FINANCIAL STABILITY r eview

in association with the Securities and Investments Board

IN THIS ISSUE The UK market for The valuation of Dealing with market high-yield debt options manipulation

There is no effective Options have transformed Manipulation damages market in high-yield risk-management markets — it distorts prices bonds (junk bonds) in the practices. However, and so may lead to United Kingdom. market participants must economic inefficiencies This article looks at the ensure they value products and the mis-allocation of role high-yield debt can appropriately. This article resources. The appropriate ISSUE THREE play and examines the reviews the Bank of regulatory response

AUTUMN 1997 development of the England’s second is necessary when sterling market. options valuation survey. addressing manipulation. Articles published in Financial Stability Review aims: Financial Stability Review, whether written by the Bank or the staff of the Securities and Investments Board or by outside • to promote the latest thinking on risk, regulation contributors, are intended to add to debate, and are and financial markets not necessarily statements of the Bank or the Securities and Investments Board. •to facilitate discussion of issues that might affect Correspondence and comments on articles are risks to the UK financial system welcome and should be addressed to:

The Editor, • to provide a forum for debate among practitioners, Financial Stability Review, Bank of England, policy-makers and academics Threadneedle Street, London EC2R 8AH. Fax: 0171 601 3668

Subscription details Copies of the Financial Stability Review may be obtained from the Bank at a cost of £10 for a single issue (£5 for students) and £18 for a year’s subscription (£10 for students). An extra charge will be made to overseas subscribers to cover postage. All enquiries regarding subscriptions should be addressed to: Subscriptions, Financial Stability Review, Bank of England, Threadneedle Street, London EC2R 8AH (Tel: 0171 601 3823). Back issues are available from the same address. FINANCIAL STABILITY r eview

Editorial Committee Financial sector issues 3

Controlling securities fraud 8 Bank of England The market may provide inadequate incentives for firms to avoid fraud Patricia Jackson , William Perraudin and Norvald Instefjord Alastair Clark Michael Foot The valuation of options 18 John Footman An account of the Bank of England’s second survey of options Howard Walwyn and Wayne Byres Garth Hewitt Patricia Jackson The genesis of regulation 29 How has regulation developed in the UK and what are the new arrangements? Oliver Page Victoria Robb Securities and Investments Board The pre-commitment approach to setting capital requirements 42 Howard Davies The pre-commitment approach has been put forward as a possible alternative Andrew Winckler to the Basle Standard and the Value at Risk methodologies Patricia Jackson, Simone Varotto and Arupratan Daripa Christopher Boyce Reflections on financial regulation 51 During the past 15 years, banking problems have been worse than in any period since the 1930s’ depression. External supervision must be improved: internal controls reinforced Charles Goodhart, David Llewellyn and Philipp Hartmann

The UK market for high-yield debt 62 The role that high-yield debt can play in the UK and the market’s development Jeremy Leake, Alex Crowe and Rupert Watson

Regulation of custody in the UK 69 Following a review by the Securities and Investments Board, anyone who provides custody services in the UK must be authorised under the Financial Services Act Liz Murall

Market manipulation 74 Market manipulation must be addressed by the appropriate regulatory response Guy Sears

Global institutions, national supervision and systemic risk 82 An industry initiative is necessary for high standards of risk management John Heimann and Lord Alexander

Published by The UK credit card market 92 Bank of England The development of the market and comparisons with the US London EC2R 8AH Collette Brooking Telephone: 0171 601 4444 Editor: Garth Hewitt Mutuality at the cross-roads 105 Assistant: Duncan Lambert The demutualisation of UK institutions raises many issues Illustrations: Christopher White Adam Boxall and Niall Gallagher Printed by Park Communications Limited 40 Clifton Street, London EC2A 4DX Regulatory developments 118

IN THE PREVIOUS ISSUES OF FINANCIAL STABILITY r eview

THE FOLLOWING TOPICS WERE COVERED:

ISSUE 1 Culture of regulation Successful supervision depends not only on having the right rules and tools, but also on getting the culture right. The Deputy Governor of the Bank of England looks at the lessons from the Bank’s review of supervision. Building society conversions Some building societies are becoming banks. What are the motives and mechanics of changing status and what supervisory issues arise? The SIB review of the metals markets The London Metal Exchange is the largest exchange trading metal derivative contracts. It is different in some key respects from other derivatives exchanges. Electronic money: public policy issues Many types of ‘electronic money’ are under development, promising more convenient and efficient ways for consumers to pay, but also with potential risks. Rating sovereign risk The Managing Director of sovereign ratings at IBCA describes the risks in sovereign lending, and how rating agencies try to measure them. Deposit protection and bank failures in the United Kingdom Bank failures are usually big news. But UK banks that have gone under were mainly small, with a low payout from the Deposit Protection Fund. International regulatory co-operation Banks and securities firms are becoming more global, and the distinction between their business more blurred. How can regulators respond? CREST: its recognition and approval London’s new equity settlement system was launched on 15 July 1996. As well as the work on design and implementation, CREST posed issues for regulators. Bancassurance: European approaches to capital adequacy Banks and insurance companies are increasingly getting involved in each other’s business. What capital treatment is appropriate for the banks?

ISSUE 2 The efficiency of regulation The Chief Executive of the SIB explains why the SIB and SROs will be giving efficiency of regulation a high priority this year. Modelling and pricing credit risk Many banks are focusing on developing credit risk models as a method of pricing loans. Remuneration and risk How does the structure of staff remuneration packages affect a firm’s risk profile and what can firms do about it? Operational risk management: where to start?: There is a growing recognition of the need for financial firms to manage their operational risks. Why has operational risk become more important and what steps can firms take to control it? Fraud: a personal view The retiring Head of the Bank’s Special Investigations Unit shares his personal view on the nature of financial fraud. UK mortgage margins How have developments in this competitive market affected mortgage margins and do the changes represent a prudential concern? Beyond Glass-Steagall: regulatory change in the United States How will changes in the regulatory framework affect the US financial scene and what issues need to be addressed? Lloyd’s: current developments and the challenges ahead Few institutions have survived losses on the scale of Lloyd’s of London in recent times — what challenges lie ahead for Lloyd’s and its regulators? Credit exposure in OTC derivatives: a risk management challenge How can firms model and manage the credit exposures that arise when they trade in over-the-counter derivatives? Regulation and market design: the Stock Exchange’s order book Forthcoming changes to the London Stock Exchange potentially herald a second ‘Big Bang’ for the City. What issues has the SIB had to address as regulator? FINANCIAL STABILITY r eview

A LIMITED NUMBER OF BACK ISSUES IS AVAILABLE SUBSCRIPTION FORM

FINANCIAL STABILITY r eview SUBSCRIPTION FORM

If you subscribed for Issues 2 & 3 of the Financial Stability Review, your subscription period has now ended. To ensure that you receive the next issue, please detach and complete this form and return to the address below together with your remittance.

