188 to 4.192). In addition, some banks continue to offer older PCAs despite having launched new PCAs with generally more favourable terms. However, although not all banks automatically transfer their customers to their new PCAs, neither do they place barriers to stop their customers switching to newer accounts. None of the banks argued that these market segments constituted separate product markets. Supply-side substitution between these segments was relatively straightforward. We therefore did not find these to be separate markets.

Conclusions on product market definition

3.30 Although a small minority of customers with relatively straightforward financial needs might regard basic bank accounts, instant access savings accounts or credit union accounts as substitutable for a PCA, there is no evidence to suggest that there are enough of these customers to constrain the terms on which PCAs are offered. We found that packaged accounts are part of the same market as other PCAs, but believe that offset/current account mortgages are in a separate market. We did not find that there was a broader market for personal financial products. We therefore conclude that the product market should include all PCAs for all customers (see paragraph 3.2), including packaged accounts, but should not be drawn more widely to include basic bank accounts, instant access savings accounts, credit union accounts, offset/current account mortgages or other personal financial products.

Geographic market definition

3.31 We considered whether Northern Ireland is in the same market for PCAs as Great Britain and/or the Republic of Ireland, and whether there are local markets for PCAs within Northern Ireland. We concluded that, due to differences in currencies and regulation, Northern Ireland PCAs are not in the same market as the Republic of Ireland, and we therefore focused on potential overlap with the market in Great Britain.

3.32 Northern Ireland consumers could switch to remote providers based in Great Britain, using the LINK ATM network and/or the Post Office to access their accounts. The Post Office offers counter services under contract to smile and cahoot. It would be possible to operate a PCA by Internet or telephone. Customers could operate PCAs

50

with branch-based banks in Great Britain in the same way. Some of the banks argued that there was an inexorable move to a wider, UK banking market, particularly due to improved technology.

3.33 However, the evidence that we received suggests that many customers are reluctant to switch to remote providers. Customers attach significance to the availability of branches and familiarity with the supplier’s brand when choosing a bank. There is no evidence of any significant numbers of Northern Ireland consumers operating accounts through bank branches in Great Britain. We therefore believe it unlikely that many customers would switch to banks based in Great Britain if prices increased by 5 per cent in Northern Ireland relative to Great Britain (see paragraph 3.1).

3.34 For the reasons set out in paragraph 3.16, we did not find that supply-side sub- stitution by banks based in Great Britain was likely.

Chain of substitution

3.35 We considered whether the non-clearers were constrained by the actions of the large banks based in Great Britain (Barclays, HSBC, Lloyds TSB and RBS), and whether the clearers in Northern Ireland would respond to changes in PCAs introduced by the non-clearers that were, in general, making a UK-wide offer. If so, this might, in theory, suggest that there was a chain of substitution from Great Britain to Northern Ireland. However, we saw no evidence either that the clearers in Northern Ireland changed their pricing directly and rapidly in response to pricing changes by non- clearers, or that the non-clearers changed their offers in response to changes by the large banks based in Great Britain. Many of the non-clearers told us that, in so far as their offers did respond to the actions of the other banks, it was likely to be in response to other non-clearers. It seems to us that, in so far as the clearers might adjust their pricing and offers in response to competitive pressures from the non- clearers, they are responding to the non-clearers’ offers in Northern Ireland. Hence it is possible that competition could develop in distinct ways in Great Britain and Northern Ireland even with the non-clearers having similar offers in both markets. We note that this is consistent with a different pricing structure, that of traditional PCAs, having continued to be offered in Northern Ireland for many years, despite not being available in Great Britain.

Local markets

3.36 Since customers open their accounts predominantly in branches, and many want easy access to a branch, there could be distinct local markets for banks. However, we found that local markets were unlikely to be relevant to our assessment of competition (see paragraphs 4.308 to 4.311). No bank in Northern Ireland appears to operate any policies on a local basis, there would be a chain of substitution between overlapping local markets throughout most of Northern Ireland, and customers may have access to banks from several locations (for example, their home, workplace, and shopping destination) which would weaken any notion of local markets.

Conclusions on geographic market definition

3.37 We conclude that Northern Ireland forms a separate geographic market for PCAs from Great Britain and from the Republic of Ireland. We did not find there to be close demand- or supply-side substitution between Northern Ireland and Great Britain. We conclude that local markets are not relevant to our assessment of competition.

51

Conclusions on market definition

3.38 We therefore conclude that the product market should include all PCAs (see para- graph 3.2), including packaged accounts, but should not be drawn more widely to include basic bank accounts, instant access savings accounts, credit union accounts, offset/current account mortgages or other personal financial products. We also conclude that the geographic market is Northern Ireland.

4. Competition in the market

4.1 This section discusses competition in the market, covering market structure, bank conduct, customer conduct, an analysis of competition and other indicators (see paragraph 1.11(c) to (g)).

4.2 We discuss market structure in paragraphs 4.3 to 4.107. We first outline the business models of the banks. We then discuss market shares and concentrations, using a variety of measures of market share. Next we discuss the characteristics and uses of branch networks, before discussing entry and expansion in the market.

Market structure

Business models

4.3 The banks’ businesses are not designed solely for the provision of PCAs. We therefore looked more broadly at the business models of the clearers and the non- clearers across a range of products and/or geographies to inform our thinking about the competitive dynamics in the market for the provision of PCAs in Northern Ireland.

4.4 We note that many of the banks are part of larger groups, allowing them to benefit from economies of scale in terms of processing costs, advertising and brand building, and product and systems development (see Table 1). Following the acquisition of Ulster (by RBS) and of Northern (by Danske), the clearers are now part of groups that are as large, on average (in terms of total group assets), as the non-clearers’ respective groups. In addition, BoI and AIB, First Trust’s parent company, are the two largest banks in the Republic of Ireland.

4.5 The clearers and many of the non-clearers offer a variety of personal banking products through various distribution channels including the branch, via the Internet, or by telephone.

4.6 The clearers’ business model is, however, based on the requirements of both personal and business banking. Each of the clearers has a more extensive branch network in Northern Ireland than each of the non-clearers, and offers a wide range of both personal and business products.

4.7 PCAs have always been a core product of the clearers. In some cases PCAs have been seen as a ‘gateway’ to their other personal financial products, such as savings or loans. They are also a significant net provider of funds for the banks due to the credit balances held. The clearers told us that they did not measure the financial profitability of PCAs as stand-alone products, and normally monitored PCA perform-

52

ance by assessing volume information (such as total balances or number of accounts) and revenues (NII76 and charges).

4.8 We note that, despite overall similarities, there are some significant differences between the individual clearers. The clearer with the most branches (Northern) has more than twice the number of branches as the clearer with the fewest (BoI).77 BoI is the only clearer to have an agreement with the Post Office to provide counter services. There are also differences in strategy and IT and operating platforms. As set out in paragraphs 2.21 and 2.23 respectively, both BoI and First Trust’s parent company, AIB, have their group headquarters in Dublin. Ulster is part of RBS, headquartered in Edinburgh, and Northern is part of Danske, headquartered in Copenhagen.

4.9 The non-clearers can be divided into three sub-groups: current and former building societies (including Abbey, A&L, Halifax, Nationwide and Woolwich); banks based in Great Britain that have opened one or more branches in Northern Ireland since 2000 (Co-operative Bank and HSBC); and banks that provide a remote service by Internet, telephone, post and/or text messaging (the remote providers including cahoot, first direct, IF and smile). All the non-clearers offer standard products across the UK and do not tailor their PCA for the Northern Ireland market.

4.10 The majority of the non-clearers’ market share is held by the first sub-group. All of these non-clearers were originally building societies, offering mortgages, savings accounts and loans. They started to offer PCAs 15 to 20 years ago (see paragraph 2.29). They do not, in general, offer extensive business banking services, although some told us that they were looking to expand into this area. They each have between 9 and 21 branches in Northern Ireland78—significantly fewer than any of the clearers. None of the non-clearers focuses on Northern Ireland in particular. In many cases, the non-clearers have acquired PCA business through cross-selling from their more traditional products, particularly their mortgage and savings businesses.

4.11 The second sub-group is that of the banks based in Great Britain that have opened one or more branches in Northern Ireland since 2000. Their overall business model is more similar to that of the clearers, although the PCA that they offer in Northern Ireland is common across the UK. The Co-operative Bank has only one branch which is located in Belfast. HSBC currently has five branches in Northern Ireland. It opened its most recent branch in November 2006, and [].

4.12 The third sub-group comprises the remote providers that offer a different business model. This is based on accessing the account without the support of a branch network by, for example, Internet, telephone, post and/or text messaging. All the remote providers are owned by, or form part of, the same group as one of the branch-based banks. Their charging structure broadly resembles that of the non- clearers, in that it is usually fee-free and carries a higher rate of credit interest than is usually offered by the large banks based in Great Britain.

4.13 As discussed in paragraph 2.15, we found it helpful at times to distinguish at a high level between the clearers and the non-clearers. We recognize, however, that, as explained in paragraph 4.9, there are at least three business models within the overall group of non-clearers, and that within the clearers and the non-clearers, individual banks have evolved differently. We discuss, as appropriate, any

76NII is earned from credit and debit balances (see paragraph 2.69). 77Northern has 101 branches, including six sub-branches, and BoI has 46, including one sub-branch. 78Up to 45 if agencies are included.

53

characteristics of individual banks that might be relevant to our assessment of competition in the market.

Market shares and concentration

4.14 Market shares and concentrations are one indicator in the assessment of the competitiveness of the market. We first consider market shares based on volume data from the MORI survey.79 This is the only data that is widely available and has been produced over a number of years on a consistent basis across the banks. It also has the advantage of being used by the clearers to manage their business. However, in terms of our investigation, the data has a number of shortcomings. In particular, there are important differences of definition between the current account as defined by MORI, and our PCA definition in our terms of reference (see paragraph 4.17). We therefore looked at a range of bases for calculating market shares and concentrations. In addition to using the MORI data, we collected data from the eight largest banks from 2002 to 2005 to allow us to calculate an alternative view of volume-based market shares (see paragraphs 4.35 to 4.40). We also used this data to calculate value-based market shares (see paragraphs 4.41 to 4.47). This allowed us to calculate a variety of measures of concentration in the market, and to draw conclusions on market shares and concentrations.

Volume-based market shares and concentrations

4.15 Relative market shares in different personal banking products largely reflect the historical focus of different banks (see Table 1 in Appendix 4.1). The clearers have the highest market shares in current and savings accounts, whilst the five largest non-clearers in Northern Ireland have higher market shares in mortgages, reflecting their building society origins.

• MORI market shares

4.16 We looked at MORI market shares for current accounts by bank. We looked in particular at two measures of market shares based on MORI data: the share of total accounts (stock), and the share of new accounts (flow). New accounts include both first accounts opened by new-to-banking customers and switchers won by the banks each year.

4.17 We note that the MORI data is based on the bank where an individual’s ‘main account’ is held,80 and includes offset/current account mortgages, basic bank accounts (in particular the Post Office Card Account) and Northern’s CashMaster account, none of which are included in our PCA definition. We also note that some PCAs may generate more revenue for banks than others (eg active compared with inactive, or accounts with high rather than low credit balances) which cannot be reflected in the MORI data.

4.18 Figure 2 shows the trend in market share based on the stock of current accounts by bank from the MORI survey. This shows that the clearers had a significant market

79The clearers all subscribe to the MORI survey. This is their primary source of information on market shares based on numbers of accounts and share of new account openings. The MORI survey reports rolling annual averages of four quarterly financial surveys of a changing sample of 1,500 Northern Ireland respondents. 80We note that MORI also collect data on all accounts, rather than just ‘main accounts’, but the data that was presented to us by the clearers was based on the individual’s main account. The ORC survey suggests that around 25 per cent of PCA holders have a second PCA with a different bank, and about 20 per cent of PCA holders had at least one extra PCA with the bank that supplied their main PCA.

54

share in total (71 per cent in 2006 Q4). This share has declined from 83 per cent in 1999 Q4.

4.19 Within these totals, individual clearers’ market shares have varied substantially over the period. Ulster’s share has remained fairly stable at around 25 per cent, and BoI’s has increased from around 10 per cent at the start of the period to around 13 per cent. On the other hand, since 1999, First Trust and Northern have both lost around seven percentage points (from around 22 per cent to under 15 per cent in the case of First Trust, and from around 27 per cent to 19 per cent for Northern).

4.20 The inclusion of Northern’s CashMaster account in these figures overstates Northern’s market share compared with its share when using our definition of the PCA market. We set out in paragraph 3.8 our reasons for excluding CashMaster from the PCA market. Northern has around [] CashMaster accounts, compared with over [] accounts which meet our definition of a PCA. In our estimates of market share based on numbers of PCAs from the eight largest banks set out in Table 5, the exclusion of Northern’s CashMaster has reduced Northern’s market share by around [] percentage points.

4.21 The total MORI market share of the five largest non-clearers (Abbey, A&L, Halifax, Nationwide and Woolwich) has increased by over ten percentage points between 1999 and 2006, from 14 to nearly 25 per cent, with each of them showing gains. A&L, Halifax and Nationwide in particular have increased their market share by over 70 per cent. New entrants to the market include the Co-operative Bank, HSBC, IF and the Post Office.

FIGURE 2

Market shares of stock of current accounts, 1999 to 2006 (per cent)

30.0%

25.0% Abbey/Abbey National Alliance & Leicester Bank of Ireland Bank of Scotland (Ireland) Cahoot 20.0% Co-operative Bank Don't know/Refused/NS First Trust Bank Halifax 15.0% HSBC Intelligent Finance Nationwide Northern Bank 10.0% Other Post Office Smile the Internet Bank Ulster Bank 5.0% Woolwich

0.0%

r1 2 3 r3 4 1 2 2 3 r3 4 1 2 3 4 Qtr Qtr Qtr Qtr Qtr Qtr Qtr Qtr Qtr 2 Qtr 3 Qtr 6 Qtr 00 Qtr401 Qtr1 02 Qtr302 Qtr4 03 Qtr404 Qtr1 05 Qtr205 Qtr3 000 Qt000 000 0 0 001 Qt001 002 00 0 0 003 00 0 0 004 Qt004 005 0 0 006 006 00 1999 Qtr42 2 2 2 2 2001 Qtr22 2 2 2 2 2 2003 Qtr12 2 2 2 2004 Qtr22 2 2 2 2 2005 Qtr42006 Qtr12 2 2

Source: MORI data.

4.22 Based on the MORI data on share of stock shown in Figure 2, the market was highly concentrated in 1999 (Herfindahl-Hirschman Index (HHI) of 1,924 in Q4 1999).

55

Concentration started to reduce in 2002; by Q4 2006, the market was concentrated (HHI of 1,502).81

4.23 We also looked at the share of flow of new accounts by bank (see Figure 3). This shows greater volatility than the stock of accounts, in part because the numbers of new account openings in the MORI survey are relatively small compared with the stock of current accounts.82 In addition, the share of flow is influenced by the inclusion of the Post Office Card Account, which falls outside our PCA market definition but accounted for nearly 8 per cent of flow in 2005. Between 1999 Q4 and 2006 Q4 the clearers’ share of flow in aggregate fell from 84 to 57 per cent, driven in particular by sizeable reductions in the share of flow achieved by First Trust and Northern. It is not yet possible to determine the impact of the introduction of Northern’s new account packages on its share of flow. Ulster’s share has generally increased over the past four years.83

FIGURE 3

Share of flow, 1999 to 2006 (per cent)

35.0%

30.0% Abbey/Abbey National Alliance & Leicester Bank of Ireland 25.0% Bank of Scotland (Ireland) Cahoot Co-operative Bank Don't know/Refused/NS 20.0% First Trust Bank Halifax HSBC 15.0% Intelligent Finance Nationwide Northern Bank Other 10.0% Post Office Smile the Internet Bank Ulster Bank Woolwich 5.0%

0.0%

4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr tr Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q 6 6 1999 2000 2000 2000 2000 2001 2001 2001 2001 2002 2002 2002 2002 2003 2003 2003 2003 2004 2004 2004 2004 2005 2005 2005 2005 200 2006 2006 200

Source: MORI data.

4.24 We looked at the relationship between the stock and the flow of current accounts. Using the data that we collected from the eight largest banks on the numbers of accounts each year, we estimated that, in 2002 to 2005, flow as a percentage of stock each year averaged between 10 and 13 per cent overall (9 to 12 per cent for

81HHI is a measure of market concentration calculated by summing the squares of the market share of all suppliers in the market. It therefore ranges from 0 to a maximum of 10,000 for perfect monopoly. In assessing mergers, several competition authorities around the world have adopted benchmark HHI figures for characterizing the degree of concentration in the market. In its guidelines, the OFT states that it is likely to regard any market with an HHI in excess of 1,800 as highly concentrated, and any market with an HHI in excess of 1,000 as concentrated. These thresholds are not directly relevant to a market investi- gation; market failure might occur in any market regardless of the degree of concentration. However, it is one factor in a wider assessment of competition (see CC3 Market Investigation References, paragraphs 3.8 to 3.11). 82Of the total survey sample of 1,500 per quarter, a small percentage (around [] per cent) will have opened a new account in the previous 12 months. 83We note that, since 1999, the level of penetration of current accounts (by proportion of households) in Northern Ireland has increased by at least ten percentage points (see paragraph 2.74); consequently, a loss of market share does not necessarily imply a reduction in the absolute number of PCA customers.

56

the clearers and 11 to 15 per cent for the non-clearers).84 This suggested to us that changes in shares of flow would feed relatively slowly into changes in the stock of accounts, and that the market shares based on stock were likely to be a better measure of a bank’s market share.

4.25 In response to our Emerging Thinking, two of the clearers challenged our view that changes in shares of flow feed relatively slowly into changes in stock. One clearer ([]) presented evidence to show that, over the period 1995 to 2004, a one percentage point increase in the flow of market share was associated with a one-third percentage point increase in stock. Thus, to achieve a one percentage point increase in current stock market share required a three percentage point increase in flow. Another clearer ([]) said that the flow rates set out in paragraph 4.24 were sufficient to lead to an increase or decrease in market share by at least 50 per cent in only five years.

4.26 We accept that, in the absence of other external factors, if an individual bank has a share of flow that is higher than its share of stock, its share of stock might be expected to rise (and vice versa) such that, over time, flow and stock would equalize. However, although the rate at which banks attract new flow (either new-to-banking customers or switchers) is an important component of their market share of stock over time, market share will also be influenced by other factors. In particular, market share of stock will also be influenced by the rate at which customers close their accounts, as well as by customers holding more than one PCA with different banks (multiple account use).

4.27 We examined the relationship between the MORI data on stock and flow of current accounts (see Figures 2 and 3). In terms of the MORI data, we observed a number of anomalies:

(a) Northern’s share of stock has been around 20 per cent since 2003 but its share of flow during this period has been much lower (around 10 per cent) in every quarter, without any marked impact on share of stock.

(b) BoI’s share of stock was stable at around 11 per cent from 2001 Q2 to 2005 Q3 despite its share of flow being considerably above these rates for much of this period (11 to 21 per cent).

(c) While Ulster’s share of flow has risen since Q3 2005 and has exceeded its share of stock for the last five quarters, its share of stock over the same period has remained unchanged.

4.28 In order to get a better understanding of the share of flow data and its implications for competition in the market, we looked at the split in the share of flow between new-to- banking customers and switchers. The data is more volatile than the overall flow data, since the numbers of new-to-banking customers and switchers within the sample is relatively low.

4.29 Figure 4 shows the share of flow achieved by each bank of new-to-banking customers. We note that the clearers in aggregate have a larger share of the new-to- banking market than for new accounts overall, although their aggregate share has

84This is calculated as new accounts for the year divided by the total number of accounts at the financial year-end and expressed as a percentage. The figures quoted are weighted averages for the clearers and non-clearers as groups. The ranges quoted correspond to the range of annual figures for each of the four years ended 2005. There are significant variations by bank within the groups; for example in 2003 Ulster’s flow as a percentage of stock was [] per cent whilst First Trust’s was [] per cent.

57

fallen significantly from 86 per cent in 1999 Q4 to 70 per cent in 2006 Q4. Results are volatile due to the small number of observations (around [] over the year) and hence must be interpreted with caution (although we note that none of the clearers claimed that these results were unrepresentative). Despite the uncertainty in the numbers associated with the individual banks, Ulster’s share is consistently high at around [] per cent. Ulster has increased its share further since 2005 Q3, possibly as a result of the introduction of its new charging structure. Some of the non-clearers, such as A&L and Halifax, have lost share of new-to-banking customers in recent years.

FIGURE 4

Share of flow of new-to-banking customers, 1999 to 2006 (per cent)

[]

Source: MORI data.

4.30 We also looked at the share of flow for switchers (see Figure 5). Once again, these results are volatile because the number of relevant respondents is low (around [] over the year). The clearers in aggregate have been less successful at attracting switchers than in attracting new-to-banking customers. The share of switchers going to the clearers in aggregate fell sharply from nearly 80 per cent in 1999 Q4 to 31 per cent by 2006 Q4. Overall, the non-clearers are attracting more customers than the clearers through switching, consistent with the non-clearers’ greater focus in the past on attracting switchers (see paragraphs 4.187 to 4.192).

4.31 However, these aggregate figures mask very considerable differences in the success or otherwise of the individual clearers in attracting switchers. []

FIGURE 5

Share of flow for switchers from other banks, 1999 to 2006 (per cent)

[]

Source: MORI data.

4.32 We looked at the other factors which might influence market share of stock over time (see paragraph 4.26). We asked the eight largest banks for gross closure rates on their accounts. For 2004 and 2005, the clearers in aggregate had higher closure rates (7.4 and 8.5 per cent respectively) than the non-clearers (6.0 and 4.4 per cent respectively). There were considerable differences between the banks, with [] generally having the highest closure rates.

4.33 Closure rates will be significantly affected by the banks’ policies on identifying and closing dormant accounts, and the actual levels of dormancy, making comparisons difficult.85 Accounts might be closed not only because of switching but also for other reasons such as bad debt and death. Although we asked the banks for details of closure rates, it was not possible for us to compare these factors on a reliable basis between banks. Some banks regularly filtered out dormant accounts while others did so only occasionally. As well as the rate at which accounts are closed or fall dormant differing between banks, there may also be differences in the rate at which main

85For example, [].

58

accounts become secondary accounts (accounts which are still open and in use, but are not the customer’s main account).

4.34 In summary, on the basis of the MORI data and the data that we analysed on PCA volumes of the eight largest banks, we found that:

(a) The experience of the individual clearers has varied over time. In particular, BoI and Ulster have maintained or increased their market shares of stock, []. This suggests that the overall loss of market share of the clearers is heavily influenced by factors specific to individual banks as well as market-wide factors. These individual differences indicate that the clearers are not homogeneous, despite historical similarities in their approach to charging.

(b) Changes in share of stock are influenced by more than just share of flow. Although it was difficult to get robust evidence, we found that account closure rates varied significantly between banks, and factors such as dormancy rates may also be important. Whilst share of flow is an important indicator, we cannot use share of flow alone to predict future share of stock.

(c) The clearers’ loss of share of flow overall is particularly marked in their relative lack of success in attracting switchers from other banks. Their share of new-to- banking customers has declined but to a far lesser extent than for switchers. This is consistent with our finding that, in the past, clearers focused to a greater extent on new-to-banking customers (see paragraph 4.188).

• Other volume-based market shares

4.35 Given some of the difficulties that we had with the MORI data, we looked at data provided by each of the eight banks with the highest volume of PCAs between 2002 and 2005 (see Table 5) to enable us to calculate alternative data on market shares by volume. There are a number of differences between our estimates and the MORI data:

(a) Our estimates are based on total PCAs in Northern Ireland for the eight largest banks. The MORI data is based on a changing quarterly survey of a sample (1,500) of the population (6,000 over the year).

(b) Our estimates align with our PCA market definition. Basic bank accounts, the Post Office Card Account and Northern’s CashMaster account are, for example, excluded from our market definition but are included in the MORI estimates.

(c) The two sets of figures refer to different time periods. The account data from the banks is based on respective financial year-end positions for each bank, whereas the MORI data is based on rolling annualized averages of four quarters of survey data.

(d) Our estimates are based on data covering a four-year period (2002 to 2005), whereas the MORI data is available over a longer time period.

(e) Our estimates reflect the fact that customers may use more than one PCA, whereas the MORI data includes only the respondent’s ‘main account’.

4.36 We looked at both the number of PCAs in total and the number of those where there had been recent transactions (active accounts). One of the clearers ([]) said that our analysis should not focus only on active accounts because the banks earn revenue from all accounts—active or otherwise. Based on evidence provided by

59

three of the banks, however, we noted that average credit balances and charges, and therefore revenues, were significantly lower for inactive accounts than active accounts (for example, total average income in 2005 from active accounts was £152 compared with £16 for inactive accounts). We thought that, in theory, market shares based on active accounts would be more informative than market shares based on all accounts, since active accounts are likely to have higher balances (either credit or debit balances) and earn the banks more revenue (both from NII and charges). However, we found that there were many practical difficulties associated with the definition and estimation of active accounts. We therefore decided that it was more appropriate to base our analysis on all accounts (see Appendix 4.10). These figures are shown in Table 5.

4.37 Another clearer ([]) said that relatively little weight should be attached to our market share estimates because of limitations in the data. We recognize that there are difficulties in ensuring that the comparisons are being made on a like-for-like basis, and that our comparisons are being made over a shorter time period than the MORI data. Nevertheless, we believe it relevant to consider this data as complementary to the MORI volume data, in particular because it aligns with our PCA market definition, and is based on the total number of PCAs rather than on a sample of the population.

TABLE 5 Share of number of PCAs and MORI share of stock for the eight largest banks in Northern Ireland in 2002 and 2005 per cent

Share of number of MORI share of Share of number of MORI share of accounts 2002 stock (2002 Q4) accounts 2005 stock (2005 Q4)

BoI First Trust  Northern* Ulster Subtotal clearers 73.3 78.5 68.6 72.4

Abbey A&L  Halifax Nationwide Subtotal non-clearers 23.1 17.6 26.9 22.7

Total 96.4 96.1 95.5‡ 95.1

Source: CC analysis of financial information provided by the banks.

*The significant difference between Northern’s market share calculated on the account numbers provided by the banks and the MORI data is that CashMaster account is included in the latter but not the former. If CashMaster were included we estimate that Northern’s market share would have been approximately [] per cent in 2005. ‡We obtained detailed financial data from the eight largest banks. We used the MORI estimate of market share for those banks for which we did not have detailed financial data (2002: 3.6 per cent and 2005: 4.5 per cent) to calculate shares of the total market based on the number of PCAs for the eight largest banks.

4.38 Market shares based on the number of accounts show that the clearers’ aggregate market share was 69 per cent in 2005. This fell from 75 per cent in 2002. These shares are a little lower than those calculated on the basis of the MORI data (MORI market share data for the clearers as a group is 72 per cent in 2005 Q4).

4.39 In response to our Emerging Thinking, one of the clearers ([]) gave us data on volume market shares for the PCA market in Great Britain. This showed that over the period 2000 to 2005, the fall in market share for the banks based in Great Britain has been less marked than the fall in Northern Ireland, and the non-clearers in Great Britain have grown less rapidly than in Northern Ireland.

60

4.40 Market shares could change at different rates in Northern Ireland and Great Britain for many reasons. These might include, for example, the stage of development of the market, and differences in competitive conditions between Great Britain and Northern Ireland. Changes in market share could result from new entry, expansion, or the development of increased competition in the market. On the other hand if, for example, a group of banks were offering less competitive products, customers might be motivated to switch to other banks, even if there were substantial barriers to switching (possibly greater than those in Great Britain). Thus, in this example, there could be a greater change in these banks’ market share in Northern Ireland than would apply to similar banks in Great Britain, despite there being intrinsically higher barriers to competition in the Northern Ireland market. In the long term this may force all banks in Northern Ireland to respond to market pressures, but switching rates by themselves do not indicate whether or not the market in Northern Ireland is more or less competitive than the market in Great Britain.

Value-based market shares and concentrations

4.41 We also looked at value-based market shares and concentrations, based on the data that we collected from the eight largest banks for 2002 to 2005. The financial benefit of an individual PCA to a bank depends on customer usage. It depends, for example, on average credit balances (on which the bank can earn interest) and the level of chargeable behaviour. In addition, the bank may take into account the extent to which the account holder has, or might have in the future, other (potentially more profitable) products.

4.42 In order to calculate market shares in the market as a whole based on the information provided to us by the eight largest banks, we had to make assumptions concerning the other banks for which we did not obtain information. We used the MORI data to estimate the total market share of the other banks (4.5 per cent). We assumed that their financial data would be consistent with that of the non-clearers and thus used non-clearers averages where applicable. We considered that the market shares are not materially affected by this assumption.

4.43 We looked at the banks’ average PCA credit balances and revenues. On average we found that, in 2005, the banks earned more than half their total income from NII on credit balances, over 40 per cent from charges and 5 per cent from NII on debit balances. This was fairly consistent across the clearers and the non-clearers and across the period we reviewed (2002 to 2005).

4.44 We found that the average credit balance of the clearers is substantially higher than that of the non-clearers (see Table 6). In 2005, the average credit balance per account was £1,796 for the clearers as a group compared with £1,184 for the non- clearers as a group, a difference of over 30 per cent. Based on these figures, we estimated that the clearers held over 76 per cent of the total credit balances for PCAs in Northern Ireland in 2005.

4.45 One of the clearers ([]) said that the data on average credit balances was inconsistent with MORI data which showed that two of the non-clearers, Nationwide and A&L, had the highest-income customers. However, we note that the average credit balance could be determined by several different factors in addition to average income, including, for example, the fact that the non-clearers have a higher proportion of younger customers, or a lower proportion of main accounts (see paragraph 2.75).

61

TABLE 6 Share of PCA credit balances for the eight largest banks in Northern Ireland

per cent

Total average Total average credit balance credit balance (2002 year end) (2005 year end)

BoI First Trust  Northern Ulster Subtotal clearers 78.3 76.6

Abbey A&L  Halifax Nationwide Subtotal non-clearers 18.3 19.8

Total 96.6* 96.5*

Source: CC analysis of financial information provided by the banks.

*We obtained detailed financial data from the eight largest banks. We assumed that the average credit balances per account for those banks for which we did not have detailed financial data was the same as the average for the four non-clearers shown in Table 3, and that the market share of the smaller banks was 3.4 per cent (2002) and 4.5 per cent (2005).

4.46 We also calculated value-based market shares based on the revenue information provided by the eight largest banks. This enabled us to consider the financial benefit of all aspects of the PCA, including credit balances, charges and use of overdrafts. Using this method of calculation, the clearers earned more than 77 per cent of the total income earned on PCAs in Northern Ireland in 2005 (see Table 7).86

TABLE 7 Share of total revenue from PCAs for the eight largest banks in Northern Ireland in 2002 and 2005

per cent

Share of total Share of total revenue 2002 revenue 2005

BoI First Trust  Northern Ulster Subtotal clearers 82.5 77.5

Abbey A&L  Halifax Nationwide Subtotal non-clearers 14.9 19.3

Total 97.4 96.7*

Source: CC analysis of financial information provided by the banks.

*We obtained detailed financial data from the eight largest banks. We used the MORI estimate of market share for those banks for which we did not have detailed financial data (3.6 per cent (2002) and 4.5 per cent (2005)) to calculate shares of the total market based on the number of PCAs for the eight largest banks. Note: In order to improve comparability between banks, we used the same interest rate for each bank to value credit balances.

4.47 However, we note that the revenue a bank will earn directly from a PCA will also be dependent on the services it offers (see paragraph 4.328). A clearer may earn more

86In calculating PCA revenue the banks had to estimate the income value of PCA credit balances. In order to standardize the valuation methods we normalized the figures by applying the overall average value of funds to the individual banks. Total income from the eight banks remained unchanged.

62

revenue from PCAs than a non-clearer because it offers more services. Furthermore, some banks offer packaged accounts for which customers pay in return for the additional benefits; the figures in Table 7 did not include the cost to the bank of providing such benefits.

Conclusions on market shares and concentrations

4.48 The MORI volume-based market shares show that, overall, the clearers have lost significant share over recent years. These losses appear to have largely been driven by the relatively poor performance of two of the four clearers, and as such may be less a reflection of market-wide pressures on the clearers, and more of particular issues associated with the two clearers in question, First Trust and Northern. We found it relevant to look at other volume-based and value-based market share data given the issues that we had with the MORI data.

4.49 Each of the bases used for calculating volume and value-based market shares has certain limitations. However, irrespective of these, we found that the clearers’ market share in 2005 was between 69 and 77 per cent. The combined market share of the clearers has fallen since 2002.

4.50 Market shares based on a measure of value rather than volume give a higher market share to the clearers. There are also significant differences between the individual clearers based on the two different measures. In particular, Northern and First Trust have preserved a significantly higher share of the market by value than would appear to be the case using volume-based market shares.

4.51 The market shares of the clearers based on credit balances have fallen only slightly—by around one or two percentage points—between 2002 and 2005. This suggests that the effect on some of the clearers’ revenues of a reduction in market share by volume might have partially been mitigated by retaining more of their high- value customers and hence increasing their average credit balance per account at a faster rate than the non-clearers.

4.52 We looked at the level of concentration in the market in 2005 based on each of these measures of market shares (see Table 8). Although the market is more concentrated based on the value of credit balances or revenue than suggested by the MORI market shares based on number of main accounts or the data provided by the banks on the number of accounts, all measures show the market in 2005 to be concentrated.87 This suggests a lower risk of banks having monopoly power than if concentration were higher, although it does not indicate whether aspects of market conduct or structure may mean that this is not a well-functioning market.

TABLE 8 HHI values for the eight largest banks in Northern Ireland based on different measures, 2005

Share of main accounts (MORI) 1,541 Number of accounts 1,418 Share of PCA credit balances 1,648 Share of PCA revenues 1,664

Source: CC calculations.

Note: Treats ‘Other’ as one entity and so slightly overestimates the true HHI. For comparison, the HHI for the market as a whole based on MORI data in 2005 Q4 was 1,554.

87The HHI values for the overall market will be slightly lower as the smaller banks account for 4 to 5 per cent of the market.

63

Branch networks

4.53 The clearers all have substantial branch networks spread across Northern Ireland, albeit of varying size. In contrast, the non-clearers have fewer branches in Northern Ireland, located primarily in the main centres of population, although they have significant UK-wide networks of branches (see Table 1). Table 9 sets out the number of branches and agencies in Northern Ireland for each bank.

TABLE 9 Number of branches and agencies in Northern Ireland in January 2007

Sub-branches Full or part-time branches branches* Agencies

BoI 45 1 0 First Trust 56 0 0 Northern 95 6 0 Ulster 87 4 0 Abbey 21 0 0 A&L 14 0 0 Halifax 15 0 30† Nationwide 14 0 12 Woolwich 9 0 0 HSBC 5 1 0 Co-operative Bank 1 0 0

Source: CC from information provided by the banks.

*Sub-branches or part-time branches may, for example, have limited opening hours. †This includes 16 independent agents and 14 Halifax estate agents which offer counter services.

4.54 The clearers have reduced their numbers of branches from 345 in 1996 to 289 in 2007. However, the bulk of these closures took place in the late 1990s. Recent closures have generally been due to merging neighbouring branches. Some non- clearers have also reduced their number of branches; A&L reduced its branches from 17 in 1996 to 14 in 2007, and, although Halifax has opened four new branches, it is now present in fewer locations having reduced its number of agencies.

4.55 Some of the non-clearers (in particular Halifax and Nationwide) supplement their full- or part-time branches with agency services. These are entities which provide branch- type services to personal customers on behalf of the bank; although some are stand- alone agencies, they are often part of larger businesses such as estate agents, solicitors and insurance brokers. Halifax told us that its agencies offered its full product range and a counter service and were located in the Belfast suburbs or in smaller towns where a full branch was not warranted. In addition, several banks (primarily the non-clearers) supplement their branch networks through an agreement with the Post Office to provide counter services for PCAs including cash withdrawal, balance enquiries, and paying in cash and cheques.88 BoI is the only clearer to have such an arrangement.

4.56 The majority, if not all, of the staff in the branches are customer facing, with back- office tasks increasingly being centralized away from the branch. Figure 6 shows that the clearers have, on average, a larger number of total staff (full-time equivalents (FTEs)) per branch than the non-clearers. The number of staff per branch is, however, dependent on business mix—clearers tend to have more loans and business banking, for example, whereas the non-clearers tend to have more mortgage business. We asked the banks to estimate the proportion of branch staff

88This includes A&L, BoI, Co-operative Bank, Nationwide and smile. cahoot customers can cash cheques free of charge up to their daily limit. The Post Office provides counter services for many more suppliers of basic bank accounts.

64

working on PCAs, but this was not information that the banks could readily provide. Two of the clearers told us that around one-third of branch effort was related to PCAs. We believe that this is likely to be higher for the non-clearers given their much lower proportion of business customers.

FIGURE 6

Distribution of FTE per branch

14 30

12 25

10 20 8 15 6 10 4

2 5

0 0 1 4 7 10131619222528313437404346495255586164 No of FTE per branch

Non-clearers (L) Clearers (R)

Source: The banks. Note: The y axes are the number of branches (the blue line is non-clearers—shown on the left hand y axis; the red is clearers—shown on the right-hand axis). Agency staff are excluded.

4.57 We looked at the movement in headcount at branch level in Northern Ireland. Table 8 in Appendix 4.12 shows the movement in branch FTEs in Northern Ireland over the period 2002 to 2004. The chart shows that with the exception of [], headcount has not reduced substantially. We found that the non-clearers were increasing the number of PCAs at their branches without increasing their branch headcount.

4.58 We considered the role of the banks’ branch networks. The ORC survey showed that a branch is very important to customers when opening an account. Some of the banks told us that customers could open PCAs remotely, but, due to money laundering regulations, many banks still require the customer to go to the branch in person to validate their identity and address. Although many customers transact much of their day-to-day business away from the branch (for example, almost two- thirds of customers transact most of their business via an ATM—see paragraph 2.123), other customers, particularly those in socio-economic groups D and E, still visit the branch fairly frequently. We also understand from our Millward Brown survey that customers regard it as important to have the option of visiting a branch to resolve any difficulties that might arise. Halifax also told us that the reassurance provided by a branch was important.

4.59 In addition, the banks consider the branches to be important marketing tools, raising customers’ brand awareness and providing an opportunity for cross-sales. However, the clearers have many branches in rural locations or small towns with a relatively limited catchment area, and the costs of providing a wide branch network are

65

significant. Although it is likely that not all of these branches are profitable on a stand- alone basis, branch closures are considered to be sensitive and difficult to achieve without significantly damaging the bank’s overall reputation and hence its business.

4.60 All banks now offer remote banking as an additional channel. Whilst few customers choose to use remote banking as the only mode of accessing their PCA, we were told that they expected to be able to access their accounts from a variety of channels including branches, the Internet, and by telephone.

4.61 We thought that a substantial branch network provides a benefit to a bank in terms of attracting and recruiting both switchers and new-to-banking customers, and in providing existing customers with access to branches. Branch networks also carry substantial costs. Banks need to consider both the costs and the benefits and decisions on opening or closing branches will depend on a variety of factors, including the bank’s strategic view and the effect that branches can have on the number and type of products sold, of which PCAs are only a part. We observed that the banks have differing views on the importance of the branch network to them. Some ([]) claimed that the branch network was key to their PCA businesses, notwithstanding the growth of remote channels such as telephone and Internet banking. Others saw the branch network as less important and saw the need for fewer branches than some of the clearers currently had. [] did not consider that a large branch network was necessary for successful entry into the market and told us that, [], it would be striving to encourage greater use by its customers of automated channels. A&L publicly described itself as ‘a direct bank with a High Street presence’ and stated that ‘… the cost of supporting a large branch network will become a hindrance rather than an advantage’.89 [] believed that multi-channel banking was becoming increasingly important in the Northern Ireland market, with less reliance on branches and greater emphasis on remote channels, although the branch was still the main channel for opening a PCA.

Entry and expansion

4.62 We looked at barriers to entry, the history of entry and exit, and the likelihood of entry in Northern Ireland (see paragraphs 4.64 to 4.94). We also looked at the barriers to expansion which might inhibit a small-scale provider from expanding its activities, and the prospects for successful expansion (see paragraphs 4.95 to 4.107).

4.63 We identified seven categories of potential barriers to entry and expansion.

(a) obtaining a banking licence from the FSA and meeting other regulatory requirements;

(b) access to payment networks and a debit card scheme;

(c) access to the means of delivering PCA services to customers (eg a branch network);

(d) ability to bring new products to potential customers’ attention (including development of successful brands);

89Chris Rhodes, Managing Director of the Retail Bank, A&L, 12 November 2004 (source A&L Press Release)—quoted by First Trust.

66

(e) access to appropriate infrastructure and skills to process transactions and deal with PCA customers;

(f) access to information regarding the credit risks posed by potential customers (this is required if overdrafts are to be offered); and

(g) overcoming customer inertia and barriers to switching in the market in order to persuade customers to switch.

Entry

4.64 Barriers to entry may impact on potential entrants in three ways: delaying or complicating entry; increasing costs of entry; and/or raising operating costs or lowering revenues.

4.65 The need to obtain regulatory authorization (see paragraphs 2.98 and 4.63(a)) could be a significant barrier to entry. We note, however, that there are many possible entrants that would not need to seek further authorization to operate in Northern Ireland (either those already authorized in Great Britain, or banks from other EU member states). All new entry in the past ten years has been by banks based in Great Britain (see paragraph 4.85). We do not consider changes in ownership to represent new entry (see paragraph 4.84).

4.66 We did not find that the need to access payment networks or a debit card scheme (see paragraph 4.63(b)) constituted a significant barrier to entry. The clearers told us that other banks could access any of the clearing systems on fair and non- discriminatory terms via an agency bank arrangement. The non-clearers that employed clearing agents said that, although there were associated costs, they did not regard these as a significant barrier or an impediment to their ability to compete. Nor did we consider there to be any difficulty obtaining information on credit risks from companies such as Experian and Equifax (see paragraph 4.63(f)).90

4.67 We looked in some detail at the entry barrier that might arise from the need to establish access to some of the means of delivering PCA services to customers, including branches in particular (see paragraph 4.63(c)). We found that establishing telephone and Internet services would be reasonably easy, especially if the entrant already provided these services in other markets. Providing access to cash would be possible through the LINK ATM system, cashback from retailers (available when using a debit card to pay for transactions), or through an agency agreement with the Post Office for counter services. However, most customers place great importance on the availability of branches, even when they actually make little use of branches day to day (see paragraph 4.58). Branches are also the main means of recruiting new customers. Consequently it appears that banks without a branch network are likely to be at a disadvantage.

4.68 The clearers said that a relatively limited branch network would be required to compete in the PCA market. One clearer ([]) said that a small network of branches would be sufficient to cover the main conurbations in Northern Ireland and ensure that a significant proportion of Northern Ireland’s population had ready access to a branch. [] submitted an analysis of the ease of access of Northern Ireland consumers to the existing larger non-clearers (specifically Abbey, A&L, Halifax and

90We note that A&L said that credit rating information was incomplete because some banks shared only negative data (breached limits, missed payments, defaults etc). This might suggest that incumbent providers have an advantage over others in assessing credit risks for existing customers of other products. However, no other bank raised similar concerns.