PLEASE RETURN TO:

Subscriptions — Financial Stability Review Bank of England — HO-1 Threadneedle Street London EC2R 8AH

Tel: 0171 601 3823 Fax: 0171 601 3668 TO ORDER/SUBSCRIBE TO THE FINANCIAL STABILITY REVIEW

UK First Class UK Student1 Rest of Europe2 Rest of World2 Subscription to £18.00 £10.001 £20.002 £24.002 Issues 4 & 5 Individual £10.00 £5.001 £11.002 £13.002 Copy(ies) of per copy per copy per copy per copy Issue 1, 2, 3 or 4

A limited number of backcopies of Issues 1 & 2 are still available. Issue 4 will be published at the end of Q1 1998 — Issue 5 is due for release during Q4 1998.

I would like to order (complete as appropriate):

Number of Description Issue Price per Copy(ies) Number Copy

Subscription(s) to Issues 4 & 5 Copy(ies) of Issue Copy(ies) of Issue Copy(ies) of Issue

I enclose (complete as appropriate): A sterling cheque (drawn on a bank in the U.K.) or postal order for a total of £ or A non-sterling cheque or money order (drawn on a bank in country of customer) for an £ equivalent amount, plus £2.00 to cover bank charges, in local currency3, for a total of made payable to “Bank of England”

Mailing address details (please complete as fully as possible):

Salutation Mr / Mrs / Miss / Ms (delete as appropriate)

First Name Surname

Job Title

Company/Institution

Address

Post/Zip Code Country

Telephone Fax

1Please enclose some proof of status, eg photocopy of your student card. Students in countries outside the UK are not eligible for concessions. 2 £2.00 should be added to cheques drawn on local banks in local currency to allow for bank charges. 3Cheques drawn in £ sterling on a bank in the country of the customer cannot be accepted as bank charges exceed publication costs.

IF IN THE EU OUTSIDE THE UK PLEASE GIVE YOUR VAT REGISTRATION NUMBER HERE:

© BANK OF ENGLAND 1997

ISSN No: 1365-7267 FINANCIAL SECTOR ISSUES

This journal, Financial Stability Review, was launched 12 months ago — The anatomy of a a joint venture between the Bank of England and the Securities and super-regulator Investments Board. At the time, neither party expected this tentative Outline plans for the new regula- flirtation to lead to anything more serious. But shortly after our second tory authority were published at issue, the new government announced a shotgun wedding. The Bank of the end of July1, and are England Bill, being introduced in the new session of parliament, will give summarised in the box overleaf. the Bank operational independence in the field of monetary policy and No fewer than nine regulatory provide for the transfer to the SIB (which will soon have a new name) of bodies are to be merged into all its supervisory powers — under the Banking Act for deposit-takers, NewRO. The administration of under Section 43 of the Financial Services Act for wholesale market this change, and the task of institutions, and under Section 171 of the Companies Act for payment combining the different statutory netting schemes. objectives and organisational cultures is complicated in that the Howard Davies, until July the new legislation will come in two Deputy Governor of the Bank stages. (and in that capacity creator of this The Bank of England Act, magazine), has become the which will transfer the Bank’s chairman of the SIB and prospec- supervisory responsibilities to tively of the super-regulator NewRO, is likely to become law in currently known by the working spring 1998; but the main regula- title “NewRO”. tory reform measure, which The Bank itself will remain provides for the incorporation of responsible for the stability of the the SROs and the supervisors of financial system as a whole, building societies, friendly soci- through its market operations and eties and insurance companies into through its central role in payment NewRO, will not be introduced and settlement systems, which until the next session of Parliament give it the ability to monitor and is unlikely to become law systemic threats and a capacity, before the end of 1999. Under where necessary, to intervene. The interim arrangements, the staffs of Bank and NewRO will need to the SROs will transfer to NewRO, work closely together, and have but NewRO staff will continue to been preparing a memorandum of operate, formally, under contract understanding with HM Treasury to the SRO boards. The staffs of which sets out the responsibilities the Building Societies of each and they will co-operate. Commission, the Friendly

3 FINANCIAL SECTOR ISSUES

Societies Commission, the to pursue a flexible and differenti- The lesson from Barings, they Registry of Friendly Societies and ated risk-based approach to setting suggest, is that “the external regu- the Insurance Directorate of the standards and to supervision, lation required to prevent such Department of Trade and Industry making appropriate distinctions cases of fraud (where it had will not transfer until the new law between the regulation of whole- escaped internal management) is in force. However, all of the sale and retail business. It will would have had to be so pervasive, bodies are co-operating closely to operate in a way which recognises so intrusive and so expensive as to ensure a smooth transition. the benefits of competition and be practically impossible”. The institutional arrange- innovation to consumers and to the So while the imposition of ments for regulation are thus economy. NewRO recognises that basic minimum standards remains likely, over the next two years, to there are risks in competitive a goal, especially in emerging absorb significant amounts of markets, including the risk of countries, in more developed management time — even though, failure: “Regulation does not markets the emphasis is switching as the article, Reflections on absolve consumers of responsi- to the identification of regulatory Financial Regulation, by Charles bility for their own decisions on incentives that will reinforce Goodhart, Philipp Hartmann and their financial affairs. However, internal management risk control David Llewellyn, which is the extent to which they can exer- mechanisms, and the development published in this issue argues, cise that responsibility and the of more sophisticated tools for institutional structure is not an types of risk that they would measuring risk of all issue of primary importance to the reasonably be expected to assume types. effectiveness of regulation. depend on their level of knowl- There has, therefore, been The authors suggest that there edge and expertise.” growing interest in academic and is no single “right” structure — regulatory circles in finding ways what matters most is the way Getting the incentives to harness market forces in support supervision is conducted. The right of regulation, and to identify design of regulatory structures, or All regulators recognise that, in market or regulatory incentives the internal structure of a mega- rapidly changing financial that go with the grain of what regulator, has to be focused on the markets, regulatory objectives of responsible managers of financial objectives of regulation. any sort, however carefully institutions ought anyway to want This is recognised in the defined, are becoming more diffi- to do. An example, discussed in SIB’s report to the Chancellor: cult to deliver. Regulators cannot the article in this issue, The Pre- this starts, not with the nuts and be everywhere. Goodhart, commitment Approach to Setting bolts of integrating regulators and Hartmann and Llewellyn note Capital Requirements, was the their different legislative frame- that “external regulation, both decision in 1995 to recognise indi- works, but with a statement of the in its regulatory mode of seeking vidual banks’ value-at-risk (VaR) high-level objectives of the new to lay down, ex cathedra, models as a basis for setting body, and an outline of NewRO’s common rules and ratios that all capital requirements against general approach to regulation. banks should follow, and in a trading books, rather than perse- NewRO commits itself to supervisory-monitoring mode of vere with clumsy rule-based pursue its regulatory objectives in checking whether banks are capital requirements under which an efficient way and to ensure that complying with such rules, is different risks are simply added the costs of regulation are propor- becoming both less effective and together with no allowance for the tionate to the benefits. It promises less feasible ...” benefits of diversification across

4 FINANCIAL SECTOR ISSUES

OUTLINE PLANS FOR NEW REGULATORY AUTHORITY

The regulatory functions of nine bodies will come together to form NewRO. They are: • The Bank of England • The Building Societies Commission • The Friendly Societies Commission • The Insurance Directorate of the Department of Trade and Industry • The Investment Management Regulatory Organisation • The Personal Investment Authority • The Registry of Friendly Societies • The Securities and Futures Authority • The Securities and Investments Board

The corporate vehicle for the merger will be the SIB, a company limited by guarantee set up in 1985 to be the designated authority under the Financial Services Act, and shortly to be renamed. The budgeted staff numbers of these regulators amount to 2,100. As far as practicable, it is intended that they will transfer to NewRO from the combining organisations once the new Bank of England Act is effective — expected to be spring 1998 — so that the process of integration into a single organisation can start. However, until the Regulatory Reform Bill takes effect, prob- ably towards the end of 1999, the legal responsibilities of the current regulatory bodies will remain unchanged. During the transitional period, therefore, existing supervisory relationships will continue. For the longer term, however, significant changes of structure are envisaged. NewRO plans to adopt a functional model of regulation. Five functions are envisaged: Authorisation. This will handle all applications for authorisation and the vetting and registration of individuals. Investigations/Enforcement/Discipline. A dedicated unit will handle cases across all regulated firms and markets, bringing together the necessary specialist skills and expertise. Relations with consumers. NewRO plans to establish a single point of entry for all consumer enquiries and complaints. During the transitional period, however, existing arrangements for the resolution of disputes will continue to operate. Existing compen- sation schemes will also continue to operate on their present basis for the time being, but will be reviewed to assess the potential for harmonisation in the future. The principle that compensation should be paid for by regulated businesses — and thus consumers — will be preserved, and for reasons of cost and moral hazard there is no intention of providing a complete safety net. Supervision, which will be sub-divided into business areas. This will cover prudential supervision and conduct of business regu- lation where relevant. In regulating financial conglomerates NewRO will build on the existing “lead supervisor” arrangements to secure more effective co-ordination. The supervision team will cover Recognised Investment Exchanges and Clearing Houses, and will oversee the wider markets including providers of infrastructure support. Policy, devolved as much as possible to individual regulatory areas but with a central policy “think-tank” supporting senior management. It is expected that NewRO will have a board of 10-15, some of them executive. They will be appointed by the Treasury which has overall responsibility for the regulation of the financial system. The Board will oversee NewRO’s exercise of its powers, deal with corporate governance issues, exercise quality control over the organisation and handle major issues of policy and standard-setting. NewRO plans to devise arrangements for securing consumer input into policy formation and decision-making, including a standing consumer panel to advise the NewRO Board. It also intends to devise arrangements for securing practitioner input. Both of these issues will be the subject of consultation later in the year.