67

Nationwide) which found that 75 per cent of the Northern Ireland population lived within 10 minutes’ drive-time of one of these banks, and 85 per cent within 15 minutes’ drive-time. Other banks said that the pull of major centres such as Belfast should not be underestimated and in rural areas customers would be prepared to travel considerable distances to access their bank.

4.69 The clearers pointed to the growth of the non-clearers in the PCA market, which had been achieved with comparatively small branch networks.91 One clearer ([]) presented an analysis showing that market share growth was greater for banks with smaller networks than larger ones over the period 2000 to 2005. However, this analysis did not take into account other factors (such as the nature of the PCAs on offer), which might determine the relative success of the banks in gaining market share (see paragraphs 4.39 and 4.40).

4.70 On the other hand, some of the non-clearers indicated that a lack of a dense branch network was an important barrier to their ability to attract new customers. For example, one non-clearer said: ‘We have to work very hard … to convince somebody that, even if they do not live close to us and they are going to have to travel, the travel is going to be worthwhile because of the additional facilities and our competitive pricing’. However, all agreed that the optimum number of branches would depend on the overall strategy of the provider, bearing in mind the mix of products and services offered.

4.71 We looked at the costs of acquiring and developing a branch network. Property prices are generally lower in Northern Ireland than much of the UK, properties can be leased rather than purchased, and only some of the costs involved (eg customizing the building for use as a bank branch) will be sunk. New branches can be located in areas with the greatest density of potential customers (eg major shopping centres). Nonetheless, some non-clearers said that the cost of developing and maintaining a branch system from scratch could be considerable. For example, one non-clearer said: ‘Establishing a branch network would appear to be the most costly barrier to entry due to the significant investment involved. Any new entrant would need to consider the return on this investment would take a number of years.’ Similarly another non-clearer said: ‘it would be a huge investment decision for us to expand our network and reach a bigger scale than we have currently got’. However, given the size of some of the financial institutions that would be likely to enter and the total costs involved in such operations, we believe that the sunk cost element of developing a small number of branches would be unlikely to be prohibitive, although those costs would have to be evaluated against potential returns across all the products that respective banks offer, of which PCAs are only a part.

4.72 Branch closures can generate significant adverse publicity for a bank (see paragraph 4.59). However, we did not believe that this potential barrier to exit was likely to cause banks to be reluctant to open branches in the first place.

4.73 We were told that the extent of the disadvantage associated with the lack of a branch network might be declining over time as customers became more comfortable with alternative distribution channels. First Trust, for example, said that while the physical presence of a branch might help attract customers, non-clearers were adopting a variety of business models which placed less reliance on branch networks.

91One clearer ([]) argued that growth in market share of the PCA suppliers, other than the four largest banks in each market, had been much more rapid in Northern Ireland than Great Britain. Based on the MORI data for Northern Ireland, the share of the banks other than the four largest grew from 16 to 27 per cent between 2000 and 2005, while in Great Britain over that period based on Mintel data it grew from 29 to 35 per cent.

68

4.74 We did not consider that entry as a remote provider only could build sufficient market share to act as a competitive constraint on the existing banks in Northern Ireland. Market shares of remote providers remains low, and is expanding slowly, if at all.92 This suggests that entrants without branches would be disadvantaged, at least in the short to medium term.

4.75 In order to attract customers, an entrant would need to promote its brand and its PCA products (see paragraph 4.63(d)). The DTI survey of switching93 found that familiarity and trust in providers was an important factor in customers’ willingness to switch, particularly for banking.

4.76 Some of the clearers told us that entrants from the British market would benefit from existing awareness of their brands and products, for example through advertising on UK-wide television. We were also told that financial providers established in other product areas could exploit familiarity with their brand if entering the market for PCAs. For example, one clearer ([]) said ‘… the non-clearers are well-established brand names among people in Northern Ireland and the leading non-clearers are very substantial financial institutions with proven experience in developing their market position and strong market shares in mortgages and savings products’.

4.77 Another clearer ([]) said that the decisions of some established providers to introduce start-up brands such as cahoot and smile suggested that recognition was not a significant barrier. Co-operative Bank, however, said that it had been necessary to invest heavily in the UK as a whole to ensure recognition of smile. [] told us that ‘Customers also need to be convinced that the new entrant is positioned to keep their money safe. Hence a strong brand name is also essential, eg [].’

4.78 Paragraph 4.63(e) refers to the need to access appropriate infrastructure and skills to process transactions and deal with PCA customers. The costs of creating systems to provide PCA services are high. However, if a potential entrant already supplied PCAs in Great Britain it might be able to transfer its existing systems,94 and there are opportunities to reduce set-up costs through using a clearing agent, using off-the- shelf banking IT systems, and/or through outsourcing. [] said that: ‘Improvements in technology and the possibility of outsourcing back-office functions mean that economies of scale can be achieved with a lower asset base. The provision of ATMs and telephone and Internet banking services, in particular, has decreased the import- ance of economies of scale in personal banking markets.’

4.79 An entrant would have to recruit and train local front-office staff. [] said that there was a pool of educated and motivated labour. The main source of experienced staff would be employees of the existing providers. We did not think that a lack of availability of suitable staff would be a barrier to entry in Northern Ireland.

4.80 Issues related to switching (see paragraph 4.63(g)) are discussed in paragraphs 4.209 to 4.260. As noted in paragraph 4.208 we observe a lack of responsiveness to changes in charges or interest rates in the market. If, as a result, it is difficult to attract switchers, it will be difficult for a new entrant to build up a stock of customers by attracting them from other banks. This is the factor which is most likely to make entry unattractive to a potential new entrant. A new entrant could target new-to- banking customers, but might be disadvantaged by lack of familiarity with the brand

92We collected data from all the banks in Northern Ireland on the number of open PCAs. By this measure, the combined share of the four Internet banks was 1.1 per cent at the end of 2006. 93Switching Suppliers, Consumer Affairs Report, 2001 (DTI 2001). 94In contrast, Danske has faced high costs in aligning Northern’s systems with its own following the acquisition of Northern.

69

and products, and would need to ensure that it attracted profitable customers. The reasons for choice of bank indicate that new-to-banking customers are likely to be less sensitive to comparative prices than customers who switch; their choice of bank is largely influenced by branch location and personal recommendation, factors which do not work in favour of a new entrant.

• History of entry and exit

4.81 We looked at the history of entry into the Northern Ireland PCA market.95 We are not aware of any exits from the PCA market in the last ten years. Entrants within the last ten years are listed in Table 10.

TABLE 10 Entrants to Northern Ireland PCA market since 1995

Current market share (2006 Q4) % Date entered cahoot June 2000 Co-operative Bank May 2003 HSBC  June 2001 IF September 2000 smile October 1999

Source: MORI.

Note: Excludes providers of products which fall outside our market definition (such as the Post Office Card Account).

4.82 The market shares in Table 10 indicate that new entrants are not able quickly to build a significant market share (although we note that none have invested in large-scale branch networks). Those that have entered using a branch network have picked up a higher market share than the Internet banks; this suggests that even a few branches create more selling opportunities than a remote provider can achieve.

4.83 We considered in particular the experience of the two recent branch-based entrants, Co-operative Bank and HSBC. Co-operative Bank told us that it had encountered entry barriers, particularly low switching behaviour leading to high acquisition costs, and the branch preference of Northern Ireland consumers requiring the investment in physical outlets. HSBC told us that its entry into the PCA market was facilitated by its existing physical presence in Northern Ireland in the form of HSBC Asset Finance.96 This provided staff and a commercial customer base, as well as a small branch network. HSBC said that its main challenge was in training and developing staff who had not previously operated a retail branch.

4.84 Some of the clearers referred to changes of ownership among the Northern Ireland banks, eg First Trust was formed when AIB acquired TSB in 1991, RBS acquired Ulster in 2000, and Danske acquired Northern in 2005. However, transfer of ownership does not represent new entry. Nonetheless, we recognize that a new owner may seek to improve performance and to adopt new competitive tactics which might stimulate competition in the market.

4.85 All new entry into the provision of PCAs in Northern Ireland has been by banks based in Great Britain. The small size of the Northern Ireland market, and the fixed costs

95The larger non-clearers entered the PCA market in the mid- to late 1980s (see paragraph 2.29), and therefore were not regarded as new entrants. 96Part of HSBC’s business banking activities.

70

involved, makes entry for a start-up bank specifically for Northern Ireland relatively unattractive. There has been entry on a UK-wide basis by remote providers, but only where supported by existing banks. There has been entry through acquisition but only by banks established elsewhere.

4.86 We were told that Halifax was an example of significant entry through developing a new branch network, albeit in the Republic of Ireland.97 Bank of Scotland purchased some of the former Electricity Supply Board’s outlets in early 2005, with the stated intention of opening 46 new branches between January 2006 and spring 2007. As at the beginning of February 2007 it had opened 24 Halifax branches. In addition to acquiring these branches, it also acquired a client base. We were told that the conversion and opening programme will take two years from agreeing the purchase, but is likely to establish Halifax as a significant operator in the Republic of Ireland. However, we found that although Halifax has offered a limited range of personal banking products since early 2006, it has not yet offered a PCA in the Republic of Ireland, although this is planned for Q2 2007. This suggests that even with a branch network, entry to the PCA market is seen as difficult. We received no evidence to suggest that such an unusual opportunity to enter a market on a significant scale was likely to occur in Northern Ireland.

• Evaluation of entry

4.87 The size of any barriers to entry would depend on the particular circumstances of each potential entrant. The anticipated profitability of entry would, in particular, be affected by the need to create a branch network and to market the brand and product to overcome any reluctance to bank with less well-known providers. We consider the most significant constraint on entry to be the low willingness of customers to switch, which means the rate of payback on any investment in branches and marketing could be slow. Further, the optimal branch density for PCA sales appears to be higher than is justified by the overall portfolio of personal financial products that a new entrant might want to sell, as these other products are, we were told, less dependent on a branch network to achieve sales, and we think it unlikely that a new entrant would choose to offer only PCAs.98

4.88 We considered who might be likely to enter the Northern Ireland PCA market. We note that three of the four largest banks in the Republic of Ireland are already in Northern Ireland (the exception is Permanent TSB).99 The fifth largest, National Irish, is owned by Danske which also owns Northern in Northern Ireland.100

4.89 The most likely new entrant would be an existing bank based in Great Britain since fewer potential barriers to entry apply. Existing banks in Great Britain that do not market their PCAs in Northern Ireland101 include some large retail banks, notably Lloyds TSB. Barclays is present in Northern Ireland through its ownership of Woolwich, whose branches are being rebranded as Barclays. Other possible entrants

97The entry was originally branded as Bank of Scotland (Ireland). In August 2006 it announced that it was rebranding its retail banking business as Halifax, as this name enjoyed strong brand recognition in the Republic of Ireland. 98We were also told that branches were important for business banking, and so an entrant that is actively targeting business banking as well as personal banking may have a greater incentive to invest in branches, although here we have addressed barriers relevant to personal banking only. 99Permanent TSB is the retail banking division of Irish Life & Permanent plc. It was formed in 2001 when Irish Permanent plc merged with TSB Bank. Although it has highlighted the current account market as a key area for future growth, it is a relatively new retail bank which is currently focused on expansion in the Republic of Ireland. 100Source for Republic of Ireland market shares: Competition in the (non-investment) Banking Sector in Ireland, Irish Competition Authority, September 2005. 101Although residents of Northern Ireland could open an account through a British branch or through distance banking.

71

include banks and specialized providers with smaller shares of the PCA market in Great Britain such as Citibank, or some current or former building societies.

4.90 There is unlikely to be completely new entry into the provision of PCAs in Northern Ireland alone because of the small size of the market. Entry would be more likely on a UK-wide basis. The possible exception to this is entry by financial institutions already present in Northern Ireland, which do not currently sell PCAs. The credit unions and Northern Ireland building societies,102 for example, might consider diversifying into the market. [] It is also possible that a Republic of Ireland building society might contemplate entry. BoI told us that some Republic of Ireland building societies have developed PCAs for the Republic of Ireland market.

4.91 Credit unions are not currently able to offer PCAs. They would encounter a number of additional barriers to entry to the PCA market, which might include the need to obtain regulatory approval to be a licensed deposit taker; the need to join banking schemes for money transfer, clearing, LINK etc; and the need to obtain the information necessary to assess the credit risks of potential customers. From November 2006, nine credit unions in Great Britain, none of which are present in Northern Ireland, began testing a scheme to offer bank accounts provided by Co-operative Bank. The accounts offer a Visa/ATM card, direct debits and standing orders, but there is no cheque book or overdraft facility.103 [] However, developing this product to the trial stage has taken three years of planning and negotiation and we understand that the credit unions do not currently intend to move into PCAs.

4.92 The Post Office told us that it had no plans to enter the PCA market. Instead it is looking to expand its business acting as an agent to provide counter services for as wide a range of banks as possible.

4.93 One entry strategy would be for a UK financial services provider to diversify into the provision of PCAs. However, such entry is rare, and would be likely to be decided with reference to the whole of the UK rather than Northern Ireland alone. Tesco and Sainsbury’s banks have a known brand name, although they do not have banking facilities in their supermarkets in Northern Ireland. While these could be introduced it would involve cost and may impact their other business. Neither has indicated any intention to develop PCA operations in Northern Ireland.

4.94 An alternative entry strategy might be through a joint venture with other institutions with an existing branch system, such as the Post Office. The Post Office offers basic counter services to lodge credits, withdraw cash and make balance enquiries. [] Bradford & Bingley has two full branches and three agents, while other building societies (eg Britannia, City of Derry, Irish Nationwide,104 and Leeds and Holbeck Building Societies) only have single branches in Northern Ireland. While these or other financial institutions could be a base for entry, this would be on a small scale that would not itself be a significant constraint on the behaviour of existing banks.

Expansion

4.95 Expansion by a non-clearer is more likely to change the levels of competition in the market than entry. An existing bank seeking to expand its activities in Northern

102The Progressive Building Society has a large number of branches and agencies across Northern Ireland. The only other Northern Ireland building society, City of Derry Building Society, has just one branch. 103Source: Credit Union News, July 2006 (www.abcul.org/lib/liDownload/88/CU%20News%20July%202006.pdf). 104Irish Nationwide Building Society is a building society based in the Republic of Ireland. It is not related to Nationwide Building Society.

72

Ireland would not encounter difficulties in regulatory authorization, access to clearing and debit card schemes, or information on credit risks (see paragraph 4.63(a), (b) and (f)). Infrastructure and skills could be scaled up as required (see paragraph 4.63(e)). The same applies to certain means of delivering services to customers, such as having Internet facilities.

4.96 The remaining potential barriers to expansion are access to customers through a branch network (see paragraph 4.63(c)), bringing products to the attention of potential customers (see paragraph 4.63(d)), and persuading customers to switch (see paragraph 4.63(g)). The issues on switching are discussed in paragraphs 4.209 to 4.260.

4.97 The larger non-clearers do not have plans to expand their branch networks in Northern Ireland, despite the opportunities that branches provide to attract new customers. Although they told us that they believed a larger branch network would allow them to pick up a greater share of the Northern Ireland PCA market (see paragraph 4.70), their business model rests to a large extent on selling mortgages, loans and savings accounts which require a lower density of branches to sell and to service.

4.98 HSBC has five branches (plus one sub-branch). HSBC told us that it did not currently have plans to expand this significantly []. HSBC said that ultimately it would like an eight- to ten-branch network serving Northern Ireland.105 HSBC told us that it had not faced significant difficulties in planning its Londonderry/Derry branch (opened November 2006), as the bank was able to recruit high-quality staff from the local workforce and to find suitable premises. The other recent branch-based entrant, Co- operative Bank, has not given any indication that it intends to expand its Northern Ireland branch network at the moment.

4.99 The clearers said that the non-clearers’ branch networks were not a barrier to expansion in that they could, and were, expanding market share on the basis of their existing branch networks, and there were no barriers to increasing the density of the network. For example, Ulster noted that A&L and Halifax had both reduced the number of branches and agencies in Northern Ireland between 1996 and 2005 but they had increased their market share significantly over this period. Similarly, Northern said that the sizes of the branch networks adopted by the non-clearers was the result of a strategic decision rather than any constraints; it noted that some of the non-clearers such as Halifax were expanding their networks in other parts of the UK and in the Republic of Ireland when they judged there to be a profitable opportunity to do so.

4.100 Some non-clearers already use significant advertising, most notably Halifax. However, we found in our Millward Brown survey that while respondents may be aware of banks’ marketing efforts, these were not perceived to relate to PCAs (see Millward Brown report, section 5.1). In the case of smaller suppliers, advertising might be ineffective if interested consumers cannot research the products and open an account in a local branch. However, Ulster said that these banks were likely to be seen as focused on personal customers, and that they could cross-sell PCAs to their personal customer base.

4.101 The remote providers have failed to pick up a large share of new PCAs, even where they have agreements with the Post Office to provide counter services, or

105HSBC Senior Manager in Belfast, quoted in Business Eye, October 2003.

73

transactions can be made through associated branch banks (eg first direct and HSBC). This suggests that Internet-only delivery remains unattractive to many customers, or banks cannot recruit customers without branches acting as sales outlets, making it difficult for remote providers to expand significantly.

• History of expansion

4.102 Changes in the market share over the last six years for the five largest non-clearers are shown in Table 11. The total market share of these five banks has increased from 14.3 to 24.8 per cent (based on MORI data).

TABLE 11 Market share for the larger non-clearers, 1999 to 2006

per cent

2006 1999 Q4

Abbey A&L Halifax  Nationwide Woolwich Total 14.3 24.8

Source: MORI data.

4.103 We note that the non-clearers, whilst actively seeking and gaining market share on a UK-wide basis, do not target Northern Ireland specifically. The nature of the PCA market is such that changes in flow of PCAs would feed relatively slowly into changes in the market share of stock (see paragraph 4.24).

• Evaluation of expansion

4.104 The clearers told us that there were few if any barriers to expansion. For example, one clearer told us that: ‘There are no material barriers to a small provider of PCAs expanding its share in Northern Ireland. The experience of many of the smaller players over the last five years shows these smaller providers can successfully compete and significantly increase their shares of PCA banking in Northern Ireland.’ Another clearer said that the potential for PCA growth might depend on the overall entry strategy of a multi-product bank. It said that PCA growth must be balanced against their other business objectives in the market (eg mortgages or business banking). One of the clearers ([]) also told us that, given the more rapid fall in market share experienced by the clearers in Northern Ireland compared with Great Britain, barriers to expansion must be low in Northern Ireland (see paragraphs 4.39 and 4.40).

4.105 We believe that the sunk costs of opening new branches, on an incremental basis, are unlikely to be prohibitive for most existing small-scale operators. However, because PCAs are only one part of their business, customers of other products place less importance on branch access, and in particular, because of the slow rate at which customers can be expected to switch from other suppliers, we consider it doubtful that existing suppliers would seek to expand through investing in new branches. They would be more likely to seek increased sales from their existing branch network.

4.106 The non-clearers indicated to us that they had no plans for expansion of their branch networks in Northern Ireland, and would be more likely to exploit growth opportunities

74

elsewhere in the UK. The potential for growth through their existing branches may provide some constraint on pricing in the PCA market but would not be as effective as an expansion programme based on new branch openings.

Conclusions on entry and expansion

4.107 We found no evidence of entry to the market in Northern Ireland on any significant scale. We also found that expansion in the market took many years. While the non- clearers collectively have achieved a significant market share, they have been providing PCAs in Northern Ireland for around 20 years (since the Building Societies Act 1986 came into force in 1987). We found that barriers to entry and expansion included, in particular, a lack of willingness to switch among existing customers, and a distrust of unfamiliar banks. In addition, the non-clearers are not focused in particular on expansion in Northern Ireland. The evidence that we received has not led us to form an expectation that the non-clearers will increase their number of branches in Northern Ireland in the foreseeable future.

Bank conduct

4.108 In paragraphs 2.14 to 2.47 we described the banks that provide PCAs in Northern Ireland. Paragraphs 4.3 to 4.13 described their different business models. Paragraphs 2.48 to 2.64 discussed the characteristics of PCAs and outlined their charging structures. In this section we discuss PCA charging structures in more detail, including how they are structured and how they are explained to customers (see paragraphs 4.109 to 4.139). We then look at recent changes to the clearers’ PCA charging structures (see paragraphs 4.140 to 4.160). Next we look at how certain charges and interest rates have changed over time (see paragraphs 4.162 to 4.179), before reaching our conclusions on charging structures and levels (see paragraphs 4.180 to 4.186). At the end of this section we consider competition for customer groups (see paragraphs 4.187 to 4.192), and competition on non-price factors (see paragraphs 4.193 to 4.206).

Charging structures and levels

4.109 We outlined the overall characteristics of PCA charging structures in paragraphs 2.50 to 2.64. We now discuss these charging structures and levels in more detail:

(a) overview of charging structures: traditional and fee-free PCAs, unauthorized overdraft charges and ancillary charges;

(b) assessment of charging structures and information provision;

(c) recent changes to charging structures;

(d) trends in interest rates and charges; and

(e) discussion of levels of interest rates and charges.

We draw some conclusions on charging structures and levels in paragraphs 4.180 to 4.186.

75

Overview of charging structures

• Traditional and fee-free PCAs

4.110 Until November 2005, the majority of PCA customers in Northern Ireland were on traditional PCAs. Customers on traditional PCAs incur both transaction and maintenance charges when operating their account in authorized or unauthorized overdraft. Customers are charged for every transaction undertaken during the quarter if they go into debit at any point during the quarter (unless charges are capped or buffer zones apply). Transactions might include, for example, payments or credits by cheque, ATM withdrawals, debit card payments, or payments by standing order or direct debit. If customers go into authorized overdraft they are charged retrospec- tively for any transactions that they have made since the start of the quarter, and they continue to be charged for every transaction for the remainder of the quarter, whether or not they return to credit. In addition, a maintenance charge is levied for the quarter.

4.111 Two of the four clearers (First Trust and Northern) continue to operate traditional PCAs. Ulster switched customers on traditional accounts to a fee-free charging structure in November 2005, as did BoI in November 2006. Charges for setting up, amending or cancelling standing orders and direct debits were removed by Northern for all PCAs in February 2005, by Ulster when it changed its PCA charging structure in November 2005, and by BoI when it changed its PCA charging structure in November 2006. On 3 May 2007, First Trust announced that it would launch a new PCA in June 2007. It also announced that it would remove charges for setting up, amending or cancelling standing orders on all PCAs from early June.

4.112 The fee-free PCAs offered by the non-clearers, BoI and Ulster do not charge customers for the day-to-day running of their accounts unless they go into unauthorized overdraft. This mirrors the basis on which PCAs are offered by the banks based in Great Britain. Northern’s new account portfolio (see paragraphs 4.145 to 4.148) is also fee-free in some respects. However, unlike the other clearers, on its new non-packaged PCAs (Northern Freedom and Northern Choice), transaction charges continue to apply for heavy users of certain services (cheques and over-the-counter cash withdrawals).106,107 First Trust’s new PCA, announced on 3 May 2007, will be fee-free subject to customers making a minimum of two automated transactions within the quarterly charging period, or keeping their account in credit during the quarter.

• Unauthorized overdrafts

4.113 By opening a PCA with a bank, a customer is bound by the terms and conditions which represent the customer’s contract with the bank. As part of this contract the customer agrees to ensure that funds are available in the account to meet payments arising from their transactions, taking into account any authorized overdraft limit. If the customer breaches this term of the contract by going into unauthorized overdraft, the bank reserves the right to refuse payment of the relevant transactions and to levy charges. As lenders licensed under the CCA 1974 (see paragraph 2.98), the banks are obliged to lend in a responsible manner. Although the banks can allow borrowing in excess of authorized overdraft limits, they generally evaluate whether to make a

106Northern estimated that [] per cent of its customers currently fell below this threshold. Charges are levied for cheques and over-the-counter cash withdrawals in excess of 13 per quarter each. 107We note, however, that the majority of Northern PCA customers remain on traditional PCAs or PCAs levying a fixed charge for authorized overdrafts.

76

payment that will create or extend an unauthorized overdraft.108 In doing so, they will take into account, for example, whether they believe the customer to have the capacity to repay the unauthorized overdraft.

4.114 The types of charges levied when a customer goes into unauthorized overdraft are outlined in paragraph 2.52 and described in more detail in Appendix 4.6. In summary, they are as follows:

(a) transaction charges (per item or per day, paid at the time the transaction is made):

(i) charges incurred when a payment is made that creates or extends an unauthorized overdraft (eg referral charge, card misuse charge). With the exception of guaranteed cheques, these payments are made at the discretion of the respective banks;

(ii) charges incurred when a payment is declined (eg unpaid charge). The bank may choose to return (unguaranteed) cheques unpaid or refuse to pay a standing order or direct debit that would otherwise create or extend an unauthorized overdraft;

(b) maintenance charges: a fixed payment for having an unauthorized overdraft; these may be monthly, quarterly or annual. They may be levied when a customer goes into unauthorized overdraft (eg charges for notification of account misuse) or on an ad-hoc basis (eg charges for letters regarding unauthorized overdraft).

For the purposes of this report, we refer to these charges collectively as unauthorized overdraft charges.

4.115 In addition, banks will generally charge a higher rate of interest on unauthorized overdrafts than on authorized overdrafts. Customers will also incur ancillary charges for any ancillary services that they require, whether they operate their account in credit or in debit.

4.116 We asked the banks for data on the incidence of unauthorized overdraft charges. Several banks did not have the necessary granularity in their systems to provide us with the percentage of their customers who had incurred unauthorized overdraft charges in a specified period. Among those banks that could respond to our request, we found that the proportion of customers paying unauthorized overdraft charges in a month varied from 3 to over 12 per cent. The annual incidence of charges was typic- ally three to four times the monthly incidence. This suggested to us that many of the customers who have incurred these charges once continue to do so (had no cus- tomers incurred repeat charges, the annual incidence of charges would be 12 times the monthly incidence). Around half of customers paying charges one year were identified as having paid similar charges in the preceding year, although there was significant variability between banks and one bank found that 90 per cent of charges by value were paid by customers who also paid charges the previous year.

4.117 We note that unauthorized overdraft charges are important to banks, particularly as they move towards a fee-free model which relies to a greater extent on income from customers in unauthorized overdraft (see paragraph 4.182). Overall, the eight largest banks earned almost 20 per cent of their PCA revenue in 2005 from unauthorized

108Banks do not always have the discretion, since cheques supported by a cheque guarantee card, for example, must be honoured.

77

overdraft charges, comprising almost half of all their PCA revenue from charges (see paragraph 4.320). Unauthorized overdraft charges are also a significant source of complaints,109 and may lead to a greater proportion than average of bad debts.

4.118 Some banks told us that they were seeking to reduce their reliance on unauthorized overdraft charges. Such charges tended to annoy customers, and revenue derived from them was inherently uncertain. Two of the clearers ([] and []), for example, told us that they provided advice to their customers on managing their finances and avoiding such charges. Another clearer ([]) told us that it often phoned customers who had exceeded their overdraft limit to warn them of a referral charge and to offer advice, [] said that it also provided a facility to send customers emails or text messages automatically if their balance dropped below a limit set by the customer. We recognize that banks have taken these actions, but also note an inherent tension since such charges are a significant source of revenue for the banks (see paragraph 4.185).

• Ancillary charges

4.119 In addition, there are a great variety of ancillary charges that may be applicable to an individual PCA. These include charges for certain services or additional adminis- tration (see paragraph 2.51). Ancillary services are generally not part of the day-to- day running of an account and are linked to services provided on request.

Assessment of charging structures and information provision

• Complexity of charging structures and practices

4.120 We note that, as set out in paragraph 2.48, a PCA services a wide variety of needs: to receive payments, hold deposits and make ATM withdrawals, for example. It may also provide an overdraft facility. This wide range of services may lead to complexity in charging structures.

4.121 In addition, we note that a bank is unusual in both providing a PCA to the customer, and also having access to the customer’s account which is used to pay for the service, either by employing the customer’s own funds, or by providing an overdraft facility. This dual role may be a factor in the banks’ charging practices (see, for example, paragraph 4.122(d)).

4.122 We looked at PCA charging structures and practices to see whether they were unduly complex; that is to say, more complex than might reasonably be necessitated solely by the wide variety of uses or services offered. Undue complexity could make it difficult for customers to understand and predict the charges incurred and interest earned or paid and to compare them between banks. We noted the following aspects which might add to complexity for the customer:

(a) There are a large number of different charges that might be levied on a particular PCA (see paragraph 2.50).

(b) The total amount payable in a particular month or quarter can be made up of several different elements (eg interest payments, transaction and maintenance charges).

109For example, one clearer ([]) said that charges accounted for 19 per cent of customer complaints, and that, of these, the majority of complaints (43 per cent (sic)) relate to unauthorized borrowing, primarily the referral charge.

78

(c) Transaction charges for operating a traditional PCA in authorized overdraft are incurred retrospectively for any transactions made since the start of a quarter and continue to be incurred for the remainder of that quarter even if the customer returns to credit.

(d) In many cases customers are not advised of charges prior to their being debited from their account. For example banks are not required to pre-notify customers of charges for services that are debited at the time the service is provided and where the customer has been notified in advance (eg in the price lists or terms and conditions), although some may choose to do so (see paragraph 4.123). This includes charges and additional interest payable for authorized or unauthorized overdrafts. Details of current requirements on pre-notification, and banks’ prac- tices on pre-notifying charges are described in detail in Appendix 4.13. This lack of pre-notification makes it difficult for customers to take appropriate action to ensure that sufficient funds (or an appropriate overdraft facility) are available to pay for the charges, or to confirm that the charges are correct, prior to their being debited.

(e) Certain charges are advised and debited a long time after they are incurred, making it more difficult for customers to understand the links between their actions and the charges incurred. For example, transaction charges for First Trust’s traditional PCA are applied more than one month after the end of the billing period.

(f) Charging practices, particularly in relation to unauthorized overdraft charges, can vary between banks. For example, some banks may levy a charge for every transaction; others may charge just once a day, or limit the number of charges in a day or in a specified period, through a charge cap. Some banks have buffer zones of a certain value or duration before charges are triggered.

(g) Banks also have individual practices on honouring or refusing payments which create or extend an unauthorized overdraft. Some banks calculate in advance an undisclosed unauthorized overdraft payment limit which determines the highest payment value that the bank is willing to honour. The limit is determined by the customer’s account behaviour, and this process may be automated. Other banks will assess whether or not to make payments on an ad-hoc basis; these assessments may be automated or manual.

(h) Banks have individual practices on refunding charges once they have been debited from the account. Customers cannot predict whether a charge will be refunded, nor will they know if they have reasonable grounds to request a refund. Not all customers will request a refund even if they have reasonable grounds to do so.

4.123 We note, however, that banks are required to give advance warning of certain charges before they are deducted from a customer’s account. In particular, banks are required to give customers 14 days’ notice of standard account charges that accumulate on the account and are deducted at the end of the billing period. Some banks also choose to give customers advance warning of certain unauthorized overdraft charges (see Appendix 4.13).

4.124 We sought to understand further the charges incurred by customers when operating their account in authorized or unauthorized overdraft. We constructed customer scenarios that, if replicated in practice, would lead to customers incurring either authorized or unauthorized overdraft charges, and going into debit either once a month or once a quarter. The customer scenarios included a number of transactions

79

intended to show the possible use of a PCA; these included a monthly salary deposit of £1,250, a monthly mortgage payment, and a number of ATM withdrawals, direct debits, cheque and debit card payments, and deposits. We recognize that there is no such person as a typical PCA customer (as each will have their own level and pattern of payments into and out of their PCA, for example), and there can therefore be no one typical scenario. We ensured, however, with input from the banks, that these customer scenarios, which included a variety of transactions to test the application of the charging structure, were plausible for a customer who incurred overdraft charges. For example, we chose to use an average monthly salary payment for Northern Ireland, we included one substantial payment each month to represent a mortgage payment and we included a reasonable number of ATM withdrawals and payments by standing order or direct debit.

4.125 We asked the banks to apply their respective charges and interest rates to these customer scenarios. Following numerous exchanges with the banks, and a further round of adjustments to the charges and the way in which they were applied, we reached the results set out in Appendix 4.7.

4.126 We note that many of the banks found it difficult to complete our customer scenarios exercise without additional time and guidance. It initially appeared to us that it should be straightforward to apply the banks’ published pricing schedules to a stated set of transactions. However, it emerged that there were complexities arising from flexibility in individual banks’ practices in applying their own charging structures. For example, a bank has some discretion as to whether payments (such as some cheque, direct debit, and debit card transactions) made in unauthorized overdraft should be honoured. It was also important to consider the application of buffer limits and charge caps. Other complexities arose, for example, from whether debit card use would be authorized without the retailer seeking authorization.

4.127 Several banks told us that the customer scenarios did not accurately reflect the charges to customers because the bank could exercise discretion by refunding charges, particularly unauthorized overdraft charges. This reinforced our finding that there is a significant element of bank discretion in the application of these charges. They also told us that they would contact customers in such situations to offer advice on account management (see paragraph 4.118).

4.128 Several banks told us that there was no clear evidence that customers found the presentation of information confusing or did not understand bank charges. They told us that customers did not complain about the complexity of charges. They also noted that the BMRB survey found that over three-quarters of customers who thought that they had been charged unauthorized overdraft charges in the first half of 2006 and who had compared rates and charges between different banks, said that they found it ‘easy’ to gather and understand the information on rates and charges, with 22 per cent saying ‘very easy’ and 55 per cent saying it was ‘fairly easy’.

4.129 However, we note that the majority of respondents to the BMRB survey had not made any comparison of rates and charges, and very few had looked at the full range of applicable rates and charges.

4.130 It is also possible that customers would find it difficult to determine whether they have properly and fully understood such information. The BMRB survey indicates that many customers appear to have a limited understanding of these charges. For example, many customers told us that they had not paid charges where banks told us they had; there was a poor knowledge of applicable charges and many who said that they were well informed about charges were unable to offer an estimate of charge levels; and some 28 per cent of those who said that they had been charged

80

claimed to have been unaware such charges existed (see Appendix 4.6, paragraphs 20, 23, 24 and 66).

4.131 Overall, we found that authorized overdraft charges for traditional PCAs, and unauthorized overdraft charges for all PCAs in particular, showed considerable, and undue, complexity. We believe that this complexity in charging structures and practices makes it more difficult for customers to understand and react to overdraft charges and build their assessment into making efficient choices of PCA should they seek to do so. We discuss the implications of these complexities further in paragraphs 4.248 to 4.251 and 4.266.

• Information provision

4.132 We looked to see how the banks explained their charging structures, practices, and levels of charges to customers. All banks are required by the Code to publish terms and conditions including interest rates and charges. These are usually available in pricing leaflets in the branch and on the banks’ websites, and, under the requirements of the Code, are provided to customers when they open accounts. Customers must be notified when interest rates, charges and terms and conditions change.110 In addition, there are general obligations on financial institutions from the OFT and the FSA for all financial advertisements and communications to be clear and fair. The Code requires the banks to write all terms and conditions in clear and intelligible language.111

4.133 We looked at the fees and charges brochures and leaflets provided by the banks in spring 2006 to assess both the ease with which they could be understood, and the ease with which they could be compared between banks. The detailed results of this exercise are set out in Appendix 4.3, Annex 1. We found a number of examples of a lack of clear information:

(a) Banks generally did not, in their fees and charges brochures, provide a full explanation of charges and interest rates, explaining clearly what they meant, and under what circumstances, and how frequently, they would be levied or paid:

(i) Some banks did not include information on when credit interest would be credited and charges levied on the PCA. Billing periods were often not clearly defined, or not defined at all.

(ii) Banks often failed to inform customers that a transaction taking a PCA into unauthorized overdraft could incur two charges (both a transaction charge and a maintenance charge).

(iii) Certain aspects of the banks’ charging policies and practices were not disclosed to customers. This included, for example, many buffers and charge caps, policies on honouring payments in unauthorized overdraft and so on.

(iv) Several banks failed to include information on the time taken for payments to clear.

110Section 6 of the Code requires banks to notify customers of any changes in terms and conditions. If the change is to the customer’s disadvantage, the bank is required to tell the customer about it personally at least 30 days before it makes the change. In other cases the change can be made immediately but the bank has to inform the customer within 30 days. 111Paragraph 6.2 of the Code states ‘All written terms and conditions will be fair and will set out your rights and responsibilities clearly, legibly and in plain language. We will only use legal or technical language where necessary’.

81

To the extent that these explanations are provided, they are often in a separate document which may not be readily available. Thus, if such explanations are contained in the terms and conditions, for example, this may not be available until a potential customer reaches the stage where they have expressed an interest in opening a PCA (see (d)).

(b) The terminology used for describing different charges, including authorized overdraft, unauthorized overdraft, and ancillary charges, could be described as ‘bank-speak’ and may be difficult for customers to understand. Examples include ‘referral charges’, ‘unpaid outwards charges’, ‘unpaid inwards charges collected on behalf of customer’, ‘letter charges’, ‘BACS’, ‘CHAPS’, and ‘bankers drafts’.

(c) Different banks often use different descriptions for similar charges (eg referral charges may be called ‘paid item charges’, ‘paid referral charges’, ‘charges for cheques guaranteed’, ‘card misuse fees’; see Table 1 in Appendix 4.6. CHAPS transfers are variously described by the clearers as ‘interbank telegraphic transfer (Chaps)’, ‘telegraphic transfers’, ‘CHAPS’, or ‘interbank transfer of funds, same day value Chaps out’). In some cases, where banks change their pricing structures, they also change the descriptions of charges.112

(d) Customers often need more than one leaflet to be able to understand the full details of the bank’s charging structure.113 Whilst this is consistent with the requirements of the Code, it does not facilitate customer understanding.

(e) Banks appeared to make relatively little effort to explain their charging structures more clearly, for example through the use of customer scenarios or comparative tables.

4.134 In addition, we found that three of the four clearers (BoI, First Trust and Northern) had presented their PCA information to customers in ways that were particularly difficult to understand. They did not, in general, explain what the charges were for and when they would be incurred, they mixed personal and business charges within the same charging brochure, they sometimes did not separate different types of PCAs and their associated charges, and they used language which was difficult for the general public to understand.

4.135 During the course of the investigation BoI and Northern updated their fees and charges leaflets to make them clearer. For example, BoI provided a detailed and non-technical explanation of the charges applicable on unauthorized overdrafts. In April 2007 BoI launched new customer information leaflets. These again simplified the presentation of charges and provided a clearer explanation of how and when charges could arise. For example, the circumstances in which unauthorized charges would be applied are set out in simple language, and a glossary of terms explains much of the technical language. There is also practical advice to customers on how to avoid incurring unauthorized overdraft charges. Similarly, Ulster introduced customer leaflets when it moved to fee-free banking that provided information in a clearer form together with advice on avoiding charges. We welcome these developments, although there are still many aspects of a PCA’s charging structure that need to be communicated to customers and it remains to be seen how the

112For example, []. 113In the case of First Trust, for example, the explanations for charges contained in ‘a guide to fees and interest for personal customers’ is included in a separate document entitled ‘Personal current accounts’ account guide’. Halifax includes its explanation for interest calculations and the payment clearance cycle in its terms and conditions which is a separate document to its fees and charges guide.

82

recent changes are received and whether they address the concerns expressed in this report.

4.136 The banks told us that, as required by the Code, they provided the customer with the applicable charges leaflet on account opening, together with relevant explanations by staff. Customers could contact branches or call centres for clarification. Some banks sent customers reminders of the schedule of charges if the customer breached the overdraft limit, and all would advise the customer if there were any changes to charges. Banks made charging details available on their websites and in the branch, and A&L said that it printed a description of charges on every PCA statement.

4.137 However, there was evidence from their internal pricing papers that banks are aware that customers find their charges confusing. One clearer ([]) said that: ‘[] The new charging structure will be clearer and more satisfactory from a customer perspective.’ Customer research referred to in the same document included the following comments on unauthorized overdrafts:

• ‘I got a shock when I was charged £80 for being overdrawn for a day.’

• ‘The charges are not clear at all. I don’t know what they are for.’

• ‘I have read through the breakdown of charges but they are still not clear.’

• ‘The charges are difficult to interpret.’

• ‘I don’t understand the jargon.’114

Another clearer’s board minutes ([]) showed that it was aware that its existing charging structure was complex. []115

4.138 Our surveys also suggested that customers experienced difficulties with the way in which information is provided. Almost half (49 per cent) of all respondents to our ORC survey said that they were fairly or completely uncertain about the amount that they paid for setting up a standing order, making a deposit, or writing a cheque. Some of those who said that they were certain quoted charges which were incorrect. Our Millward Brown survey found that customers had a very low awareness of alternative offers, and over 50 per cent of respondents to the ORC survey considered themselves uninformed about charges on other PCAs. The BMRB survey also found that many customers were unable to estimate levels of unauthorized overdraft charges and authorized overdraft charges, and many gave incorrect estimates (see paragraph 4.232).

4.139 We found that, in general, and across the banks as a whole, customers are not provided with the necessary information to enable them to have a sufficient understanding of the charges and interest rates that might apply to their PCA. This was particularly true of unauthorized overdraft charges. We also thought that ancillary charges could be explained in terms that were more customer-friendly, although we note that these are not applicable to the day-to-day running of an account and, in many cases, are explained in more detail when a particular service is requested. We return to a discussion of the implications of a lack of clarity of explanation of charges in paragraphs 4.266 to 4.269.

114[] 115[]

83

Recent changes to charging structures

4.140 There have been a number of changes in PCA offers in Northern Ireland over the past two years (see paragraph 2.63). We discuss each of the clearers in turn, taking their changes, or proposed changes, to their PCAs, in chronological order. Further details are set out in Appendix 4.8. Figure 7 sets out an overview of timing of changes to charging structures.

FIGURE 7

Timeline of changes to clearers’ charging structures

Northern removes SO & DD Northern launches Feb 05 set up, amendment fees on Current Account Plus all remaining PCAs

May 05 Reference to CC Northern removes charges for setting up/renewing Jul 05 overdrafts Northern removes 'card mis-use' charge Oct 05 Nov 05 Ulster launches new pricing structure

Northern launches Apr 06 new suite of PCAs

Jul 06 Northern reduces 'unpaid outward' charge BoI announces plans to launch new pricing Sep 06 structure in November BoI launches Nov 06 new pricing structure

First Trust announces CC statutory reporting May 07 new PCA to be launched deadline

Source: CC analysis of market developments.

• Northern

4.141 Northern was the first of the clearers to introduce a fee-free PCA alongside its older PCAs. In February 2005 it introduced Current Account Plus, which did not levy transaction or maintenance charges on authorized overdrafts. Existing customers were offered the opportunity to switch to the new account, but were not transferred automatically. The decision to introduce Current Account Plus was taken prior to the supercomplaint, and was, we were told, a reaction to Northern’s decline in market share.

84

4.142 Northern also removed set-up, amendment and cancellation charges for standing orders and direct debits for all customers in February 2005.116 It told us that this was in response to customer dissatisfaction with the charges and loss of customers to the non-clearers. It said that []. In July 2005 it removed overdraft arrangement and renewal charges on all its PCAs.

4.143 Current Account Plus was introduced by Northern’s former owners, NAB, as part of a UK-wide policy, which was also applicable to Yorkshire and Clydesdale Banks. Northern said that its introduction in Northern Ireland was a response to competitive pressures from the non-clearers and that its introduction at Yorkshire and Clydesdale banks was a response to competitive pressures in Great Britain. When launched, Current Account Plus levied lower authorized overdraft interest rates and higher unauthorized overdraft interest rates than Northern’s older accounts. Credit interest rates were in line with most other Northern PCAs at 0.1 per cent and unauthorized overdraft charges were in line with those on Northern’s other accounts. Current Account Plus attracted [] new customers over the first year,117 [] of whom moved from an existing Northern product.

4.144 Appendix 4.8 sets out the financial appraisal undertaken by NAB on behalf of Northern to look at the impact of the removal of standing order and direct debit charges and the introduction of Current Account Plus. Table 1 suggests that Northern expected these changes to result in a significant net profit to the bank [], although this included ambitious assumptions on acquisition and retention rates []. Within this, the removal of standing order and direct debit set-up, amendment and cancellation charges was forecast to cost Northern [].

4.145 In April 2006 Northern removed Current Account Plus from sale and launched its current range of PCAs. Northern told us that its new account portfolio was driven by Danske, with pricing reflecting Northern Ireland conditions. The charging structure is summarized in Appendix 4.8. The five accounts are Northern Access; Northern Freedom; Northern Choice; Northern Choice Plus; and Northern Prestige. We do not consider the Northern Access account further in this report since it falls outside our market definition.118

4.146 We note the following key differences in Northern’s new suite of products compared with its older PCAs:

(a) The new accounts pay more credit interest (2 to 4 per cent) than its older PCAs (0 to 0.1 per cent). Northern told us that [].

(b) On Northern Freedom and Northern Choice there are no transaction charges within certain limits (13 branch cash withdrawals and 13 cheque transactions per quarter).

(c) There is a quarterly charge on Northern Choice Plus and Northern Prestige (£30 for Northern Choice Plus and £50 for Northern Prestige).

(d) All have a lower authorized overdraft interest rate than Current Account Plus and there is no change in unauthorized overdraft interest rate.

116These charges had already been removed for students, graduates and retirees in 2002/03. 117Representing [] per cent of all Northern’s accounts. 118[]

85

(e) There is no change in unauthorized overdraft charges compared with Current Account Plus.

(f) Northern’s Freedom account is on offer to all 18- to 27-year-olds (rather than limited to students) and offers a zero per cent overdraft rate.

4.147 Northern’s target is to attract up to [] customers to the new PCA products in the first year, whether from other Northern accounts or new customers. Existing customers were notified of the new products, but Northern told us that it could not switch customers automatically as some aspects of the new PCAs (eg limits on free cheques) may not be considered cheaper for the customer. From June to December 2006, [] existing Northern customers switched to its new PCAs, representing, on average, around [] per cent of its existing customer base per month.

4.148 In contrast to Current Account Plus, Northern told us that []. The new accounts were very similar to successful Danske accounts available in Scandinavia. Northern had looked at both their own charges and those of competitors, as well as ‘Best Buy’ tables published by Which?, in determining the new charging structure. However, as set out in Appendix 4.8, Northern expects a [] reduction in income arising from the overall changes to its PCAs in 2005 and 2006. []

• Ulster

4.149 In November 2005, Ulster removed all transaction and maintenance charges on authorized overdrafts, removed direct debit and standing order charges, and made certain changes to unauthorized overdraft charges.119 Since these were changes to existing terms and conditions, all of Ulster’s PCA customers, with the exception of those on packaged accounts, moved automatically to the new charging structure. We were told that the introduction of Ulster’s new pricing structure, first conceived in July 2004, was prompted by the need to [].

4.150 Appendix 4.8 sets out the financial analysis undertaken by Ulster. Its key assumption was an ‘out-of-order’ percentage (ie customers that would operate their accounts outside their authorized overdraft limit) of [] per cent a month, [], with sensitivity analyses based on out-of-order percentages of [] and [] per cent a month. Using an out-of-order assumption of [] per cent, Ulster forecast that the new charging structure would be broadly revenue neutral, given the growth in PCA volume that the new charging structure was expected to generate. However, it said that the new charging structure presented significant risks, uncertainty and complexity for Ulster.

4.151 The new charging structure has been revenue enhancing to date because the out-of- order percentage has been higher than forecast ([] per cent prior to launch; [] per cent in September 2006). In February 2007, Ulster told us that, since November 2005, current account openings with Ulster had increased [].120 Switching-in (via the switching service) has [].121 Most switchers-in have come from other clearers, whereas most switchers-out have gone to the non-clearers.

4.152 Although the percentage of customers incurring unauthorized borrowing charges is greater than originally envisaged, Ulster told us that it expected the level of income

119Customers in unauthorized overdraft incur monthly charges of £30 for breaching the limit (previously no charge), a paid referral charge of £30 per item up to £90 a month, compared with £13 per excess previously, a card misuse fee of £30 and unpaid outwards charge of £30 per item (previously £38). The £6 unpaid inwards fee was removed. 120[] per cent uplift week-on-week in the first 41 weeks of 2006 compared with 2005 (Note: excludes student PCAs). 121Run rate has increased from [] per week in Q4 2005 to a run rate of [] per week in the first nine months of 2006.

86

from unauthorized overdraft charges to reduce significantly over the coming months as customers become increasingly familiar with the new charging structure. Ulster told us that it was taking a number of steps to ensure that its customers and potential customers understood the new charging structure and more customers avoided incurring charges, such as providing advice on account management and how to avoid charges. It also told us that unauthorized borrowing was generally higher around the Christmas period than at other times of the year. []

4.153 In March 2007, Ulster told us that it was repositioning its packaged accounts, U First and U First Gold, by increasing buffer limits, offering interest free overdrafts and preferential overdraft rates, and offering mobile phone insurance. []

• BoI

4.154 BoI said that [].

4.155 [] BoI introduced a fee-free PCA for both new and existing customers on 24 November 2006, whereby BoI:

(a) removed all transaction, maintenance and overdraft arrangement and renewal charges;

(b) removed set-up, amendment and cancellation charges for direct debit and standing orders for all customers;

(c) increased the referral item charge from £14.50 to £19 for all customers;

(d) reduced the maximum number of daily referral item charges from five to three; and

(e) introduced a new unauthorized overdraft usage charge for personal customers of £19 a calendar month from April 2007. The unauthorized usage charge will not apply where the unauthorized balance is less than £20.

4.156 BoI’s financial analysis is set out in Appendix 4.8. It estimates that if customer behaviour does not change, the new model will earn total income from charges which is broadly revenue neutral. Income foregone is forecast to be broadly offset by new charges or increased charges relating to unauthorized overdrafts. []

• First Trust

4.157 First Trust said that it had been losing [] market share []. It said that [].

4.158 Very shortly before finalizing our report, on 3 May 2007, First Trust announced that it intended to launch a new PCA in June 2007. It told us that this was designed to reduce the loss of PCA customers, attract switchers-in, increase market share, []. The new PCA is based on a fee-free model, subject to customers depositing a minimum amount (£1,200) in their account each month. Customers will be required to use at least two automated transactions per quarter, or keep their account in credit, to avoid transaction charges. Customers would earn 4 per cent credit interest on balances up to £1,200. Existing customers would need to apply for the new PCA. First Trust will also offer free standing order and direct debit set up, amendment and cancellation on all PCAs.

4.159 []

87

4.160 []

• Non-clearers

4.161 There have been no changes to the non-clearers’ PCAs in the last two years of the same magnitude as the changes discussed in paragraphs 4.141 to 4.160 in relation to the clearers. The non-clearers’ each offer the same PCA throughout the UK, and all are based on the fee-free model. Since around 2000 there has been widespread promotion of credit interest rates. Some have also sought to promote low authorized overdraft interest rates. Charges for unauthorized overdrafts have increased since 2000. Halifax launched a packaged account, Ultimate reward, in February 2007.

Trends in interest rates and charges

4.162 We looked at trends in PCA interest rates and charges. We compared credit interest rates, certain transaction and maintenance charges, certain ancillary charges, and the charges for setting up standing orders and direct debits for the PCAs offered by the clearers and three of the largest non-clearers between June 2001 and January 2007. We also compared certain unauthorized overdraft charges since 2001 for the clearers and since 2003 for the non-clearers. Further details are set out in Appendix 4.4. Our analysis is based on information provided by the banks in their responses to our financial questionnaires and on the clearers’ internal pricing papers. Charges and interest rates for comparative purposes were selected because they were levied by several banks in the same way and we could obtain a reasonably long time series of data. Our analysis does not extend beyond January 2007, and therefore does not include First Trust’s new PCA which it announced is to be launched in June 2007.

4.163 We note that any comparison of this nature is difficult. A PCA’s charging structure is made up of a variety of terms and conditions, and there is some risk in taking an individual charge or interest rate out of context. In addition, individual charges can be applied in different ways. Nevertheless, we found it useful to compare a selection of interest rates and charges across the banks.

• Credit interest rates

4.164 Table 1 in Appendix 4.4 lists the credit interest rates offered by the clearers and non- clearers in January 2007. We found that most PCAs offered by the clearers paid very low rates of interest (0.1 per cent) or zero interest. Some of the non-clearers offered better credit interest rates, although in these cases there may be conditions attached, such as minimum levels of monthly funding, ceilings on the value of funds receiving the higher interest rate, use of remote banking channels, or time limits.122 We note that, based on the data we collected on revenue, the average rate of credit interest paid to customers by the non-clearers in 2005 was 1.02 per cent (see Table 3).

4.165 We also looked at the credit interest rates on packaged accounts. Ulster and First Trust were the only two clearers to offer packaged accounts until Northern launched its new accounts in April 2006. The introduction of First Trust’s account in December 2003 coincided with Ulster raising the interest rate on U-First. First Trust then appeared to respond with an improved credit interest rate with Ulster subsequently

122As at 28 February 2007, Abbey, for example, offered 6.3 per cent interest AER (annual equivalent rate) for the first 12 months on the first £1,000 if the account is credited with at least £1,000 a month (the account pays 2.5 per AER monthly which is then increased up to 6.3 per cent AER after 12 months). The A&L Premier Direct account offers 6.1 per cent interest AER until the end of April 2008, after which the interest rate reverts to the BoE base rate minus 1 per cent.

88

improving its offer, again resulting in a response from First Trust. For a period of 11 months from September 2004 to July 2005 all the packaged accounts were offering the same credit interest rate of 1 per cent, although U-First’s rate was then halved. Northern’s new portfolio was introduced in April 2006, offering credit interest rates of 2.5 per cent on Northern Choice Plus and 3 per cent on Northern Prestige. In early August 2006, First Trust increased its credit interest rate on amounts below £2,000 to 1.25 per cent, and Ulster increased the rate on U First Gold to 1.5 per cent in mid-August. Northern increased the credit interest rate on its packaged PCAs by 1 per cent in November 2006. First Trust increased the rate on its packaged account by 0.25 per cent in December to match U First Gold on 1.5 per cent, but Ulster increased the rate on its U First Gold account by 0.5 per cent to 2 per cent in January 2007. It also increased the rate on U First by 0.5 per cent at the same time.

• Charges on traditional PCAs

4.166 We considered charges on traditional PCAs, which accounted for the great majority of PCAs in Northern Ireland until the clearers began to move away from this charging structure (see Figure 7). We looked at three different types of transaction and maintenance charges: automated transaction charges (eg on debit card transactions and ATM withdrawals), manual transaction charges (eg for cheques), and maintenance charges. We also looked at direct debit set-up charges, standing order set-up charges, and card replacement charges. In all cases, the non-clearers did not charge for these transactions. The details of our comparisons are set out in Figures 1 to 7 in Appendix 4.4.

4.167 We found a significant degree of similarity in the levels of clearers’ transaction, maintenance and other charges, and the timing of any changes to them. When one clearer increased prices, the others tended to follow within a short period. Northern appears in most cases to have initiated price increases on automated transactions and direct debit and standing order set-up charges.123 First Trust was generally the first to increase prices of manual transactions; prices were not quite as closely aligned as they were on automated transactions but each of the clearers raised prices to a new band shortly after First Trust increased its charges. Such behaviour did not always apply: for example, on card replacement charges, during the period we reviewed, the clearers at times charged identical charges but at other times had very different prices. There was no evidence of parallel pricing on transaction or other charges on packaged accounts, where these charges have been set at different levels or removed altogether.

4.168 Until the adoption of fee-free accounts by Northern, Ulster and BoI, prices of auto- mated and manual transaction charges and maintenance charges tended to con- verge to similar levels.124 We did not observe clearers reducing transaction charges on their traditional PCAs until fee-free PCAs were introduced. These charges also converged to a large extent, although in some cases they have been removed altogether. Ulster dropped card replacement charges in July 2004, as did Northern in January 2007. Northern removed standing order and direct debit set-up charges in February 2005. As part of its revised charging structure, Ulster removed standing

123Northern said that it did not lead any price increases and pointed to the fact that no consistent pattern of leadership was observed on other charges and rates. However, we found that other banks looked to Northern’s pricing decisions in forming their own decisions on some charges (see paragraph 4.301). Northern removed all remaining standing order and direct debit set-up charges in February 2005. 124These charges are shown from 2001 to the start of 2007 in Figures 1 to 4 of Appendix 4.4. At times prices have diverged by 5 to 10 per cent, but these divergences have not been sustained. Price converged on very similar levels for all the clearers for each of these charges by October 2004.

89

order and direct debit charges in November 2005; BoI similarly removed standing order and direct debit charges when it launched its revised charging structure in November 2006.

• Charges and interest rates for unauthorized overdraft charges

4.169 We looked at certain unauthorized overdraft charges since 2001 for the clearers, and since 2003 for the non-clearers. We considered trends in referral charges, unpaid charges, and unauthorized overdraft interest rates. Some unauthorized overdraft interest rates were negotiable and we did not include these in our comparison (unauthorized overdraft interest rates on First Trust, some Northern and some Ulster PCAs).

4.170 The trends in unauthorized overdraft charges and interest rates are set out in Figures 9 to 11 in Appendix 4.4. In contrast to transaction and maintenance charges on traditional PCAs, there is considerable divergence in the level of specific charges and unauthorized overdraft rates between banks, and there is no obvious pattern of price following. We note that, before November 2005, there were no examples of banks cutting charges. We also observe a considerable spread of unauthorized overdraft interest rates.

Discussion of levels of interest rates and charges

4.171 With the exception of Northern’s new PCA portfolio which offers 2 per cent credit interest on its Northern Choice account, we found, in general, no evidence to suggest that the clearers were competing on credit interest rates during the period covered by our analysis. The pattern and levels of credit interest rates on packaged accounts, however, suggest a greater degree of competitive interaction between the clearers on these products (see Figure 8 of Appendix 4.4). This might be due to the greater visibility of the monthly fixed charge which might draw customers’ attention to readily comparable figures such as credit interest rates. Some of the non-clearers were also promoting their credit interest rates, although we note that, in many cases, there were conditions attached to these. Table 3 shows that, on average, the four largest non-clearers paid 1.02 per cent credit interest in 2005, compared with an average of 0.10 per cent paid by the clearers.

4.172 We considered the reasons for the observed similarity in some of the clearers’ charges, particularly on traditional PCAs. We note that strong competition on charges would be likely to cause prices to converge over time, provided the products were homogeneous. However, we do not think that the banks compete for customers primarily on the basis of their charges. We have not seen these charges heavily advertised. Nor have we found that customers are particularly aware of, or sensitive to, differences in charges between banks. In addition, the clearers told us that the main competitive threat—which has forced them in recent months to review their PCA products and associated pricing—has come from the non-clearers. However, the evidence that we received from the clearers’ internal pricing papers over a longer time period suggests that, at least until recently, they have monitored and set charges primarily to be in line with the other clearers.

4.173 We note that in most cases the banks do not know their costs of providing particular PCA services, and they have stated that many such costs are not allocated to particular financial products (see paragraph 2.66). We therefore thought it unlikely that similarity of costs could account for the similarity of pricing behaviour that we observed.

90

4.174 We looked at the way in which banks set the level of unauthorized overdraft charges and interest rates to assess whether they appeared subject to competitive constraint. Some of the banks, in modelling the impact of increased unauthorized overdraft charges, anticipated some limited response from customers, or considered the sensi- tivity of their assumptions to changes in customer behaviour.125 We also note that the pricing decisions often refer to the charges set by competitors.126 However, this does not seem to have constrained the level of charges set. Some banks have substan- tially increased charges despite being aware of competitor levels. In addition, the shift towards a fee-free model by the clearers suggests that they believe it to be more profitable to rely on revenue from unauthorized overdraft charges and debit interest than to retain their traditional PCA, suggesting to us that customers are generally less sensitive to these charges (see paragraph 4.243).

4.175 Banks could revise their charging structures or the way that they are applied, to cut the overall cost of charges to customers. This has happened to some extent when banks have introduced buffers and charge caps, and we found through the scenarios exercise that these can have a substantial effect on total charges payable (see paragraphs 4.124 to 4.126 and Appendix 4.7). [] told us that complaints about charges had led it to introduce an unpublicized buffer zone for referral charges. This approach appears to be designed to manage discontent with charges among existing customers.

4.176 We think it unlikely that costs could account for the levels of unauthorized overdraft charges. Once again, in most cases the banks told us that they did not know their costs of providing particular services, and they have stated that many such costs are not allocated to particular products or individual PCA services. In addition, it is unlikely that costs would vary substantially between banks, or that the doubling, for example, of a charge could be explained by a change in cost.

4.177 We considered various arguments which suggested that banks did not compete on unauthorized overdraft charges and debit interest rates. It was put to us that banks could not be expected to compete actively and openly on unauthorized overdraft charges and debit interest, since customers could not be encouraged to breach the terms of their contract. Second, promoting better unauthorized overdraft charges and debit interest may be against the interests of the customer, if it influences them to increase their debt, possibly to unsupportable levels, and increases the number of times they are charged for unauthorized overdrafts. Third, encouraging use of unauthorized overdrafts might increase credit risk to the banks, as unauthorized overdrafts carry a higher risk of default because customers’ ability to repay the overdraft has not been assessed and agreed.127 We accept that there are reasons why banks would not be expected actively to promote their unauthorized overdraft policies and charges. Nonetheless, we believe that greater transparency and understanding among customers would increase pressures on the banks to constrain these charges.

125We note that one of [] pricing papers included an assumed response from customers leaving ([]) because of an increase in charges, but judged the increase to be profitable, while Ulster’s assessment of its new charging structure included some sensitivity analysis depending on the rate of out-of-order accounts being incurred following implementation of new charges. 126Pricing papers often refer to the headline charges imposed by other banks, and so will compare levels of referral charges, levels of unpaid charges etc. None of them look at the total cost of all charges together, and none appear to consider or measure the incidence of charges making allowance for application of the different policies on charging and exemptions, such as buffer zones. Banks are unlikely to be aware of their competitors’ precise policies in this respect. 127However, [] said that there was a core difference between a bad credit risk and someone going into unauthorized overdraft. Any increased credit risk could be controlled in other ways; apart from guaranteed cheques, and some debit card payments, banks have discretion on whether or not they will honour a requested payment, and so in this way they can control the extent to which they increase their credit risk. The banks currently with lower charges have not told us that they find credit risk increases as a consequence.

91

4.178 Several of the banks told us that they did not seek to use unauthorized overdraft charges as a source of profits, because this might alienate customers and reduce their potential to build a long-term profitable relationship with a customer, including cross-sales. One clearer ([]) said that: ‘If we simply end up charging those 10 per cent of customers who do not enjoy fee-free banking in any quarter with excessive delinquency charges, on a reputational basis, that takes you nowhere ultimately.’ Several of the banks said that they would contact customers and seek to educate them to change their behaviour so as to reduce the likelihood of them going into unauthorized overdraft again;128 this may take the form of explaining charging structures, giving advice on budgeting, and suggesting more suitable financial products such as loans. [] said that it had succeeded in reducing the incidence of charges following the introduction of its new charging structure, suggesting that there were behavioural aspects leading some people into unauthorized overdraft which could be corrected through education. []

4.179 The arguments set out in paragraphs 4.176 to 4.178 suggest that there could be a lack of competition on unauthorized overdraft charges and interest rates for understandable and rational reasons. Nevertheless, given the lack of competitive constraint on these charges and interest rates, as well as charges on traditional PCAs, we believe that some charges and debit interest rates are likely to be higher, and some credit interest rates lower, than would apply in a well-functioning market. We discuss this further in paragraphs 4.262 to 4.287.

Conclusions on charging structures and levels of charges and interest rates

4.180 We found that there is an inherent complexity in PCA charging structures that arises in part because PCAs service a wide variety of needs. However, we also found that PCA charging structures and practices are unduly complex, particularly for author- ized overdraft charges associated with traditional PCAs, and unauthorized overdraft charges levied on all PCAs. In addition, we note that the dual role of the bank in providing a PCA to the customer and having access to the customer’s funds may be a factor in the banks’ charging practices (see paragraphs 4.120 to 4.131), notably in the way certain charges are deducted from a PCA without pre-notification.

4.181 We also found that, in general and for the banks as a whole, customers are not provided with the necessary information to enable them to have a sufficient understanding of the charges and interest rates that might apply to their PCA. This was particularly true for information on unauthorized overdraft charges. We also thought that ancillary charges could be explained in terms that were more customer- friendly (see paragraphs 4.132 to 4.139).

4.182 The new charging structures that have been introduced over the past year show that the clearers are, in general, moving towards a fee-free business model.129 However, we note that the undue complexity associated, in particular, with unauthorized overdraft charges, remains. Furthermore, almost 20 per cent of customers in Northern Ireland remain on traditional PCAs (see Table 2).

128[], for example, both told us that they contacted and worked with customers who had incurred unauthorized overdraft charges to help avoid the situation arising again. 129We note that Northern’s Choice and Freedom PCAs are generally fee-free, with the exception of heavy users of cheques or cash withdrawals from branches; its other accounts, Choice Plus and Prestige are packaged accounts carrying a quarterly subscription fee. First Trust’s new PCA to be launched in June 2007 is fee-free, subject to customers making a minimum of two automated transactions or by keeping their account in credit during the quarter.

92

4.183 The financial analysis that we carried out on the clearers’ new PCAs suggests that most of the new charging structures are forecast to be broadly revenue neutral for each bank, albeit the banks’ estimates are subject to considerable uncertainty. The forecasts represent only limited reductions in charges on average to the customer. We recognize, however, that, depending on individual customer usage, some customers will benefit from the new charging structures whilst others will incur higher charges (see paragraphs 4.140 to 4.161).

4.184 The trends in PCA charges and interest rates that we looked at suggested similarities in pricing among the clearers’ traditional PCAs that we do not think can be explained by similarity in costs or by high levels of competition in the market. We did not find such similarities in the pricing of packaged accounts. We note that, of the three clearers that have introduced new charging structures to January 2007, Northern is the only bank that has chosen to increase substantially the levels of credit interest offered.

4.185 With regard to unauthorized overdraft charges, we found that some banks are able to maintain unauthorized overdraft charges or interest rates significantly higher than their competitors for extended periods. We recognize that the banks have found that these charges are a major source of customer discontent, and that some have taken measures to encourage customers to avoid these charges (see paragraph 4.178), or have adopted more lenient policies on implementing such charges (eg by using buffer zones). Nonetheless, there is an inherent tension because unauthorized over- draft charges are a significant source of revenue to banks (see paragraph 4.117). The customer scenarios suggested that when all the charges were aggregated, total charges varied significantly by bank. In the unauthorized overdraft customer scenarios, total charges were generally higher for fee-free PCAs than for traditional PCAs. We found, therefore, that there is a lack of competitive constraint on banks’ unauthorized overdraft charges and debit interest rates. This was also supported by the evidence that we had on the way in which banks set unauthorized overdraft charges.

4.186 We discuss further the implications of the banks’ charging structures and practices for competition in the market, and the resulting levels of charges and interest rates, in paragraphs 4.262 to 4.287.

Competition for customer groups

4.187 Paragraphs 3.27 to 3.29 discussed market segmentation with respect to market definition and concluded that, primarily due to supply-side substitution, different market segments did not represent separate markets. In this section we consider if competition between the banks varies for different customer groups, and if so in what way, in order to inform our thinking about competition within the market. Details of our analysis are set out in Appendix 4.5.

4.188 The banks told us that they focused primarily on the mass market, and did not consider the market to be segmented. We found that in the past, the clearers targeted the new-to-banking segment of the market and, in particular, students. They offer their most advantageous terms to students (or, in the case of Northern, the youth segment), and have usually, prior to the recent launch of fee-free accounts, spent the majority of their PCA-specific marketing spend on student and graduate campaigns.

4.189 The non-clearers have had a greater focus on switchers and the mass market. In particular, many offer inducements to switchers, for example offering cash rewards, introductory rates of credit interest, or specific offers to ease the switching process

93

(see Appendix 4.5 for more details). This suggests that they do not believe switching barriers to be prohibitive. We note, however, that since the larger non-clearers’ offers are UK-wide, their strategy is not necessarily reflective of conditions in Northern Ireland. They also have fewer existing customers compared with the potential new customers that they could win and hence have less of an incentive to differentiate between them.

4.190 The clearers have recently increased their efforts to attract switchers. Most notably, both Ulster and Northern as part of the launch of their new charging structures or product portfolios, have been targeting switchers. BoI launched a marketing campaign to attract switchers on 1 March 2007. Ulster announced on 2 May 2007 that it would offer a £150 cash incentive to customers who switched their PCA from another bank to Ulster during the month of May. Prior to these initiatives, the outcome of which is uncertain at this stage, the clearers continued to win a higher proportion of new-to-banking customers than they did of new customers in total (see paragraph 4.30).

4.191 The banks said that students and graduates were offered better terms because they were likely to become high-value customers in the future. However, this does not explain why students have received terms that are more advantageous than those offered to existing high-value customers, nor why they are such a focus of marketing spend. There are some indications that the clearers have recently placed less emphasis on recruiting students in particular than in the past. For example, Northern’s new product structure does not include a specific student account, and some of the clearers have recently targeted their advertising at switchers.

4.192 We consider that the clearers’ greater focus on new-to-banking customers prior to the launch of the clearers’ fee-free PCAs, was based on the existence of barriers to switching, since a strategy of offering preferential terms to new-to-banking customers is only rational if such customers, once won, face barriers to switching (see paragraphs 4.217 to 4.255). We observe that the clearers who have introduced new PCAs since 2005 are, to some extent, displaying a different focus of interest and are actively seeking switchers from other banks. It remains to be seen whether the clearers’ behaviour has changed permanently or whether this is a temporary change following the substantial revisions to their PCAs. We believe that substantial barriers to switching remain in the market (see paragraphs 4.256 to 4.260).

Non-price competition

4.193 We looked at various possible indicators of non-price competition. These were chosen because they were likely to be affected by the level of competition in the market. They were also readily comparable between banks. We looked at how these indicators compared between the clearers, the non-clearers, and some of the large banks based in Great Britain (Barclays, Lloyds TSB, HSBC and RBS/NatWest):

(a) processing of cheques (in branch or centralized);

(b) product innovation over time;

(c) introduction of Internet banking;

(d) introduction of telephone banking;

(e) using ATM cards outside the UK and Republic of Ireland; and

(f) opening hours.

94

4.194 We also considered investment in customer service reviews, levels of marketing spend, and the provision of banknotes as further possible indicators of non-price competition.

4.195 We summarize the results of our analysis in Appendix 4.9. We saw no clear differences between the clearers, the non-clearers, and the large banks based in Great Britain on processing of cheques or the introduction of telephone banking (see paragraph 4.193(a) and (d)). We discuss each of the other aspects in turn.

4.196 Table 1 in Appendix 4.9 lists the number of new PCA products introduced by the banks in the last five years. We found no clear difference between the clearers as a group and the other banks. In terms of product innovation, we also looked at how quickly particular innovations spread between rivals. For the purposes of this exercise, we looked in particular at the introduction of packaged accounts.

4.197 Packaged accounts are offered by three of the clearers (Northern, Ulster and First Trust) but not by most of the non-clearers.130 The large banks based in Great Britain also tend to offer packaged accounts: Barclays, HSBC, Lloyds TSB and RBS all offer packaged accounts. However, the large banks based in Great Britain appear to have introduced packaged accounts slightly earlier than the clearers, with the majority introducing these accounts more than five years ago whereas Ulster and First Trust introduced theirs in 2001 and 2003 respectively, and Northern introduced its packaged account in 2006.

4.198 Table 2 in Appendix 4.9 outlines the responses we received from the banks about the introduction of Internet banking. We found that the clearers as a group do not offer the same degree of functionality as the large banks based in Great Britain and the non-clearers. All the banks based in Great Britain and all the non-clearers that we surveyed offered the ability for customers to make third-party payments, set up and amend standing orders, and cancel direct debits through their Internet service. By contrast BoI, First Trust and Ulster do not offer this functionality. [] Northern, however, already offers full-function Internet banking.131

4.199 Almost all the banks charge their customers for using ATMs outside the UK and the Republic of Ireland.132 There appears to be no clear difference in the level of charges applied by the clearers in comparison with the other banks surveyed.133 BoI introduced chip-and-pin cards to customers in November 2006 which can be used in ATMs both at home and abroad. Since late 2006, and following its migration to the RBS Group technology platform, Ulster has been issuing customers with ATM cards that can be used outside the UK and the Republic of Ireland, whereas previously they had to apply for a free Travel Cash Card.

4.200 Finally, we looked at branch opening hours (see Table 4 of Appendix 4.9). There are two key differences in branch opening hours between the clearers and the large banks based in Great Britain. First, the large banks based in Great Britain and the non-clearers appear to open for more hours a week, and second, of the clearers,

130Halifax introduced a packaged account in February 2007, it had previously offered a different packaged account which was withdrawn from sale in 2004; HSBC, Co-operative Bank, and smile also offer packaged accounts. 131Northern introduced the ability to set up and amend standing orders and to cancel some direct debits in 2003. Further functionality in relation to cancellation of direct debits was introduced in 2006. 132Nationwide is the only exception. It levies no up-front charge for ATM withdrawals abroad. 133A direct comparison of charges is in any case problematic as the final cost to the customers will also be affected by the exchange rate applied.

95

only Northern opens a significant number of branches on a Saturday134 whereas the majority of banks based in Great Britain135 and all of the non-clearers have branches open on a Saturday.

4.201 In addition, the clearers pointed to investment in customer service as evidence of non-price competition. Northern launched a customer experience programme in 2002, consisting of enhanced customer service training for staff and redesigning Northern’s branches. Ulster reviewed its service delivery as part of its development programme from 2000 to 2002, and made significant changes including introducing more customer advisers, investing in staff training, and introducing customer service reviews. First Trust has also introduced customer service reviews. It is difficult to assess the impact of these initiatives on both customers and banks.

4.202 Based on the evidence set out in paragraphs 4.193 to 4.199, product innovations in the PCA market have largely consisted of changes in pricing structures (such as packaged and fee-free PCAs). The fundamental functionality of the PCA has remained unchanged. This is likely to be a consequence of the nature of the product. Innovation in Internet banking in Northern Ireland has lagged behind Great Britain. We therefore infer that innovation has not been a significant driver of competition in the Northern Ireland PCA market.

4.203 We also looked at whether PCAs were heavily advertised in Northern Ireland. In general, we were told that this was not the case, and the ORC survey showed that there was low customer awareness of PCA advertising. The non-clearers, on the other hand, benefited from UK-wide advertising. However, there have been recent changes in advertising strategy by several of the clearers. For example, after initial promotion of its fee-free PCAs, Ulster has continued actively to market its PCAs. It has carried out two campaigns between September 2006 to February 2007, costing £[], stressing the fee-free offer and the ease of switching. It told us that the campaigns were targeted at all customers rather than being targeted specifically at students, new-to-banking customers or switchers. Northern similarly has continued to advertise its new packages, spending £[] on campaigns targeted particularly at switchers. It stressed its advantageous credit interest rates compared to those of other banks, publicized its email and text balance alert service, and publicized that its Freedom account was open to all young customers (aged 18 to 27) rather than just students. BoI said that initially it marketed its fee-free charging structure primarily to its existing customers. However, BoI said it had launched a campaign from March 2007 targeted at attracting switchers to its PCAs.

4.204 It may be too early to see the impact of these changes in pricing and strategy. However, we note that the increased spends by Ulster and Northern are consistent with the pattern of expenditure in the past—that is to say, increased levels of spend accompanying product launch—and we have received no evidence to suggest that this will lead to a higher level of marketing expenditure on PCAs in the medium to long term.

4.205 In paragraphs 2.119 and 2.120 we discussed the financial advantage to the clearers of issuing their own banknotes. Neither the clearers nor the non-clearers suggested that the current arrangements on banknotes in Northern Ireland created a competi-

134Northern opens 23 per cent of its branches on a Saturday. However, it offers significantly shorter weekday opening hours (typically 10.00 to 15.30) so that overall its weekly opening hours are low compared with the non-clearers and the large banks based in Great Britain. 135Only 19 per cent of RBS branches are open on a Saturday.

96

tive advantage for the clearers, nor did we receive any other evidence supporting such a conclusion.

4.206 Overall, we have identified several indicators of a relative lack of competition in Northern Ireland, particularly among the clearers in relation, for example, to branch opening hours and functionality of Internet banking. Product innovation in Northern Ireland appears to have been relatively limited in the past, and the continued existence of some traditional PCAs may reinforce our finding of a lack of competition in the market.

Customer conduct

Overview

4.207 We next look at the behaviour of existing and future PCA customers, taking into account various characteristics already discussed in the report: customer character- istics and levels of switching (see paragraphs 2.72 to 2.92), market structure (see paragraphs 4.3 to 4.107) and bank conduct (see paragraphs 4.108 to 4.206). We consider customers’ reasons for searching and switching and possible barriers to searching and switching, including a discussion of customers’ views on unauthorized overdraft charges.

4.208 In section 2 we discussed the levels of switching that we observed in the Northern Ireland PCA market. We considered the banks’ and survey data on levels of switching, as well as switching rates for other products and relative rates of switching in Northern Ireland and Great Britain. Whilst evidence was mixed on the absolute levels of switching, given the customer attitudes and barriers to searching and switching discussed in paragraphs 4.209 to 4.260, there seemed to us to be a lack of responsiveness to changes in charges or interest rates in the market.

Reasons for searching and switching

4.209 We looked at the reasons underlying switching and the choice of new provider. The decision to switch may be taken either because a better offer is seen elsewhere (a ‘pull’ factor) or because of dissatisfaction with the existing provider (such as poor service or because the customer feels charges are unfair or rates are uncompetitive; a ‘push’ factor). In the latter case a separate and subsequent decision is made as to the preferred alternative bank.

4.210 Both the ORC and MORI surveys asked respondents for their reasons for switching.136 The detailed results are set out in Appendix 4.3. We found that switch- ing is primarily influenced by push factors. When the MORI survey asked customers the main reason why they had closed their account, 15 per cent cited high charges on their existing account, 10 per cent mentioned a poor level of service, and others mentioned errors with the account and unhelpful staff (all push factors). The percent- age citing high charges had doubled since 1999. We also found that there was an

136We also looked at evidence in the UK-wide survey Mintel Report Current Accounts, Finance Intelligence (June 2005), and the Datamonitor report UK current accounts-Service not included? (June 2004), which supported our finding that dissatisfaction with the existing provider was the main reason for switching, although pull factors were found to be cumulatively significant. The Mintel report Customer retention, to switch or not to switch (January 2006), found that customers who had not switched and were not considering switching placed a much greater emphasis on service quality as an important attribute of operating a PCA than those who had switched or were possible movers. This suggests that those who are happy with their bank’s service are less likely to switch.

97

increasing awareness of better offers elsewhere, although it remained low overall: 13 per cent of customers who closed their accounts stated that the main reason was that there was a better account available at a different institution, up from 8 per cent in 1999 Q4. When asked for all reasons for closing an account, the percentage who stated that there was a better account elsewhere increased from 12 per cent in 1999 Q4 to 25 per cent in 2006 Q4. The ORC survey showed that the main reason respondents decided to change bank was that charges were too high or were lower elsewhere (29 per cent of respondents who had switched). Poor customer service (12 per cent) and a more convenient branch location (13 per cent) were also important.

4.211 The BMRB survey found similar results among customers who said that they had paid charges. For those who had paid unauthorized overdraft charges, the two main unprompted137 reasons given for switching were ‘charges too high at old bank’ (35 per cent) and ‘poor customer service’ (14 per cent). The same reasons applied for customers who said that they had paid authorized overdraft charges; ‘charges too high at old bank’ (25 per cent) and ‘poor customer service’ (15 per cent). On an unprompted basis, few respondents referred to a better deal, lower charges or better rates at a different bank as a reason for switching, suggesting that pull factors from other banks with a more favourable PCA were relatively unimportant in triggering switching. Some 7 per cent of those charged for unauthorized overdrafts gave ‘could get a better deal elsewhere’ as a reason, and 8 per cent of those charged authorized overdraft charges referred to a lack of services and facilities in comparison with a new bank. When customers were prompted to consider better deals and pricing at other banks, they were much more likely to agree that this was a factor.

4.212 Once an individual has decided to move to a new bank, the single most significant factor in the choice of bank appears to be access to a branch (25 per cent of respondents to the MORI survey in 2005 Q4 who had switched bank quoted ‘near where I live’ or ‘near where I work/study’ as their main reason for choice of bank). Recommendations (20 per cent), levels of charges on the new account (7 per cent), and interest rates (7 per cent) play a greater role in the choice of alternative bank than in the decision to switch. For new-to-banking customers, proximity to home was the most important factor (around 25 per cent for the major reason), followed by ‘parents banked there’ (14 per cent) and ‘recommended to me’ (10 per cent).

4.213 The BMRB survey found that the main two prompted reasons given for choosing a new bank for customers who believed that they had paid charges and switched bank in 2006, were ‘good reputation’ and ‘well known name’.138 Location, recommendation and family and friends using them (70 per cent, 56 per cent and 52 per cent respectively of charged customers) were also significant reasons, but ‘whole offer was better’ was given as a reason by 73 per cent of unauthorized overdraft charge payers and 60 per cent of authorized overdraft charge payers, and ‘generally charges or rates were better than other banks’ was given by 67 per cent of unauthorized overdraft charge payers and 45 per cent of authorized overdraft charge payers. While this may reflect some post-rationalization by respondents it suggests that those who have paid charges are more inclined to consider relative rates and charges in choosing their new bank.

137In some questions respondents were presented with a list of possible responses or reasons from which they could choose. Answers to such questions are described as prompted. Unprompted responses were those where the respondent was not given any guidance on possible answers. 13878 per cent for each for those who had paid authorized overdraft charges, and 86 per cent and 76 per cent respectively for those paying unauthorized overdraft charges.

98

4.214 A clearer ([]) submitted an analysis of MORI data which attempted to demonstrate that market share was responsive to respondents’ average level of satisfaction with their bank’s charges (a push factor, if satisfaction is falling), and the average satisfaction score for all other banks’ charges (a pull factor, if the satisfaction with other banks’ charges is seen to be increasing). Such a relationship was detected for the clearers but not for the non-clearers. However, our analysis of the data suggested that market shares are influenced by a number of factors, and the effect of changes in satisfaction with charges on customer behaviour is, in fact, small. Thus, declines in customer satisfaction can, through this model, only explain a small proportion of First Trust’s and Northern’s loss of customers over time, while market shares over the last five years have increased for BoI and Ulster []. We therefore do not believe that relative satisfaction with charges is a major driver of the original decision to switch.

4.215 Mintel (2006) asked respondents in Northern Ireland what factors might prompt them to switch to a different bank.139 The question was hypothetical as it did not reflect actual behaviour or experience. 15 options were offered, of which 12 referred to ways in which a new PCA might provide a better service or pricing, and one referred to push factors—‘if I was dissatisfied with the poor service of my current provider’. The others were ‘none of these’ or ‘would not change’. Nearly 40 per cent of respondents referred to dissatisfaction with the current provider as a reason for considering switching. The next highest reason for switching was if a better credit interest rate was offered, at 25 per cent, but over 27 per cent said nothing would motivate them to switch.

4.216 In summary, we found that awareness of better offers is increasing as a reason for a customer to decide to switch from their old bank and in the reasons for choice of their new bank. Nonetheless we found that overall there was a lack of responsiveness to changes in charges or interest rates in the market (see paragraph 4.208) leading to the observed overall rates of switching (see paragraphs 2.72 to 2.92).

Barriers to searching and switching

4.217 Survey evidence shows that customers perceive switching to be problematic. Concerns include anticipation of hassle and delay, as well as risk and costs of switching. The ORC survey found that around 40 to 50 per cent of respondents who had not switched anticipated likely problems in terms of the amount of paperwork; time taken to switch; moving direct debits and standing orders; moving salary or benefit payments; and incurring charges for switching. The perception seemed far more negative than the actual experience of those who had switched, although 20 per cent experienced problems with moving direct debits or standing orders, due to difficulties with either one of the banks, or with a third party. We note that the banks do not have control over the entire switching process, since, for example, third parties are involved in amending direct debit instructions and employers must alter the destination of salary payments.

4.218 The Which? survey of current accounts (2006) showed that 29 per cent of UK respondents had considered switching but had not done so. The three main concerns given were worries about standing orders and direct debits not being paid (39 per cent), a perception that the process was too complicated (34 per cent), and ‘customer couldn’t be bothered’ (31 per cent). 76 per cent of those who had switched (23 per cent of the sample) found it easy overall, although 42 per cent experienced

139Customer retention, to switch or not to switch, Special Report, January 2006.

99

one or more problems. This was a UK-wide survey of Which? subscribers who, if anything, might be expected to be more financially literate than the average customer. These results appear to reinforce our finding that there are both perceived and actual difficulties associated with switching.

4.219 The BMRB survey also gave some indication as to why customers who believed that they had paid authorized overdraft charges on traditional PCAs and had considered switching had chosen not to do so. The main reasons for not switching were related to it being perceived as a difficult process. 30 per cent of respondents said it was ‘too much hassle’ to move things such as salary payments, direct debits and standing orders between banks, and 13 per cent said it would take too long to switch. Similar answers in the top ten reasons include ‘haven’t had time’, ‘inconvenient’, ‘lots of paperwork’ and ‘still thinking about it’. 6 per cent believed fees for switching were too high.

4.220 Although switching is free under the Code, self-switchers may be subject to charges when switching from or to traditional PCAs, as charges may be levied for cancelling or setting up standing orders and direct debits where the bank does not know that the customer is switching. In the past, the clearers imposed a charge of around £30 for transferring an account, in addition to cancellation costs for standing orders and direct debits. The clearers removed the charges for transferring an account in late 2002 or early 2003. The ORC survey showed, however, that more than half of all respondents who had not switched anticipated charges for switching, suggesting that customers might be unaware that these charges have been removed.