5 FINANCIAL SECTOR ISSUES

markets. VaR models had been penalty regime, which is essential Patricia Jackson, William developed to give bank manage- to the credibility of the scheme, Perraudin and Norvald Instefjord ments a consistent view of the may itself deliver perverse incen- explore possible incentives for overall portfolio risk taken by the tives if not very carefully management to maintain an effec- bank, and with some additional designed. Three possible types of tive control environment. safeguards regulators have agreed penalty have been advanced: addi- Regulatory incentives include that these can in some cases form tional capital requirements, fines setting higher capital requirements the basis of regulatory capital and public disclosure. However, for firms whose systems are requirements as well. concern about possible reputa- found wanting, an approach Federal Reserve Board econo- tional damage if there were a hitherto used mainly by bank mists have proposed a more breach could deter large conserva- regulators; increasing the radical approach under which, tive banks from running intensity of supervision for rather than have the regulators lay significant trading book risk, or perceived higher-risk players; down rules regarding the amount encourage them to pre-commit ex-post fines, which may be of capital a firm should hold excessive amounts of capital. This publicised — an approach adopted against its exposures, or setting is because any reputational by many securities supervisors; or parameters determining how the damage would affect the larger requiring, in a more or less public output of a VaR model is banking book as well as the way, the removal of management converted into a capital require- trading book. A further issue is at various levels. ment, the firm itself specifies in that the penalties would fall on the In New Zealand attempts have advance the maximum amount shareholders who, for the most been made to pin clear public that it expects to lose in a given part, do not determine the size of responsibilities on bank directors period from trading activities, and the exposure being run; or, at the to certify the adequacy of control is then judged against that promise extreme, on counterparties who systems, with heavy criminal and required to pay fines or face may simply find that the bank has and civil penalties if they get it other penalties if it exceeds the defaulted before the regulator gets wrong. Regulators in the UK limit it has set for itself. round to fining it. But the have also have been placing This approach, known as “pre- approach is now being tried in greater emphasis on board commitment”, allows the firms to New York, and the effect on firms’ responsibility. set their own capital requirements. risk-taking behaviour will be of The ultimate rationale for The pre-commitment great interest to regulators. regulation, as Goodhart et al approach avoids some of the draw- observe, is to correct for various backs of the regulators themselves Securities fraud forms of market imperfection or setting detailed parameters to Incentive structures can be market failure. So it is worth calculate capital requirements for relevant to other aspects of firms’ asking, in this case and others, trading books on the basis of VaR controls. An article in the April why the market itself cannot models. It also avoids the “one- 1997 issue of Financial Stability provide the necessary incentives. size-fits-all” approach, which Review2 showed how the design One reason is the ignores the differences between of traders’ remuneration packages principal/agent problem: firms in terms of their skills and had the potential to add to the risks managers in a firm, who take market positions. they took on their employers’ all the key operating decisions, But pre-commitment carries behalf. In their article in this issue, do not necessarily have the its own risks. One is that the Controlling Securities Fraud, same risk/reward trade-off as

6 FINANCIAL SECTOR ISSUES

do the shareholders. Typically New files will be archived or management will have a will be erased. These and other greater tolerance of risk logic problems can cause serious because they bear less of the problems for debt collection, cost of failure. interest calculation and ageing of information. They could disrupt Millennium risk: business significantly. a survival issue For many firms, achieving No responsible manager can Year 2000 compliance will be ignore a risk to the entire complex and resource intensive. operating system of his firm. The Bank for International The Bank for International Settlements calls for each bank to Settlements has released a report3, have an action plan establishing the latest of many, on the risks to compliance as a strategic objec- banks and their supervisors of the tive at the highest level, ensuring millennium problem. This is the recognition throughout the firm risk that many IT applications will that Year 2000 compliance may cease to function normally because be a survival issue, quantifying date fields do not recognise in the risks, setting priorities and 2000 the change of century. This is preparing inventories of an important issue across the corrective measures. On a whole economy but for banks and critical path, all of this should securities firms, dealing and already have happened, leaving settling in fast-moving markets time for adjusting systems across the world, the potential and testing them before costs of a system failure are mid-1999. This mirrors very immense, threatening the entire closely the approach taken firm and the financial system as a by the Bank and the whole. Financial Services Act While new applications tend regulators; NewRO has to be Year 2000-compliant, many now established a task-force old applications continue to run, to co-ordinate and new applications or operating their action.■ systems often depend on them. The problem exists because NOTES the two-digit identifier of the year 1 Report to the Chancellor on the Reform of the Financial Regulatory System, July used in older programs — 67 for 1997. Available from the SIB. 1967 — breaks down when it is 2Financial Stability Review, Issue Two, Howard Davies, until July the Spring 1997. Remuneration and Risk, by 00, which could be interpreted to Deputy Governor of the Bank Daniel Davies mean 1900 rather than 2000. So (and in that capacity creator of 3 The Year 2000: A Challenge for Financial systems may treat transactions as this magazine), has become the Institutions and Bank Supervisors: chairman of the SIB and Bank For International Settlements having been open for 100 years (or prospectively of the Website http:\\www.bank for international terminated before they began). super-regulator settlement.org

7 CONTROLLING SECURITIES FRAUD bank’s principal risk activity is credit, not market-risk, activity of this kind could make a sizeable dent By Patricia Jackson and William Perraudin, Bank of England, in the balance sheet. and Norvald Instefjord, Birkbeck College The accompanying boxes set out details of cases which resulted In recent years, several prominent financial firms have been seriously in substantial losses for three firms: affected by the fraudulent or irregular activity of single traders or groups Barings, Daiwa and Kidder Peabody. of traders. Often, control failures by senior management have paved the These cases have common features: way for these problems. Several important cases and their implications there were weaknesses in control for regulatory policy are explored in detail in a forthcoming Bank of systems which meant that losses England discussion paper, Securities Fraud by Norvald Instefjord, were not uncovered and spurious Patricia Jackson and William Perraudin. Drawing on that analysis, this profits were recorded; in two of article asks why the market may provide inadequate incentives for firms the cases spurious profits boosted to avoid problems of this kind and how regulators can motivate manage- dealers’ bonuses; and manage- ment within financial firms to exert adequate control over traders and ments’efforts to monitor the dealers other key employees. This issue is the matter of active debate among regu- were inadequate. lators and this article should be seen as an analytical contribution to that There are many ways in which debate rather than an anticipation of the final outcome. those who work for financial firms may attempt to deceive them. For As banks have become increasingly example, in 1996 Morgan Grenfell involved in securities activities, lost a great deal of money because their exposure to sudden damage a fund manager created a network from illicit or irregular activity by of shell companies to conceal his their employees has increased. highly speculative and, ultimately, Although banks have always heavily loss-making investments. been open to fraud by employees, In a second broad category of the scope for frauds amounting to cases, individuals in financial firms hundreds of millions of pounds is attempt to deceive or manipulate limited in traditional lending banks, other market participants to gain an given the checks and balances in the advantage. Between 1989 and 1991, granting of loans. In securities activ- traders in Salomon Brothers broke ities the scope can be much larger. rules limiting the fraction of a US Traders may be in a position to take Treasury auction for which a firm substantial exposures for the firm in could bid to 35 per cent. fast moving markets. In the May 1991 auction of The problem can be especially two-year notes, for example, by great in markets where highly-geared placing bogus orders on behalf of exposures can be created with little its clients, Salomons ended up up-front expenditure or where the controlling $10.6bn of the $11.3bn complexity of some of the transac- issue. The initial cost to Salomons’ tions increases the potential for reputation of these rule breaches illicit or irregular activity. Even if a was very substantial. For a period,