4.221 Customers also believe that switching bank may damage their longer-term banking relationship, jeopardizing, in particular, future availability of credit. The BMRB survey also found that some respondents perceived that someone who had paid unauthorized charges would be less attractive to a new bank and may receive worse terms. Although we found that customers perceive this risk to be greater than it is in reality, we understand that some banks take into account customer behaviour over time in decisions, for example, on whether to refund disputed charges (see paragraph 4.122). In general, customers tend to be distrustful of new providers, particularly in connection with their financial affairs.140

4.222 The clearers said that the Code, and the banks’ switching services, provided an assurance that the hassle factor in switching was minimized. The banks have switching packs which explain the switching process. Ulster said that it was spending significant amounts ([]) on promoting ease of switching. However, the Millward Brown survey found that most respondents felt uninformed about the switching process, and we believe that it is likely to take time to change public perceptions about switching, particularly given the potentially significant impact to an individual of mistakes being made (eg salary not received or mortgage not paid). The evidence that we received suggested that measures such as the Code are working to ease the switching process, but that customers either are not aware of this prior to switching, or are unwilling to rely on it working smoothly in their own case.

4.223 We also found that customer indifference,141 lack of interest in PCAs, and ignorance of some charges (see paragraphs 4.231 and 4.232) are important reasons for lack of switching. This may be due to a perceived lack of financial incentive to switch, to

140See the UK-wide report on switching across a range of industries, Switching Suppliers, Consumer Affairs Report 2001 (DTI, 2001). 141We chose to use the term ‘customer indifference’ as a neutral term which is not intended to imply any fault on the part of the customer.

100

complexity in charging structures, or to a lack of transparency in the market making searching and switching difficult. We were told that levels of financial capability generally were low (see paragraph 2.72), and that many customers felt that PCAs were ‘all the same’. The Millward Brown survey also suggested that customers in Northern Ireland feel intimidated by the banks, both for opening accounts and for ongoing servicing, and considered it a ‘privilege’ to hold a PCA, which might increase the barriers to searching and switching.

4.224 We sought to estimate a customer’s financial incentives to switch. We estimated that, on average and across all customers of the eight largest banks, the average customer might incur a benefit from switching banks of up to £72 in 2005.142 However, we recognize that there is considerable variation in the financial incentive to switch depending on an individual customer’s circumstances.

4.225 For the majority of customers who pay no transaction or maintenance charges, the financial incentive to switch is likely to be low (associated primarily with differences in credit interest).143 However, some of the non-clearers compete by offering attractive headline rates of credit interest.144 We were told by these banks that customers were responsive to credit interest rates. They told us that the promotion of attractive rates of credit interest resulted in increased levels of switching-in. There is other evidence that customers are more responsive to credit interest rates than debit interest rates. A Bank of England study145 found that market shares in PCAs in the UK from 1996 to 2001 were moderately sensitive to differences in the credit interest rate across banks. Customers were found to be much less responsive to differences in the debit interest rates.

4.226 As illustrated by our work on customer scenarios (see Appendix 4.7), a customer who regularly operates their account in overdraft, and particularly in unauthorized overdraft, might have a financial incentive to switch of several hundred pounds over a 12-month period. In addition, the clearers do not, in general, offer financial incentives, such as lower charges, access to better interest rates on savings accounts, or finan- cial inducements to switch, although such incentives are more commonly offered by banks based in Great Britain.146 We note that where customers have a low propensity to switch, and switching is not driven primarily by financial incentives, the banks will have a reduced incentive to offer financial advantages over their competitors.

4.227 In early 2007, Halifax and first direct both increased their incentive payments payable to switchers throughout the UK who meet funding requirements from £50 to £100.

142Source: CC analysis of financial information provided by the banks (calculated as the charges and interest paid less the interest received). This was based on our analysis which showed that, after normalizing for differences in average credit and debit balances between banks, the average customer of one bank paid up to £72 more than the average customer of another bank in 2005 (2004: £67, 2003: £63 and 2002: £57). In our provisional findings, we presented a similar analysis which was based on active accounts only. Based on active accounts, the comparable figures were 2005: £87, 2004: £80, 2003: £83 and 2002: £59. 143For example, a customer paying a basic rate of tax who maintains an average credit balance of £2,000 over the year would earn additional credit interest of around £30 in switching from a bank paying 0.1 per cent credit interest to one paying 2 per cent credit interest. 144Some of the clearers are also offering credit interest rates substantially higher than used to apply on traditional PCAs, particularly on packaged accounts—see Appendix 2.4. 145Bank of England Working Paper 292—Switching costs in the market for personal current accounts: some evidence for the UK Celine Gondat-Larralde and Erlend Nier. 146Halifax offered a £100 incentive payment for switchers; as at April 2007 first direct offered a £100 incentive payment for switchers; Nationwide has recently offered existing customers of Nationwide’s non-PCA products of at least six months’ stand- ing a £100 incentive to switch their PCA to Nationwide (this offer was open to 5 May 2007); Lloyds TSB offers switchers access to accounts with better terms such as credit interest rates; and A&L limits its preferential savings account offer to new PCA customers. A&L also offers a £25 ‘recommend a friend’ payment to existing account holders where a recommended friend opens a Premier Direct Current Account or a Premier Current Account. These incentives tend to be conditional on minimum monthly credits.

101

Nationwide offered a £100 switching incentive to some customers for a limited period in early 2007. Ulster announced on 2 May 2007 that it would offer a £150 cash incentive to customers who switched their PCA from another bank to Ulster during the month of May. We note that these banks consider it worthwhile to offer such incentives. We also note that Halifax subsequently withdrew its incentive payments and Ulster’s offer was for a period of one month only. It remains to be seen whether such incentive payments continue to be offered, and if so whether this substantial increase in the level of incentive payments will overcome some of the reluctance to switch in the market. It is possible that if such incentives are sustained, they may in time impact on general attitudes to switching.

Customer attitudes to unauthorized overdraft charges

4.228 We looked at the evidence that we received on customers’ attitudes to unauthorized overdraft charges, whether customers were aware of and considered unauthorized overdraft charges, and whether or not these factors were likely to constitute a barrier to searching or switching.

4.229 Various pieces of evidence suggested that customers were not particularly aware of or responsive to unauthorized overdraft charges.

4.230 In the BMRB survey, many customers who the banks identified as having paid unauthorized overdraft charges in the first half of 2006 told us that they had not paid charges.147 It is possible that some customers could not accurately recall when charges were levied, and that some customers might be reluctant to admit that they had had an unauthorized overdraft. However, it is also plausible that some customers were not aware that they had incurred charges. Such charges are more commonly incurred by higher income customers, who may be less likely to monitor balances closely and charges on their accounts than those on lower incomes.

4.231 We then sought to assess whether customers were aware of their bank’s unauthorized overdraft charges and whether they knew how these compared to those of other banks. The ORC survey found that switchers did not consider themselves to be more or less informed than non-switchers about the charges associated with other banks’ PCAs, suggesting that charging did not stimulate customers into researching PCAs in the market. However, the ORC survey did not distinguish between authorized and unauthorized overdraft charges. Therefore responses to questions about charges will, in the case of those on traditional PCAs, reflect charges arising from both authorized and unauthorized overdrafts.

4.232 The BMRB survey asked specifically about customers’ levels of knowledge of unauthorized overdraft charges and whether experience of charging affected this. Just under half of customers who said that they had paid unauthorized charges felt that they were well informed or very well informed about how the level of their bank’s charges compared to other banks (the same applied to 35 per cent of those who had not paid charges). This indicates that recent experience of being charged is associated with a greater knowledge of charges.148

4.233 However, the levels of awareness, and the accuracy of that awareness were questionable. Of those who said that they were well or very well informed, 16 per

147142 respondents out of 340 charge payers identified by the banks—see Appendix 4.6. 148BoI said that the inclusion in the group of charged customers of customers who told us that they had paid charges where the banks said they had not may bias downwards these results. This is possible although where such customers were referring to charges incurred just before or after the sample period their responses are unlikely to show such bias.

102

cent of the charged group and 72 per cent of the uncharged group were unable to give an estimate of an unpaid item charge, and 27 per cent of the charged group and 81 per cent of the uncharged group were unable to estimate the relevant referral charge. Consequently it appears many customers have limited knowledge of applic- able charge levels. Estimates offered varied, although most gave answers broadly similar to actual charges (eg in the range of £20 to £40). Recent experience of being charged was strongly associated with a higher level of knowledge, although charge levels are much less relevant to those who rarely if ever go into unauthorized overdraft. Some 28 per cent of those who had been charged told us they had not known such charges existed before they were applied.

4.234 The clearers argued that the BMRB results indicated that there was a substantial degree of awareness of unauthorized overdraft charges in the market. In support they particularly noted that: around 70 per cent of respondents were aware of the existence of these charges before they were charged; over half of customers who had been charged who had switched bank had compared charges a lot or a little; those who were charged were more likely to have compared interest rates and charges than those who had not been charged. We disagree that these results indicate a high level of awareness. For example we consider it notable that 28 per cent of respondents were not aware of the existence of charges, and that although many of those who were charged and switch bank do compare charges, overall the comparative level of charges is not a primary reason in the choice of new bank and taken across all evidence received responsiveness to comparative charges appears to be low.

4.235 Evidence drawn from the banks suggested that they perceived that customers were unlikely to switch as a result of changes in these charges. A pricing paper from one of the clearers ([]) said that ‘the vast majority of [unauthorized overdraft] fees … are believed to be unshopped149 at point of sale, and relatively inelastic in comparison to interest rate’. One of the non-clearers ([]) said that the vast majority of customers who found themselves in unauthorized overdraft were ‘at the very far end of the apathy scale’. Another internal pricing paper from the same non-clearer ([]) said that: ‘Increasing charges will have less impact on our marketing proposition than a credit interest change due to its lower visibility.’ In discussing the impact of increasing the cap on paid items and the unpaid charge, the same paper said: ‘Banking have no concerns that sales will be significantly affected by the fee or funding changes. Sales generation for primary accounts focuses on credit rate and [] as the positive account features.’ However, other banks ([]) said that customers who had had unauthorized overdraft charges in the past would be much more likely to be aware of them.

4.236 We also note that the banks’ evaluations that we have seen of changes to charging structures and proposed increases in unauthorized overdraft charges (see paragraphs 4.140 to 4.161 and Appendix 4.8) make only limited allowance for changes in customer behaviour or switching resulting from increases in these charges (see paragraph 4.174). If customers were sensitive to unauthorized overdraft charges, banks might not choose to refocus on unauthorized overdraft charges in order to help retain existing customers and attract new customers. []

4.237 We looked at survey evidence on the relationship between switching and charges to see if there were indications as to whether customers were responsive to these

149We take ‘unshopped’ to mean that, when choosing a PCA, the customer does not compare unauthorized overdraft charges with the same or similar products offered by different providers.

103

charges.150 The ORC survey suggested that switchers were no more likely on average to have experienced charges (of any sort) than those who had not switched. However, the BMRB survey found that 8 per cent of customers who said that they were charged unauthorized overdraft charges in the first half of 2006 had switched banks in 2006. This was significantly higher than the 2 per cent of customers who had not been charged in this period who had switched. Customers who had never switched their main bank were then asked whether they had considered switching. 29 per cent of the charged group, and 13 per cent of the uncharged group said that they had considered switching in 2006. In contrast to the ORC results, the BMRB survey suggests experience of being charged leads to higher switching rates and a greater likelihood of considering switching.

4.238 The ORC survey found that charges were a major reason given by customers for choosing to switch. In the BMRB survey, among charged customers who had switched banks, over one-third of customers (35 per cent) cited ‘charges too high at old bank’ as their main reason for switching, and when prompted 62 per cent said this had been a reason for switching (a marginally higher proportion referred to authorized overdraft charges on traditional PCAs as a reason for switching).

4.239 It therefore appears to us that a decision to switch often results from a customer being dissatisfied with charges they have experienced. However, even though account charges are found to be one of the main push factors in persuading customers to switch, comparative charges are only a very small factor in the reasons switchers give for their choice of provider.151 This suggests that customers are pushed into switching because they feel charges are excessive or unfair rather than because they considered them to be uncompetitive with other banks.

4.240 The BMRB survey indicated that switchers took more account of unauthorized overdraft charges in their choice of bank. Some 50 per cent of switchers who thought that they had been charged agreed that lower unauthorized overdraft charges was a factor in the choice of new bank, although this was the ninth most popular response. This appears high and some other responses such as ‘whole offer was better’ may have reflected such charges; however reputation and location were the most important aspects in the choice of new bank. The low consideration given to unauthorized overdraft charges is supported by the fact that most switching has been from clearers to the non-clearers, who in many cases have had higher unauthorized overdraft charges. Less than one-quarter of charged customers who had switched provider said in the BMRB survey that they had made any comparison of un- authorized overdraft charges between banks.152 The proportions were lower for customers who had not been charged. Therefore it appears that comparative charges are not a strong influence on the choice of new bank.

4.241 We now consider why customers do not appear to be strongly influenced by pull factors in relation to unauthorized overdraft charges and are disinclined to compare charges between banks.

150We also note the evidence from []. 151The uncharged group sample was matched to the charged group so results would not reflect demographic differences. 8 per cent of switchers gave lower or no charges as their reason for choice of new bank in the MORI survey for the year to 2005Q2. Even when prompted, only one-third of respondents in the BMRB survey who had not switched banks referred to charges and rates as a factor in their choice of bank. 15256 per cent of switchers said that they had compared current accounts a little or a lot before switching. 76 per cent of these, who had paid charges, said that they had compared interest and charges before switching, and of those 57 per cent said that they had compared unauthorized overdraft charges.

104

4.242 The BMRB survey found that most customers who believed that they had incurred unauthorized overdraft charges subsequently took actions to try to avoid charges arising in the future. Some 88 per cent of those who had paid unauthorized overdraft charges in the first six months of 2006 said that they resolved to monitor their balances more closely in future; some 81 per cent said that they kept more money in the PCA; 22 per cent said that they arranged a higher authorized overdraft limit; and 8 per cent said that they took out a loan. This suggests to us that customers may anticipate that, even where they have incurred a charge in the past, they would not do so again in the future. [] also pointed to a TNS survey conducted on behalf of the OFT (report date May 2006) that found that the reason many customers incurred unauthorized overdraft charges was because they either forgot a payment was going to be made from their account or a payment into their account was delayed or not made. It is possible that customers would not anticipate that this would occur again, and [] said that this was consistent with a finding that customers chose to take simple steps to manage their account more carefully.

4.243 One clearer ([]) said that if customers responded to unauthorized overdraft charges by modifying their behaviour so as to avoid future charges, this represented a constraint on the banks. We accept that such actions will impact on the banks’ abilities to raise revenues from such charges, but this is not equivalent to a competitive constraint on charging. In this case the customer’s reaction is not influenced by how their bank’s charges compare to those of rival banks and no competitive advantage would be achieved by lowering charges below those of rivals. Rather, banks would individually decide on appropriate charge levels given their own customers’ reactions instead of taking account of competitors’ charges. There has been no indication that the extent to which customers take action to avoid repeat charges and the success with which this is achieved is responsive to the relative level of charges. Evidence from the banks (see Appendix 4.6 annex 1) also shows that many charges are paid by customers who had incurred charges the previous year, suggesting customers do not effectively implement or maintain their strategies for avoiding future charges.

4.244 We consider that there are several possible reasons for customers failing to compare unauthorized overdraft charges when switching bank, or ignoring overdraft charges when making their initial choice of bank:

(a) Customers do not believe that they will go into unauthorized overdraft. This may reflect optimism, but also if customers anticipate the need for additional funds, it would be more rational for them to seek a loan or to create or extend their overdraft limit rather than incur unauthorized overdraft charges. However, the incidence data that we have seen suggests that many charges are incurred repeatedly by the same individuals, and so while it may be rational to seek to avoid charges where possible, some customers repeatedly fail to manage this, and would have an incentive to seek to minimize the charges they face (see paragraph 4.116).

(b) Customers find that they cannot understand the overdraft charges, either because they are too complex or are not clearly explained (see paragraphs 4.121 to 4.139).

(c) Customers seek to compare more immediately visible aspects of the banks’ charging structure (eg credit interest rates).

(d) Switchers wish to ‘punish’ their old bank for imposing a charge which they consider to be unfair.

105

4.245 The BMRB survey asked respondents whether the presence or level of unauthorized overdraft charges were a consideration when choosing their current bank. Some 79 per cent of the charged group and 84 per cent of the uncharged group said that they had not been a consideration. Most customers said that this was because they did not think unauthorized overdraft charges would apply to them. For example 23 per cent of those in the charged group said ‘I did not believe I would breach an overdraft limit’. Another 11 per cent of the group who said they had been charged for an overdraft said that they had not considered unauthorized overdraft charges in choosing their bank because they never had an overdraft (compared with 48 per cent for those who had not paid charges in the first half of 2006). Some 24 per cent of charge payers said they just did not think about them. Similar reasons for not looking at unauthorized overdraft charges were prevalent among customers who had switched banks. For example, the most common responses among customers who had paid charges in the first half of 2006 were: ‘not relevant to me’ (34 per cent compared with 69 per cent of those who had not paid charges in the first half of 2006); ‘not interested in comparing charges’ (9 per cent); and ‘did not have the time’ (9 per cent).153

4.246 Overall this indicates to us that many customers, even if they have incurred unauthorized overdraft charges, either do not think about these charges at all or do not consider them to be relevant to them because they do not believe they will incur these charges again.

4.247 The banks told us that there was nothing in their policies that prevented customers who had had unauthorized overdrafts from switching. A history of unauthorized overdraft should not be a barrier to switching, although customers may believe that their history would debar them from acceptance by a new bank; and/or they would be offered worse terms and conditions; and/or their application would require additional credit checks or other assessments. The customer may be embarrassed about asking for a PCA to be opened when they have a history of unauthorized overdraft. In addition, the responses to the ORC survey show that many customers believe that there are advantages in staying with one bank, and we found that there were some instances where an established relationship has an impact on the bank’s treatment of its customers (see paragraph 4.122(g)). The BMRB survey found many customers believed having had an unauthorized overdraft would mean that any new bank would offer inferior terms to such a person switching to them, although very few customers who had incurred charges said that this made them less likely to switch bank.

Complexity and lack of clarity

4.248 Issues of undue complexity and lack of clarity of charging structures and practices are discussed in paragraphs 4.121 to 4.139. Notwithstanding the requirement for the PCA to service a wide variety of needs in a single product, we found it difficult to understand PCA charging structures and practices and to compare accounts between different providers. The incidence and frequency of charging is not always

153We noted that the BMRB survey found similar results for authorized overdraft charges. Of customers who had paid authorized overdraft charges on traditional accounts and who had switched accounts, 15 per cent said that they had compared accounts before switching a lot and 20 per cent a little. 16 per cent of charged customers and 7 per cent of uncharged customers said that they had compared authorized overdraft charges between banks. Those who had not made comparisons were asked why not: 28 per of charge payers and 62 per cent of non-payers said that charges were not relevant to them; 17 per cent of payers and 5 per cent of non-payers said that they were too busy; and 16 per cent of payers and 7 per cent of non- payers said that they were not interested in comparing charges. 83 per cent of customers said that they did not consider authorized overdraft charges when choosing their traditional account; reasons given were that they just did not think about it (34 per cent); and did not believe they would need an overdraft (30 per cent). This therefore suggests that in a similar way to unauthorized overdraft charges, many customers have not thought about these charges when choosing a bank.

106

clear to customers. The fact that customers are charged both retrospectively (ie for transactions that have already taken place previous to this point in the quarter) and for the remainder of the quarter makes it difficult for customers to understand the costs of running a traditional PCA in authorized overdraft and to compare the costs between different accounts with any certainty, raising barriers to switching. Customers often do not have the opportunity to see the charges incurred before they are debited to their account, adding further to the difficulty in understanding and managing them appropriately. The terminology used for charges is not always clear and is often inconsistent between banks, particularly in relation to unauthorized overdrafts, and ancillary charges.

4.249 We note that several independent comparisons of PCAs in Northern Ireland are available. These include Which? and price comparison websites such as uSwitch and moneysupermarket.com. However, it is not easy to apply the advice available from such websites to individual circumstances; costs of operating in unauthorized overdraft are usually not included and none offer a facility to search for best-buy accounts on this basis; and accounts which are no longer being offered by the bank to new customers are often not included, despite there being a large percentage of customers on such accounts.

4.250 We therefore thought it likely that, should customers seek to compare PCAs between banks in order to decide whether to switch, where to switch to, or to choose a new PCA, they would find it difficult to do so. We believe it plausible that customers’ reliance on factors such as branch location and personal recommendation for choice of bank may well be due to the difficulty that customers have in comparing the price of PCAs between banks.

4.251 The BMRB survey found that many customers were unable to estimate the unauthorized overdraft charges applying to their account, and therefore it is unlikely that such customers would be aware of how charges compared to other banks (see Appendix 4.6). Awareness was higher, but still low, among those who had paid charges. Although the BMRB survey found that those customers who tried to compare rates and charges between banks largely found this process easy (see paragraph 4.128), we note that many respondents had made no such comparison and very few respondents had examined unauthorized overdraft charges. We also note that the BMRB survey found a higher degree of ignorance or misunderstanding of authorized overdraft charges on traditional PCAs than applied to unauthorized overdraft charges.

Customer satisfaction

4.252 Several banks told us that an alternative explanation for the relatively low switching rates in Northern Ireland was that customers were satisfied with the service that the banks provided. Few customers therefore experience the push factors that are most likely to cause switching. We note that the MORI survey shows high levels of satisfaction with PCAs—for the period December 2005 to December 2006, 89 per cent of all respondents were fairly or very satisfied with their PCAs.

4.253 Our Millward Brown survey, however, did not support this evidence. The report states, for example, that ‘the range and scope of expressed dissatisfaction within the groups was considerable and unprompted by the moderator’. The dissatisfaction was evident in all the groups interviewed.154 Which? told us that in its survey of current

154Millward Brown Ulster, January 2006, p16.

107

accounts in 2006, the clearers occupied the bottom four places out of 23 banks operating in the UK in terms of overall satisfaction. We also note that the proportion of respondents in the MORI survey who were very satisfied with their PCA fell from 54 per cent at the end of 1999 to 41 per cent by the end of 2006, although the proportion of respondents reporting that they were dissatisfied or very dissatisfied remained low (rising from 3 to 7 per cent during this period).

4.254 The 2001 DTI switching study found customers in several industries reporting satisfaction in quantitative surveys which were at odds with findings in qualitative work. It examined the apparent contradiction between quantitative and qualitative research that can arise. It found that quantitative findings are often contradicted by in-depth qualitative research, suggesting that consumers rationalize greatly in the context of quantitative surveys about subjects in which they have little interest. It suggests that, in certain contexts, quantitative survey research alone needs to be analysed carefully and must be treated with caution.

4.255 Taking these results in conjunction with the other evidence that we have on levels of charges and interest rates and customer behaviour, such as inertia and a lack of searching for better deals, we thought it unlikely that the lack of responsiveness to changes in charges or interest rates in the market are caused by high levels of customer satisfaction; rather we consider that the satisfaction ratings in the MORI survey reflect an absence of active and informed customers.

Conclusions on searching and switching

4.256 We found (see paragraph 4.210) that a customer’s decision to switch is more often prompted by dissatisfaction with their bank than the recognition of a better offer elsewhere. Some of the banks argued that push factors driving switching are an important constraint on banks. We did not see that most customers were likely to evaluate pricing or service relative to other banks in taking the decision to switch. In particular there was evidence that customers were unlikely to consider and compare either authorized or unauthorized overdraft charges, partly because many customers believed that such charges would not apply to them. This might be because customers were optimistic that they would not get into financial difficulties or they believed that they would avoid such overdrafts arising (eg by monitoring their accounts); alternatively it might reflect an unwillingness to think about the issue or a lack of understanding of charging structures.

4.257 If a customer is pushed into switching, the decision about the choice of bank may be based on factors that can easily be observed (eg credit interest rates), or on information that can readily be obtained (eg personal recommendation), rather than on a rational decision based on the costs and benefits (financial or otherwise) of switching. It is therefore not clear to us that banks would gain a significant competitive advantage over their rivals by improving factors that cannot easily be observed. Similarly, for new-to-banking customers the difficulty in searching may result in these customers making wrong, or inefficient, choices (ie choosing an account which does not best meet their needs).

4.258 We therefore conclude that the obstacles to searching and switching set out in paragraphs 4.217 to 4.253 are significant for many customers.155 Whilst some of the

155The literature on competition in the presence of consumer switching costs is reviewed in a paper by Paul Klemperer, Competition when consumers have switching costs: an overview with applications to industrial organization, macroeconomics and international trade, Review of Economic Studies, 62,515–539. This demonstrates that, whilst suppliers may compete strongly for customers entering the market (such as students), they will also compete less vigorously for existing customers,

108

issues with switching are perceived rather than real, we believe that these perceptions affect customer behaviour, leading to a high level of customer indifference. This results in a lack of searching and switching which we believe contributes to a lack of competition in the market.

4.259 We found that in the case of unauthorized overdraft charges, customers who had incurred such charges were more likely to switch bank. However, few customers make comparisons of these charges between banks, and customers tend not to think about such charges when choosing a new bank, largely because they do not believe that they will be affected by such charges. This could be because they intend to avoid unauthorized overdrafts in future. This suggests that customers object to paying such charges but banks would receive little competitive benefit in terms of attracting switchers from lowering these charges.

4.260 We found that there was a level of complexity in PCA charging structures and practices and a lack of clarity of information that would be likely to make it difficult for customers to search and switch. As competitive price offers are not easily observed or understood by customers, competitive pressure on banks is likely to be reduced, giving banks an incentive to raise prices above those that would exist in a well- functioning market. This would be in the interest of banks since it would result in increased revenue from existing customers, with lower levels of switchers than would be the case in a well-functioning market. Similarly, banks would have less of an incentive to reduce charges since this would result in fewer new customers than would be the case in a well-functioning market. Similar considerations apply to service and quality levels.

Analysis of competition

4.261 In this section we set out our assessment of unilateral and/or coordinated effects in the market, before briefly considering local competition. This is based on the evidence and discussion set out in the sections on market structure, bank conduct and customer conduct.

Unilateral effects

4.262 We looked at whether competitive conditions in the market might give individual banks unilateral market power, which might result in, or contribute to, for example, higher charges or debit interest rates, or lower credit interest rates, service quality and/or innovation, than might be the case in a well-functioning market. If customers are unwilling to search for a better PCA and switch to it, banks may be able profitably to maintain charges and debit interest rates at higher levels, and credit interest rates or levels of service and/or innovation at lower levels, than would otherwise apply (see paragraph 4.258). This is only sustainable in the long term if banks can attract new customers. First-time customers will not face switching costs and so banks would need to discriminate in favour of new customers, or rely on other competitive aspects such as brand and branch networks if they are to sustain high prices persistently. Banks also need to weigh up the potential benefits of attracting switchers.

4.263 Given customers’ reluctance to switch, both clearers and non-clearers can enjoy unilateral market power. In particular, given the lack of responsiveness of customers to changes in charges or interest rates, their lack of interest in PCAs and the lack of

who face searching and switching costs. The overall effect is likely to be that prices are higher than they would be in the absence of searching and switching costs.

109

financial capability in the market, and the customer perception of the difficulties and risks of switching, we found that there is scope for all banks to exercise their market power. In addition, for many customers, particularly those who do not incur overdraft- related charges (or who do not believe that they are going to incur such charges), the financial gains from switching are low. The undue complexity of charging structures and practices and the lack of clarity of terminology also contribute to a lack of competition in the market.

4.264 Fee-free PCAs tend to have higher charges and debit interest rates associated with operating in unauthorized overdraft than traditional PCAs, but lower costs to the customer of operating in authorized overdraft. This was illustrated by the work that we carried out on the costs to a customer of operating in authorized or unauthorized overdraft (see paragraphs 4.124 to 4.126 and Appendix 4.7). Table 3 shows that, overall, the clearers’ customers, most of whom were on traditional PCAs, paid more than the non-clearers to operate their PCAs, as evidenced by their higher total revenue earned directly from PCAs per account in 2005.

4.265 The move from a traditional PCA pricing structure to a fee-free model could be consistent with a well-functioning market. However, this would depend on customers being able to make efficient choices of PCA. This might be the case, for example, if they were readily able to evaluate the full-life cost of different PCAs—covering the different ways in which a PCA may be used (in credit, and in both authorized and unauthorized overdraft)—or if competition on charges that could readily be observed and compared by customers (visible charges) was sufficient to offset any detriment on less visible charges.

4.266 We looked in particular at unauthorized overdraft charges as a key part of the charging structure for PCAs on which there was little visibility. A banks’ decision to move to a fee-free PCA charging structure often involves an increase in unauthorized overdraft charges. As set out in paragraph 4.180, charging structures and practices relating to unauthorized overdrafts are unduly complex and difficult to understand and apply. We found that customers have little awareness of charges and how they compare to competitors, and tend not to consider such charges when choosing a bank even if switching as a result of having been charged. We have seen evidence that there are perceived barriers to searching and to switching; banks in general do not appear to set prices to take account of the negative impact arising from increasing their unauthorized overdraft charges, and we have not, until recently, seen banks seeking unilaterally to undercut competitors.

4.267 The main constraint on unauthorized overdraft charges appears to be possible adverse publicity arising from comparatively high charges, or from applying charges where breaches of overdraft limits are small or short-lived. It appears that banks have increased the use of buffer zones and charge capping, and have used discretionary refunds, to limit customer dissatisfaction arising from unauthorized overdraft charges. These pricing practices, and the extent to which banks use their discretion, are not normally known to customers.

4.268 We note that there are additional administrative costs for the bank in managing a customer in unauthorized overdraft. The banks must have the people and systems in place to make the necessary decisions. For example, it may require manual intervention to decide whether to honour a cheque, and to return it unpaid if deemed necessary. In addition, unauthorized balances are likely to pose a greater risk of bad debt to the bank. However, as set out in paragraph 4.176, we think it unlikely that costs could account for the observed levels of unauthorized overdraft charges, both because the detailed costs are not known by the banks, and because the substantial

110

changes in charges that we have seen are unlikely to correspond to similar changes in costs.

4.269 The banks told us that they sought to avoid customers incurring unauthorized overdraft charges. Their objective was to foster a positive customer relationship and increase customer satisfaction so as to maximize the possibility of cross-sales. Nonetheless, the banks’ business plans that we have seen rely on unauthorized overdraft charges for generating PCA revenues. In addition, the incidence data set out in paragraph 4.116 shows that whilst some customers may be able to modify their behaviour to avoid incurring charges, a significant proportion go repeatedly into unauthorized overdraft and are likely to continue to incur charges, whatever their level.

4.270 We found that customers had a similar lack of understanding of charges, lack of knowledge of charging levels, and a lack of inclination to consider and compare charges between banks in relation to authorized overdraft charges on traditional PCAs. Because of this, customer behaviour is unlikely to incentivize banks to compete vigorously on both authorized overdraft charges and unauthorized overdraft charges. This suggests that banks will have the opportunity to exploit unilateral market power under a range of different charging structures.

4.271 We therefore believe that, as a result in particular of the barriers to searching and switching that customers face arising from the undue complexity of charging structures and practices and lack of clarity of terminology, customers may be paying higher charges or levels of debit interest, or receiving lower levels of credit interest, lower levels of service, or there may be less innovation, than would be likely to occur in a well-functioning market.

4.272 We believe that both clearers and non-clearers have the ability to exercise unilateral market power. However, their incentives to do so might vary due, for example, to their different strategic focus and business models (see paragraphs 4.3 to 4.13), their different customer bases, or their different operating scale or costs, such as the size of their branch network. Consequently it is possible that the way in which banks seek to exploit unilateral power and the extent to which they try to do so might vary. A bank with a smaller base of existing customers would, for example, have relatively less to gain in the long run from pricing high to maximize revenue from customers who were reluctant to switch, and more to gain from offering attractive prices to switchers.156

4.273 We observed several indicators that suggest that the incentive to exercise unilateral market power was, and to some extent still is, greater for clearers than for non- clearers:

(a) continued existence of certain charges on traditional PCAs, relatively low levels of credit interest, and the higher observed average total revenues earned directly from PCAs per account of the clearers (see Table 3). More than one-third of customers in Northern Ireland were still on traditional PCAs in September 2006 although since the introduction of the BoI’s fee-free package, this has fallen to around 19 per cent;

156That banks with a larger market share have a greater incentive to exercise unilateral market power is supported by the findings of a Bank of England study (Bank of England Working Paper 292—Switching costs in the market for personal current accounts: some evidence for the UK Celine Gondat-Larralde and Erlend Nier), which found a positive relationship between levels of market share and the price of PCAs in the UK from 1996 to 2001. It also found the marked dispersion in price was persisting through time.

111

(b) relative lack of competition between the clearers and the non-clearers on non- price factors (eg opening hours and introduction of full-functionality Internet services) (see paragraph 4.206);

(c) clearers’ focus on competition for students and graduates, and relative lack of competition for switchers, particularly prior to the launch of the clearers’ fee-free PCAs (see paragraphs 4.187 to 4.192); and

(d) ability to continue to attract a relatively high proportion of new-to-banking customers because of the clearers’ well-established branch networks and strong brands (see paragraphs 4.58 and 4.59).

4.274 The clearers said that overall market share would be significantly impacted by a failure to recruit either new-to-banking customers or switchers, and pointed to the reduction in market share (by number of main accounts) suffered by some of the clearers in the last five years as evidence that they could not afford to exploit unilateral market power in this manner. In addition, they said that those customers that held multiple PCAs could easily switch their credit balances and transactions between providers without opening and closing their accounts. However, in so far as customers switched their account usage between banks, it should be captured in survey evidence where customers are asked to report changes in the bank providing their main PCA. We saw no evidence that customers switched account usage between banks in response to differences in the offer in a way that would not be captured in the survey evidence showing how customers switched their main PCA. The clearers also said that the non-clearers had attracted a much higher proportion of new accounts than suggested by their stock of PCAs.

4.275 As noted in the section on market shares and concentrations, it appears that the non- clearers’ share of flow does not translate directly into their share of stock. Even if customers are being lost to other banks, it may still be the most profitable course of action for the clearers to continue to seek to take advantage of market power over existing customers than to compete directly on price for new ones.

4.276 One clearer ([]) put it to us in response to provisional findings that unilateral power was inconsistent with the fact that most customers pay no charges, that charges for authorized overdrafts are being removed by clearers, and those with higher charges for unauthorized overdrafts are expanding market share. We do not agree; banks have chosen their charging structures to reflect the characteristics of the market and the way customers behave. We found that bank charges are being rebalanced towards charges (for example unauthorized overdraft charges) that are less visible to customers and so banks may be able to attract customers despite such charges. We also note the lack of significant credit interest paid on many PCAs.

4.277 Overall, we found that both the clearers and the non-clearers have the ability to exercise unilateral market power. The documentary evidence from the clearers in particular did not suggest that charges and interest rates were subject to the competitive constraints that would exist in a well-functioning market.

4.278 We also looked specifically at whether competitive conditions in the market since the supercomplaint was made had changed sufficiently to suggest that the banks no longer had the incentive or the ability to exercise market power.

4.279 Competitive conditions in the PCA market in Northern Ireland have changed over the past few years. In particular, there is a growing awareness of the option to switch coupled with reducing dependence on branches for undertaking transactions; increasing awareness of the Code’s provisions on switching and switching services

112

offered by the banks; change in Northern’s ownership and the launch of new PCAs by the clearers; increasing publicity about PCAs and the clearers; growth in market share of the non-clearers; increasing publicity about unauthorized overdraft charges; the recent introduction of substantial cash incentives to switch by some banks; and increasing marketing of PCAs by the banks.

4.280 Despite the evidence of increased publicity around switching and the experience of most switchers that it was a straightforward process, we found that some switchers still experienced problems in switching (see paragraphs 4.217 and 4.218), switching was still widely perceived by those who had not yet switched to be risky and costly, and switching rates, although increasing, were still relatively low.157

4.281 Historically Ulster and Northern have been operated as locally distinct banks with locally determined pricing and products. We thought that, under the ownership of RBS and Danske respectively, Ulster and Northern would still take pricing decisions based on local conditions (as, for example, was the case with their new PCA portfolios and charging structures), although we recognize that the changes in ownership may lead to a change in competitive tactics and positioning. In particular, there are likely to be permanent changes as a result of the introduction of new IT systems.

4.282 We thought that the growth in market share of the non-clearers might be more significant as an indicator of changed market conditions, suggesting increasing external constraints on the clearers’ market power. Nevertheless, we thought that barriers to expansion by the non-clearers remain which may explain why their market share remains lower than that of the clearers despite having been in the market for many years and the fact that, in certain respects, they continue to offer better terms than those offered by some of the clearers. These barriers include, in particular, customers’ reluctance to switch, marketing costs, and the advantages of an extensive branch network that many customers find attractive. In addition, the non- clearers are not focused on expansion in Northern Ireland. The evidence that we received has not led us to form an expectation that the non-clearers will increase their number of branches in Northern Ireland in the foreseeable future. We thought it unlikely that the non-clearers would exert a stronger competitive constraint in the foreseeable future than they do today (see paragraphs 4.299 and 4.300).

4.283 Unauthorized overdraft charges have recently been the subject of increased publicity both in Northern Ireland and in the UK more generally. Which? and GCCNI have launched campaigns to raise public awareness and encourage customers to challenge the charges made. The OFT published its findings under the Unfair Terms in Consumer Contracts Regulations 1999 on credit card default charges on 5 April 2006.158 This concluded that credit card default charges had generally been set at a significantly higher level than it considered legally fair.159 It stated that default charges should only be used to recover certain limited administrative costs; these may include postage and stationery costs and staff costs and also a proportionate share of the costs of maintaining premises and IT systems. The OFT said that these broad principles were likely to be relevant to other default charges in standard

157As noted in paragraph 2.88, patterns of switching, which appear to have been primarily from the clearers to the non-clearers in the past, may be changing. 158OFT press release 68/06. 159The OFT findings are based on its view of the law, which is in essence that default charge provisions are open to challenge on grounds of unfairness if they have the object of raising more in revenue than is reasonably expected to be necessary to recover certain limited administrative costs incurred by the credit card issuer. The OFT acknowledged that only a court could finally decide whether a charge was unfair or not, and reported that its views had not generally been accepted by most of the eight credit card issuers consulted.

113

agreements with consumers such as those for bank accounts. On 7 September 2006 the OFT published a press release saying that, in response to the OFT’s statement of principles, credit card issuers had agreed to reduce their default charges—the majority by almost half. It also said that it would undertake further work on the application of these principles to PCAs, working closely with the FSA and the BBA, and taking into account the concerns about the PCA market expressed by the GCCNI. Following further fact finding by the OFT into unauthorized overdraft charges, the OFT announced on 26 April 2007 the launch of a market study into personal bank current account pricing, alongside a formal investigation into the fairness of charges for unauthorized overdrafts and returned items. The OFT said that the study was designed to help it to consider the current level and incidence of the charges in the broader context of efficiency, transparency, value and consumer choice within personal current accounts.160

4.284 It has been suggested that recent media attention, the campaigns from Which? and GCCNI, and the OFT’s work on credit card default charges and overdraft charges, have raised public awareness of unauthorized overdraft charges, and promoted customer resistance to such charges, and increased customer complaints, and switching.161 This might cause banks to be particularly sensitive to public perceptions of unauthorized overdraft charges, and high charges may lead to adverse publicity for the banks. However, we note that so far such charges have not generally been cut or eliminated, nor have banks told us that they plan to do so, suggesting that this factor is not acting as a significant constraint on charging levels. As discussed in 4.239 it appears that unauthorized overdraft charges are a push factor in causing switching, but we have seen no indications that customer perceptions are driven by how their bank’s charges compare to those of competitors, and nor do customers seem to consider charges when choosing who to bank with, thus reducing any incentive for banks to compete on charge levels. BoI has increased its unauthorized overdraft charges, albeit to levels that remain below those of many competitors, as part of its new PCA charging structure launched in November 2006. It remains to be seen whether there is a permanent change in the public perception and acceptance of these charges and whether adverse publicity may motivate any changes in charging levels and policies.

4.285 We discussed the recent changes to PCA charging structures in paragraphs 4.140 to 4.161. The changes in charging structure have made the Northern Ireland PCA market more similar to the PCA market in Great Britain; differences have also been partly eroded with the growth of the non-clearers, and the entry of a major British high street bank in the form of HSBC and the rebranding of Woolwich branches as Barclays.162 However, we have not taken the market in Great Britain as representing a competitive benchmark since we have not undertaken an investigation of the PCA market in Great Britain.

4.286 Some of the changes in pricing structure may be responses to the investigation itself, or to greater customer awareness associated with it, rather than necessarily a response to more permanent changes in market conditions. We did not find that

160OFT press releases 54/07 and 67/07. 161[] in its second hearing said that: The research seems to suggest that customers have been tolerant of unauthorized charges because they are regarded as penalizing them legitimately for things they have done wrong and they have been prepared to tolerate it, but I think that there is a growing head of steam from consumer groups and in the press that unauthorized charges are at too high a level from many providers … If we are looking forward, our view would be that the customer perception of the fairness or appropriateness of the penalty charges is beginning to turn. 162Barclays told us it did not accept that the rebranding of Woolwich branches as Barclays would contribute to the erosion of differences between the markets in Northern Ireland and Great Britain, as Woolwich had previously offered the same PCA proposition throughout the UK.

114

there has been a significant, widespread and permanent change in customer behaviour to make them more aware of competitive offers in the market and more inclined to switch bank. However, we recognize that the clearers are starting to compete more actively, and it is unlikely that the clearers will be able to return to the traditional pricing model having introduced fee-free accounts. We note that the changes are designed to align the clearers’ pricing structure more closely with the non-clearers by moving to a fee-free model. Thus charges will, in the revised models, be focused on a smaller group of customers who run unauthorized overdrafts, rather than a larger number of customers who run authorized or unauthorized overdrafts. Only one clearer, Northern, has so far increased its level of credit interest above the standard 0.1 per cent offered by the other clearers. Northern’s new pricing strategies appear to differ from those of the other clearers, Northern has adopted products similar to those offered by Danske elsewhere in Europe; it seeks to offer customers packages of products, there are some limits on free use of non-automated transactions, and credit interest rates can be above those of rivals.163

4.287 In summary, it appears to us that, despite the recent changes in behaviour that represent a shift in competitive positioning by the clearers, there remain several characteristics of the market which mean that all the banks, clearers and non- clearers alike, continue to have the ability to exercise unilateral market power. While the changes in the clearers’ behaviour are substantial, we do not think they are sufficient to demonstrate that the market is now functioning well, because we do not consider the linkage between bank conduct and customer switching to yet be sufficiently strong. Whether and in what way the banks choose to exercise their market power will depend on the particular incentives facing them. However, where it is profitable to do so, we expect banks to take advantage of such market power, whether by imposing higher charges or levels of debit interest, paying lower levels of credit interest, and/or by offering lower levels of service or less innovation than would otherwise apply.