8 SECURITIES FRAUD

many counterparties reduced their ensure firms’ controls are sound at trading with Salomons which forced every level. Illicit or irregular activ- it to shrink its balance sheet by a ities generally constitute a tiny third. fraction of the huge volume of bona A more extreme example of the fide transactions carried out by a vulnerability of firms to outside firm and hence are very difficult to fraud by their employees is that of spot from outside the firm. Drexel Burnham Lambert. In 1988, Furthermore, lax control envi- Drexels agreed to pay fines of ronments are seldom characterised $650m because of frauds carried by the absence of key control out by Michael Milken involving systems (easier to observe from activity in the junk bond market. outside) but by controls being over- The fines, although large, were ridden or set aside (much harder to outweighed by the reputational observe). costs. The effect on counterparties’ The key issue for regulators is, willingness to deal with the firm therefore, how to provide incentives was one of the contributory factors to encourage management to put behind Drexel’s later insolvency. appropriate control systems in place and to ensure they are effective at Market incentives Leeson was conducting an all times. Before discussing these It is not possible for regulators to elaborate deception, made incentives, it is worth asking why possible because he had control check and cross-check activities in of both dealing and back-office the market alone does not provide minute detail in their efforts to operations sufficient incentives for the manage-

BARINGS AND LEESON Box 1 Barings was declared insolvent on 26 February 1995, due to the activities of one trader — Nick Leeson — in Barings Futures Singapore (BFS), a Barings subsidiary. The Barings business was subsequently taken over by International Nederlanden Group NV (ING). From late 1992 until early 1995 Leeson had reported increasingly large profits on apparently riskless arbitrage trading, where Nikkei exposures taken on the Singapore Futures Exchange, SIMEX were supposedly being completely hedged by offsetting contracts taken on the Japanese exchanges. In fact, Leeson was conducting an elaborate deception. From the outset, rather than making profits as reported to the firm, he had incurred losses. To conceal the losses, Leeson had employed a hidden SIMEX account numbered 88888. He had persuaded a back-office programmer to alter Barings’ accounting information so that information about the trading on 88888 would not be reported to London. Information on margin requirements for this account were, however, sent to London but were not reviewed. On 27 February the losses on Leeson’s dealing positions amounted to £827m. The fact that Leeson was in charge of both dealing and back-office operations in the Barings Singapore subsidiary was crucial in facilitating the deception. The problem of lack of segregation between the back and front offices in BFS had been spotted in an internal audit of the company in August 1994 but Barings’ management had not implemented the internal audit recommendations. The transfer of huge volumes of funds from Barings to Singapore to cover the margin calls on SIMEX, created by Leeson’s illicit trading positions, led to a marked discrepancy between the requests for margin and the margin report. Instead of fully examining this discrepancy at an early stage, assumptions were made about the calls being generated by customer positions. This led to the exposures being uncovered only at a very late stage.

9 SECURITIES FRAUD

Box 2 KIDDER PEABODY AND JOSEPH JETT

The case of Kidder Peabody involved are called reconstitutions. In the Kidder Peabody system, where the creation of false profits, allegedly a strip was held, the reconstitution could be entered as a by a single trader, Joseph Jett. forward and generate an immediate apparent profit. As the Although the trader appeared to be settlement date approached, the apparent profit would disappear making substantial profits ($32.5m in as current and forward prices converged. In order to maintain a 1992 and $151m in 1993) he was in given level of profit, it would therefore be necessary to fact making losses. progressively increase the volume of forward reconstitutions. The recorded “profits” enabled The management of Kidder Peabody became concerned Joseph Jett to earn a bonus of $9m. about the swings in the total balance sheet caused by these When the action was uncovered there was a balance sheet gap positions and the size of the reported profits and carried out an of $350m. investigation which uncovered the problem. Kidder Peabody claimed that Joseph Jett had exploited One of the issues in the case was weakness in the Kidder a weakness in its accounting system with regard to strips. The Peabody control systems. The Securities and Exchange Comm- New York Fed offers bond dealers a service whereby it ission brought cases against Jett’s two superiors for failure to exchanges conventional government bonds for a basket of strips, supervise, which they settled. A case was also brought against representing claims to the bond’s coupon and principal payments. Joseph Jett for securities fraud. Jett is contesting the case and has The Fed will also exchange strips for the original bonds — these claimed he was acting under orders.

ment to put in place an effective led to substantial losses, but the bonuses — or even sacking — those control environment. management might be affected to a who are responsible for monitoring First, it may be true that the much smaller degree. the fraudster. private cost of failure to share- Indeed, the management may Change of management may holders is lower than the social cost well have reasons not to ask not be a very effective sanction, and, therefore, there may be social searching questions of star traders. however, given the likelihood that reasons for tightening controls further One feature of some of the recent individuals, particularly if they were than shareholders might think cases is the extent to which the not seen as being directly involved necessary. bonuses of senior management in the problem, could be hired by This would certainly be the case depended on the profits, or apparent competitors. where the failure of a large firm profits, which were generated by Furthermore, the potential dis- would disrupt the market and star traders. ruption which would result from the perhaps affect confidence in other Although shareholders have sacking or penalising management institutions within that market. limited scope to influence the may be quite serious (for example Second, for large banks, the share- detailed control environment, they Morgan Grenfell faced problems in holders do not normally take decisions could, in principle, insist on the wake of the Peter Young affair). about controls. These decisions are contracts for senior managers which Therefore, it may not be possible for taken by the management of the would provide appropriate incen- firms ex-ante to commit to take such bank, creating what is known as a tives by penalising managers after action (vis-à-vis key managers) if a principal-agent problem. The share- problems have been discovered. fraud is discovered as the result holders would suffer if lax controls Possibilities include withholding could be further damage to the firm.