Coordinated effects

4.288 We also looked at the possibility of coordinated effects in the market. We considered whether the conditions existed to facilitate tacit coordination between the banks. Where markets are sufficiently concentrated, the actions of individual firms can have identifiable effects on their competitors, such that firms recognize their interdepen- dence and act accordingly. Thus, through mutual self-awareness of interdependence, banks might find that if each of them adopts a less competitive strategy, they could all enjoy higher prices or lower costs (eg because of a reduction in innovation, marketing spend or service quality). This would lead to higher profits than would otherwise have been the case, or, alternatively, greater scope to sustain inefficient operating practices.

Conditions for coordination

4.289 The three conditions necessary for coordination are set out in our guidelines.164 We believed that any potential coordinating group would be likely to consist of the clearers, given the different business models and incentives of the non-clearers. We also noted that while the clearers told us that the main competitive pressures came

163We note that, at a very late stage in the investigation, First Trust announced that it intended to launch a new PCA in June 2007, offering customers the opportunity to earn credit interest of 4 per cent on credit balances up to £1,200 subject to a monthly minimum deposit of £1,200. 164Competition Commission Guidelines on Market Investigations References, CC3, June 2003, paragraphs 3.62 to 3.64.

115

from the non-clearers, the pricing papers we saw indicated that, at least until recently, the clearers monitored and set prices to be in line with the other clearers.

4.290 The clearers argued that there was no clear distinction between the clearers and non-clearers, in particular in terms of MORI market share between the smallest of the clearers (BoI, 12 per cent in 2006 Q4) and the largest of the non-clearers (Halifax, 9 per cent in 2006 Q4). BoI told us that, in many respects, it considered itself to resemble the non-clearers rather than the clearers. We found, however, that despite the relatively small difference in MORI market share between BoI and Halifax, significant differences remain between the clearers and the non-clearers in terms of business model and incentives (see paragraphs 4.3 to 4.13).

4.291 The first condition necessary for coordination requires the market to be sufficiently concentrated and transparent for the clearers to be aware of the high interdependence of their strategies and be able to monitor each other’s behaviour.

4.292 The clearers said that the market for PCAs in Northern Ireland was insufficiently concentrated for this first condition to be satisfied. They noted that the market was only moderately concentrated (HHI of 1,526 based on MORI data) and that concentration had fallen over time as the non-clearers had made significant gains in recent years. In addition, two of the clearers ([]) said that the market was not transparent due to the large number of different charges and interest rates and the existence of different products such as packaged accounts.

4.293 However, the potential coordinating group consisted only of the four clearers. The HHI measure quoted in paragraph 4.292 is for the market overall. We found that the clearers are aware of each other’s prices and behaviour, and they can, and do, monitor each other’s behaviour. We have seen that competitors’ levels of individual charges and rates are carefully considered when they make their own pricing decisions. Despite the difficulty of understanding the way in which the charges are applied, we found that headline charging rates were well-known and we conclude that the market is sufficiently concentrated and sufficiently transparent for the first condition to be satisfied.

4.294 The second condition requires the clearers to have sufficient incentive to coordinate and to sustain coordination over time. We looked at whether the clearers’ incentives were aligned, and the possible costs of deviation.

4.295 The clearers said that their incentives were not aligned due to differences in their size and in their market performance (with Northern and First Trust in particular losing market share).

4.296 Whilst there are differences in size between the clearers, and two of the clearers have experienced a decline in market share, it is not clear that these differences are sufficient to prevent parallel behaviour. The clearers have sufficiently similar business models to lead them to have similar incentives (see paragraph 4.6).

4.297 We recognize, however, that customers are relatively unresponsive to changes in charges or interest rates in the market (see paragraph 4.208). Given the low propen- sity of customers to switch between different banks, the cost of deviation from a coordinated equilibrium is likely to be low as the remaining firms would have little incentive to cut prices in response. Low switching rates also mean that banks would not have to build the likely response of their competitors into consideration when choosing their competitive behaviour, as their ability to retain customers would not be significantly affected by whether competitors adopted an aggressive or a conciliatory response. Therefore there appears to be no incentive to coordinate in the first place

116

as banks can exploit their market power in a unilateral manner without dependence on a coordinated response from competitors. As a result we conclude that the second condition for coordination is not likely to be met.

4.298 The third condition requires there to be relatively weak competitive constraints from outside the coordinating group such that coordination is sustainable. In this market, we believe that the expansion of the non-clearers is the primary competitive con- straint to the clearers.

4.299 The clearers said that the non-clearers represented a serious competitive threat and had made substantial gains in market share in recent years. Their view was that attempts at coordination would provide the non-clearers with further opportunities for expansion. We note that the market share of the non-clearers has been growing. However, we thought that there were barriers to expansion of the non-clearers which may limit the competitive constraint that they pose in the future. These include, in particular, customers’ reluctance to switch and their preference for a known brand. In addition, the non-clearers are not focused on expansion in Northern Ireland (see paragraph 4.107).

4.300 Many of the non-clearers have been offering PCAs in Northern Ireland for almost 20 years and their growth has not been particularly rapid. We note that the clearers only began substantially to revise their traditional PCA charging structures in 2005.165 This suggests that past competitive pressure from the non-clearers was not an effective constraint on the clearers’ actions. It is not clear whether recent changes in pricing structure by Northern, Ulster and by BoI are primarily due to competitive pressure from the non-clearers, although we recognize that this may be a contributory factor. We cannot yet assess the full impact of these changes in charging structure. Since we concluded that the second condition was not met, we did not find it necessary to reach a firm view as to whether the third condition is met.

Evidence of past parallel behaviour

4.301 We note that there is some evidence of parallel behaviour by the clearers in the past and in particular prior to the supercomplaint. Our pricing analysis found a significant degree of similarity in some of the clearers’ charges on traditional PCAs, demon- strated by prices moving in parallel, the convergence of prices on similar levels, and a general absence of price reductions (see paragraphs 4.166 to 4.168). We also found documentary evidence of an awareness of the need for price alignment with the other clearers and conscious price-following behaviour. In an internal pricing paper, one clearer ([]) said, in relation to proposed pricing changes for 2003/04: ‘[] traditionally lead the price changes and I would recommend that as in the past we follow their lead. This makes us joint highest with the [] on some fee and service income lines.’ Prices were not set with similar regard to the non-clearers, even though we have been told by the clearers that these banks posed a major competitive constraint.

4.302 However, we note that there was no consistent price leader. While clearers seem to have followed price increases on transaction charges quite closely, they did not always follow price increases on other charges for a substantial period of time, and

165Some banks said that they had introduced changes to pricing structures before 2005. For example, Ulster and First Trust had introduced packaged accounts. Northern introduced ChequeMaster in early 1990s; this has no transaction charges but has a fixed charge for overdrawn accounts. Northern also introduced the Principal account (1999, for high net-worth customers), the student package in 2002 and the graduate package in 2003, none of which incurred transaction charges. The retiree, student and graduate packages did not charge for the set-up, amendment or cancellation of direct debits or standing orders.

117

the originator of the price increase did not respond by reversing the increase (for example, First Trust’s increase in maintenance charges in 2001, and Northern’s increase of direct debit and standing order set-up charges in 2001). It may be that price following only occurred on certain charges, but it is not apparent why the patterns of parallel pricing behaviour were more evident on some charges than on others. The financial information available to the banks and provided to us was not sufficiently detailed to enable us to assess whether price following was more evident on those charges of greater financial significance to the banks. In addition, we found no evidence of parallel behaviour on other PCAs, such as packaged PCAs.

Recent behaviour

4.303 More recent changes in PCAs by some of the clearers suggest that any parallel behaviour that may have occurred in the past has broken down. These recent changes include:

(a) the introduction in February 2005 by Northern of Current Account Plus and the subsequent introduction in April 2006 of its new account portfolio;

(b) the introduction by Ulster of a fee-free charging structure;

(c) the introduction by BoI of a fee-free charging structure;

(d) the announcement by First Trust of the launch of a fee-free PCA in June 2007.

4.304 The implementation of these changes means that, from June 2007, the clearers will, in general, no longer levy transaction and maintenance charges when a customer on one of their new PCAs is in authorized overdraft.166 The majority of those charges on which there may have been evidence of parallel behaviour in the past are therefore no longer charged by these banks, at least to new customers.

4.305 It is possible that, if the underlying conditions which facilitated parallel behaviour in the first place remain unchanged, parallel behaviour may re-emerge, possibly on different charges. However, we believe this to be unlikely, given our conclusion that the second condition for coordination is not met because of the low propensity of customers to switch.

Conclusions on coordinated effects

4.306 In summary, we conclude that overall the conditions for sustained coordination are not met, principally because the clearers do not have an incentive to coordinate due to the low propensity of customers to switch between banks. For coordination to be possible it is necessary that all three conditions set out in the CC guidance are met.

4.307 There is evidence of some prices moving in parallel in the past, in particular in the period prior to the supercomplaint. However, recent changes in charging structures by the clearers suggest that any parallel pricing that might have taken place in the past has now broken down.

166Although those customers on older Northern accounts (IFCA and ChequeMaster) continue to pay charges when in authorized overdraft, and heavy users of certain services (cheques, cash withdrawal over the counter) continue to incur charges under Northern’s new account portfolio. Similarly, customers on all of First Trust’s older accounts will continue to pay charges when in authorized overdraft, and there are conditions associated with fee-free banking on First Trust’s new PCA.

118

Local competition

4.308 In our Emerging Thinking and the associated working paper on local competition, we considered whether there might be local competition concerns in the PCA market in Northern Ireland. We looked at whether there were many areas where there were few or only one of the banks represented; the extent to which the clearers face each other in local markets; and the extent to which the clearers face the non-clearers in local markets.

4.309 We did not identify particular concerns in any of these areas, and have received no evidence since then to cause us to change our view. Our findings on local competition are as follows:

(a) Northern and Ulster have a considerable number of branches without local competitors because of their denser branch networks.167 However, we found no evidence to suggest that the clearers choose to locate their branches to avoid head-to-head competition.

(b) Many of the clearers’ branches are located close to those of non-clearers. BoI and First Trust branches face a Halifax and/or a Nationwide branch in more than half of their branch locations. Nearly half their branches face an Abbey branch as well.168 Proportionate exposure to a specific non-clearer is lower for Ulster and Northern as a result of their greater number of branches.

(c) The market shares of the five largest non-clearers varied from 30 per cent in Greater Belfast to 16 per cent in Londonderry/Derry (based on MORI data), with an average of 24 per cent in urban areas and 19 per cent in rural areas. GCCNI noted that 41 per cent of the Northern Ireland population lived in smaller towns and rural areas where there were not likely to be any of the non-clearers, and 26 per cent of the population did not have access to a car. However, Ulster submitted evidence to suggest that most Northern Ireland customers can reach one of the five largest non-clearers within a 10- to 15-minute drive-time, and that only 11.6 per cent of PCA customers in Northern Ireland do not have access to a car.169 In addition, some PCAs can be accessed through Post Offices. In May 2004 there were 568 Post Office branches in Northern Ireland.

(d) The banks’ strategy and offers are set primarily on a national basis. Although there is some local flexibility, particularly with regard to setting overdraft limits and rates and waiving or refunding certain charges, we saw no evidence to suggest that banks alter their behaviour on a local basis according to the nature of local competition.

4.310 We also considered whether the market may be segmented along religious lines. We found that the extent to which the different banks are in head-to-head competition does not appear to depend on any possible religious associations. We considered whether branch location correlated with population distribution by religious grouping. Our analysis suggested that in most cases there are likely to be few instances of religious divisions strongly separating local markets and impacting on local market

167Northern faces no local competitors in 28 per cent of its locations, and Ulster in 17 per cent of its locations. 168BoI faces Halifax in 22, Nationwide in 21 and Abbey in 18 out of its 37 locations. First Trust faces Halifax in 26, Nationwide in 22 and Abbey in 19 of its 41 locations. 169Two of the clearers ([]) suggested that 10- to 15-minute drive-times were likely to be conservative estimates of the time customers were willing to travel to reach their bank of choice. We note that a trip to the bank will often be part of another journey (work, shopping etc), which may also increase customers’ options for accessing branches. Data on car ownership from MORI.

119

boundaries. Although the historic location of branches means that certain branches will be the local provider to certain sections of the community, we found no evidence to suggest that this had any significance for overall competition in the market. Our Millward Brown survey found that while all respondents were aware of perceived religious associations for the individual banks, none attached any weight to this in their choice of bank. The ORC survey suggested that religious associations are unlikely to influence customers’ choice of bank.

4.311 We conclude that there are no local market issues in the Northern Ireland PCA market that raise any competition concerns.

Other indicators of the extent of competition

Financial performance and profitability

4.312 We look at profitability in market investigations as a possible indicator of the extent of competition in the market. Although high profits can occur in competitive markets, competitive pressure should result in profit levels moving towards the cost of capital in the medium to long run. Therefore profits which are persistently and substantially in excess of the cost of capital for firms which represent a substantial part of the market could be an indication of limitations in the competitive process. Alternatively, normal or low profits may conceal ineffective competition where firms with market power are able to sustain higher costs than would be possible in a more competitive market.170

4.313 We reviewed the financial performance of the eight largest banks in Northern Ireland: the four clearers—BoI, First Trust, Northern and Ulster—and the four largest non- clearers—Abbey, A&L, Halifax and Nationwide. Together these provide over 95 per cent of PCAs in Northern Ireland. We note that Nationwide is a building society with mutual ownership and it passes benefits on to its members via pricing benefits.

4.314 A significant proportion of each bank’s operating processes and costs, including the operation of its branch network and centralized head office and group functions, is shared across its businesses and products. The clearers do not routinely assess the profitability of their PCA business alone, looking instead at the profitability of a wider range of their retail business in Northern Ireland. The non-clearers assess the profitability of their PCAs, but for the UK overall. All banks were asked to allocate revenues and shared costs to their PCA businesses in Northern Ireland. Whilst the allocation of revenues presented the banks with some difficulties, these were not as significant as the difficulties the banks faced when allocating costs. We recognized that there are significant conceptual and practical difficulties in establishing an appropriate basis for allocating the many shared costs that may be attributable to the banks’ PCA business in Northern Ireland, and that any allocation and apportionment of costs is subjective and dependent on the assumptions made. Nevertheless, we examined the revenues, costs, and resultant profits, allocated to the Northern Ireland PCA business.

4.315 Initially we considered asking the banks for a financial analysis of PCAs for the five years to 2004. However, many of the banks told us that, due to system changes and limitations, they would be unlikely to be able to provide the information we required for the earlier years. We therefore asked the banks to estimate the revenues, costs and capital employed in their PCA business in Northern Ireland for the three years to

170CC3: Market Investigation References: Competition Guidelines, June 2003, paragraphs 3.81 to 3.85.

120

2004. This required the banks to apportion or allocate revenues, costs and capital to the Northern Ireland PCA business in order to estimate profitability and understand financial performance. We subsequently asked them to update the revenue information for 2005. Appendix 4.10 sets out the financial performance of the eight banks’ PCA businesses. Appendix 4.11 discusses cost allocation and return on capital. In drawing conclusions based on the information provided by the banks, we note the considerable difficulties that they had in providing robust information on the PCA business in Northern Ireland (see paragraph 2.66). Later in the investigation we requested certain information for 2006 in order to understand the impact of the changes to PCA charging structures. However, the 2006 information does not cap- ture the effect of all the changes (for example, BoI changes took place in November 2006). Furthermore, First Trust was unable to provide the requested information in the required timescales.

Financial performance

4.316 Customers in Northern Ireland held 1.3 million PCAs at these eight banks in 2005, of which 1.0 million were active accounts. The number of accounts has grown at an average of 3 per cent a year since 2002. However, whilst the number of accounts at the clearers as a group has remained relatively stable, the non-clearers as a group have experienced average growth of 9 per cent a year over this period. This is reflected in relative market shares (see paragraphs 4.14 to 4.40).

4.317 Credit balances provide a bank with funds that it uses elsewhere in the bank to earn income. Average credit balances in 2005 were £2.1 billion, which represents average annual growth of over 10 per cent since 2002. Individual banks experienced average annual growth in credit balances of between 8 and 23 per cent over the same period. Average credit balances per account are higher for the clearers as a group (2005: £1,796) than for the non-clearers (2005: £1,184). Average overdraft balances in 2005 were £155 million, equivalent to approximately 7 per cent of the value of credit balances. Average debit balances per account are higher for the clearers as a group (2005: £142) than for the non-clearers (2005: £65).171

4.318 In 2005 the eight banks earned income in total of £167 million from providing PCAs, which represents nearly 8 per cent a year average increase since 2002. During this period the clearers experienced a growth in revenue of over 5 per cent on average a year whilst the non-clearers experienced growth of 19 per cent on average a year.

4.319 On average the banks earn more than half their total income from NII on credit balances, over 40 per cent from charges and 5 per cent from NII on debit balances. The relatively high proportion of total income represented by NII on credit balances reflects the significant benefit to the banks of paying low average credit interest rates. In 2005 the eight banks earned £92 million from NII on credit balances, £66 million from charges and £8 million from NII on debit balances. On average, the clearers earned total revenue per account of £145. The non-clearers earned significantly less; total average revenue per account was £92.

4.320 We asked the banks for the breakdown of their income from charges, but they had difficulties in allocating their revenues consistently. We sought, in particular, to understand the importance of unauthorized overdraft charges as a source of income

171Average balance per account is calculated by dividing the total balance by the total number of PCAs. We do not have information on the average number of accounts which are in credit and those which are in debit. The credit balance per account does not, therefore, represent the average credit balance on the accounts which are in credit and, similarly, the debit balance per account does not represent the average debit balance on the accounts which are in debit.

121

for the banks as a whole. Although there is some uncertainty surrounding the estimate, it was clear that unauthorized overdraft charges are a key source of income. Based on the information provided by the banks, it appears that the eight banks earned £31 million in 2005 from unauthorized overdraft charges, representing 47 per cent of all charges. This is likely to be an underestimate because the analysis from some of the banks allocated no income to unauthorized overdraft charges. We estimate that a more realistic figure for 2005 might be around £35 million, represent- ing over half of all income from charges.172

4.321 We looked at the income that banks earn directly from PCAs (charges, gross debit (overdraft) interest paid by customers less credit interest paid to customers), compared with imputed income earned from the provision of PCAs but from non-PCA products. This income is earned from using the net credit balances from PCAs173 elsewhere in the bank. In 2005 the eight banks in total earned 45 per cent of the total income earned from the provision of PCAs directly from PCAs and 55 per cent indirectly by using the net credit balances elsewhere in the bank.

4.322 We sought to understand the revenues that a bank earned directly from PCA customers and whether and to what extent these varied between banks. The revenues a bank earns directly from customers comprises revenues from charges, gross interest income on debit balances less the interest paid to customers on credit balances.174 Because the size of the banks’ PCA business varies significantly, we examined the direct income from PCAs per account. This analysis does not attempt to represent what an individual customer pays for their PCA as this will vary significantly depending upon how the customer uses the PCA. It does, however, show the cash flow between a bank and its customers, and therefore how much, on average, customers pay for their PCA.

4.323 Initially we considered the revenues as submitted by the banks, and noted that from 2002 to 2004 there was a distinct difference between the unadjusted average income earned directly from a PCA by the clearers as a group (£62 to £68 per account) compared with the non-clearers as a group (£25 to £26 per account). In 2005 the clearers on average earned £66 compared with £39 for the non-clearers, but this masked some differences between the individual banks, and some non-clearers earned the same as some clearers.

4.324 However, we note that there may be a difference in customer behaviour between banks, and in particular average credit and debit balances differ between banks. We therefore ‘normalized’ the individual banks’ average credit and debit balances per account by adjusting them to the overall average for the eight banks. We then recalculated the interest debit and credit balances by applying the individual banks’ average interest rates to the normalized balances. The results of this are shown in Table 12.

172We note that, given the recent and future changes in the clearers’ charging structures, unauthorized overdraft charges are likely to become a more important source of income in the future. 173Credit less debit balances. 174We consider the gross interest paid on credit balances. Depending on the customers’ tax situation this income may be taxed.

122

TABLE 12 Average direct revenues earned per account (interest adjusted for differences in average balances)

£ 2002 2003 2004 2005 2005 analysis Debit Credit Total Total revenue Charges interest interest revenue

Clearers 65 66 66 64 54 12 –2 64 Non-clearers 28 26 27 41 43 14 –16 41 Weighted average 56 56 56 58 51 12 –5 58

Source: CC analysis of responses and updates to Questions 4 and 5 of the Financial and Accounting Questionnaire.

4.325 We noted that from 2002 to 2004 there was a distinct difference between the average income earned directly from a PCA by the clearers as a group (£65 to £66 per account) compared with the non-clearers as a group (£26 to £28 per account). In 2005 the clearers on average earned £64 compared with £41 for the non-clearers, but this masked some differences between the individual banks, and some non- clearers earned more than some clearers.

4.326 We note that there are some issues associated with using the analysis of average revenue to assess the effectiveness of competition in the PCA market in Northern Ireland:

(a) Differences in average revenues per account between banks may reflect factors other than price. For example average revenues may reflect banks attracting different types of customers, or some banks offering accounts with maintenance charges and additional benefits. Customers also use their accounts in different ways.

(b) The concept of average revenue per account does not attempt to reflect the situation of an individual customer. At any point in time, a customer does not operate both a debit and credit balance on a single PCA; therefore to attach a revenue for both in the calculation does not reflect the ‘average’ customer. For example, the majority of customers operate their account in credit and therefore do not pay any charges in relation to overdrafts or debit interest.

4.327 However, we considered that analysis of average revenues was appropriate because after adjusting for a significant difference in customer behaviour between banks (the average account balance) and omitting the subjective internal valuation of funds, there remained notable variations between banks and groups of banks. We therefore considered that despite the limitations and in the absence of an alternative, this analysis of average revenues was informative.

4.328 Although we were able to adjust interest for differences in average balances, we were unable to adjust charges for differences in services offered or customer behaviour. We note that, whilst there are many similarities between the banks’ PCAs, there are some important differences between banks:

(a) Services differ—some accounts are ‘packaged’ with other benefits, for example. The size of the branch network differs significantly between banks (see paragraph 4.53). The clearers offer services not provided by the non-clearers such as night safes and safe custody. Some banks also offer free ‘customer reviews’.

(b) Different PCAs generate income in different ways including transaction charges, maintenance charges and unauthorized overdraft charges.

123

(c) Customer profiles (in terms, for example, of age or socio-economic group) differ between banks. For example, we understand that one clearer ([]) has a greater percentage of older customers than the banks’ customers as a whole. Different customer groups are likely to use their PCAs in different ways.

4.329 Much of this historical analysis pre-dates the recent changes to charging structures (Ulster’s new charging structure, Northern’s new account portfolio and BoI’s new charging structure). Furthermore the 2005 analysis contains eight months of Northern’s Current Account Plus.

Operating costs

4.330 The banks faced significant issues when trying to estimate the costs of providing PCAs in Northern Ireland:

(a) The majority, if not all, of the operating costs associated with providing PCAs are shared with other services and customers. For example, clearers’ branches generally serve both personal and business customers, and provide a wide range of financial services, of which PCAs are only one part.

(b) The banks viewed their banking business as one with a fixed cost base to which revenues from a portfolio of products contribute. Conceptually, PCAs are not a stand-alone banking product—the essence of banking is to take deposits and then lend the money out.

(c) Customers normally use multiple banking products, for example savings accounts, mortgages, personal loans, credit cards as well as PCAs. While customers can source these products from different suppliers, the general business model of banks is to seek to cross-sell products to customers.

Similar issues arose when we considered the allocation and apportionment of assets.

Return on capital and profitability

4.331 We intended to compare the return on capital (RoC) to the cost of equity. This con- trasts with the return on capital employed approach (ROCE) in which returns to both debt and equity holders are considered and compared with the cost of holding debt and equity, ie the weighted average cost of capital (WACC). Although the two approaches are mathematically equivalent, it is normal to assess financial insti- tutions’ profitability on the basis of return on equity capital.175 We prefer this approach conceptually as we consider that borrowing and lending money is at the core of these institutions’ business operations and is not just a pure financing function as it is for most non-banking businesses.176 Their main source of revenue is interest income and one of their expenses is interest payable. This is reflected in the presentation of interest expense as an operating item in the statutory accounts of banks. Aside from our own conceptual preferences, return on equity capital is widely used within the sector as a performance measure.

4.332 The analysis presented in Appendix 4.11 suggests that for some banks in some years PCAs are loss-making. However, given the interdependency of PCAs and

175For example, the CC adopted this approach in its report on The supply of banking services by clearing banks to small and medium-sized enterprises within the UK, March 2002, Cm 5319, and in its report on the Store Cards Market Investigation, March 2006.These can be found on our website at www.competition-commission.org.uk/inquiries/completed/index.htm. 176See also Copeland, Thomas E, Tim Koller and Jack Murin, Valuation (1996).

124

other banking products, and the practical difficulties in identifying appropriate and accurate bases for cost and capital allocations, we are unable to measure profitability with a sufficient degree of accuracy and hence were unable to conclude as to whether the banks are making returns in excess of their cost of capital for their PCA businesses.

4.333 It is the usual CC approach to calculate a cost of capital in order to assess the return on capital. However, given the difficulties associated with calculating a return on capital, we have not sought to calculate the cost of capital.

Efficiencies

4.334 An analysis of efficiency would normally focus on an analysis of costs. In the banking industry, a commonly used indicator of efficiency is the cost/income ratio. However, given that the PCA cost figures in particular rely to a large extent on subjective assumptions about allocations, we attempted to reach more reliable conclusions by focusing on PCA revenues, and other volume-based data.

4.335 Our analysis focused on how the banks monitor and improve their efficiency. In doing so we have looked at how the banks operate their branches in Northern Ireland and in particular the effect that any differences between the banks’ branch structures has on their efficiency.

4.336 We sought to understand how the banks’ management monitor the efficiency of their PCA businesses in Northern Ireland. In the absence of reliable data on cost/income ratios, we sought the following data:

(a) benchmarking of their Northern Ireland PCA businesses against other regions in the UK;

(b) comparison of the efficiency of their different products;

(c) activity-based costing of their PCA activities;

(d) the split of their activities between PCA/non-PCA business, and between back- office and front-office operations; and

(e) revenue or contribution from branches for both their PCA and retail banking businesses.

4.337 The banks were unable to provide us with answers to detailed questions on efficiency sufficient for us to draw conclusions from them. We noted that, whilst the banks may monitor their efficiency for their retail business as a whole, they do not separately monitor the efficiency of their PCA business in Northern Ireland on a routine basis.177 In particular, there is no benchmarking of their efficiency either internally within the banks’ Northern Ireland PCA businesses or against other entities except at the highest level of bank cost/income ratios. However, the clearers are all seeking to improve the efficiency of their PCA businesses: [].

4.338 Northern Ireland is less important to the non-clearers due to its relatively small size compared with the UK as a whole. Although some of the non-clearers monitor their

177[] told us that whilst it may not have been able to respond to some of our specific questions, it does monitor the efficiency of its PCA business. In particular, it highlighted KPIs for [].

125

PCA performance on a regional basis, in all cases their Northern Ireland PCA business is included within a larger region for management purposes.

4.339 We also considered whether the lack of cost information might be an indicator of ineffective competition. We noted that if the banks had better information on costs they may be able to allocate resources within the bank more efficiently. Banks might also seek to gain a competitive advantage by lowering prices where prices were above costs. However, PCAs perform many functions which are charged for in a variety of ways, and some functions are provided without charge. Therefore it would be unrealistic to expect every aspect of charging necessarily closely to reflect the costs of performing the associated service, even if it could be calculated. Further- more, we also noted that detailed cost information might be costly to obtain. Further details of our analysis of efficiencies are set out in Appendix 4.12.

5. Findings and features

5.1 In this section we summarize the key elements of our findings, cross-referring to the sections of our report where we have discussed these elements in more detail. We first summarize our findings on the competitive position in the market. We then set out some common features that, we believe, are the underlying causes of some of our findings.

5.2 We set out the context for these findings in section 2 (overview of the sector), section 3 (market definition), and the first part of section 4 (market structure) of our report.

Findings

5.3 In summary, our findings are as follows:

(a) There is a general lack of customer interest in PCAs and customers tend to view PCAs as ‘all the same’ (see paragraphs 2.73 and 4.223).

(b) Customer perception is that switching PCAs is much more difficult and risky than it is in practice. Some problems do nevertheless arise, despite the success of the Code in relation to switching (see paragraphs 4.217 to 4.222).

(c) In the present state of competition in the market, the financial incentives to switch are unlikely to outweigh the perceived risks for most customers (see paragraph 4.224).

(d) Customer understanding of unauthorized overdraft charges appears to be low; some are not aware of when they are charged; and few know the levels of their banks’ charges or how they compare with competitors (see paragraphs 4.228 to 4.247).

(e) Annual rates of switching in the PCA market in Northern Ireland as a whole are between 1.5 and 4 per cent. There is a lack of responsiveness to changes in charges or interest rates in the market (see paragraphs 2.78 to 2.93).

(f) A customer’s decision to switch is more often prompted by dissatisfaction with their existing bank than the recognition of a better offer elsewhere. Dissatisfaction often arises from the bank levying unauthorized overdraft charges; poor service, unhelpful staff, and bank errors are examples of other causes of dissatisfaction which may prompt switching (see paragraphs 4.209 to 4.211).

126

(g) The single most important factor in the choice of new bank is access to a branch; personal recommendation and, to some extent, levels of interest and charges are also important (see paragraphs 4.209 to 4.216). However, customers have, up to now, been unlikely to compare and assess such charges when choosing a new bank (see paragraphs 4.241 to 4.247).

(h) There is an inherent complexity in PCA charging structures, in part because PCAs service a wide variety of needs. However, even recognizing this, PCA charging structures and practices are unduly complex, particularly for authorized overdraft charges associated with traditional PCAs and for unauthorized overdraft charges levied on all PCAs. The incidence and frequency of charging on PCAs is not clear. There are often exceptions or discretionary items which are not visible to customers and add to complexity (see paragraphs 4.120 to 4.131).

(i) A bank is unusual in both providing a PCA service to the customer, and also having access to the customer’s account which is used to pay for the service (see paragraph 4.121).

(j) The banks, particularly the clearers, describe their charges using terminology that is unclear and, in many cases, inconsistent between banks. This is particularly true of unauthorized overdraft charges and, to some extent, ancillary charges. In general, customers are not provided with the necessary information to enable them to have a sufficient understanding of the charges and interest rates that might apply to their PCA (see paragraphs 4.132 to 4.139).

(k) Charges and credit interest rates on traditional PCAs showed similarities in pricing that are unlikely to be explained by costs since, in most cases, banks do not know their costs of providing particular PCA services. Nor are they likely to be explained by high levels of competition in the market (see paragraphs 4.162 to 4.173 and 4.184).

(l) Unauthorized overdraft charges and debit interest rates are unlikely to be explained by costs. Charges and debit interest rates are likely to be above the levels that would apply in a well-functioning market (see paragraphs 4.174 to 4.178 and 4.185).

(m) Clearers’ customers pay more, on average, than non-clearers’ customers to operate their PCA, as evidenced by the total revenues earned directly per account (see Table 3 and paragraphs 4.322 to 4.327).

(n) There is a relative lack of competition, particularly among the clearers, on several non-price factors such as branch opening hours and the introduction of full- function Internet banking (see paragraphs 4.193 to 4.206).

(o) Clearers, in particular, whilst monitoring their business at a broader level, and despite some recent improvements, have limited management and financial information specific to their PCA business in Northern Ireland. This includes limited information on numbers and destinations of switchers; incidence of particular charges; activity-based costings; costs of acquiring new customers; benchmarking of the efficiency of different products; and levels of cross-sales (see paragraphs 2.66, 2.81 to 2.88, 4.116, 4.314 and 4.315, 4.337 to 4.339 and Appendix 4.3).

(p) Whilst we recognize the costs of an extensive branch network, clearers benefit from their networks, as well as their strong brands. They continue to attract a relatively high proportion of new-to-banking customers in relation to their share of

127

flow of all new accounts. Non-clearers are not looking to expand their branch networks in Northern Ireland in the foreseeable future (see paragraphs 4.29, 4.58 and 4.59, and 4.107).

(q) There was evidence of some prices moving in parallel in the past. However, the conditions for sustained coordination in the market are not met, principally because the clearers do not have an incentive to coordinate due to the low propensity of customers to switch between banks (see paragraphs 4.288 to 4.307).

(r) Any competition concerns are unlikely to have a local market focus (see paragraphs 4.308 to 4.311).

5.4 In many cases these findings are indicators or outcomes of a lack of competition in the market. The banks as a whole continue to be able to impose higher charges or levels of debit interest, pay lower levels of credit interest, and/or offer lower levels of service and innovation, in particular to existing customers, than might be the case if switching were more prevalent. Although the findings are caused by market-wide features (see paragraph 5.7), the findings may differ among individual banks depend- ing, for example, on the bank’s business model, ownership or strategy.

5.5 We note that many of the findings of the European Commission’s report from its Sector Inquiry on retail banking under Article 17 of Regulation 1/2003 (see paragraph 2.2) are consistent with the findings of our investigation. However, we did not use this report in forming our views since it was not specific to PCAs in Northern Ireland.

5.6 This investigation has addressed the PCA market in Northern Ireland. We recognize that many banks adopt UK-wide strategies, but we have not looked at whether similar issues may apply in the rest of the UK since this is outside the terms of reference of our investigation.

Features

5.7 A feature can take the form of the structure of the market and/or conduct on the part of the banks or customers (see paragraph 1.4). ‘Conduct’ includes any failure to act (whether or not intentional) and any other unintentional conduct. We consider if any feature prevents, restricts or distorts competition in the market for the provision of PCAs in Northern Ireland thereby having an adverse effect on competition.

5.8 We can consider either individual features or a combination of features of the market. In identifying any such features of the market, we seek to compare what we have found with those levels of competition which we might reasonably expect to find in a well-functioning market.

5.9 We found that the features within the meaning of the Act which prevent, restrict or distort competition in the PCA market in Northern Ireland are as follows:

(a) banks have unduly complex charging structures and practices;

(b) banks do not fully or sufficiently explain their charging structures and practices; and

(c) customers generally do not actively search for alternative PCAs or switch bank.

5.10 The first two of these features lead to difficulties in customers making properly informed choices. This applies both to new-to-banking customers in searching for an

128

appropriate PCA, and to existing customers who might otherwise switch to a bank that might provide a PCA to meet their needs at lower cost. This is exacerbated by the third feature: the perception and, to some extent, experience of difficulties in switching, as well as by customer indifference. Therefore banks compete less hard on price and service because they do not believe that they will lose a significant volume of customers, and customers pay higher charges and interest or receive less credit interest and/or lower levels of service or less innovation than would be expected in a well-functioning market.

5.11 We therefore find that, on the statutory questions that we have to decide pursuant to section 134(1) of the Act, there are features of the relevant market, either alone or in combination with each other, that prevent, restrict or distort competition in connection with the supply of PCAs in Northern Ireland, and hence that there is an AEC within the meaning of section 134(2). The features are those that we identified in para- graph 5.9.

6. Remedies

Overview

6.1 We now consider measures to remedy, mitigate or prevent the AEC, or resulting detrimental effect on customers, as set out in section 5. We first consider the AEC to be addressed and the framework for the remedies review. We then set out the details of each of the remedy options, including the costs associated with each option. Next, we consider the scope of the remedies, issues of implementation and the monitoring and enforcement of remedies. Finally, we discuss the package of remedies as a whole and the proportionality of that package in relation to the scale of the AEC, as well as relevant customer benefits.

6.2 On 12 October 2006, we issued the Notice, which invited comments on the actions that might be taken by the CC, or recommended for implementation by others, to remedy, mitigate or prevent the AEC or resulting detrimental effect on customers, identified in our provisional findings in relation to the market for the supply of PCA banking services in Northern Ireland. In the light of evidence received in response to the Notice, and our further analysis, we published our provisional decisions on remedies on 6 March 2007. This section of the report updates our provisional decisions on remedies to take into account the responses received to it.

6.3 We welcome many of the developments that are ongoing in the UK banking industry and which will affect Northern Ireland. We have taken care to understand those developments and to consider how they might interact with any package of remedies. Specifically, we are aware that the independent review of the Code is considering changes to the Code that may help to address our concerns (see paragraphs 2.3 and 2.107). Whilst recognizing that no decisions have yet been taken, we have worked closely with the BCSB. We have taken account of the possible changes to the Code that may result from the review in deciding what remedies to put in place and in deciding how our package of remedies should be implemented. The independent Code reviewer, Mike Young, on the basis of our provisional decisions on remedies, has indicated to us that he is currently minded to recommend that all our remedies should be incorporated into the revised Code. This is subject to him seeing our final report and any further representations that he might receive. We are also aware that the CCA 2006, the final provisions of which are expected to be implemented in October 2008, will bring changes to the banking industry—many (but not all) of which are expected to be incorporated into the Code. We have also taken the effects of

129

these changes, as far as they are known, into account in considering possible remedies.

Framework for decisions on remedies

6.4 Having identified a number of features of the market for the supply of PCA banking services in Northern Ireland that have an adverse effect on competition, we are required to decide the following questions under section 134(4) of the Act:

(a) whether action should be taken by the CC for the purpose of remedying, miti- gating or preventing the AEC concerned or any detrimental effect on customers so far as it has resulted, or may be expected to result, from the AEC;

(b) whether we should recommend the taking of action by others for the purpose outlined in (a) above; and

(c) in either case, if action should be taken, what action should be taken and what is to be remedied, mitigated or prevented.

6.5 In choosing appropriate remedial action, we have a statutory obligation to achieve as comprehensive a solution to the AEC and any resulting detrimental effect on cus- tomers as is reasonable and practicable.178

6.6 As noted in our guidance (CC3, paragraph 4.9), we will consider the effectiveness of different remedies and their associated costs and will have regard to the principle of proportionality when deciding on appropriate remedies. In considering whether a remedy is reasonable and practicable, we will consider the cost associated with implementing the remedy (CC3, paragraph 4.10). We will endeavour to minimize any ongoing compliance costs to the parties, provided the effectiveness of the remedy is not reduced (CC3, paragraph 4.12). However, we will balance those costs against the benefit to the economy and to customers in particular.

6.7 We will also seek to implement remedies or a package of remedies which are not disproportionate in relation to the AEC and any resulting detrimental effect on customers. If we are choosing between two remedies or packages of remedies which we consider would be equally effective, we will choose that which imposes the least cost or that is the least restrictive (CC3, paragraph 4.10).

6.8 We will also have regard to the effects of any remedial action on any relevant cus- tomer benefits within the meaning of section 134(8) of the Act arising from the adverse feature or features of the market concerned. Such benefits comprise lower prices, higher quality or greater choice of goods or services or greater innovation in relation to such goods and services. To qualify within the meaning of section 134(8), the CC must believe that the benefit would be unlikely to accrue without the relevant feature or features.

6.9 In general, we will seek to implement (or recommend) remedies that address the AEC, though we may also choose to address the detrimental effect on customers in addition or as an alternative (CC3, paragraph 4.6).179

178Section 134(6) of the Act. 179The CC has said (CC3, paragraph 4.22) that it ‘will first look for a remedy that would be effective in dealing with the adverse effects on competition of the market features rather than seeking to deal with any detrimental effect on customers’. However, the CC is prevented from taking action to address future (rather than existing) detrimental effect on customers if it is not also remedying the AEC (section 138(6) of the Act).

130

The need for remedial action

6.10 We recognize that, as set out in paragraphs 2.63 and 2.64 and 4.140 to 4.161, there have been significant changes in the PCA market in Northern Ireland in recent years. These include the following:

(a) The two largest clearers, Ulster and Northern, have both changed ownership since 2000 (now being owned by RBS and Danske respectively).

(b) Ulster, Northern and BoI have made significant changes to their PCAs in the last two years. On 3 May 2007, First Trust announced that it intended to launch a new PCA in June 2007.

(c) There is increased publicity about the possibility of switching and, in particular, increased awareness of issues surrounding unauthorized overdraft charges as a result of recent campaigns.

6.11 We conclude that, despite these significant recent changes in the market, the features outlined in paragraph 5.9 still persist and prevent, restrict or distort compe- tition in the market and that, without taking remedial action, the market as a whole will remain uncompetitive for the foreseeable future and that there will, as a conse- quence, be a detrimental effect on customers. We set out a range of possible remedies in our Notice and in our provisional decisions on remedies to address the AEC and detrimental effect on customers. We discuss each remedy option in turn in paragraphs 6.12 to 6.223.

Assessment of remedy options

6.12 Remedies 1 to 5 listed in the Notice dealt broadly with the quality, clarity and time- liness of the information provided to customers to enable them to take informed decisions. Remedies 6 to 10 listed in the Notice were aimed at addressing the difficulties perceived or experienced by customers in switching and/or prompting customers to consider whether switching would be appropriate. However, we note that those remedies that improve the quality, clarity and timeliness of information provided to customers may also help to address some of the difficulties customers perceive or experience in switching, if these result from a lack of understanding of how a customer’s PCA, or those offered by other banks, might work.

6.13 We believe that customers should have information that is sufficiently complete, clear and timely for them to make appropriate decisions. This might be particularly relevant, for example, when opening or switching PCAs or when in unauthorized overdraft. One of our key aims in deciding on these informational remedies is to enable customers to take informed decisions. This could influence the way in which customers operate their PCA or enable them to decide to switch their PCA. The remedies that we include in our package in relation to the switching process are intended to address customers’ experience or perception of the switching process as being risky or burdensome.

6.14 We discuss each of the remedy options set out in our Notice and further options put to us since its publication. Paragraphs 6.16 to 6.181 set out our discussion and decisions on the remedies that we have decided to pursue. We discuss the remedies that we have decided not to pursue in paragraphs 6.182 to 6.223. Taken together, our package of remedies, summarized in section 7, helps address the features that we have identified that:

(a) banks have unduly complex charging structures and practices;

131

(b) banks do not fully or sufficiently explain their charging structures and practices; and

(c) customers generally do not actively search for alternative PCAs or switch bank.

6.15 However, in addition to action by the banks, customers must themselves actively engage if they are to benefit from the emergence of a more competitive market for PCAs. They must use the additional information that these remedies will provide to understand better the cost of running their PCAs, to manage their PCAs better, and actively to consider whether they might benefit from switching their PCA. Consumer bodies also have an important role to play in influencing and empowering customers to choose and switch to the PCA that best suits their needs.

Remedies that we have decided to pursue

6.16 We have decided to pursue the following remedies:

(a) Remedy (a): Easy-to-understand terminology and descriptions of PCA services (see paragraphs 6.18 to 6.42).

(b) Remedy (b): Explanations of the levels of charges and interest rates and how and when they are applied (see paragraphs 6.43 to 6.65).

(c) Remedy (c): Information on statements (see paragraphs 6.66 to 6.86).

(d) Remedy (d): Summary and breakdown of charges and interest (see paragraphs 6.87 to 6.110).

(e) Remedy (e): Advance notice of charges and debit interest incurred (see paragraphs 6.111 to 6.146).

(f) Remedy (f): Regular ‘rights reminder’ (see paragraphs 6.147 to 6.160).

(g) Remedy (g): Changes to the switching process (see paragraphs 6.161 to 6.181).

6.17 We discuss each of these remedies in turn before discussing the remedies package as a whole. For each remedy option, we outline the proposed remedy, responses to the proposal, the reasoning which has led us to our decision, our decision, and the way in which the proposal addresses the AEC or the customer detriment. We discuss monitoring and enforcement in relation to the overall remedies package.

Remedy (a): Easy-to-understand terminology and descriptions of PCA services

6.18 In the Notice, we proposed that banks might be required to provide an easy-to- understand description of each technical term used (remedy 1). This would cover the specific service offered and/or the activity performed by the bank or the customer, and the related charges and interest rates, for each PCA.

6.19 In our provisional decisions on remedies, we clarified this proposal. We proposed that banks operating in Northern Ireland must satisfy the BCSB that all their customer communications in relation to PCAs are easy for customers to understand. In order to do so, we proposed that banks should ensure that all their written communications with customers with respect to the charges and interest rates that apply to PCAs were either:

(a) certified by an independent organization specializing in plain English; or

132

(b) otherwise tested with customers and found to be easily understandable.

This obligation would apply to all written communications with customers, whether in hard copy or electronic form, including all pre-contractual, contractual and post- contractual information in relation to PCAs.

6.20 We considered in particular:

(a) how best to ensure that the language used was genuinely easy for all customers to understand;

(b) whether it would be reasonable or desirable to require banks to adopt standard terminology to describe the services offered or the activities performed, and the charges and interest rates applied;

(c) how best to ensure that customers have ready access to these descriptions of services, charges and interest; and

(d) possible costs of implementation.

• The views of parties

6.21 Overall, parties expressed support for the provision of clear and easy-to-understand information written in plain English. GCCNI considered that easy-to-understand terminology could be achieved by requiring banks to provide evidence that they had consulted with user groups and had taken into account average literacy and numeracy levels of customers as well as the legibility of written information. Several parties ([]) suggested that banks could seek certification from third parties180 to ensure that their written customer communications were clear.

6.22 Although some parties supported the standardization of certain ‘core terms’, most parties were against the standardization of terminology on the basis that it could be very costly and time-consuming. They said that it would involve banks in Northern Ireland agreeing on a standard terminology whenever they developed a new service or changed some elements of an existing PCA service. Parties also argued that standardization of terminology could stifle innovation, acting as an impediment to the development and launch of new PCAs and other services. In this way, they argued that standardization of terminology might be anti-competitive.

6.23 With regard to accessing the information, one bank ([]) noted that a number of publicly available sources such as the FSA’s consumer website, the Code, and commercial websites (eg moneyfacts.co.uk) contained explanations of key financial terms.

6.24 We discuss the banks’ views on costs of implementation of this and the next remedy in paragraphs 6.50 and 6.51.

6.25 In their responses to our provisional decisions on remedies, three banks ([]) raised questions about which of the bank’s communications should be included in the scope of this remedy. One of them ([]) suggested that it should include only those documents and literature whose main purpose was to provide customers with

180Third parties proposed included the Word Centre, the Plain English Campaign and the Educational Guidance Service for Adults.

133

information on the level and application of charges, rather than all written communications with customers. This would cover: fee and interest brochures, PCA brochures, summary box information, the annual summary of interest and charges, and pre-notification of charges. Another of them ([]) similarly suggested that a series of broadly standard documents could be approved, with a carve-out for minor variations within those documents. Two of them ([]) said that requiring every piece of customer communication to be reviewed formally would be cumbersome, time consuming and expensive.

6.26 Also in response to our provisional decisions on remedies, another bank ([]) expressed concern that the process of approval for communications could delay the launch of new or revised products. It also suggested that the process of certification and/or consumer testing could result in breaches of confidentiality, possibly to the competitive disadvantage of the affected bank.

6.27 Most parties indicated that they regarded the Code and the BCSB as the most appropriate tool and forum to support the implementation of this remedy. [], for example, said that the BCSB should play an important part in the implementation and monitoring of this remedy because of its experience in monitoring banks’ compliance with the Code requirement for plain language.

• Our views

6.28 We noted that the Code already requires banks to give clear information to customers in ‘plain language’. This requirement has been in place for some time and we were told that many banks reviewed their customer literature to improve its clarity. Notwithstanding this, there is evidence that customers do not understand the charges and interest rates that apply to PCAs, and we found that there was scope for improvement in this area.

6.29 We considered whether to require standardization of at least certain key terms used to describe charges and interest rates in relation to PCAs. We thought that standardization could make it easier for customers to compare different PCAs. However, given the evidence set out in paragraph 6.22, we found that it was neither necessary nor desirable to require the standardization of descriptions of certain key terms. In our view, provided the charges and interest rates that apply to different PCAs are described in language that is easy to understand, customers should be able to make comparisons without terms being standardized.

6.30 However, we identified a number of key terms which we thought were particularly important for customers’ understanding of PCAs. These included terms relating to authorized and unauthorized overdraft services and charges. These might include, for example, terms indicating charges levied when the bank makes a payment which creates or extends an unauthorized overdraft (such as paid fee or item, paid referral, referral fee); terms used to indicate charges levied by the bank for returning cheques unpaid, or declining to make a payment that, if paid, would create or extend an unauthorized overdraft (such as unpaid fee or item, failed item); BACS; and CHAPS.

6.31 We considered how best to ensure that descriptions of charges and interest rates that apply to PCAs are in language that is genuinely easy for customers to under- stand. Whilst we did not wish to be overly prescriptive in how each bank should ensure clarity in its written communications with customers, we thought that these communications should be subjected to testing with customers or with bodies from outside the banking community. We considered the suggestion from one bank ([]) that the process of consumer testing communications might lead to a breach of confidentiality. However, we noted that some banks already test brochures and other

134

materials on consumers or with third party bodies and no one else had raised this concern with us. We considered that it would be possible for the banks to put in place sufficient safeguards to ensure that confidentiality was not breached.

6.32 Taking into account the responses received to our provisional decisions on remedies, we considered what communications should be covered by this remedy. We accepted that banks should not be required to satisfy the BCSB that every individual communication with customers is easy to understand. However, we were concerned that, by specifying particular types of communication (eg PCA brochures), we could leave scope for disputes about whether a particular communication was of the specified type.

6.33 We noted that we set out in relation to remedy (b) a number of instances when customers should be provided with these explanations: when choosing a PCA, when opening a PCA, on statements (see also remedy (c)), and when pre-notified of charges and interest payments. We thought that the requirement on the banks to ensure that their communications with consumers in relation to PCAs were easy to understand must apply to written communications with customers at these times. This would ensure that consumers would have information on the cost of running a PCA, and would have this information at key times. Although not required by this remedy, we would strongly recommend that the BCSB should encourage the banks to use the same process that they would use to comply with this remedy, to ensure that other communications with customers met the same standard.

6.34 We understood from the banks’ cost estimates (see paragraphs 6.50 and 6.51) that the costs associated with implementing this remedy would account for a relatively small proportion of the total cost of the package of remedies. For example, we note that life corporate membership of one of the organizations that provides services such as document editing to a standard of ‘plain English’ costs £15,000 a year including VAT.

6.35 The costs of implementation of this remedy would be lower if banks were not required to use standard terminology. We also thought that the costs of implementing this remedy would be mitigated by building on the existing arrangements for monitoring compliance with the ‘plain language’ requirement of the Code by the BCSB.

6.36 We noted that, in response to our provisional decisions on remedies, one bank ([]) expressed concern about the cost and administrative burden that would result from this remedy. We considered that, by clarifying the scope of this remedy and limiting the communications covered by this remedy to those written communications with customers when choosing a PCA, when opening a PCA, on statements and when pre-notified of charges and interest payments, this remedy would not impose an undue administrative burden or cost on banks.

6.37 In addition, we noted that the industry may move in the direction suggested by this remedy in any case. Recommendations have been made to the Code review (eg by the OFT) that the Code should be amended to encourage banks to do more to make their communications with customers easy to understand.181

181OFT Response to Banking Codes Review 2007, February 2007.

135

• Our decision

6.38 We therefore conclude that the banks must satisfy the BCSB that all information provided to customers when choosing a PCA, when opening a PCA, on statements, and when pre-notified of charges and interest payments, is easy to understand. The banks must ensure that all such communications are:

(a) certified by an independent organization specializing in plain English; or

(b) otherwise tested with customers (eg by means of focus groups or customer sur- veys) and found to be easily understandable.

We should emphasize that we are not suggesting that if the banks ensure that their communications are tested in one of the ways set out in (a) and (b) above, they will necessarily be acceptable to the BCSB. Ultimately, the BCSB must be satisfied that the banks’ written communications are easy to understand.

6.39 We are not requiring the use of standardization of descriptions of key terms. However, we consider that the banks must ensure that the terms that are important for customers’ understanding of PCAs are explained in a way that is easy for customers to understand. These terms include: BACS, CHAPS, ‘item’ and terms relating to authorized and unauthorized overdraft services and charges.

6.40 The requirement on banks to satisfy the BCSB that all bank communications with customers in relation to PCAs are easy for customers to understand will equally apply to all our other remedy options.

6.41 In addition, we strongly recommend that the BCSB should encourage the banks to use the same process that they would use to comply with this remedy to ensure that other communications with customers are equally easy to understand.

6.42 This remedy will address the features of the market that we found that banks have unduly complex charging structures and practices and that they do not fully or sufficiently explain their charging structures and practices. These features make it more difficult for customers to understand the charges applied to their PCA by their current bank. This remedy will also help to address the difficulties customers perceive and experience in searching for and switching PCAs. It will therefore address the feature that customers generally do not actively search for alternative PCAs or switch bank.

Remedy (b): Explanations of the levels of charges and interest rates and how and when they are applied

6.43 In the Notice, we proposed that banks might be required to make available to customers information on the level of charges and interest rates, the circumstances in which the customer would incur these charges and/or receive interest, and the way in which, and when, these charges and interest rates would be applied (remedy 2). In our provisional decisions on remedies, we clarified the relevant charges and interest rates:

(a) credit interest rates applicable to relevant balance limits;

(b) current account charges;

(c) charges for standard account services (eg charges for setting up direct debit or standing orders);

136

(d) authorized overdraft debit interest rates;

(e) authorized overdraft charges;

(f) unauthorized overdraft debit interest rates; and

(g) unauthorized overdraft charges (including maintenance charges, paid and unpaid items charges and transaction charges).

We proposed that the banks should make this information freely available to customers when choosing a PCA, when opening a PCA, on statements (see paragraph 6.83) and with any pre-notification of charges and interest payments that is not sent together with a statement (see paragraph 6.143). We also proposed that the banks should inform customers of the existence of discretionary policies in relation to charges or interest.

• The views of parties

6.44 Most parties were supportive of this remedy. Some parties noted that most of the information was already provided to customers in various leaflets. The BCSB said that the best way to implement this remedy would be through the introduction of summary boxes on marketing and pre-sale material for PCAs. It provided, for illustrative purposes, an example of a summary box that it had prepared in the context of the independent review of the Code in 2004 (see Figure 8).

137

FIGURE 8

Illustrative summary box

SUMMARY BOX: YOUR CURRENT ACCOUNT

The information in this box summarises key product features and is not intended to replace our Terms and Conditions

Interest rates We will pay interest of 0.1% gross (0.079% after deduction of tax at the standard rate) a year on your cleared credit balances. Interest is paid annually on 1 January.

We will start to pay you interest as follows: Cash The same day as you pay in Cheques Three days after they are paid in BACS credits The same day as they are received Minimum balance There is no minimum balance required Borrowing Your account should not go overdrawn without first making arrangements with us. Authorised borrowing is charged at 1% per month (EAR* 12.8%) When you are overdrawn beyond your agreed limit, if any, by more than £20, we will charge a monthly usage fee of £20 and a monthly interest rate of 2.05% (EAR* 27.5%) Statements We will send you a statement every month When you can withdraw your You can take your money out as follows: money If you have paid in cash: on the same day If you have paid in cheques: four days afterwards If you have received BACS credits: on the same day How you can withdraw cash You can take it out by card or cheque across our branch counters or through any UK cash machine Charges and penalties If you pay in a cheque which ‘bounces’, we will charge you £25. If you do not have enough money on your account for us to pay a cheque or a direct debit, we will charge you £25 for each one. If you ask us to ‘stop’ one of your cheques, we will charge you £10. If you ask for a duplicate statement, we will charge you £5. Other charges are detailed in our Terms and Conditions. Who can open an account? Up to three individuals, including young people Clubs, associations and trusts What is needed to open an Evidence of who you are and where you live. For instance, a passport and a account? utility bill How long will it take to open an Usually about three days account?

*EAR = Equivalent Annual Rate. Source: BCSB.

6.45 In responding to our provisional decisions on remedies, one bank ([]) said that the sample summary box should contain information about ATM and currency transaction charges, together with a reminder that customers should review their statement entries and advise the bank of any discrepancies as soon as possible. However, other respondents to our provisional decisions on remedies thought that the sample summary box contained too much information. The BBA noted that, both in relation to this remedy and remedy (c), independent research on summary boxes had shown that customers felt that the level of information that they already had for their current accounts, both pre- and post-sale, was sufficient. This research had suggested that, before account opening, customers valued information in relation to charges, the basis of operation of the account and cheque clearance times. Its research had also suggested that, after they had opened the account, customers found summary box information in relation to fees, other account charges and cheque usage fees to be the most useful.

138

6.46 Another bank responding to our provisional decisions on remedies ([]) thought that, although customers needed an adequate explanation of the level of charges and interest rates and how and when they were applied, the proposed frequency of the communications might result in ‘information overload’.

6.47 We explored whether to require banks to disclose information on buffers, fee caps and pencil limits which are applied at the discretion of the bank (see paragraph 4.267). The exercise of discretion takes into account various elements including the customer’s credit status, past usage of the PCA, and the bank’s approach to risk management. We noted an important distinction between the existence of a policy allowing discretion to be exercised in relation to certain charges and interest payments and the details of how any such policy might operate, eg the factors that would be taken into account in the exercise of discretion and the way in which they would affect any decision.

6.48 Most parties did not support the disclosure of the detail of banks’ discretionary policies on buffers, fee caps and pencil limits. They argued that such policies benefited customers by allowing flexibility to take account of a customer’s individual circumstances. They said that the disclosure of these policies would remove the banks’ ability to exercise discretion. The way in which these policies operated would become codified, effectively establishing the new limits within which customers would expect to be able to operate their PCAs. These policies might then lose their purpose and the banks might cease to use them, to the detriment of customers. One bank ([]), in response to our provisional decisions on remedies, noted that, strictly speaking, all its charges and interest rates were subject to its discretion. This bank therefore argued that informing customers of the existence of discretion would lead to confusion.

6.49 The BCSB, on the other hand, said that the fact that banks may exercise discretion in respect of some charges should be revealed to the customer since this was part of the PCA offer the customer received from the bank. However, the BCSB said that this disclosure must avoid leaving customers with a false impression of the way in which charges would be levied.

6.50 The banks estimated the costs of implementing remedies (a) and (b). Estimates of the one-off costs of implementation varied greatly. One bank’s estimates ([]) suggested that there would be no one-off cost associated with these remedies. Four banks ([]) suggested one-off costs of between £15,000 and £35,000. The other three banks which provided estimates ([]) suggested one-off implementation costs of between £115,000 and £350,000. We estimated the total one-off cost of imple- mentation for all those banks to which these remedies would apply182 at around £1 million.

6.51 Based on estimates provided by two banks, and on our own assumptions,183 the estimated annual cost per bank was around £50,000. This suggests an annual cost of implementing these remedies for all banks of around £500,000. For both the one-

182This is calculated on the basis that a threshold would be applied so that only banking groups with more than 10,000 PCAs held by customers with a postal address in Northern Ireland, and individual banks within such banking groups with more than 5,000 PCAs held by customers with a postal address in Northern Ireland, would apply the remedies. This would mean that currently ten banking groups would apply the remedies. This is discussed in more detail in paragraphs 6.224 to 6.237. Calculations of the costs of implementing each remedy across the banks as a whole have all been undertaken on this basis, with the exception of remedy (c), as explained in paragraph 6.74. This is in line with our approach to the assessment of the costs associated with the package of remedies overall, set out in paragraphs 6.257 and 6.258. 183In estimating the annual cost, we have taken into account the requirements for banks to finance additional resources needed for monitoring these remedies.

139

off and annual costs, the banks suggested that the bulk of the costs were associated with the second of these two remedies.

6.52 In responding to our provisional decisions on remedies, one bank ([]) suggested that this remedy (and remedy (c)) could stifle innovation because it would increase the costs to banks of introducing personalized rates and charges if banks were required to provide information on these instead of or in addition to standardized rates and charges. It argued that the remedy could therefore have a negative impact on the competitive dynamics in the provision of PCAs in Northern Ireland.

• Our views

6.53 We considered the evidence that we had received in relation to the scope and nature of the information to be provided to customers in explaining the level of charges and interest rates and how and when they are applied. We remained convinced that the banks must ensure that customers are provided with explanations of the levels of charges and interest rates and how and when they are applied. We also remained convinced that, although prepared for a different purpose and hence containing different information, the sample summary box in Figure 8 is a good example of how banks might provide this information to customers. We noted that if some banks wished to go further than this and include additional information, they would not be prevented from doing so provided that the additional information did not obscure or detract from the explanations of charges and interest rates required by this remedy. As with remedy (a), we thought that the best way to implement this remedy would be to require the banks to satisfy the BCSB that explanations of the levels of charges and interest rates and how and when they are applied had been provided to customers.

6.54 In considering our remedies, we took the potential for ‘information overload’ very seriously. However, we thought that, in order to take an informed decision on which PCA product is best for them, information on the costs of running PCAs must be readily available to customers in a form that is easy to understand at the appropriate time. Thus, we remained convinced that banks must, at certain key stages (see paragraph 6.56), provide customers with explanations of the level of charges and interest rates and how and when they are applied. We did not accept that requiring the provision of such fundamentally important information would contribute to ‘information overload’.

6.55 We found that the discretion exercised in relation to the application of charges and interest rates contributed to a lack of clarity among customers about their PCAs, but also benefited customers in that it gave banks the scope to take into account individual circumstances. Overall, given the evidence set out in paragraphs 6.47 to 6.49, we concluded that we should not pursue a mandatory disclosure of the detail of discretionary policies. We accepted that, to some extent, banks may exercise discretion over whether to apply any charge or interest payment. We thought, however, that customers should be made aware of the existence of discretionary policies. Our concern focused on charges and interest rates relating to unauthorized overdrafts. Thus, we considered that banks must make customers aware of the existence of discretionary policies in the banks’ application of charges and interest only in relation to unauthorized overdrafts.

6.56 We looked in particular at when customers should be given information explaining the charges and interest rates that could be applied to their PCAs. We recognized the risk of ‘information overload’ for customers and unnecessary cost for the banks. However, we also thought it important to ensure that customers have information about charges and interest rates at those times when they are likely to be receptive

140

to such information, and likely to use it. We therefore found that, in order to be effective in addressing the features of the market it has identified, the information set out in paragraph 6.43 would need to be available to customers:

(a) when they were choosing between PCAs (so that they could compare the likely costs of running a PCA with different operators);

(b) at the time a PCA is opened (to ensure that the customer is fully aware of the interest rates and charges that will apply to the PCA);

(c) on customer statements (see paragraph 6.86); and

(d) when customers are pre-notified of charges and interest payments to be deducted from their PCAs (see paragraph 6.146).

6.57 Banks must take particular care to ensure that it is clear where a single transaction can trigger more than one charge (eg that a cheque that takes a customer into unauthorized overdraft may trigger both an unauthorized overdraft maintenance charge and a paid item charge). Similarly, banks must ensure that it is clear where charging is applied retrospectively (for example, authorized overdraft charges on traditional PCAs). In this respect, the use of examples on the way charges are applied could be useful.

6.58 We also considered whether to require that customers receive a complete set of terms and conditions each time a change is made to them. The Code requires banks to ensure that customers have a copy of the terms and conditions that apply to their PCA when the PCA is opened. It also requires banks to ensure that any changes to those terms and conditions are communicated to the customer in a way that allows customers to understand the changes made. Although the way in which customers are informed of changes to their terms and conditions varies between banks, we did not conclude that customers should receive a new set of terms and conditions each time a change is made to them. If we had reached this conclusion, it might have resulted in customers receiving several copies of their terms and conditions each year, making it more difficult for the customer to see the changes that have been made, and resulting in additional cost to the banks. However, we found that, in order to ensure that customers have clear information, any change in terms and conditions should be accompanied by sufficient explanation and context to allow customers to understand the implications of the change for them.

6.59 Paragraphs 6.50 and 6.51 set out the banks’ cost estimates for the implementation of this remedy (together with remedy (a)). We considered that not all the costs associ- ated with implementing this remedy may be incremental. Some of the recommen- dations to the Code review suggest that the industry as a whole should do more to ensure visibility of charges and interest rates. In addition, there are likely to be synergies between the implementation of remedies (a) and (b) and remedies (c) and (d).

6.60 We noted the suggestion from one bank ([]) that this remedy could stifle innovation by increasing the costs associated with introducing personalized charges and interest rates. We considered that, in order to make well-informed decisions about PCA providers and products, customers must have explanations of the level of charges and interest rates and how and when they are applied. We accepted that, depending on how this remedy is implemented, it may increase the cost to banks of introducing such charges and interest rates. However, we also noted that our package of remedies, by increasing the extent to which customers will shop around for the best deal on PCAs, will also increase the incentive for banks to offer innovative products.

141

We therefore did not consider that we should be concerned about the effect of this remedy on innovation.

• Our decision

6.61 We conclude that banks operating in Northern Ireland must ensure that their Northern Ireland customers receive explanations of the levels of charges and interest rates on their PCA, the circumstances in which the customer would incur these charges and/or receive interest, and the way in which, and when, these charges and interest rates would be applied. Explanations must be provided of at least the following charges and interest:

(a) credit interest rates applicable to relevant balance limits;

(b) current account charges;

(c) charges for standard account services (eg charges for setting up direct debit or standing orders);

(d) authorized overdraft debit interest rates;

(e) authorized overdraft charges;

(f) unauthorized overdraft debit interest rates; and

(g) unauthorized overdraft charges (including maintenance charges, paid and unpaid items charges and transaction charges).

6.62 We conclude that banks operating in Northern Ireland must also make clear to customers the existence, but not the detail, of discretionary policies in the banks’ application of charges and interest in relation to unauthorized overdrafts.

6.63 We conclude that this information must be made freely available to customers when choosing a PCA, when opening a PCA, on statements, and when pre-notified of charges and interest payments to be deducted from their PCAs.

6.64 This remedy is complemented by remedy (a), which will ensure that this information (together with any other written communication supplied to customers by banks) is easy for customers to understand. Together, we found that these two remedies, working in conjunction with other remedies in our package, would be effective in addressing the features that banks have unduly complex charging structures and practices and banks do not fully or sufficiently explain their charging structures and practices.

6.65 One of the key commitments to customers set out in the Code is to provide clear information about how an account or service chosen by a customer works, its terms and conditions and the interest rates which apply to it. We view these two remedies as reinforcing and extending the Code requirements by putting emphasis on the clarity and completeness of the information to be given to customers about PCA services, interest rates and charges and their application.

Remedy (c): Information on statements

6.66 In the Notice and in our provisional decisions on remedies, we proposed that banks should be required to provide information on each PCA statement on the levels of

142

charges and interest rates, and how and when they are applied, together with other information on switching and on the key terms and conditions of the PCA (remedy 3).

6.67 We considered the following:

(a) what information should be displayed on PCA statements;

(b) how such information should be displayed;

(c) how often the information should be included on the statement;

(d) how best any such remedy might be implemented; and

(e) the extent to which the summary box considered by the BCSB following the previous Code review might provide a basis for this remedy.

• The views of parties

6.68 Although all parties agreed that a certain amount of information on charges and interest rates should be included on statements, most parties said that they already provided this information to customers. They expressed concern that additional infor- mation on statements might become excessive and repetitive and result in customers becoming confused or impervious to the information.

6.69 The BCSB consulted on the inclusion on statements of summary boxes containing key account information in the context of the last review of the Code in 2004 (see illustrative summary box in Figure 8). The BCSB noted that, while the detailed content might need to be revisited, it saw no reason why such a summary box should not be included on statements. Some banks ([]) said that the remedy would be easier to implement, and in their view more effective, if the information were provided on a separate leaflet rather than on statements.

6.70 Many parties said that this remedy would overlap with the evaluation and possible introduction of summary boxes for PCAs currently being conducted by the BBA as a result of a recommendation by the independent reviewer at the last Code review. The parties said that any change to statements should be implemented in a consistent manner and only once. One bank ([]) also noted that the survey commissioned by the BBA on summary boxes indicated that customers had less interest in post-sale summary boxes (ie summary boxes on statements) than pre-sale summary boxes (ie on marketing material).

6.71 We received a variety of suggestions regarding the information that should be included on statements. This included not only details of interest rates, authorized and unauthorized overdraft charges, but also contact details of how to obtain further information, seek help if in financial difficulty, and how to lodge a complaint.

6.72 As discussed in relation to our remedy on explanations of charges and interest rates (remedy (b)), in responding to our provisional decisions on remedies, one bank ([]) thought that the sample summary box contained too little information, while others ([], BBA) thought that it contained more information than felt necessary by customers, based on independent research into summary boxes carried out in 2006. One bank ([]) raised concerns in relation to information overload. Another bank ([]) said that there would not be sufficient room on the front of its statements to provide the information required and that it would need to provide this information on an additional page.

143

6.73 Similarly, as discussed in relation to remedy (b), one bank ([]) argued that this remedy would raise the costs to banks of introducing personalized charges and interest rates and could thereby stifle innovation.

6.74 Estimates of implementation costs varied, with the extent of the variation depending at least in part on the banks’ statement systems and the information already provided. In its response to our provisional decisions on remedies, one bank ([]) revised upwards the cost estimates it had originally given, because it believed that IT changes would be more extensive and complex than previously envisaged and that the additional information to be put on statements would increase their size. One bank ([]), that previously had not given us cost estimates for any of the remedies, provided an estimate of one-off implementation and annual costs only in relation to this remedy. These costs assumed that the bank would implement remedy (c) on a UK-wide basis, which it said would be disproportionate for UK-wide banks with a small proportion of their UK customer base in Northern Ireland. Taking into account all these revised cost estimates, the banks’ estimates suggested a range of one-off costs of implementing the remedy (assuming that information could be located either on the front or the back of the statement) from around £180,000 ([]) to £5.0 million ([]). The banks’ estimates suggest a total one-off implementation cost for all banks required to implement this remedy in the range of £11 million to £14 million.184 The two estimates of the annual cost of this remedy were zero ([]) and £1 million ([]). We estimated an annual implementation cost for all banks of around £2 million.185 One bank ([]) noted that a requirement to include credit and debit interest rates in the information to be provided on statements would be particularly expensive. This was because these interest rates changed relatively frequently, meaning that the information on statements would need to be updated more frequently than would be the case if these interest rates did not need to be included. It was suggested that, to reduce costs, banks could instead make clear on statements where customers could get information on credit and debit interest rates.

• Our views

6.75 We believed that the availability of information on statements on PCA key terms would enhance customers’ ability to understand the services provided by a bank, and the charges and interest rates applied, on their own PCA. This would make it easier for them to compare the PCAs offered by other banks and to make decisions about the best PCA to meet their needs. We thought it essential that information on key terms was provided on the statement itself to ensure that customers could make sense of the transaction information on their statement and relate this to their behaviour. We found that the illustrative BCSB summary box (see Figure 8) repre- sented an appropriate starting point for the banks’ consideration of how best to provide information on statements in line with this remedy (subject to including the information set out in paragraph 6.61).

6.76 If customers want information on interest rates and charges that apply to their PCAs, they currently have to request it from their bank, or refer to documents given to them either when they opened the PCA or at the time of the latest changes in charges. This could mean referring back to documents received months or years before.

184This is calculated by adding the cost estimate for a UK-wide implementation of this remedy provided by one of the ten banking groups that would apply the remedies ([]) (see footnote to paragraph 6.37) to the cost estimates calculated by applying the average estimated cost for the banks that provided us with cost estimates for all the remedies to nine of the ten banking groups that would apply the remedies in Northern Ireland. 185The annual cost was calculated on the same basis as the one-off implementation cost.

144

6.77 We considered in particular whether banks should be required to include credit and debit interest rates as part of the information to be provided on the statement. We recognized that it may be more costly, at least for some banks, to include this information on the statement. However, we noted that credit and debit interest repre- sent a very important element in the cost of running a PCA. We noted that for the majority of customers, who do not go into unauthorized overdraft, the credit interest that they have forgone by keeping money in a PCA rather than, say, in a savings account is likely to represent a particularly important element in the cost of running their PCA. We considered whether it would be sufficient for banks to make clear on statements where information on credit and debit interest could be found. However, we thought that the effectiveness of this remedy would depend on customers having on their statement all the information that they would need to understand the transactions included on that statement, thereby removing the need for customers to search several sources of information in order to understand their PCA. Requiring customers to go elsewhere for such important information as credit and debit interest rates would undermine the effectiveness of the remedy.

6.78 We wished to ensure that information provided to customers did not result in information overload. We did not think that the additional value to customers of including ancillary charges on statements (such as charges for money transfers or for copies of statements) would be sufficient to merit requiring their inclusion. Because ancillary services are only provided on request, the level of ancillary charges and how they will be deducted from accounts will be made clear to customers by banks at the time of the request, as required by the Code. Banks may, however, choose to include additional information on the statement, including telling the customer where to find information on ancillary charges or on the clearing cycle, provided that this does not obscure or detract from the information that we are requiring them to provide.

6.79 Many banks already provide other information on their PCA statements such as web- site addresses and telephone helpline numbers for customer enquiries; where to seek advice if in financial difficulties; the FSA consumer website address and tele- phone helpline; and the complaints process. While such information may well be helpful to customers, we decided not to require its provision on statements (or else- where) as it would not directly address the AEC, or any detrimental effect on customers, arising from the features of the market it has identified. Again, banks would be free to include such information on statements, provided that it does not obscure or detract from the information we are requiring them to provide.

6.80 We looked at whether to specify how the information should be provided: on either the front or the back of the statement, split between the two, or otherwise provided on a separate sheet but still as an integral part of the statement. We thought that a requirement to provide the information on the front of the statement could make it crowded, could lessen the impact of the information on transactions and might reduce the likelihood of a customer reading the statement. In addition, there are significant differences between banks in the way in which statements are created, leading to differences in costs estimates between banks, depending on where the information is located. We therefore did not find it necessary to prescribe the location of information on statements.

6.81 We noted that one bank ([]) had said that there would be insufficient space on the front of statements to provide this information and that it would need to provide it on an additional page. It is open to banks to provide the information required by this remedy on an additional page, provided that any such additional pages clearly form part of the statement. Similarly, we wished to ensure that where this information is provided as part of an electronic statement, the information should be clearly visible

145

to anyone accessing that electronic statement, ie it should not require customers to read a ‘pop up’, click on an icon, or scroll down well beyond the end of the transactional information.

6.82 We noted the estimates of the costs of implementing this remedy set out in paragraph 6.74.

• Our decision

6.83 We conclude that banks must provide customers with key account information as part of their PCA statements. As set out in paragraph 6.61, we decided that banks should provide as part of the statement (whether in hard copy or electronically) explanations of the level and application of at least the following:

(a) credit interest rates applicable to relevant balance limits;

(b) current account charges;

(c) charges for standard account services (eg charges for setting up direct debit or standing orders);

(d) authorized overdraft debit interest rates;

(e) authorized overdraft charges;

(f) unauthorized overdraft debit interest rates; and

(g) unauthorized overdraft charges (including maintenance charges, paid and unpaid items charges and transaction charges).

6.84 The information should be provided to each PCA customer, in the same form as a customer receives their statement, unless the customer elects otherwise. It may be provided on the front or the back of the statement, on a separate sheet on ‘statement paper’, or as part of an electronic communication, provided the information is clearly part of the statement.

6.85 This remedy seeks to address the following features we have identified: that banks have unduly complex charging structures and practices and banks do not fully or sufficiently explain their charging structures and practices.

6.86 There is a link between this remedy and remedy (b). Remedy (b) requires the banks to provide comprehensive information on the PCA services available, the level of charges and interest and how charges and interest are applied to the PCA. This remedy requires that information to be provided as part of a customer’s statement.

Remedy (d): Summary and breakdown of charges and interest

6.87 In the Notice, we proposed that banks should provide on statements an easy-to- understand analysis of the charges and interest rates that have been applied to each PCA both in the period covered by the statement and in the previous 12 months (remedy 4). This would allow customers to see the amount of charges and interest they have paid in the last year, both as a total and by category of charges and interest.

6.88 We sought the parties’ views on:

146

(a) what information could usefully be included in any breakdown of charges and interest in order to maximize their usefulness to customers (eg the amount of each individual charge or interest rate; how many times during the period the charge or interest rate had been applied to the customer’s PCA; the total amount of each charge or type of interest the customer had paid in the period; the total amount of all charges in the period);

(b) how any such breakdown should be presented in order to maximize its impact;

(c) how often it should be provided (eg once a year or each month); and

(d) whether the period covered by the summary should be a period other than the last 12 months, and if so, whether it should be accompanied by a rolling total covering the previous 12 months.

• The views of parties

6.89 None of the banks supported this remedy. They said that historical information was of very limited use to PCA holders, that a summary or rolling total of charges and interest would create confusion for customers, and that there was no customer demand for more information on charges and interest on statements. Several banks said that there was a risk of information overload. One bank ([]) suggested that an annual statement or leaflet with a breakdown of charges and interest incurred during the tax year (to coincide with interest statements) would be a simpler and clearer solution for customers. Another bank ([]) said that if we were minded to pursue this remedy, annual summaries should be provided as a stand-alone document that could be sent with the statement rather than appearing on the statement.

6.90 In their responses to our provisional decisions on remedies, some banks ([]) noted that we had proposed to include ancillary charges in the annual summary. One bank ([]) argued that the inclusion of ancillary charges in the annual summary would add £[] million to the cost of implementing the remedy. Another bank ([]) said that different banks had different definitions of ancillary charges and that we would therefore need to define quite carefully what was meant by ‘ancillary charges’. A third bank ([]) said that many ancillary charges did not relate to the main functions of the customer’s PCA and argued that the effectiveness of the remedy would not be compromised if their inclusion in the annual summary was not required.

6.91 Also in responding to our provisional decisions on remedies, one bank ([]) proposed that we should not mandate the provision of the annual summary in the same form (electronic or hard copy) as the customer routinely received their state- ments but rather leave banks the option as to whether to send the annual summary in electronic or hard copy form. It noted that this would enable the banks to choose the most cost-effective way of conveying this information to customers.

6.92 The same bank ([]) also suggested that we should not require banks to send the annual summary together with a statement. It suggested that banks should be given the choice of whether to send the annual summary with a statement or separately.

6.93 The same bank ([]) said that we should require all banks to provide summaries of charges and interest payments in the same format to ensure consistency of appli- cation. It suggested a list of the charges and interest payments that all banks should be required to include in the annual summary.

6.94 Some parties (Which?, Advice NI, GCCNI) expressed support for this remedy. Proposals for the information to be contained in any such summary included the

147

charges and interest incurred and earned; the market average of each charge; the authorized and unauthorized amount borrowed; the number of days in unauthorized overdraft; information on customers’ rights; and how to make a complaint. Most parties suggested that the summary should be annual, although Advice NI said that the breakdown of charges and interest should be provided to customers on a monthly and six-monthly basis and GCCNI said that it should be included on each statement for both the period of the statement and the preceding year.

6.95 Many banks said that this remedy would be particularly costly for them to implement. Not all banks keep cumulative totals for the different types of charges and interest payments applied to PCAs. Where this information exists, some banks argued that this remedy would require them to design new software, or adapt existing software to bring the information together before a written communication could be prepared for the customer.

6.96 The banks’ cost estimates, as revised in their responses to our provisional decisions on remedies, suggested that the one-off costs to the banks of implementing this remedy would range from £380,000 ([]) to £4 million (the upper end of the range provided by []). The total one-off implementation cost for all banks required to implement this remedy was between £10 million and £16 million. The banks’ esti- mates suggested an annual cost of implementing this remedy of around £240,000 per bank, or around £2.4 million for all banks.

• Our views

6.97 Banks already inform customers, either on statements or through separate communi- cations, of the amount borrowed by authorized overdraft, the amount of unauthorized overdraft and the number of days in unauthorized overdraft. Although providing summaries of this information may go some way to increasing awareness of those charges and interest payments, we found that banks’ current practice did not go far enough. Northern Ireland PCA customers must be provided with information that makes the cost of their PCA clear to them, in the same way as the costs of other financial services products, such as mortgages, are made clear. We thought that an annual summary and breakdown by category would have greater impact and would clarify the cost to customers of running the PCA. In addition, a 12-month total would be a larger figure than a monthly total and could therefore be more effective in stimulating customers to consider switching.

6.98 We also considered whether the remedy would be more effective if it required the banks to provide the information on a different basis or more frequently. Specifically, we looked at whether banks should be required to provide each month, or each quarter, the total charges and interest payments that had been deducted from the PCA over the last 12 months (ie a 12-month rolling total). We thought that quarterly or monthly rolling totals or year-to-date accumulations may provide information that some customers would find helpful. However, we also took into account other factors including the possible adverse consequences in terms of information overload and the additional costs that would be associated with such a requirement. We therefore decided not to require summaries and breakdowns more than once a year.

6.99 We decided that banks should be required each year to provide an annual total of the charges, credit interest payments and debit interest payments that have been deducted by the bank from the PCA over a specified 12-month period (eg 1 January to 31 December or the tax year). We did not find it necessary to specify the 12-month period that should be covered by the annual summary. Nor did we find it necessary to require the banks to send annual summaries to all their customers covering the

148

same 12-month period. We believe that this flexibility should help to minimize banks’ compliance costs.

6.100 We noted the suggestion that we should allow banks the choice of whether to send the annual summary with a statement or separately. We believed that the effec- tiveness of this remedy did not depend on the annual summary being sent with a statement. We therefore decided that banks should be allowed the flexibility to send the annual summary with a customer’s statement or separately as they wished.

6.101 We also considered the suggestion put forward by one bank ([]) that we should allow the banks to choose the form—hard copy or electronic—in which the annual summary should be sent to customers. We accepted that this could allow banks to choose the most cost-effective way of providing this information. We thought that allowing banks this choice could undermine the effectiveness of the remedy by allowing banks to send annual statements in a form that customers would not find easy to access. However, we noted that it may be the case that a customer would like to receive their annual summary in a different form than that in which they received statements. We therefore decided that banks should provide customers with a choice as to the form in which they would like to receive their annual statement. If customers do not actively choose to receive their annual summary in a different form than that in which they routinely receive statements, those customers must be provided with their annual summaries in the same form as that in which they routinely receive statements.

6.102 Banks will not be required to include ancillary charges in the annual summary. However, if they do not include these charges, they must include a clear and prominent statement informing the customer that they may also have incurred ancil- lary charges during the year and that these charges can be found on the customer’s regular statements. Furthermore, in implementing this remedy we will ensure that the banks are not permitted to redefine other charges and ancillary charges and thereby remove them from the scope of this remedy.

6.103 We designed the remedy so as to mitigate its costs. In particular, we have not stipulated when the summary should be provided and we have not required the banks to include ancillary charges in the annual summary, although they will be free to do so if they wish. Although there might have been some value in all customers receiving these summaries and breakdowns at more or less the same time, to create a ‘switching season’, we left it open to the banks to choose a date to fit in with their business cycles. They may also choose not to send the annual summaries and breakdowns to all their customers at the same time, possibly staggering them over the year. There will be no requirement on the banks to send a customer’s annual summary with or at the same time as they send that customer a statement, although banks may do this if they so wish.

6.104 We considered whether there would be value in requiring all banks to provide summaries of charges and interest payments in the same format to ensure con- sistency across banks. However, beyond the need for all annual summaries to be easy for customers to understand, and to include explanations of the level of charges and interest rates and how and when they are applied (in line with our other remedies), we do not find that the effectiveness of this remedy would be enhanced if banks were to adopt the same approach to these summaries. We therefore did not believe it necessary for us to be more prescriptive in setting out a standard approach to these summaries beyond those discussed above.

6.105 We are aware that for customers who remain in credit the total cost of operating a PCA (not taking account of credit interest forgone) over a year may be very low.

149

However, we noted that it is not necessary for all PCA customers actively to search for alternatives to make the market more competitive, and for some customers— notably those who have incurred unauthorized overdraft charges—the cost of operating a PCA over a year may be considerable. In conjunction with remedies (f) and (g), we found that this would be key to the effectiveness of our remedies package, despite the relatively high costs of implementation.

• Our decision

6.106 We conclude that the banks must provide PCA customers once a year with an annual summary and breakdown of charges and interest. In order to be effective, this information should be broken down to show the total of each individual type of charge or interest payment made over the relevant period relating to the customer’s PCA but need not include ancillary charges. The summary must include an explanation of the level and application of all those charges included on the statement (including those set out in paragraph 6.83).

6.107 We decided that the charges and interest payments that should be covered in the summary include the following:

(a) any regular account charge (eg monthly charge for maintaining the PCA) payable when the PCA is in credit;

(b) any transaction charges (eg for setting up a standing order) payable when the PCA is in credit;

(c) credit interest;186

(d) authorized overdraft:

(i) debit interest;

(ii) annual or monthly maintenance charges; and

(iii) other relevant charges; and

(e) unauthorized overdraft:

(i) debit interest;

(ii) maintenance charges;

(iii) transaction charges (eg referral or unpaid charges); and

(iv) other relevant associated charges (eg letter to customers).

If the summary does not include ancillary charges, banks must include a clear and prominent statement informing the customer that they may also have incurred ancillary charges during the year and that these charges can be found on the customer’s regular statements.

186Banks must show credit interest and debit interest separately. They will not be permitted to provide a single net figure comprising debit interest minus credit interest.

150

6.108 The summary should be sent to the customer in the same form (electronic or hard copy) as the customer routinely receives their statements, unless the customer elects otherwise. We will not mandate the date on which the summary must be sent. Neither will we mandate the period to be covered by the summary, provided it is a 12-month period. Banks will be free to send summaries more often than once a year if they so wish.

6.109 We thought that this remedy would help to address all three features of the market that we identified as resulting in the AEC, making it a particularly powerful element of the package. This remedy would effectively—in conjunction with the first three remedies ((a) to (c))—help to address the features of the market that banks have unduly complex charging structures and practices and do not fully or sufficiently explain their charging structures and practices. Furthermore, we thought that this remedy, in conjunction with remedies (f) and (g), would, by presenting customers with the total cost of operating their PCA over a year, increase the scope for them actively to consider different banks and hence address the third feature that we identified, that customers generally do not actively search for alternative PCAs or switch bank.

6.110 There is a link between this remedy and remedy (b). This remedy requires an explanation of the levels of charges and how and when they are applied to be included on the annual summary.