10 SECURITIES FRAUD

Regulatory incentives inspections of securities firms, sys- ments met by the banks are varied Thus, ensuring control systems are tems and controls. according to the risks they face and effective is an important objective Once problems have been iden- quality of controls rather than being for financial regulators. To pursue tified, regulators have at their a fixed minimum. this objective, they must first disposal various levers to encourage The risk-asset ratio is not made inspect firms to check that key firms to tighten the control environ- public and therefore does not affect controls are in place. ment. In the UK, in the case of a bank’s reputation. However, it is In the UK, inspections of the banks, an important lever is the risk- an effective lever — if a bank is banks’ systems and controls are asset ratio. If there are concerns close to its target ratio it has to seek organised in two different ways. about the control environment, and more capital from shareholders to The banking supervisors set the therefore about the risks faced by maintain the same profile of book, scope for a report by the firm’s the firm, the supervisors will which may well be regarded as reporting accountants which goes to request remedial action and, if not costly by the bank’s management. the firm and to the regulators. The satisfied, will increase the bank’s Traditionally the securities regula- supervisors also carry out their own capital requirements by raising the tors have not tried to use capital as team visits to look at systems and risk-asset ratio. This is possible a lever in this way but have had controls. The Securities and Futures because the UK is one of the few fixed minimum capital require- Authority (SFA) carries out on-site countries where capital require- ments.

Box 3 DAIWA AND IGUCHI

The Daiwa case also involved the disguise of loss-making positions. Iguchi, who had been appointed Head of Daiwa’s New York bond trading operations in 1983, used fraudulent activity to disguise trading losses. Over an 11-year period (from 1984 to 1995) the total loss amounted to $1.1bn by way of an estimated 30,000 unauthorised trades — an average loss of $400,000 for every working day. The losses were hidden by selling securities held on behalf of customers. These sales were concealed by hiding the trade confir- mations and forging statements of securities holdings as well as falsifying the statements of securities held for customers. As in the case of Barings, disguising the losses was possible because Iguchi, until 1993, was in charge of both the front and back offices. Even after separation of the two offices, the discrepancy between actual securities holdings and Iguchi’s forged statements was not picked up. When announcing the losses, Daiwa said that the pay of senior management would be temporarily cut to reflect their responsibility. On 2 October 1995, the Federal Reserve Board and New York State Banking Department issued a cease and desist order against the New York branch. In a joint statement in November 1995, by the Federal Reserve, the FDIC, NYSBD and several other state banking departments, it was announced that a joint order had been made terminating the US banking operations of Daiwa Bank. Orders were also made requiring the termination of the Daiwa Bank Trust Company operation in the US. The actions were taken on the basis that Daiwa Bank and its officials had engaged in unsafe and unsound banking practices and violations of law over an extended period of time that were most serious in nature. Daiwa agreed to terminate or sell its business in the US by 2 February 1996. In addition, the bank was faced with a 24-count indictment claiming criminal fraud. One issue of concern to the US authorities was that although Daiwa’s senior management had learnt of the losses from Iguchi in July 1995, they had concealed those losses from the US authorities for almost two months. Also in 1992 and 1993, the management of the firm had misled the Federal Reserve examiners regarding their trading activities and the separation of trading from the back office function. In February 1996, Daiwa pleaded guilty in a US district court to a conspiracy to conceal the trading losses. It was required to pay a fine of $340m.

11 SECURITIES FRAUD

However, the UK’s SFA has gaze of the regulators, might be less public sanctions including fines. started to move in this direction so. It is noticeable that in both Banking supervisors have tended to by introducing variable capital the Barings and Daiwa cases the concentrate on private remedial requirements for credit exposures of problem lay in overseas operations, action. securities firms, enabling it to and in the case of Drexel in an oper- Several banking supervisors, increase substantially the require- ation geographically separated from such as those in the UK, do not have ments for firms perceived as having the main business. the power to fine banks (see poor credit control procedures. Table 1) and others which do have Among other levers available the power rarely use it. Banking to regulators, intensity of supervi- supervisors have focused on the sion can be used both as a response removal of management and, in to perceived weaknesses and a way Even if a firm extreme cases, removal of authori- of encouraging management to sation. However, if possible the improve controls. A number of regu- does possess latter is avoided and remedial action lators are currently in the process of is undertaken. Actions taken short putting in place systems of this the key control of de-authorisation are usually not kind. For example,the UK banking publicised to prevent loss of confi- supervisors are developing an app- processes the dence. Where managers are no roach called RATE which will longer regarded as fit and proper attempt to assess the risk profile of crucial question they would, in the UK, not be able banks and the adequacy of their to take a senior position in another control environments. This will r emains whether bank or investment firm, although enable supervision to be tailored they may still be able to seek a posi- towards higher-risk players. The it is operating tion as a consultant. SFA has introduced a similar effectively Securities regulators’ willing- approach and the fund management ness to announce penalties publicly and retail conduct of business regu- at all times reflects in part the fact that securi- lators, IMRO and the PIA have ties firms, because of the liquid moved in this direction as well. nature of their balance sheets, are Even if a firm does possess the easier to close, if confidence in them key control processes the crucial wanes, without substantial losses to question remains whether it is oper- creditors. Drexels eventually failed ating effectively at all times and Ex-post penalties and was liquidated with little loss to whether, for example, in the case of The design of appropriate ex-post creditors. The same would not be internal audit, its recommendations penalties is, therefore, also impor- true of a bank because the loan book are faithfully implemented. For tant. It is noticeable that banking can only be sold quickly at a outside regulators, or auditors acting and securities regulators have devel- substantial discount. Recently, secu- on their behalf, these questions may oped rather different approaches to rities firms may well have become be difficult to answer. This point is such penalties. The securities regu- somewhat less liquid than in the particularly important in the case of lators have used a combination of past because of the volume of swaps large diversified institutions. Core approaches: private remedial action, that they carry, but even these can activities may be well run, but far for example, to deal with inadequate be sold much more easily than a flung operations, further from the management, and the imposition of loan book because counterparties

12 SECURITIES FRAUD

Table 1 FINES — BANKING SUPERVISORS Country Power to Frequency Amount Exercise Germany Yes Usually 10 to 20 cases of enforcement fines and about Current enforcement fines DM10,000 (£3,500) up to 4 to 10 cases of administrative fines per annum. DM50,000(a) (£17,500). Administrative fines not exceeding DM100,000(b) (£35,000) and should reflect benefit gained from infringement. In practice, usually between DM200 (£70) and DM25,000 (£8,500).* Norway Yes Authority not used so far. No maximum. The amount would depend on the size of the institutions and the severity of the issue. Austria Yes N/A Enforcement fines up to ASch300,000 (£14,850) Greece Yes N/A N/A Iceland Yes Seldom used. N/A Luxembourg Yes Not frequently used. Luxfr5,000 (£85) to Luxfr 5,000,000 (£85,000) and/or 3-8 days’ imprisonment depending on the nature of the infringement. Liechtenstein No. Power No cases within last few years. Depending on the specific case up to SFr50,000 rests with the (£21,00) or SFr100,000 (£42,000). Government and the district court Belgium Yes Not used so far. Fines under administrative law are BFr10,000 (£170) to BFr1m (£17,000 ) per calendar day and a global upper limit BFr50 m (£850,000). France Yes Rare. In most cases fines are imposed for late trans- Fines may not exceed the minimum amount of capital mission of financial information to supervisors or required for the institutions. They have ranged between failure to comply with money laundering regulations. FFr10,000 (£1,000) & FFr300,000 (£30,000). Ireland No Italy Yes Some 80-85 cases on average over the three years Breaches of prudential rules, integrity require- 1994-96 ments etc. L1m (£360) to L50m (£18,000). Corporate officers who breach the rules regarding disclosure to customers of terms and conditions of contracts L2m (£720) to L25m (£9,000). UK No Spain Yes Authority used frequently but usually for not meeting Maximum amount for very serious infringements, complex reserve requirements. 1 per cent of own funds or Pta5m (£20,500), for serious infringements 0.5 per cent of own funds or Pta2.5m (£10,250). Maximum fine for a manager Pta 10m (£41,000). Netherlands No USA OCC — Yes This type of action usually reserved for individuals The amount depends on the seriousness of the except where a bank has filed a false or misleading case but most fines for individuals are $10,000 report. Atotal of 614 fines has been assessed in the last (£6,200) to $25,000 (£15,500). Federal six years. Powers to fine firms for law violations, Maximum $1m or 1 per cent of assets per day Reserve — Yes eg misreporting. 5-10 cases a year all foreign banks. for a firm. Largest fine BCCI $200m (£124m). Individuals can be fined for their actions — perhaps Fines for individuals usually $10,000 to two cases a year. $100,000 (£6,200 to £62,000). Japan Ministry of Only one case of a bank being fined. Other penalties Maximum of Y500,000 (£2,600), likely to be Finance — Yes are used as well such as restrictions on business activ- revised upwards. Bank of ities. Japan — No Denmark Yes Used only once for failure to submit annual accounts. There is no limit on fines. * Germany — A new law will take effect on 1 January 1998 increasing (a) to DM500,000 (£175,000) and (b) to DM1m (£350,000)

13 SECURITIES FRAUD

Table 2 ANALYSIS OF SFA FINES, 1991 TO DATE

(i) (ii) (iii) (iv) (v) Control Failures Fraud or Breach of Exceeding Misleading Deception Reporting Dealing Information (Theft) Requirements Authority Advertisements No Cases 58 95 17 10 6 Max Fine (£) 240,000 200,000 25,000 18,000 200,000 Min Fine (£) 2,000 1,000 1,500 5,000 9,000

Notes: (i) Fines on firms or senior management Source: SFA notices (ii) Usually individuals who are also often banned from the industry (iii) Usually Firms (iv) Sometimes individuals, sometimes firms (v) Usually firms

are generally of high quality and Changes in the Table 3). Given the potential prin- hence exposures are more easily cipal agent problem between the priced. UK regulatory shareholders and the managers, this The proposed changes in the may not always be appropriate. regulatory structure in the UK will structure will Where the shareholders are domi- mean that, in future, policy on nant and in a position to influence penalties and regulatory incentives mean that policy the way management runs the busi- for the management of financial ness, fines on firms may provide firms will be developed by a single on penalties and appropriate incentives. If share- regulator spanning financial services holding is widely dispersed, penalties firms — banks, building societies, r egulatory for individuals within the firms may friendly societies, credit unions and incentives for have more effect. insurance companies. One of the Instefjord, Jackson and Perr- issues which will be considered is the management audin (1997) provide a theoretical whether the approaches adopted for analysis of how fines on individual different types of firms should be of financial dealers and on those charged with consistent or whether there are good monitoring them, affect dealers’ reasons for differences. firms will be incentives to report profits accu- rately. Those acting as monitors Analysis of penalties developed by a may either be management or share- An analysis of SFA fines from 1991 holders depending on whether the to date (Table 2) shows that a rela- single regulator interests of these two groups are tively high proportion (over 30 per aligned. The analysis suggests that cent) have been for control failures. spanning penalising the manager decreases It is interesting to note that fines financial the prevalence of false reporting as are frequently levied on securities one might expect. Simply raising firms themselves rather than on services firms the level of dealer fines may have the individual within them (see the unintended effect of lowering

14 SECURITIES FRAUD

the amount of monitoring by the years — the numbers rising from firm since managers could rely too one and six in 1991 and 1992 to 12 much on the deterrent effect of the Raising the and 17 in 1995 and 1996 respec- dealer penalty. tively (see Table 3). An important supplement for level of dealer An important question facing direct pecuniary penalties such as regulators is: how many levels of fines is to ban those associated with fines may have management, above an individual fraud or who have been negligent committing securities fraud, should regarding the control environment. the unintended be sacked or banned? Instefjord, Bans may be hard to enforce since effect of Jackson and Perraudin (1997) look at individuals may resume their activ- this question in a theoretical model ities under other guises. However, it lowering the of a hierarchy in which individuals is likely to impair the future earn- decide whether or not to commit ings of the individual concerned amount of fraud and whether to monitor indi- significantly. In the UK, the SFA viduals below them in the hierarchy. has designated increasing numbers monitoring The model explores how regu- of individuals “not fit and proper” to lators may influence the agents’ work for regulated firms in recent incentives to monitor or commit fraud