Remedy (e): Advance notice of charges and debit interest incurred

6.111 In the Notice we proposed a requirement for banks to provide customers with advance notice of charges and debit interest before they are debited to their PCA (remedy 5). This would improve the transparency of PCA charges.

6.112 In particular, we considered:

(a) the notice period banks should be required to provide to customers (eg 14 days, 30 days or some other period) before an amount is deducted from their PCA; and

(b) whether such notice should be provided on customers’ statements or by means of a separate communication.

• The views of parties

6.113 The Code (section 5.4 and 5.5) already requires banks to give customers at least 14 days’ notice of any charges and interest for standard account services that accumulates to the PCA. Examples of charges where 14 days’ prior notification is already required are set out in the Code:

(a) usage fees for authorized overdrafts;

(b) debit interest; and

(c) charges for standard account services such as processing cheques, standing orders, direct debits and UK debit card transactions.

6.114 Under the Code there is currently no requirement to pre-notify:

(a) charges for services that are debited at the time the service is provided and where the customer has been notified in advance (eg in the price lists or terms

151

and conditions). These include, for example, charges and additional interest payable for authorized or unauthorized borrowing; and

(b) charges for services that are debited at the time the service is provided and where the customer is informed at the time the service is provided, for example charges for stopped cheques or a money transfer.

6.115 Most respondents to the Notice supported a requirement on banks to pre-notify interest payments and charges. Many parties noted that it would be in line with the current Code requirement for banks to give customers 14 days’ notice of the appli- cation of charges for standard account charges that accumulate on the PCA and were charged at the end of the billing period. Some parties ([]) said that they already exceeded the Code requirements and pre-notified customers of unauthorized overdraft charges. GCCNI recommended a 30-day pre-notification period while Advice NI said that the length of the pre-notification should be tailored to the amount to be charged: the higher the amount, the longer the pre-notification. Further details of the banks’ existing pre-notification arrangements are set out in Appendix 4.13.

6.116 We note that the CCA 2006 will require banks to give the customer notice when a sum becomes payable due to a breach of the PCA agreement (default sum). An example of such a sum is a charge imposed for entering into unauthorized overdraft. The term default sum does not apply to interest, and therefore this provision would not apply to differential interest rates applied to unauthorized overdrafts. The CCA 2006 does not, however, provide for a particular notice period before the imposition of the default sum. The bank may only require the customer to pay interest in connection with a default sum 28 days after the day the notice was given. The notice may be issued separately or incorporated in a statement or another notice that is required under the CCA 1974 or the CCA 2006.

6.117 Several parties ([]) recommended that charges and interest be pre-notified on statements rather than on a separate communication. One bank ([]) suggested, however, that a separate pre-notification letter might have more impact than the statement itself. Advice NI also suggested that advance notice of charges and interest could be included on statements obtained from ATMs (mini-statements).

6.118 Two banks ([]) did not support a requirement that banks should pre-notify interest payments and charges. [] said that it pre-notified customers of debit interest on authorized and unauthorized overdrafts and it contacted customers when their PCA went into unauthorized overdraft. [] also said that in its view the implementation costs of this remedy would be disproportionate to the benefits it was likely to achieve, since the majority of customers did not pay these charges. It argued that in order for a bank to pre-advise unauthorized borrowing charges, it would be likely to have to accumulate charges and apply them once a month. It noted that this could result in a time delay of five to six weeks between the transaction or activity that gave rise to the charge and the actual application of the charge and customers would need to keep sufficient funds to pay the accumulated charges in the following month.

6.119 [] said that this remedy would mean that it was not able to deduct charges from a customer’s account immediately after they had been incurred, and that this would increase the credit risk associated with the customer. This could lead it to apply a more stringent credit assessment and lending policy with more unauthorized transactions being rejected.

6.120 In responding to our provisional decisions on remedies, three banks ([]) argued against any requirement that banks should provide customers with a statement for any month in which the customer incurred a charge or debit interest. One of them

152

([]) noted that whereas statements were sent regularly, charges and debit interest might be incurred irregularly, and that it might be more important for a customer to receive immediate notification of a charge than to have notification linked to a statement. Another bank ([]) said that the link with statements would add to the implementation costs of the remedy and that it was unclear that customers would benefit from the link.

6.121 The third ([]) said that linking pre-notification and the deduction of charges and debit interest with monthly statements would mean that it would not be able to retain its quarterly charging cycles. This, the bank said, would require it to change the pricing and features of some of its PCAs. It would also disadvantage its customers, as debit interest, which was currently accumulated over a three-month period, would be debited to the customer’s account earlier (ie monthly) and form part of the account debit balance for further interest calculation. It requested that it be allowed flexibility to retain its quarterly charging cycle and argued that linking pre-notification with statements was not necessary because the pre-notification communication itself could include the account balance. It also noted that the statement would not predict the transactions that would be completed on the customer’s account between the date of the statement and the application of the charge or interest payment that had been pre-notified.

6.122 Similar concerns were raised by another bank ([]) which calculates charges and interest quarterly. It said that the requirement to link the pre-notification to the issuance of a monthly statement would require significant changes to its IT systems. It would also remove the bank’s ability to offer offset products (ie products by which personal customers were able to use the credit balances in their PCAs to offset the interest that accrues on personal loans), as interest on personal loans accrued quarterly.

6.123 One bank ([]) proposed that rather than being required to send pre-notification in the same form (electronic or hard copy) as that in which a customer received their statement, banks should be allowed to choose the form of pre-notification.

6.124 The banks provided two estimates of the cost of implementing this remedy. The first estimate was based on there being no requirement to link pre-notification and the sending of a statement. The second required pre-notification on a statement.187 The one-off costs of implementing the remedy, without any link between pre-notification and statements, ranged from £0 ([]) to £4 million ([]), with a total one-off implementation cost for all banks in the range of £9 million to around £15 million. The estimates for the annual costs provided by one bank ([]) suggested an annual cost of implementing this remedy of around £150,000 for all the banks required to implement this remedy.

6.125 The banks suggested that it would be more costly to pre-notify on statements than it would have been to pre-notify without requiring a link to statements. In their response to the provisional decisions on remedies, three banks ([]) revised their initial cost estimates. Taking these revisions into account, the one-off implementation costs for all banks would be in the range of £14 million to £23 million. The annual cost of implementing the remedy for all banks could be around £4.4 million.

187We are aware that neither of these forms precisely conforms to this remedy as expressed here. However, we thought that the cost estimates for this remedy were likely to fall within the same range.

153

6.126 The BCSB said that the independent directors had recommended that the Code should be amended to require all banks to pre-notify default charges in their submission to the current review of the Code.

• Our views

6.127 We noted that, whilst all banks adhered to the provisions of the Code, pre-notification practices varied considerably between them. Some banks pre-notify customers of unauthorized overdraft charges a number of days before these charges are taken from the PCA. Other banks contact customers to alert them that their PCA is in unauthorized overdraft and charges are being levied, or apply the unauthorized overdraft charges to customers’ PCAs on the day they are incurred, without any communication to customers except the regular account statement.

6.128 Some banks, such as [], routinely contact customers who have incurred charges and debit interest to discuss their situation. We welcome this and believe that our remedy should not preclude banks from taking a pro-active approach to raising customers’ awareness of the link between their behaviour and the charges and interest payments deducted from their PCA.

6.129 Given the limits of the current requirements on pre-notification and the different practices of the banks, we thought that there was scope to improve pre-notification. This remedy would also be consistent with the banks’ responsible lending requirement, and would facilitate discussion between the banks and their customers about how best to manage their finances.

6.130 We believed that a remedy requiring banks to pre-notify charges (excluding ancillary charges) and interest should increase the transparency of the banks’ charging structures and practices for customers. By increasing customers’ understanding of the banks’ charging structures and practices, customers would be more aware of the need to look for the most appropriate PCA to meet their needs. The pre-notification requirement would cover all authorized and unauthorized overdraft charges, including any that do not accrue to the PCA. We did not think that ancillary charges should be included in the scope of this remedy, since the customer would be advised of these charges on requesting the service and should therefore understand why the charge had been deducted from their PCA.

6.131 We noted that, in their responses to our provisional decisions on remedies, some banks had argued against the requirement not to deduct pre-notified charges and interest payments from a customer’s account until at least 14 days after the statement date, and the requirement to provide a customer with a statement covering the month in which the charges were incurred. We believed that it is important to ensure that customers receive a statement covering the month in which the charge or interest payment was incurred at least 14 days before the charge or payment is deducted from their account. This is because the statement details would allow the customer to understand the link between their behaviour and the charge incurred. In addition, we believed that the account information provided on the statement—in particular, the account balance and the transactions on the account—would ensure that the customer had sufficient information to take appropriate action before the charge was deducted from their account. The information on the statement would also allow the customer to challenge the charge if, for example, it was levied as a result of some error on the account. In order for the customer to make use of this information, we thought that the customer should receive it in good time before the charge was deducted from the PCA. We therefore conclude that it is important that banks must not deduct a pre-notified charge or interest payment from a PCA until at least 14 days after the date of the statement. This should allow the statement to be

154

prepared, posted and received by the customer in time for any appropriate action to be taken.

6.132 We believed that the customer should receive a clear and easily understandable explanation of the level of the charges and interest they have incurred and why they have been incurred with the pre-notification. This will ensure that the customer can understand the link between their behaviour and the charge or interest payable. If pre-notification takes place with a statement, this information can be the information required on statements in line with remedy (c). However, if pre-notification does not take place with a statement, this information must be provided additionally.

6.133 Similarly, in order that the customer can link the charge or interest payable with their behaviour, we thought it important to ensure that charges and interest do not accumulate over a long period of time before the customer is notified of them and before they are deducted from the PCA. For this reason, we decided that banks should not allow charges and interest to accumulate to a PCA for more than one month before the customer is pre-notified of the bank’s intention to deduct the charge or interest. The bank must send the customer a statement at the end of any month in which authorized or unauthorized overdraft charges are incurred.

6.134 We understand that not all PCA customers in Northern Ireland receive monthly statements. Some banks (such as [] and []) generally send statements quarterly. Others allow their customers to choose when to receive statements, with the proviso that they must receive at least one statement annually. We do not wish to preclude this from happening. However, we thought it important not only that customers receive a statement at around the same time as they are pre-notified of charges, but also that customers can relate the charge to particular transactions. Thus, we decided that, if in a particular month a customer incurs a charge, the bank must send a statement covering the month in which the charge was incurred and must do so no later than five days after the end of that month.

6.135 We looked at when customers should be pre-notified. We believed that customers should have sufficient time to take action so that they do not incur additional unauthorized overdraft charges. However, it was equally important that the period that elapsed between incurring the charge and taking the charge from a customer’s PCA was not too long. The longer the period, the more difficult it would be for the customer to manage their PCA. We did not believe that the length of time between incurring and deducting the charges should be dependent on the level of the charges. Whilst customers might like to have longer to ensure that sufficient funds were in their PCA to pay for a larger charge, we thought that varying the period for pre-notification according to the level of the charge could lead to customer confusion. We therefore thought that a reasonable balance would be struck by requiring banks to pre-notify customers of charges or interest to be deducted from the PCA at least 14 days before they are deducted from the PCA.188

6.136 We also considered the argument that this remedy would increase the credit risk associated with the customer. We recognize that the implementation of this remedy would result in changes to the current charging processes of banks and in a time delay between the activity giving rise to the charge and the actual application of the charge. We did not, however, believe that this remedy would increase a bank’s credit risk to the extent that it would lead to banks significantly reconsidering their current lending approach. We did not receive any evidence showing that the lending prac-

188Assuming that the bank deducted the charge from the PCA 14 days after the date of the statement, a charge could be taken from a customer’s PCA up to six weeks after it had been incurred.

155

tices of banks that currently pre-notify customers of unauthorized overdraft charges are different or more restrictive than those of the banks that do not pre-notify these charges.

6.137 We considered whether to require banks to pre-notify together with a statement. However, we noted the argument that a separate pre-notification letter may have more impact than the statement itself. We therefore decided that banks should be permitted either to pre-notify with a statement or at a time before the statement is sent. However, we decided that, whether pre-notification occurs with a statement or in advance of a statement, the bank must not deduct the pre-notified charge from the PCA until at least 14 days from the date of the statement.

6.138 We considered in particular the points raised by two banks ([]) about the impli- cations of this remedy and in particular of the link between pre-notification and statements for their quarterly charging cycles. In particular, we noted that it would be difficult under this remedy to accumulate charges over a quarter, and the bank would need to send its customers a statement at the end of each month in which the cus- tomer incurred an authorized or unauthorized overdraft charge or interest payment. For those customers who do not go into overdraft, [] would not necessarily need to adjust their statement and charging cycles, but we accept that where customers do go into authorized or unauthorized overdraft, changes will be required. Overall, however, we remained of the view that the link with statements was an important element of this remedy, and one which should be retained.

6.139 We looked at whether it might be sufficient to ensure that customers had ready access to their account balance to allow the customer to take appropriate action in response to charges. We also specifically considered the suggestion made by [] that it might be sufficient for banks to include the account balance on the pre- notification communication. We accepted that a statement covering the month in which the charge or interest payment had been incurred would not provide a definitive list of the transactions that would be completed on the customer’s account before the charge or interest payment was deducted, but we thought that such a statement would provide the customer with a reasonable basis on which to take a view as to whether it was likely that they would need to take action to ensure that there were sufficient funds in the account to meet the payment.

6.140 We considered the suggestion made by one bank ([]) that the banks should be allowed to choose the most cost-effective form (electronic or hard copy) in which to pre-notify customers. We were concerned that allowing banks this choice could result in customers receiving pre-notification in a form that they could not easily access or with which they were not familiar. However, we accepted that customers may wish to receive pre-notification in a form different from the one in which they routinely receive their statements. Thus, we decided that banks should pre-notify a customer in the same form as that in which the customer routinely received their statement, unless the customer has chosen to receive the pre-notification in a different form.

6.141 This remedy mirrors the requirement being recommended by the BCSB’s indepen- dent directors to extend pre-notification to cover default charges (see paragraph 6.126), but goes further by ensuring that customers have a recent statement available shortly before they are faced with the charge or interest payment being deducted from their account. Since the Code already includes a requirement for pre- notification, and since this may be extended following the review, we thought that the BCSB would be well placed to monitor compliance with this remedy (see paragraphs 6.244 to 6.249).

156

6.142 Paragraphs 6.124 and 6.125 set out the cost estimates provided by the banks. We noted the additional costs that the banks estimated would result from a requirement to pre-notify on statements. However, if this remedy is to be effective in empowering customers to take appropriate action to manage their PCA, we thought it vital that customers have relevant account information to hand when, or shortly after, they receive pre-notification of charges to be deducted from their PCA. We therefore thought it important to mandate a link between pre-notification and statements. In addition, given the requirements of the CCA 2006, and the changes to the Code that are being considered, we were not persuaded that the costs attributed by the banks to the implementation of a pre-notification remedy will be stand-alone.

• Our decision

6.143 We conclude that banks must pre-notify customers of all authorized and unauthorized overdraft charges. Pre-notification may be sent together with a state- ment or in advance of a statement and in the same form unless the customer elects otherwise. However, the banks must not deduct a pre-notified charge from a customer’s PCA until at least 14 days after the date of the statement. If the pre- notification is not sent together with a statement, the bank must provide the key account information that is required under remedy (b) on the pre-notification communication.

6.144 Banks must not allow charges and interest to accumulate to a PCA for more than one month before the customer is pre-notified of the bank’s intention to deduct the charge or interest from the PCA. The bank must send the customer a statement at the end of any month in which authorized or unauthorized overdraft charges are incurred.

6.145 We believe that this remedy would help to address the feature that banks do not fully or sufficiently explain their charging structures and practices which leads to difficulties in customers making informed choices.

6.146 There is a link between this remedy and remedies (b) and (c). This remedy requires banks to provide with the pre-notification an explanation of the level and of why the charge or interest payment that is being pre-notified has been applied. If a bank pre- notifies together with a statement, the information that is provided on the statement could provide this explanation.

Remedy (f): Regular ‘rights reminder’

6.147 In the Notice, we proposed that banks might be required to remind customers regularly, prominently and clearly that they have a right to terminate the agreement under which their PCA is provided by their existing bank and to switch PCA provider (remedy 8). In our provisional decisions on remedies, we clarified this proposal, suggesting that this should be provided as part of the summary and breakdown of charges and interest. We also proposed that it should be accompanied by a general communication to explain the switching process.

• The views of parties

6.148 The views of parties on this remedy were mixed. Some parties did not believe that such a reminder was necessary. They thought that such a reminder would be intrusive, run contrary to the normal business objectives of banks and risk creating information overload. Other parties expressed support for this remedy but they thought that the most effective way to implement it would be for the reminder to refer to the switching process. In particular, [] and [] said that they believed that a

157

standard industry leaflet should be developed to include a message on the customer’s ability to switch bank and the detail of the switching service available to them. Such a leaflet might have more authority and credibility with customers than a bank-specific leaflet.

6.149 Ulster said that it used to include the following text in its ‘switcher brochure’:

We’re confident you’ll never have a reason to leave. At Ulster Bank we believe in making life easy for you. We’ll endeavour to give you excellent banking service and hope that you will never have any reason to leave us. But should you ever decide that you want to leave we will:

• Provide your new bank with a list of your direct debits and standing orders by close of business on the third working day following your request

• Cancel your direct debits and standing orders on the second day following your request

• Transfer your account balance by the third working day, if you want us to

• Close your account within 4 working days, once we have enough funds to clear any overdraft you have (including interest and costs).

6.150 Ulster said that it would not object to reintroducing such text in its marketing literature.189 Although, in its view, it might be counter-productive to developing a long- term relationship with customers, and such a message might be better conveyed by means of a standard leaflet produced by the BBA, Ulster said that, in principle, it supported the remedy.

6.151 Another bank ([]) echoed Ulster’s position, saying that any rights reminder should be included in marketing material but not on the annual summary of charges and interest. This bank also supported the idea of a generic leaflet explaining the switching process.

6.152 A further bank ([]) expressed support for this remedy and also suggested that the reminder should be provided in a positive way. It said that it was considering a state- ment along the following lines: ‘Remember it’s easy to move banks. Contact us at [] if you have any questions about switching bank.’

6.153 The BBA argued that reminding customers that they could close a PCA and switch to another provider went against customer relationship management. The BBA and one of the banks ([]) said that there no similar precedents in any other industry. However, the BBA said that if there were to be a requirement to remind customers of their ability to switch accounts, we should allow each bank to develop its own text, which could alert customers to the Code’s switching requirements and to alternative accounts available from that bank. It suggested that we should permit banks to send this to customers as part of their generic correspondence and not require it as part of the annual summary of charges and interest.

189The text was removed when the literature was relaunched in October 2006.

158

6.154 The banks provided estimates of the costs of implementing the ‘rights reminder’ remedy, including the distribution to customers of a leaflet setting out the switching process. Six banks provided estimates of one-off costs; these varied considerably, ranging from £7,000 ([]) to £210,000 ([]). The banks’ estimates suggested that the one-off costs of implementing this remedy would be around £700,000 for all banks required to implement it. The banks’ estimates suggested ongoing annual costs of implementing this remedy of, on average, around £25,000 per bank, sug- gesting a total annual cost for the implementation of this remedy of around £250,000 for all those banks required to implement it.

• Our views

6.155 We noted the banks’ reluctance to include a reminder that customers can switch banks in their communications with existing customers. However, we did not believe that it would be sufficient to include such a message only on marketing communi- cations. Although customers may be aware of marketing communications when they open a PCA, customers with existing PCAs—many of which will have been opened some time ago—will not necessarily be aware of their bank’s current marketing materials. An important element in the value of this remedy comes from its effect in stimulating customers, and because of this it may well be more important for customers who have held their PCA for a long time.

6.156 We also noted that—as demonstrated by the Ulster example—it is possible for banks to word such a reminder in a way that presents a positive message to the customer. We did not think that including such a reminder with the annual summary of charges and interest would necessarily damage the banks’ customer relationships. Indeed it could be portrayed as a statement reflecting confidence in the services provided by the bank and an awareness of the choices that customers can make in a competitive market.

6.157 It is our view that the ‘rights reminder’ should be accompanied by a generic communication, such as a leaflet, that explains clearly and simply what is involved in the switching process, making clear what steps customers should take and what obligations on banks exist. This communication could be developed by the banks working together, possibly through the BBA.

6.158 We formed the view that the rights reminder would be most effective if it were provided as part of the same communication as the annual summary and breakdown of charges and interest (remedy (d)). This would mean that customers would receive a reminder of their ability to switch providers at the same time as they had received information on the costs of operating their PCA. We noted that, while maximising the effectiveness of the remedy, this would also minimise its cost. We believed that the additional costs for the banks of including ‘rights reminder’ wording on the annual summary and breakdown of interest and charges and providing a leaflet or electronic communication explaining the switching process would be negligible. In our view, the costs associated with developing the generic communication explaining the switching process should be borne collectively by all those banks that are subject to our remedies.

• Our decision

6.159 We conclude that banks must include wording on the annual summary of charges and interest (see remedy (d)) that makes it clear that the customer can close their PCA with that bank and seek to obtain PCA services from a different bank. It should also specify where the customer can find further information on how to do this. This

159

communication should be accompanied by a generic leaflet or electronic communi- cation which explains the switching process, the fact that it is easy to switch using the banks’ switching services, and that customers are not charged for switching. We believe that this will help to ensure that customers know how to switch and would help to challenge the perception that switching is difficult. This communication could be developed by the banks working together, possibly through the BBA. It may be in electronic form or hard copy, but should be sent to customers in the same form as they receive their summary and breakdown of charges and interest, unless the customer elects otherwise.

6.160 We believe that a regular reminder to customers of their ability to switch, the availability of switching services, and the ease of switching will address the feature that customers do not actively search for alternative PCAs or switch bank. In our view, this remedy complements remedy (d). Together, these two remedies would ensure that customers are aware of how much their PCA had cost them in the last year, and remind customers that they could switch bank.

Remedy (g): Changes to the switching process

6.161 In the Notice, we considered whether it would be appropriate to include a remedy to improve the effectiveness of the switching process in order to address both actual and perceived difficulties in the process (remedy 10).

6.162 In the Notice we consulted on whether it would be possible to devise a remedy to make PCA numbers portable190 within Northern Ireland, allowing customers to switch PCAs between banks without changing their account number. Paragraph 6.215 sets out our views on this remedy. However, following publication of our Notice and remedies hearings with parties, we began to consider other ways of improving the switching process and customer perception of it. In our provisional decisions on remedies, we focused on ways to hold the customer harmless from any errors that might occur during the switching process.

6.163 We noted that most of the problems that occur on switching PCAs between banks relate to the transfer of direct debits, direct credits and standing orders. Although there are relatively few problems, the fact that difficulties occur, and the risk that an individual customer will face them, reinforces the customer perception that switching is difficult and risky. In further consultation with the banks, we therefore explored whether there was scope to improve the process by which direct debits, direct credits and standing orders were transferred between PCAs held with different banks.

• The views of parties

6.164 Some banks ([]) provided anecdotal evidence that certain originators of direct debits and direct credits were often late in transferring transactions to the new bank. However, the data provided on actual failures showed them to be relatively few. In particular, [] provided the results of a study that it had conducted in summer 2006 of non-compliance with the process that originators of direct debits are required to follow when a customer switches from one bank to another. This showed that the lowest level of compliance by any individual originator was 98.99 per cent.

190The concept of ‘number portability’ has been applied in the telecommunications industry as a regulated facility which enables subscribers of publicly available telephone services (including mobile services) to change their service provider whilst keeping their existing telephone number.

160

6.165 From discussions with the banks, we identified three reasons underlying most of the failures to transfer direct debits, direct credits and standing orders in good time:

(a) the customer does not confirm to the new bank which direct debit, direct credit or standing order they wish to transfer;

(b) the originator will seek confirmation from the customer before acting on the instruction of the new bank to transfer the direct debit and direct credit;191 and

(c) at least one of the participants in the process does not use a fully automated system for the transfer of direct debits, direct credits and standing orders. This results in details being transferred manually, creating scope for human error.

6.166 BACS said that there was a greater likelihood of an error in the transfer of a direct debit, direct credit or standing order where one of the banks involved did not use a fully automated system for such transfers. BACS noted that the fully automated system (known as TODDASO4) was not currently available to banks that accessed the BACS system through agency arrangements with settlement members. We understand that two of the clearers ([]) do not use the fully automated system. BACS said that there would be costs involved in moving to a fully automated version of the system.

6.167 In further consultation with the banks, we looked at the extent to which the cus- tomer’s experience of the switching process could be improved without changing the switching process itself, for example by holding the customer harmless against any problems arising as a result of the switching process. This would provide re- assurance to those thinking of switching that the process was not as risky as they might have thought and would ensure that those who did decide to switch would not suffer as a result of any failures in the process.

6.168 We consulted on a proposal to require banks to offer customers an interest- and charge-free overdraft facility for a limited period after switching to them. One bank ([]) said that interest-free overdraft facilities for new account holders were routinely offered by some banks as a means of competing for new customers. [] also pointed out that some customers had personal, ethical or religious objections to over- drafts so that compulsory provision of an overdraft on switching might deter these customers from switching. Others, such as [] and [], said that the Code currently required banks to lend responsibly. Paragraph 13.1 of the Code required banks to assess the customer’s ability to repay credit before offering it (with the implication that customers who were unable to repay should not be offered credit). They said that a mandatory requirement to extend an overdraft to all customers switching to them would place them in breach of the Code. Several banks ([]) also said that some customers might misuse overdraft facilities, increasing the risk of fraud.

6.169 In responding to our provisional decisions on remedies, one bank ([]) also said that the requirement to offer an interest- and charge-free overdraft to a customer would result in that customer undergoing a credit-scoring check each time they switched account. This bank noted that if a customer left many credit-scoring ‘footprints’ in a relatively short period of time, this could negatively affect their credit score, which might in turn act as an impediment to switching not only PCAs but also credit cards, personal loans and possibly mortgages.

191BACS told us that if an originator refused to accept the instruction of the new bank, this constituted a breach of the BACS scheme rules. Such cases were, according to BACS, unusual and were addressed by BACS as they came to its attention.

161

6.170 One bank ([]) proposed an alternative remedy, which it said would address any negative customer perception of the switching process. It suggested that banks could be required automatically to refund any costs incurred by customers within a defined period as a direct result of a failure in the switching process. Paragraph 7.4 of the Code already requires banks to cancel any charges the customer incurs during switching which are the fault of the bank, and to refund any charges already taken from a customer’s PCA where an error by the bank has been identified. However, this suggestion extends the scope of the refund to cover any charges incurred as a result of a direct failure in the switching process, whether through the fault of the new bank, the old bank, direct debit or direct credit originators, or the customer. [] said that any such requirement should be accompanied by a statement from the CC recognizing that there are parties outside the control of the new bank who must discharge their obligations if the switching process is to complete in a successful and timely manner. [] noted that this included direct debit and direct credit originators, as well as customers themselves.

6.171 In responding to our provisional decisions on remedies, one bank ([]) argued that requiring the banks to hold new customers harmless for a period of three months in all cases was inappropriate. It said that 98 per cent of its switches were completed within 70 days and argued that banks should be entitled to return customers to normal account conditions once the switch was complete, if it was completed within the three-month period.

• Our views

6.172 Having considered the likely causes of failure in the switching process set out in paragraph 6.165, we formed the view that it would not be possible to eliminate all possibility of error from the process. We therefore focused on the possibility of holding the customer harmless for any errors that occurred.

6.173 We thought that requiring banks to offer overdrafts, without any obligation on customers to accept the offer, would address the concerns of some customers about overdrafts. We thought that it would be important to ensure that the size of the overdraft facility provided the customer with meaningful protection against deficits in the switching process. For example, a customer who pays in a monthly salary of £1,500 and who has a monthly mortgage payment of £600 would require a greater overdraft than a customer with a monthly salary payment of £700 and a monthly mortgage payment of £300. The size of the overdraft should therefore reflect the customer’s usual level of transactions, while being consistent with the bank’s credit- scoring policy.

6.174 However, we acknowledged that a requirement on banks to offer an overdraft to any customer that requested it might be in breach of the banks’ requirement for responsible lending. We took the view that allowing banks to offer overdrafts in line with their usual credit-scoring policies should ensure that the banks are not required to violate the principles of responsible lending. We accepted that not every customer could be offered an overdraft or would wish to have one. We therefore also looked at the proposal set out in paragraph 6.170 and noted that this would go a long way to address the actual and perceived difficulties associated with the switching process. Where a customer is not eligible for such an overdraft or does not wish to have one, we thought that the banks could instead be required automatically to refund any costs incurred by the customers within a defined period as a direct result of a failure in the switching process. In our view, this would not provide the same level of protection as an overdraft facility, since a bank may still decline an important payment if the salary or benefit has not been paid into the PCA. But it would provide those customers who do not want or would not be offered an overdraft with re-

162

assurance that they would not incur costs as a result of any problem with the switching process.

6.175 We looked at the period during which customers should be ‘held harmless’. The Code currently suggests that a switch should be completed within 14 days. Given that most key transactions on a PCA happen no less frequently than monthly, we noted that it could be six weeks before a problem in the switching process became visible. The ‘hold harmless’ period must allow sufficient time for any problem to be resolved. This was why our provisional decisions on remedies set out that the ‘hold harmless’ period should be three months.

6.176 We considered whether banks should be permitted to return switching customers to normal account conditions within three months if the switch is complete. We accept that once a switch has been successfully completed, a customer will no longer have need of any ‘hold harmless’ safeguards. However, it is important to note that one of the main aims of this remedy is to address the perception among customers that switching is risky, so that the effectiveness of this remedy depends on a strong and simple message being sent to customers. We therefore take the view that it is important to stipulate an absolute period for any interest- or charge-free overdraft as well as for any indemnity against costs incurred as a result of direct failures in the switching process. We therefore decided to specify the ‘hold harmless’ period as a fixed period of three months.

6.177 We found that the cost associated with any additional credit risk in relation to the ‘hold harmless’ remedy was unlikely to be significant. On the basis that the switching process generally operates smoothly, we did not expect a requirement to refund charges resulting from errors in the process to be significant either.

• Our decision

6.178 We conclude that, for each Northern Ireland customer who uses the switching service to open a new PCA, banks must offer an interest- and charge-free overdraft facility, the amount of which is commensurate with the expected transactions on the PCA, for a minimum of three months after the PCA is opened, provided the customer would be eligible for such a facility under the bank’s usual credit-scoring policy. Where a customer is not eligible for such an overdraft or does not wish to have one, banks must guarantee to refund the customer any charges and interest which are incurred within a minimum period of three months after the PCA is opened as a result of a failure in the switching process. This is regardless of whether the charges and interest were incurred as a result of an error by the new bank.

6.179 This remedy would represent a significant change to the switching process. In order to address negative customer perceptions about switching, it would need to be communicated effectively. We therefore conclude that banks must publicize the ‘hold harmless’ provisions in their switching literature.

6.180 This remedy will address the feature that customers generally do not actively search for alternative PCAs or switch bank. It would address both actual and perceived difficulties in the switching process, thereby adjusting the balance between the ‘costs’ and the ‘benefits’ from switching.

6.181 In addition, whilst we did not decide on a remedy to address the switching process itself, we recommend that BACS should review the switching process, with a view to identifying and addressing any outstanding impediments to switching direct credits, direct debits and standing orders.

163

Remedies that we have decided not to pursue

6.182 We have decided not to pursue the following remedies:

(a) Remedy (h): Provision of personal indicative ‘quotes’ for PCAs (see paragraphs 6.184 to 6.190).

(b) Remedy (i): Provision of ‘typical customer’ indicative quotes for PCAs (see para- graphs 6.191 to 6.202).

(c) Remedy (j): Switching statistics (see paragraphs 6.203 to 6.212).

(d) Remedy (k): Changes to the switching process: number portability (see para- graphs 6.213 to 6.215).

We also discuss the direct regulation of PCA prices (see paragraphs 6.216 to 6.219) and the additional remedies proposed by third parties (see paragraphs 6.220 to 6.223).

6.183 We discuss each of these remedies in turn. For each remedy option, we outline the proposed remedy, responses to the proposal, and our views on the proposal.

Remedy (h): Provision of personal indicative ‘quotes’ for PCAs

6.184 In the Notice, we proposed a remedy under which a bank would be required to provide a prospective customer with personalized information on the charges and interest rates that might apply to their new PCA (remedy 6). This would be based on their existing PCA usage over a certain period. We thought that the provision of such personal indicative quotes would make it easier for customers to compare PCAs and support switching, thereby making it more likely that customers would actively consider switching.

• The views of parties

6.185 Some parties expressed strong concerns that the provision of such quotes might be misleading as customers’ past behaviour would not necessarily be an indication of their future behaviour. The BCSB said that banks would need to surround such quotes with caveats to avoid liability to compensation where PCA usage differed in practice from that assumed in the quote, or where the bank exercised its right to vary its charging structure. Such caveats could, in its view, undermine customers’ trust in the quotes, thereby reducing their benefit.

6.186 Which? and [] said that this remedy could increase customers’ search costs. They said that in order to get comparative information sufficient to inform a switching decision, customers would need to get personal indicative quotes from a number of different banks. Rather than personal indicative quotes reducing the inconvenience of switching, they could therefore make the process more arduous.

6.187 Most banks also said that this remedy would be very expensive to implement because of the information technology resources required to build a model for the development of the provisional quotes and the additional staff required to provide these quotes to potential customers.

164

• Our views

6.188 We noted that customers considering switching bank must research the charges and interest rates offered by other banks and then calculate the way in which these may be expected to apply to them given their PCA usage. We noted that comparison websites, such as uSwitch and moneysupermarket.com, require the customer to collect and input a substantial amount of information and do not compare all aspects of PCAs, for example charges for unauthorized overdrafts.

6.189 We were not persuaded that the provision of personal indicative quotes need impose a significant burden on customers. However, we acknowledged that personal indica- tive quotes may mislead customers, since customers’ past usage may not be a good guide to future usage. In particular, a large part of the cost of running a PCA can be associated with going into unauthorized overdraft, and this is likely to be un- predictable. Furthermore, customers may be reluctant to admit that they will go into unauthorized overdraft, at least with any frequency or persistence.

6.190 In any event, we reached the view that, if remedies were implemented that had the effect of making information on the interest and charges applied to PCAs more easily understandable, customers would find it easier to compare the cost to them of operating a PCA with different banks. This would reduce the need for additional measures to address this. We also noted the apparently high implementation costs of this remedy. On balance, we decided not to pursue a remedy based on personal indicative quotes.

Remedy (i): Provision of ‘typical customer’ indicative quotes for PCAs

6.191 In the Notice, we looked at whether banks should be required to provide indicative quotes based on a number of ‘typical customer’ profiles as an alternative or an adjunct to the provision of personal indicative quotes (remedy 7). This remedy would address the feature that customers generally do not actively search for alternative PCAs or switch bank by reducing search costs. The information provided would allow customers to identify which ‘typical customer’ they most closely resembled, and then to see quickly and easily what sort of charges and interest they would expect to pay in operating their PCA with different banks.

• The views of parties

6.192 Generally, parties thought this remedy to be more feasible and less costly than a remedy based on personal indicative quotes.

6.193 Many parties said that, if we were minded to pursue this remedy, the number of ‘typical customer profiles’ should be limited. Some parties suggested a maximum of three profiles:

(a) customers who are always in credit;

(b) customers who operate in credit or within an authorized overdraft; and

(c) customers who go overdrawn without permission or exceed an authorized overdraft.

6.194 In contrast, GCCNI recommended developing indicative quotes for the following six customer profiles:

165

(a) customers regularly in unauthorized overdraft;

(b) customers occasionally in unauthorized overdraft;

(c) customers regularly in authorized overdraft;

(d) customers occasionally in authorized overdraft;

(e) customers always in credit; and

(f) customers with large credit balance.

6.195 Several parties suggested that the FSA could play a role in collecting the information on typical profiles and in publicizing them through the media and on its website, thus helping to increase the credibility of the typical quotes.

6.196 Many parties nevertheless expressed strong reservations about the provision of ‘typical customer’ indicative quotes for PCAs, questioning whether it would be poss- ible to establish realistic customer profiles and therefore provide customers with a reasonable estimate of the costs of running a PCA. One bank ([]) noted that the quotes would need to contain so many caveats that they would be difficult for cus- tomers to use. Other banks said that the provision of typical customer quotes might lead to account engineering, such that banks adjusted the design of their PCA to ensure that the quotes were as advantageous as possible, or failed to innovate because the quotes did not show other unique account attributes and/or banks’ service levels.

6.197 The FSA noted that typical quotes for PCAs were not straightforward, as the gains depended on individual customer usage. It questioned the extent to which typical quotes would add value to what was already publicly available. It said that the priority should be to increase customer awareness of existing opportunities for customers to compare PCAs (for example, using organizations such as uSwitch or moneysupermarket.com) rather than developing new opportunities. The FSA added that it had considered establishing comparative tables for PCAs, but it had identified other priority areas.

6.198 Which? said that customers might not be sufficiently aware of the existence of the typical quotes for this remedy to be effective. As an alternative it proposed that banks should be required to send customers an annual document comparing the costs of PCAs offered by all banks. In its response to our provisional decisions on remedies, Which? said that although capturing each consumer’s consumption pattern would be an ‘impossible ideal’, there would be value in having quotes based on particular scenarios, which could give consumers a useful indication of the advantages and disadvantages of different products. Which? said that it had experience with PCA comparisons and was not convinced that they would mislead consumers. It noted that, as a general rule, the PCA products that did best under one ‘best buy’ scenario, did well in all scenarios. It suggested that banks should be required to send con- sumers scenario-based comparisons on a yearly basis, preferably with the annual summary of charges and interest and the rights reminder. Which? also suggested that, if we declined to pursue typical customer quotes, we could require the inclusion of references to comparison websites such as Which?, moneysupermarket.com or uSwitch on statements, and on the annual summary of interest and charges.

166

• Our views

6.199 In order to define typical customer profiles, various assumptions would need to be made. These would include the amount credited to the PCA each month, the amount of authorized overdraft, the amount of unauthorized overdraft, the frequency with which the PCA goes into unauthorized overdraft during a year, the number of paid transactions and the number of unpaid transactions. Furthermore, the profiles would require frequent updating to reflect changes in interest rates and charges, changes in the structure of charges and the introduction of new PCAs.

6.200 Even if we accepted that, as suggested by Which?, those PCA products that do best in ‘best buy’ tables may attract customers, ‘best buy’ tables do not tell individual consumers which PCA is best for them on the basis of their actual usage. Unless a customer’s PCA usage largely conformed to the assumptions underlying the profiles—and it is unlikely that many would—they would find it difficult to use the typical quotes to get a reasonable view of what it might cost them to run different PCAs. Given the evidence that we received in response to our Notice, we formed the view that typical quotes might mislead customers as to the actual costs they would face in running a PCA, and this could work against some of the remedies aimed at ensuring that customers have clearer information about PCA interest and charges.

6.201 We thought that such a remedy would be of limited value, and decided not to pursue it further.

6.202 We considered the suggestion from Which? that we require banks to include references to comparison websites on their statements or on the annual summary of interest and charges. However, we noted that there any many ways in which customers may make comparisons between PCA products. We believe that, once customers are reminded of the possibility of switching PCAs, and have been prompted to do so by greater awareness of the cost of running their PCAs, they will be able to judge for themselves the best way in which to compare different products.

Remedy (j): Switching statistics

6.203 In the Notice, we considered whether each bank should be required regularly to publish statistics on the number of customers that have switched from and to it in a given period (remedy 9). This would address the feature that customers generally do not actively search for alternative PCAs or switch bank. We thought that, if these statistics were publicized, they might raise customers’ awareness of their ability to switch bank and also encourage banks operating in Northern Ireland to compete in attracting customers. We also considered whether, as an alternative to the publication of switching statistics, an industry-funded advertising campaign on the switching process and switching services would be more effective in raising customers’ awareness of the possibility and the ease of switching bank.

• The views of parties

6.204 Most banks expressed concerns about a remedy mandating the publication of switching statistics of individual banks. They argued that its implementation could be hindered by practical considerations such as how to deal with statistics for customers who opened new PCAs but did not close their old ones, or who might run two or more PCAs in parallel for some considerable time. Generally, the banks felt that it would be difficult to devise statistics that would give a complete and accurate representation of the actual level of switching in the market.

167

6.205 Several banks ([]) expressed doubts about whether the publication of switching statistics would be effective in raising customers’ awareness of the switching process and in changing their perception of it. One bank ([]) said that a requirement for banks to publish statistics on their performance against certain defined industry benchmarks might be more effective to reassure potential switchers that any appli- cation to switch would be dealt with in a timely and professional manner. Which? said that it might be difficult for customers to interpret statistics accurately and said that switching in itself did not guarantee that a customer would benefit.

6.206 In contrast, GCCNI and Advice NI supported the publication of switching statistics. GCCNI suggested that the FSA could collate and publish the statistics, which should be published annually. It suggested the publication of both statistics on the number of complaints about the switching process received by each bank and on the number of complaints that had to be passed on to the Financial Ombudsman Service.

• Our views

6.207 We identified the following types of switching statistics that banks could collect relatively easily and make available for publication:

(a) number of customers that have switched from and to each of the banks during a year and have used the switching services; and

(b) data on banks’ shares of the PCA market provided by third parties (eg MORI).

6.208 We accepted that this data would not capture customers who switched by opening a new PCA without closing their old PCA, nor would it capture customers who switched PCA without using the switching service. However, we thought that the statistics could still raise awareness of the possibility of switching PCA and of using the switching service.

6.209 However, we noted that the effectiveness of this remedy would depend critically on media interest in these statistics; if the statistics were not widely reported, they would not raise awareness of switching. We thought that the generally low level of interest in PCAs among customers made it unlikely that the statistics would attract sufficient media attention. Furthermore, if the number of switchers was relatively low, the statistics might reinforce a view that switching PCAs was burdensome or risky. We therefore decided not to pursue this remedy.

6.210 We looked at the proposal to publish statistics on the number of complaints about the switching process received by each bank. We thought that the publication of numbers of complaints would be beset by similar difficulties as the publication of switching statistics, for example in relation to the accuracy and interpretation of data. Furthermore, we did not believe that, even if implemented successfully, this would address the feature that customers generally do not actively search for alternative PCAs or switch bank. We therefore concluded that the remedy mandating the publication of switching statistics should not be pursued further.

6.211 We noted the argument made by some banks ([]) against a remedy involving an industry-sponsored advertising campaign to raise awareness of the possibility of switching, and the ease of switching. Although, in theory, an industry-sponsored advertising campaign might raise awareness of switching and the ease of switching, much would depend on the way in which the campaign was devised and imple- mented. We saw no merit in being overly prescriptive about the content of such a campaign. However, we were concerned that the banks would have little incentive to promote switching generally as opposed to switching to their PCAs. Overall, we

168

thought that our package of remedies should be sufficient to encourage banks to promote switching as part of their own advertising in a more competitive market.

6.212 We therefore decided not to pursue a remedy that would require the banks to fund an industry-wide campaign in relation to switching.

Remedy (k): Changes to the switching process: number portability

6.213 In the Notice we consulted on whether it would be possible to devise a remedy to make PCA numbers portable within Northern Ireland allowing customers to switch PCAs between banks without changing their account number (remedy 10). We noted that a similar solution had been implemented in the Netherlands and wished to explore whether it would be possible in Northern Ireland, and, if so, what might be involved.

• The views of parties

6.214 All parties viewed a ‘number portability’ remedy as very expensive, cumbersome to implement, and disproportionate. Most parties pointed out that PCA number port- ability could not be implemented on the current UK clearing infrastructure. The current infrastructure operates on the basis of a unique bank/branch identifier number () which is part of the PCA number. The remedy, therefore, would require changing the use of sort code and account number across the whole of the UK clearing system. It might also increase the risk of fraud.

• Our views

6.215 We decided that the introduction of bank account number portability in Northern Ireland would not be practicable. We also noted that when number portability had been introduced in the Netherlands, it had been as part of a fundamental redesign of the clearing system as a whole. We therefore decided not to pursue further the idea of improving the switching process by introducing number portability.

Direct regulation of PCA prices

6.216 We found it likely that, as a result of the adverse effect on competition of the features we identified, the charges and debit interest paid by customers for PCAs in Northern Ireland would be higher, and the levels of credit interest lower, than they might be in a well-functioning market.

6.217 The CC has stated in its guidance192 that in considering remedies it will first look for a remedy that would be effective in dealing with the AEC of the market features rather than seeking to deal with any detrimental effect on customers. We sought remedies that would stimulate and facilitate the emergence of a more competitive market for PCAs in Northern Ireland, in preference to remedies that would deal only with the symptoms of a lack of effective competition.

6.218 Furthermore, we noted that PCAs are complex multi-functional products that consist of a number of different components. It is therefore not clear which prices would need to be controlled adequately to address the AEC or any resulting detrimental effect on customers or how they could effectively be measured. We thought that there was a

192CC3, paragraph 4.22.

169

risk that price controls in this market could result in a ‘waterbed effect’ whereby controls on some elements of the PCA could see banks raise the prices of other elements to maintain revenues overall. We also noted that it would have been necessary to control not only the prices of PCAs, but also the terms and conditions of those PCAs and the standards of service, to ensure that the banks did not adjust these other aspects of the PCA to compensate for controls on prices.

6.219 In response to the Notice and to our provisional decisions on remedies, none of the parties argued that price regulation would address the AEC that we had identified. We decided not to pursue further any remedy based on price regulation.

Additional remedies proposed by third parties

6.220 GCCNI, Which? and Advice NI proposed additional remedies to those set out in the Notice. GCCNI suggested that the CC:

(a) require the banks to adhere to the law and set their penalty charges accordingly;

(b) require the banks in Northern Ireland to comply with the Unfair Commercial Practices Directive; and

(c) establish a minimum competency standard for bank staff offering information on PCAs.

6.221 GCCNI said that (a) and (b) in paragraph 6.220 would address our concern that banks might be charging disproportionately, and that its third proposed remedy (paragraph 6.220 (c)), would address the complexity of charges and cost structures applied by banks. After discussion, however, GCCNI agreed that, since there was no clear link between our findings and GCCNI’s proposed remedies, it would not be appropriate for us to pursue its suggested additional remedies.

6.222 Which? said that it considered access to branches to be an important feature of the market. It therefore suggested that we consider a remedy requiring banks to share branches to facilitate entry and expansion for parties that had limited branch presence. Advice NI said that PCAs should be made accessible through Post Offices to increase access to banking for vulnerable and financially excluded customers. It also said that banks should include links to advice centres on their websites so that customers have information on where to go if they get into financial difficulties.

6.223 We did not find that a lack of access to the branch network, or lack of information on assistance available to customers in financial difficulty, were features of the market that gave rise to an AEC. It was not apparent to us, therefore, that either of the additional remedies proposed by Advice NI would facilitate switching or otherwise address the features that we had identified.

Scope of remedies

6.224 We found features that apply across the market for PCAs in Northern Ireland which prevent, restrict or distort competition and therefore result in an AEC. Most of our package of remedies (in particular, remedies (a), (b), (c), (d) and (e)) is designed to ensure that customers have good-quality information about their PCA in order to take

170

well-informed decisions at appropriate times.193 As far as possible, any customer in Northern Ireland should have access to this information.

6.225 However, we considered whether every bank providing PCAs in Northern Ireland should fall within the scope of our remedies or whether, for reasons of practicality or proportionality, the scope of our remedies should be limited in some way by intro- ducing a threshold below which banks would not be required to implement the remedies package. There are a number of banks with a very small presence in Northern Ireland. These include some of the Great Britain-based clearing banks (such as Lloyds TSB) which have no branch presence in Northern Ireland and only a small number of PCA customers with addresses in Northern Ireland. We noted that applying our remedies to all banks with PCA customers in Northern Ireland could lead to a bank falling within the scope of the remedies simply as a result of some of their customers from Great Britain moving (perhaps only temporarily) to Northern Ireland.

6.226 The banks said that much of the estimated cost of implementing the proposed remedies was fixed and therefore did not vary significantly according to the number of PCA customers. Many banks said that it would be difficult for them to implement the remedies in Northern Ireland alone, as changes would be required to systems that operate across a wider geographical area (for example, the UK as a whole). As noted in paragraph 6.74, one bank ([]) said that the remedies package, and remedy (c) in particular, would be disproportionate for UK-wide banks with a small proportion of their UK customer base in Northern Ireland. Another bank ([]) said that any remedies package should be applied across the UK as a whole. We noted that the terms of reference of our investigation related to the market for the supply of PCA banking services in Northern Ireland. However, whilst we could not mandate that remedies were applied more widely, banks might choose to implement the remedies across a wider geographical area, to the benefit of all their customers. These costs are considered in more detail in the discussion of proportionality (see paragraphs 6.256 to 6.270).

6.227 We considered what form any threshold might take. We thought it important that our remedies should cover the vast majority of PCAs in Northern Ireland. We considered whether the threshold might be expressed in terms of market share or number of PCA customers. Market shares can vary from year to year. More importantly, banks cannot easily control or predict their market share, since market shares are dependent, in part, on the actions of other banks. We believed, therefore, that market shares would not be an appropriate basis for any threshold.

6.228 By contrast, the absolute number of PCAs that any bank has where the customers have a Northern Ireland address will be known to each bank. We recognized that basing the threshold on the number of PCAs may result in one customer being counted more than once if they held multiple PCAs with the same bank. However, on the basis of the information provided by the banks for our investigation, the number of PCAs held by customers with a postal address in Northern Ireland was likely to be easier for the banks to determine than the number of customers they have in Northern Ireland. We therefore decided that, on practical grounds, any threshold should be based on the number of PCAs that a bank has for which it has customers with postal addresses in Northern Ireland, even though certain customers might be counted more than once and might in this sense inflate the number of customers of a particular bank or banks.

193Remedies (a) and (b) are also aimed at ensuring provision of good-quality information for customers before opening a PCA.

171

6.229 We considered whether accounts held by different bank ‘brands’ within the same corporate group should be counted together for the purpose of these calculations. We noted that banks within the same corporate group have the same sort code and, often, the same systems. We also noted that it would be undesirable if, by creating new ‘brands’, banks could evade the scope of the remedies and thereby undermine their effectiveness. We therefore thought that PCAs held by banks within the same corporate group (such as Abbey, cahoot and Cater Allen) should be counted together for the purpose of the threshold calculations.194

6.230 In response to our provisional decisions on remedies, [] asked for clarification as to whether, if a corporate group as a whole had more than the specified number of PCAs held by customers with a Northern Ireland postal address, all UK banks within that corporate group would automatically fall within the scope of the remedies.

6.231 We accept that it would be disproportionate to require a bank with a very small number of PCAs held by customers with a Northern Ireland postal address to imple- ment our remedies. In addition to applying a threshold in relation to the number of PCAs held by customers with a postal address in Northern Ireland served by each corporate group, we therefore thought we should also determine the scope of our remedies taking into account the number of PCAs held by customers with a postal address in Northern Ireland served by each individual bank within the corporate group. Thus, a corporate group that had more than a specified number of PCAs held by customers with a postal address in Northern Ireland would fall within the scope of our remedies and a bank within that group would have to apply those remedies, unless it could demonstrate that it had fewer than some specified lower number of such customers. In short, we looked at possible thresholds at both the group and the individual bank levels.

6.232 Considering first the threshold that might be applied at the level of the corporate group, we found that excluding those corporate groups with fewer than 10,000 PCAs held by customers with a postal address in Northern Ireland would result in the application of the remedies to well over 99 per cent of PCAs in Northern Ireland.195 We also noted that there is a significant difference between the number of Northern Ireland PCAs served by the corporate group with the smallest number of PCAs above the proposed threshold and the corporate group with the largest number of PCAs which fell below the threshold. []

6.233 We thought that a threshold at this level would safeguard the ongoing effectiveness of our remedies. With a threshold of 10,000 PCAs it seemed likely, for example, that, even were there to be a high level of market entry or expansion, the vast majority of PCAs in Northern Ireland would be covered by our remedies.

6.234 To assess the reasonableness of a 10,000 PCA threshold at the corporate group level, we considered whether the threshold might be higher. Taking into account current PCA numbers, we examined the effect of applying our remedies only to those banks with more than 30,000 PCAs.196 The use of a 30,000 threshold would reduce the current coverage of our remedies from more than 99 per cent of Northern Ireland PCAs to just over 98 per cent. We noted that 30,000 customers represents more than 2 per cent of the total number of PCAs in Northern Ireland. This higher threshold

194This is consistent with the general approach in competition law that companies within the same corporate group are generally considered as forming part of the same ‘undertaking’. 195It was estimated that such a threshold would result in the application of remedies to 99.7 per cent of PCAs in Northern Ireland. 196The difference in the number of PCAs between [] and [] means that any threshold between 10,000 and 20,000 PCAs would result in the remedies having the same coverage.

172

therefore has the potential to exclude a relatively large absolute number of PCAs from the scope of the remedies. If there were significant market entry or expansion, a substantial proportion of PCAs might be excluded from the scope of the remedies. We therefore thought that a threshold at this level might unreasonably weaken the effectiveness of our remedies over time. We also considered whether the threshold at the corporate group level should be lower than 10,000. However, given that we were satisfied that a threshold set at 10,000 PCAs would ensure the effectiveness of the remedy, we saw no need to set a lower threshold at the corporate group level.

6.235 In considering the level of the threshold that might be applied at the bank level, we noted that applying a threshold so that banks with fewer than 5,000 PCA customers with a postal address in Northern Ireland did not have to apply the remedies would slightly reduce the application of the remedies, but would still result in coverage of more than 98 per cent of PCAs in Northern Ireland. We therefore decided that an individual bank threshold of 5,000 PCA customers would not compromise the effec- tiveness of our remedies. We also noted that, especially where those banks share systems with sister banks that must apply the remedies, corporate groups may choose to apply the remedies even to those banks with fewer than 5,000 PCAs held by customers with a postal address in Northern Ireland. We considered whether the threshold at the individual bank level should be set at less than 5,000 PCAs. However, given that we were satisfied that a threshold of 5,000 PCAs would ensure the effectiveness of our remedies, we saw no need to set a lower threshold. Having considered thresholds at the individual bank level of above and below 5,000, we concluded that a threshold at 5,000 struck the right balance between the need to ensure the effectiveness of our remedies and our wish not to burden banks with a very limited presence in Northern Ireland with undue cost.

6.236 We were mindful of the need to ensure that determining the threshold should not become overly burdensome. We were therefore keen to establish a threshold that would provide a clear basis for the application of the remedies without entailing significant additional cost for the banks. We therefore decided that the banks should be assessed on their numbers of PCAs held by customers with a postal address in Northern Ireland on a specified date each year (the specified date). A bank that is in a corporate group that has 10,000 PCAs held by customers with postal addresses in Northern Ireland and which itself has more than 5,000 such PCAs on that date would be considered as being covered by the remedies for that year. The date specified would depend on the date of implementation of the remedies order.197

• Our decision

6.237 Taking into account the practicalities and proportionality of the application of our remedies, we decided the following:

(a) It is appropriate to establish a minimum threshold. Banks with very few PCAs in Northern Ireland will therefore not fall within the scope of our remedies.

(b) This threshold will be based on the absolute number of a bank’s PCAs held by customers with a postal address in Northern Ireland.

197By way of example, given that the earliest date that we have required for implementation of elements of our package of remedies is 1 April 2008, the specified date might be 30 September 2007 (and likewise 30 September in subsequent years) to provide time for the banks to know in advance whether the threshold would apply to them and to make appropriate changes to their systems.

173

(c) A bank that is within a corporate group that has more than 10,000 PCAs held by customers with a postal address in Northern Ireland will be required to apply the remedies, unless that bank can demonstrate that it has fewer than 5,000 PCAs held by customers with a postal address in Northern Ireland.

Implementation of remedies

6.238 Seven banks ([]) provided us with estimates of the time it would take them to implement the remedies package.198 The implementation time varied between the different remedy options driven largely by the systems changes a particular bank would need to make to implement each of the remedies.

6.239 Having considered the detailed proposals in our provisional decisions on remedies, some banks produced further estimates of the time needed to implement our remedies. Two banks in particular raised concerns. One of these banks ([]) noted that we had taken into account the interaction of our remedies with other regulatory changes, such as the CCA 2006, but said that the number of regulatory changes that it would be required to make in the next 12 to 18 months meant that it would not be able to commit sufficient resources to implement remedies (b) and (c) until some- where between the first and third quarters of 2009 at the earliest. The same bank noted that it would not be able to begin storing the data that our annual summary of charges and interest remedy would require until Q1–Q3 2009, which would mean that the first annual summaries would not be sent out to customers until somewhere between the first and third quarters of 2010. Given the link between this remedy and remedy (f), this would also mean that customers would not receive a rights reminder until this time. This bank said that it would envisage considerable difficulties in implementing our pre-notification remedy before the fourth quarter of 2008.

6.240 Another bank ([]) said that we should postpone the implementation of our remedies until such a time as wider regulatory developments had become clearer and more certain, thereby enabling more efficient implementation of system develop- ments. It had particular concerns in relation to two of our remedies. In relation to remedy (d) it expected that two to three years would be required for implementation, with IT requirements alone amounting to around seven to ten man-years. In relation to remedy (e), this bank said that the link between pre-notification and statements, and the requirement for statements to be sent for each month in which charges are incurred, meant that implementing the remedy would entail a change in its entire banking platform and business model. This bank estimated that the implementation of remedy (e) would require 11 to 15 man-years development and testing time. [] This bank said that regulatory developments and other group projects were relevant to the implementation of our package of remedies. It noted that it had more than 150 projects scheduled for the next two to three years, and that these projects required prioritization and were dependent on IT resources being available.

6.241 As discussed in relation to the individual remedy options, we thought that many of our remedy options would build on existing requirements under the Code, and also that they were in line with many of the views on the desirable changes or additions to the Code submitted in the course of the current review. Others will be required under the CCA 2006. We understood that the BBA was likely to indicate in autumn 2007 what changes will be made to the Code, and that these changes would be expected to be implemented from April 2008. We noted that the Code review and the CCA 2006 would be likely to require banks to make changes to their systems irrespective

198With the exception of remedy (g).

174

of the remedies package set out in this report. We thought that we should allow the banks scope to implement the changes required as a result of our remedies alongside those changes that are likely to be required as a result of the Code review and the implementation of the CCA 2006. Noting the need to ensure that our remedies are implemented in a timely fashion, we therefore said in our provisional decisions on remedies that our package of remedies should be implemented by April 2008.

6.242 In their responses to our provisional decisions on remedies, many banks ([]) said that our remedies requiring the banks to ensure that communications with customers are easy to understand and to provide explanations of the levels of charges and interest rates and how and when they are applied (ie remedies (a) and (b)) could be implemented by April 2008. One bank ([]) said that it could work to implement remedy (a) by the proposed date of April 2008, but would not be able to implement remedy (b) before 2009. After our provisional decisions on remedies had been published, we learned that the changes required under the CCA 2006 would not need to be in place until 1 October 2008.199 Some banks drew our attention to this in their responses, and noted that it would be helpful if the process of implementing our remedies could be adjusted to reflect this. We noted that many banks had said that they could implement remedies (a) and (b) by April 2008. However, we accepted that other remedies within our package might involve systems changes that would be best implemented alongside any changes necessary following the Code review and the requirements of the CCA 2006, certain of which are not now expected to be implemented until October 2008. We noted the arguments put forward by some banks that we should allow additional time for the implementation of our remedies beyond October 2008. However, we did not believe that consumers should be deprived of the information they need to make the right choice about their PCA and bank as a result of decisions made by the banks about the priority assigned to different IT projects and allocation of resources.

• Our decision

6.243 We conclude that we should require banks to implement remedies (a) and (b) by 1 April 2008, and require them to implement the other remedies within our package by 1 October 2008. We believe that banks will in practice be able to implement many of the changes required by these other remedies within our package before October 2008, and we therefore formed the view that this was a ‘long-stop’ date for the implementation of our remedies.

Monitoring and enforcement of remedies

6.244 We believe that monitoring arrangements are important in ensuring the effectiveness of our remedy package. The OFT has a statutory role in relation to the monitoring200 and enforcement201 of remedies. However, in line with the principles of better regu- lation, we do not wish to impose any unnecessary additional regulatory structure on an already highly regulated industry, provided we are satisfied that existing regulatory structures and processes could be used to monitor and enforce compliance effectively.

199DTI responses of March 2007 (URN 07/698) to the consultation on the draft CCA statutory instruments. 200Section 92 of the Act. 201Section 94(6) of the Act. The CC also has a role in respect of enforcing remedies under section 94(7) of the Act.

175

6.245 Most of our package of remedies relates to the provision of information by banks to customers in Northern Ireland. The Code already places requirements on banks to provide certain information to customers, and the BCSB monitors compliance with the Code. We also note that the remedy which requires banks to ‘hold harmless’ customers during the first three months of the switching process could be seen as an extension of the existing requirement of the Code (section 7.4) on a bank to waive any charges that a customer incurs as a result of an error by it during the switching process. We therefore believed that the BCSB would be well placed effectively to monitor compliance with these remedies alongside its monitoring of compliance with the Code.

6.246 Some elements of our package of remedies dovetail with new requirements under the CCA 2006; in particular, the requirement to extend existing pre-notification requirements to cover unauthorized overdraft charges. The BBA said that the next set of revisions to the Code will include changes that take account of the new requirements under the CCA 2006. We thought that this strengthened the case for the involvement of the BCSB in monitoring our remedies because it affords the opportunity to bring together monitoring of the banks’ compliance with the Code, with the CCA 2006 and our remedies.

6.247 The BCSB monitors compliance with the Code across the UK and said that it allocated resources using a risk-based approach. We noted that the BCSB monitoring team is relatively small (just five or six people) and that this could mean that only a relatively small part of its monitoring resource is focused on Northern Ireland. We therefore explored with the BCSB the possibility that the banks could be required to fund additional resource for the BCSB that would be dedicated to monitoring compliance with our remedies in Northern Ireland. The BCSB said that it was not clear to it at this stage what additional work it would need to do in order to monitor compliance with our remedies. It said that it would support a requirement on the banks to fund such additional resource as it reasonably required in order to fulfil any additional role. It could not be more specific at this stage about whether and what additional resourcing it would need without knowing the overall requirements of each of our investigation, the Code review and the CCA 2006.

6.248 We thought that the BCSB would need to work closely with the OFT in order to ensure that the BCSB is effective in its monitoring work. In addition to its statutory responsibilities, we believe that the OFT’s involvement in this way would alleviate any concerns that might exist about the independence of the BCSB from the banks that it would be monitoring. This is not to doubt the independence of the BCSB, whose board is comprised of seven independent directors and three representative directors of banking industry stakeholders. The OFT’s and the CC’s powers under the Act to enforce the package of remedies would be unaffected by the nature of these monitoring arrangements.

• Our decision

6.249 We conclude that the BCSB should monitor compliance with our remedies. The banks should fund additional BCSB resource as appropriate to enable it effectively to monitor compliance with the remedies package. The BCSB should provide the OFT with regular written reports to the OFT’s specification. At a minimum, it is likely that the OFT would require the reports to set out what the BCSB has done to monitor compliance, whether it has found any instances of non-compliance and, if so, whether it has managed to secure compliance.

176

Effectiveness and proportionality

6.250 On the basis of the above consideration of different remedy options, we decided to pursue the following package of remedies:

(a) easy-to-understand terminology and descriptions of PCA services;

(b) explanations of the levels of charges and interest rates and how and when they are applied;

(c) information on statements;

(d) summary and breakdown of charges and interest;

(e) advance notice of charges and debit interest incurred;

(f) regular ‘rights reminder’; and

(g) changes to the switching process.

The least cost, least intrusive package of effective remedies

6.251 We noted that in the CC’s guidance it is stated that:

The Commission must have regard to the reasonableness of any remedy and will aim to ensure that no remedy is disproportionate in relation to the adverse effect on competition and any adverse effects on customers. Part of its consideration will include an assessment of the costs of implementing a remedy …; and the costs of complying with a remedy, for example, providing the OFT with periodic information on prices or margins. However, the Commission must consider the wider picture. Adverse effects on competition are likely to result in a cost or dis- advantage to the UK economy in general and customers in particular. Where significant, these costs might usually be expected to outweigh the costs incurred by any person on whom remedies are imposed. If the Commission is choosing between two remedies which it considers would be equally effective, it will choose the remedy that imposes the least cost or that is least restrictive.202

6.252 As part of our consideration of each of the remedy options, we looked at the costs of implementation. We ensured that each individual remedy was no more intrusive or costly than required to remedy, mitigate or prevent the AEC or any resulting detrimental effect on customers. We also noted our statutory obligation to have regard to the need to achieve as comprehensive a solution to the AEC and any resulting detrimental effect on customers as is reasonable and practicable.

6.253 In our view, each element of this package of remedies is needed effectively to address the features of the market that we have identified as preventing, restricting or distorting competition thus resulting in an AEC. We thought that, in order for the PCA market in Northern Ireland to become more competitive, banks should take action to ensure that customers should: have a better understanding of the costs of running their existing PCA; understand that there may be gains from considering alternative PCAs; have a better appreciation of alternative PCA offers and be able to

202CC3, paragraph 4.10.

177

compare them with their own; know how to switch PCA; and understand that switching PCA is not particularly difficult or risky. We thought that our chosen package of remedies would achieve these objectives.

6.254 Remedies (a) to (e) will significantly improve customers’ understanding of PCAs. Together, they will ensure that each PCA customer has access to the information they need to understand the cost of running their account. Remedy (e) will also empower customers to manage their account better and to avoid unnecessary charges. Remedies (a) and (b) together will ensure that customers have access to information to assess the cost of operating a PCA with an alternative provider.

6.255 Remedies (d) and (f) together will help to stimulate customers’ awareness, ensuring that they are aware of the cost of running their PCA over the year, and at the same time reminding them that they are able to close their PCA and switch to another bank. Remedy (f) will also remind customers of the availability of switching services to facilitate the switching process. The ‘hold harmless’ provision in remedy (g) will address the perception that switching is risky, and the banks will be required to inform customers of this safeguard.

Proportionality to the scale of the AEC

6.256 In addition to considering whether the package represents the least cost, least intrusive package of remedies that would be effective in addressing the AEC, we also considered whether this package of remedies is proportionate to the scale of the AEC.

6.257 Eight banks provided estimates of the costs of implementing remedies (a) and (b); seven provided estimates of the costs of implementing remedies (c), (d) and (e); and six provided estimates of the costs of implementing remedy (f). Two banks provided estimates of the annual costs for remedies (a), (b), (c) and (d), and one bank provided an estimate of annual costs for remedies (e) and (f). In response to our provisional decisions on remedies, three banks revised their estimates for one-off implementation costs and/or annual costs for some of the remedies. The banks had an opportunity to provide estimates of the costs of implementing changes to the switching process, but none did so. Using the estimates provided by individual banks, as revised, we estimated the total costs that all banks would face in the first year of implementation, including one-off costs, and the annual costs. Applying a threshold of 10,000 PCAs held by customers with a Northern Ireland postal address would result in ten banking groups being required to implement these remedies. These calculations suggest that the cost of implementing the package of remedies in the first year, including one-off costs, would be in the range £37 million to £55 million. The annual cost of implementing the package of remedies thereafter would be around £9 million.

6.258 We could not verify the banks’ cost estimates by undertaking a detailed assessment of the changes required to the banks’ IT systems and business practices. We recognized that this lack of verification was a potential shortcoming in our assessment of the proportionality of the remedies package. Nevertheless, we also believed that there were other more significant factors which enabled us to make a reasoned judgement on proportionality. These factors are:

(a) The banks provided only high-level estimates of implementation costs, and we thought it likely that these would be cautious estimates.

(b) It would be misleading to arrive at a total cost for the remedies package by aggregating the costs of each individual remedy on a stand-alone basis. Signifi-

178

cant synergies might reasonably be expected from implementing the different remedy options as a package.

(c) The assessment of cost proportionality should take account only of costs which would be incurred incrementally as a result of the implementation of this remedies package. The assessment should therefore exclude any costs which could be expected to arise from either the Code review or the implementation of the CCA 2006, for example the requirement to notify customers when they have incurred a default sum (see paragraph 6.116).

6.259 There are two further factors which we have not explicitly taken into account in assessing proportionality. The first, and more important, is that practically all of the costs relate to the provision of information which any bank customer might reasonably require, and expect to be provided with, as a matter of course, for there to be a balanced commercial relationship between a bank and its customers. Second, there is the question of the legacy costs associated with an inefficient market. We thought that, as a matter of principle, the nature and extent of remedies intended to benefit customers and consumers more generally should not be attenuated by virtue of a high cost of implementation which was attributable to a prolonged period of market inefficiency. We thought that customers should not continue to be deprived of the information they need to make the right choice about their PCA and bank simply as a result of the banks’ need to update systems, which, had they been operating in a well functioning market, may have required updating earlier.

6.260 One bank ([]) has noted that, as a bank operating across the UK, the additional costs it would incur as a result of implementing these remedies in Northern Ireland would place it at a competitive disadvantage in competing for customers in the market in Great Britain. We noted that it will be competing in Great Britain with banks that do not have PCAs in Northern Ireland and which will therefore not incur the costs associated with implementing this remedy package. However, considering the costs of implementing our package of remedies in the context of the banks’ turnover in Great Britain, we did not find that the need to incur these costs would place those banks that must implement our remedies at a disadvantage when competing with banks that do not have to implement our remedies.

6.261 In reaching our decision on proportionality, we also took into account the estimated size of the consumer detriment resulting from the features of the market that we identified. We recognized that there was no precise way in which this might be done and therefore made what were, in our view, reasonable estimates.

6.262 Banks charge customers for providing a PCA in a variety of ways and there is no single ‘price’. Costs and revenues flow between banks and customers in three forms: charges, interest on credit balances and interest on debit (overdraft) balances. Table 13 presents estimates of the customer detriment on the basis of two sets of assumptions.

6.263 We looked at the market as a whole using 2005 figures as the starting point. The first set of assumptions we used provided an estimate of the upper bound of customer detriment from the AEC we had identified. In estimating this upper bound we looked at charges on an absolute basis.203 We considered the implications of total charges

203We also considered basing the lower bound estimate of customer detriment on the assumption that total charges and debit interest paid by customers might be 5 per cent higher, and credit interest 5 per cent lower, than it might otherwise be in a competitive market. However, we considered that this assumption was unrealistic and we preferred to base our lower bound estimate of detriment on the assumptions set out in paragraph 6.264.

179

per account being £46, compared with the average across the banks in 2005 of £51; average credit interest being 0.7 per cent compared with a 2005 average across the banks of 0.29 per cent; and average debit interest being 9 per cent compared with a 2005 average across the banks of 9.81 per cent. We thought it was more realistic to consider the impact of raising average credit interest to 0.5 per cent rather than raising it by 5 per cent (which equates to an average increase from 0.29 per cent to only 0.30 per cent). The values generated using these assumptions were in each case within the range of average charges per account, average credit interest rate and average debit interest rate exhibited by the eight largest banks in Northern Ireland.

6.264 In our provisional decisions on remedies we used a second set of assumptions that reduced total charges paid by customers by 5 per cent, but reduced debit interest rates to the nearest 0.5 per cent and increased credit interest rates to the nearest 0.5 per cent to estimate a lower bound for the customer detriment from the AEC. We consider that the 5 per cent reduction in total charges paid by customers still represents a lower bound of the detriment to customers in respect of charges. However, we believe that reducing debit interest and increasing credit interest rates to the nearest 0.5 per cent is a more realistic assumption than reducing the amount of debit interest paid by customers by 5 per cent and increasing the amount of credit interest paid to customers by 5 per cent. This is because such adjustments could produce what might be regarded as a somewhat arbitrary interest rate which would be unlikely to be applied in practice (eg a 5 per cent reduction in a 6.5 per cent interest rate would generate a new rate of 6.175 per cent). Also, for credit interest, where rates are already low, the actual changes could be so small as to be immaterial.

6.265 We recognized that such calculations could only provide an indication of the scale of the customer detriment that might result from the lack of competition in the market. Basing the lower bound of the estimate on a 5 per cent decrease in charges, but reducing debit interest to the nearest 0.5 per cent and increasing credit interest to the nearest 0.5 per cent, the range of possible customer detriment is estimated to be between £12.1 million and £24 million a year. For the reasons set out in paragraph 6.264, we believe this range to be the more reasonable.

TABLE 13 Estimates of customer detriment

Lower bound Upper bound

Assume that detriment 5% of total charges paid by customers (£3.3m) Reduction in average charge per account by equals £5 per year to £46 per year (£6.6m) Increase in average credit interest by 0.21 per- centage points to 0.5% (£10.4m) Increase in average credit interest by 0.41 per- centage points to 0.7% (£20.3m) Decrease in average debit interest rate by 0.31 percentage points to 9.5% (£0.5m) Decrease in average debit interest rate by 0.81 percentage points to 9% (£1.2m)

Annual detriment pre-tax £14.2 million £28.1 million

Annual detriment post-tax* £12.1 million £24.0 million

Source: CC analysis.

*In the absence of information on the tax position of the banks’ customers we have assumed an average income tax rate of 20 per cent on the element of the detriment that relates to credit interest paid. We note that customers pay income tax on credit interest at nil, 20 or 40 per cent depending on their personal tax position.

6.266 As noted in paragraph 6.257, the cost estimates that we have calculated based on the banks’ estimates suggest first year implementation cost in the range of £37 million to £55 million and an annual running cost of around £9 million. We would expect our remedies to remain in place over time, and/or to be replaced or supple-

180

mented by further improvements in the banks’ normal practices and terms and conditions as the market becomes more competitive. The benefits to customers will therefore continue to accrue over a prolonged period. In our provisional decisions on remedies, we calculated the difference between the cost of the remedies and the benefits to customers over a ten-year period. Two banks ([]) in their response told us that as the benefits accrued over time, we should calculate the discounted present value of the costs and customer benefits. We agreed with their suggestion and have calculated the net present value (NPV) of the post-tax costs204 and customer benefits over a ten-year period. We have applied a post-tax discount rate of 9 per cent to calculate the NPV of the banks’ estimated costs. This is a conservative assumption based on the average cost of equity of three banks ([])205 over the period 2002 to 2005. We have used the discount rate of 3.5 per cent recommended by HMT for the appraisal and valuation of policies, programmes and projects206 to calculate the NPV of the customer detriment.

6.267 Based on the NPV calculation, the gain to customers over a ten-year period could be in the range £100 million to £200 million. The NPV calculation of the cost estimates suggests that costs to the banks of implementing the remedies over the same period would be in the range of £60 million to £75 million.

Conclusions on effectiveness and proportionality

6.268 We have considered the proportionality of our remedies by considering two questions:

(a) Is our package of remedies the least cost, least intrusive package of remedies that would be effective in addressing the AEC?

(b) Is our package of remedies proportionate to the scale of the AEC.

6.269 As discussed in paragraphs 6.253 to 6.255, we conclude that each of the remedies in our package is necessary effectively to address the AEC. As discussed in relation to each of the remedies, we have considered what each remedy in our package needs to achieve, and we have considered how best to achieve this without going further than is necessary. On the basis of these considerations, we conclude that our package of remedies is the least cost, least intrusive package of remedies that would be effective in addressing the AEC.

6.270 We have considered the proportionality of our package of remedies to the scale of the AEC by assessing the costs associated with the implementation of the package against the scale of the customer detriment that the package will address. This assessment is set out in paragraphs 6.256 to 6.267 above. As noted in that discussion, we were unable to verify the cost estimates provided by the banks, and we believe that the estimates provided by at least some of the banks as to the cost of implementing our package of remedies may represent the upper end of a range of possible costs, especially taking into account costs that are likely to be incurred in any case following the Code review and the implementation of the CCA 2006. On this basis, it is likely that the incremental costs of implementing our package of remedies will be significantly lower than the estimates provided by the banks. However, even

204To calculate the post tax implementation and annual costs we have reduced by 30 per cent (ie the corporate tax rate) the first year implementation cost and the annual ongoing costs. In the NPV calculation we have assumed that the banks’ one-off implementation costs are incurred all in the first year. 205Banks’ responses to the financial questionnaire. 206HM Treasury, The Green Book, Appraisal and Evaluation in Central Government.

181

taking those costs at face value, our view is that over time, the benefits from the remedies could reasonably be expected to exceed the costs by a substantial margin. We therefore conclude that our package of remedies is proportionate to the scale of the AEC.

Relevant customer benefits

Framework for assessment of relevant customer benefits

6.271 In deciding the question of remedies, the CC may ‘in particular have regard to the effect of any action on any relevant customer benefits of the feature or features of the market concerned’.207 As the Act makes clear, the CC will be concerned with the effect of any remedy on relevant customer benefits; it will not seek to weigh the adverse effect of the feature or features of the market concerned against any resulting relevant customer benefits in deciding whether remedies are appropriate.208

6.272 Relevant customer benefits are limited to benefits to relevant customers in the form of:

(a) lower prices, higher quality or greater choice of goods or services in any market in the UK (whether or not the market to which the feature of features concerned relate); or

(b) greater innovation in relation to such goods or services.209

6.273 A benefit is only a relevant customer benefit if the CC believes that:

(a) the benefit has accrued as a result (whether wholly or partly) of the feature or features concerned or may be expected to accrue within a reasonable period of time as a result (whether wholly or partly) of that feature or those features; and

(b) the benefit was, or is, unlikely to accrue without the feature or features con- cerned.210

6.274 If the CC is satisfied that there are relevant customer benefits deriving from a market feature that also has adverse effects on competition, it will consider whether to modify the remedy that it might otherwise have imposed or recommended. When deciding whether to modify a remedy, the CC will consider a number of factors, including the size and nature of the expected benefit, how long the benefit is to be sustained, and the impact of the benefit on different customers. It is possible that the CC will decide that the benefits are of such significance compared with the effects of the market feature(s) on competition that no remedy is called for. However, such cases are likely to be rare.

Conclusions on relevant customer benefits

6.275 None of the parties said that there are relevant customer benefits deriving from any of the features of the market for the supply of PCA banking services in Northern Ireland that we had identified. We have not identified any relevant customer benefits

207Section 134(7) of the Act. 208The CC’s approach to the assessment of relevant customer benefits is discussed in more detail in CC3, Market Investigation References, paragraph 4.26 and following. 209Section 134(8) of the Act. 210Section 134(8) of the Act.

182

on which the proposed remedies, alone or in combination, would have an effect. We therefore did not consider relevant customer benefits further.

7. Our decisions

7.1 We are required to decide various statutory questions pursuant to section 134 of the Act (see paragraphs 5.11 and 6.4).

7.2 We conclude that there are three features of the relevant market, either alone or in combination with each other, that prevent, restrict or distort competition in connection with the supply of PCAs in Northern Ireland, and hence that there is an AEC with the meaning of section 134(2). The three features that we identified are:

(a) banks have unduly complex charging structures and practices;

(b) banks do not fully or sufficiently explain their charging structures and practices; and

(c) customers generally do not actively search for alternative PCAs or switch bank.

7.3 We have decided that we should take action for the purpose of remedying, mitigating or preventing the AEC concerned and any detrimental effect on customers so far as it has resulted or may be expected to result from the AEC. We have also decided that we should recommend the taking of action by others for this purpose, in this case by BACS.

7.4 On the basis of the analysis set out in section 6, we decided that the following package of remedies would represent a comprehensive, reasonable and practical solution to the AEC and any resulting detrimental effect on customers that we have identified:

(a) Remedy (a): Easy-to-understand terminology and descriptions of PCA services. Banks operating in Northern Ireland must satisfy the BCSB that all information provided to customers when choosing a PCA, when opening a PCA, on statements, and when pre-notified of charges and interest payments, is easy to understand. The banks must ensure that all such communications are:

(i) certified by an independent organization specializing in plain English; or

(ii) otherwise tested with customers and found to be easily understandable.

(b) Remedy (b): Explanations of the levels of charges and interest rates and how and when they are applied. Banks operating in Northern Ireland must ensure that their Northern Ireland customers receive explanations of the levels of charges and interest rates on their PCA, the circumstances in which the customer would incur these charges and/or receive interest, and the way in which, and when, these charges and interest rates would be applied. Explanations must be provided of at least the following charges and interest rates:

(i) credit interest rates applicable to relevant balance limits;

(ii) current account charges;

(iii) charges for standard account services (eg charges for setting up direct debit or standing orders);

183

(iv) authorized overdraft debit interest rates;

(v) authorized overdraft charges;

(vi) unauthorized overdraft debit interest rates; and

(vii) unauthorized overdraft charges (including maintenance charges, paid and unpaid items charges and transaction charges).

Banks operating in Northern Ireland must also make clear to customers the existence, but not the detail, of discretionary policies in the banks’ application of charges and interest in relation to unauthorized overdrafts.

The banks must make this information freely available to customers when choosing a PCA, when opening a PCA, on statements, and when pre-notified of charges and interest payments.

(c) Remedy (c): Information on statements. Banks must provide customers with key account information, as part of their PCA statements. Banks must provide explanations of the level and application of at least the charges and interest rates set out in remedy (b). The information should be provided to each PCA customer, in the same form as a customer receives their statement unless the customer elects otherwise. It may be provided on the front or the back of the statement, on a separate sheet on ‘statement paper’, or as part of an electronic communication, provided the information is clearly part of the statement.

(d) Remedy (d): Summary and breakdown of charges and interest. Banks operating in Northern Ireland must provide PCA customers once a year with an annual summary and breakdown of charges and interest. This summary should cover all the charges and interest payments relating to the customer’s PCA but need not include ancillary charges. This summary must include an explanation of the level and application of all those charges included on the statement (see remedy (b)). The summary should be sent to the customer in the same form (electronic or hard copy) as the customer routinely receives their statements, unless the customer elects otherwise.

(e) Remedy (e): Advance notice of charges and debit interest incurred. Banks must pre-notify customers of all authorized and unauthorized overdraft charges. Pre-notification may be sent together with a statement or in advance of a statement and should be in the same form unless the customer elects otherwise. However, the banks must not deduct a pre-notified charge from a customer’s PCA until at least 14 days after the date of the statement. If the pre-notification is not sent together with a statement, the bank must provide the key account information that is required under remedy (b) on the pre-notification communication. Banks must not allow charges and interest to accumulate to a PCA for more than one month before the customer is pre-notified of the bank’s intention to deduct the charge or interest from the PCA. The bank must send the customer a statement at the end of any month in which authorized or unauthorized overdraft charges are incurred.

(f) Remedy (f): Regular ‘rights reminder’. Banks must include wording on the annual summary and breakdown of charges and interest that makes it clear that the customer can close their PCA with that bank and seek to obtain PCA services from a different bank. It should also specify where the customer can find further information on how to do this. This communication should be accompanied by a generic leaflet or electronic communication which explains the switching process,

184

the fact that it is easy to switch using the banks’ switching services and that customers are not charged for switching. This communication should be developed by the banks working together, possibly through the BBA. It may be in electronic form or hard copy but should be sent to customers in the same form as they receive their summary and breakdown of charges and interest unless the customer elects otherwise.

(g) Remedy (g): Changes to the switching process. For each Northern Ireland customer who uses the switching service to open a new PCA, banks must offer an interest- and charge-free overdraft facility, the amount of which is commensurate with the expected transactions on the PCA, for a minimum of three months after the PCA is opened, provided the customer would be eligible for such a facility under the bank’s usual credit scoring policy. Where a customer is not eligible for such an overdraft or does not wish to have one, banks must guarantee to refund the customer any charges and interest which are incurred within a minimum period of three months after the PCA is opened as a result of a failure in the switching process. This is regardless of whether the charges and interest were incurred as a result of an error by the new bank. Banks operating in Northern Ireland must publicize these ‘hold harmless’ provisions in their switching literature.

7.5 We decided that it would be appropriate to establish a minimum threshold for the application of our package of remedies. We decided that this threshold should be based on the absolute number of a bank’s PCAs held by customers with a postal address in Northern Ireland. A bank that is within a corporate group with more than 10,000 PCAs held by customers with a postal address in Northern Ireland will be required to apply the remedies, unless that bank can demonstrate that it has fewer than 5,000 PCAs held by customers with a postal address in Northern Ireland.

7.6 We decided that these remedies should be implemented by means of an order. We decided to require the implementation of remedies (a) and (b) by 1 April 2008, and to require the implementation of the other remedies within our package by 1 October 2008 at the latest. This latter date should allow the banks scope to implement the changes required as a result of our remedies alongside those changes that are likely to be required as a result of the Code review and the implementation of the CCA 2006.

7.7 The OFT has a statutory role in relation to the monitoring and enforcement of remedies. However, in line with the principles of better regulation, we do not wish to impose any unnecessary additional regulatory structure on an already highly regu- lated industry. We concluded that the BCSB should monitor compliance with our remedies. The banks should fund additional resource for the BCSB as appropriate to enable it effectively to monitor compliance with the remedies package. The BCSB should provide the OFT with regular written reports to the OFT’s specification.

7.8 We conclude that our package of remedies is the least cost, least intrusive package of remedies that would be effective in addressing the AEC. We believed that the estimates provided by at least some of the banks as to the cost of implementing our package of remedies may represent the upper end of a range of possible costs, especially taking into account costs that are likely to be incurred in any case following the Code review and the implementation of the CCA 2006. Even taking these costs at face value, we found that over time, the benefits from the remedies could reasonably be expected to exceed the costs by a substantial margin. We therefore conclude that our package of remedies is proportionate to the scale of the AEC.

185

7.9 None of the parties said that there are relevant customer benefits deriving from any of the features of the market for the supply of PCA banking services in Northern Ireland that we had identified. We have not identified any relevant customer benefit on which the remedies, alone or in combination, would have an effect. We therefore did not consider relevant customer benefits further.

7.10 As set out in paragraph 6.181, we are recommending that BACS review the switching process, with a view to identifying and addressing any outstanding impediments to switching direct credits, direct debits and standing orders.

186