Table 3 SFA TYPES OF PENALTY — JAN 1991 TO MAR 1997

Individuals: Year No of Fines Costs Not Fit & Reprimand Cases Awarded Proper 1991 11 0 1 1 11 1992 11 4 4 6 3 1993 10 3 3 10 0 1994 34 28 25 12 5 1995 22 10 11 12 8 1996 37 18 34 17 16 1997 53414 Total 130 66 82 59 47 Firms: Year No of Fines Costs Expelled Reprimand Restrictions Not Cases Awarded Imposed Permitted 1991 19 19 7 0 1 0 0 1992 11 5 1 5110 1993 3110101 1994 4322 000 1995 6570101 1996 11 11 11 0 4 0 0 1997 1110100 Total 55 45 30 7 9 1 2

15 SECURITIES FRAUD

by altering penalties imposed on culture of a firm and can affect the fraudsters and fines levied on costs for individual managers of managers found not to have moni- monitoring their subordinates. tored those below them in the Recently, regulators in several hierarchy. It also explores the action countries have considered clarifying which could be taken by the firms to the responsibilities of senior manage- encourage greater monitoring. ment as a way of influencing firms’ The model shows that reliance control arrangements. on heavy penalties alone might not The example of New Zealand have the intended effect. Managers is interesting in this respect. In might even monitor staff less because 1996, the Reserve Bank of New of a belief that subordinates would Senior Zealand introduced a new approach be deterred from committing fraud to supervision for banks operating due to the heavy external penalties. management in New Zealand, all of which are Although penalties for the managers foreign-owned. Prudential supervi- in the hierarchy above the fraudster strongly sion of banks by the authorities was might help to reduce fraud, it is reduced and a new public disclosure possible that they could be made influences the regime introduced. A prime objec- more effective by increasing the compliance tive was to strengthen the incentives number of levels of management for the management of banks to that are held responsible. culture of a maintain sound banking practices The most effective way to and to facilitate monitoring of bank encourage tighter controls is for firm and can financial performance and risk firms to reward tight monitoring management by other participants and the spotting of potential or affect the costs in the market. actual problems. This is all the more An important element in the the case because these rewards for individual New Zealand approach is a strength- must be large enough to outweigh ening of the role of bank directors. the costs of monitoring which could managers of Such directors are required to sign be perceived as large. For example, the disclosure statements to certify if a senior manager has a particular monitoring their that “the information contained in star trader in his group he might be subordinates the statements is not false or mis- concerned that tight monitoring leading”, and to attest to their bank’s might lead to the loss or demotiva- financial soundness and to adequacy tion of that individual, affecting the of its risk management systems. returns for the group and the indi- Directors must state that the bank vidual bonuses, including his own. has systems in place to monitor and control adequately the group’s Senior management material risk, including credit risk, It is clear from the cases examined concentration of credit risk, interest in Instefjord, Jackson and Perraudin rate risk, currency risk, liquidity (1996) that senior management risk, and other business risks and strongly influences the compliance whether those systems are being

16 SECURITIES FRAUD

properly applied. Where the disclo- math of problems, individuals not sure statement is found to be false meeting these responsibilities could or misleading, banks and their be subject to disciplinary procedures. directors face potentially severe Another issue for the firms is penalties, both criminal and civil. whether there should be an execu- In the UK, the Bank of England tive director with sole responsibility is exploring whether an annual for the control function and with no statement should be sought from the business line responsibilities. In the chief executive and chief financial Barings episode, it was noticeable officer of banks, affirming that that the recommendations of internal effective systems and controls are in Boards should audit were overridden by the busi- place and that Banking Act require- ness line management, begging the ments and policy guidelines have provide a question whether the control func- been complied with. Following tion reported to a high enough level difficulties winning disciplinary statement that within the company. However, it cases against senior managers, the would be important that the pres- SFA proposed a new rule imposing on the basis of ence of such a director were not a direct duty on senior executive enquiry they used to absolve other executive officers (SEOs) to take all reason- directors of responsibility for fail- able steps to ensure that employees are satisfied ures of controls and therefore there of the firm act so as to avoid serious would have to be a wider board damage to its reputation as a regu- that the responsibility. lated firm. An SEO would be liable The banking supervisors are of to disciplinary proceedings if this institution the view that the whole board duty were not met. This proposal should accept responsibility. The has considerable resistance from the complies with board should provide a statement firms because of an implicit reversal ’ that on the basis of enquiry they of the burden of proof. However, the Bank s have satisfied themselves that the the proposal has now been modified authorised institution (or branch) is to deal with this concern. The fitness and in compliance with the Bank’s burden of proving failure by an properness fitness and properness criteria. The SEO to comply with his duties boards might also be asked to under the rules will rest with the criteria provide a statement as to whether SFA, as at present. the institution has systems in place The SIB has also carried out a to monitor and control adequately review of these issues and released the institution’s material risks and a consultative paper in July, entitled whether those systems are being Responsibilities of Senior Manag- properly applied. ement proposing a different approach. The development of a single Every firm would have to prepare regulator in the UK will mean that a statement setting out the manage- these issues will be carried forward ment structure and defining where by a body which spans all forms of responsibilities rested. In the after- financial activity.■

17 designed to follow up some of the THE VALUATION OF OPTIONS issues raised — and to explore other By Howard Walwyn, Bank of England, and issues that were not covered — in this earlier work. More specifically, Wayne Byres, Reserve Bank of Australia the objectives of the second survey Options have become widely traded products in financial markets. were: By their very nature — an option, rather than an obligation — they • to supplement the process of have transformed risk-management practices and have allowed the model recognition under the development of new financial products. These are used by a wide Capital Adequacy Directive range of investors and consumers. This is the Bank of England’s (CAD)2 by ensuring that banks second survey of options valuation: the first was conducted in 1995. that had received model recog- nition were valuing their options The pricing and valuation of options portfolios appropriately; is not straightforward. It involves • improve further the Bank of modelling a complicated interrela- England’s own knowledge of the tionship between a number of market complexities of valuing options, factors which are not always easily particularly in relation to some measurable. Consequently, the valu- of the more exotic products that ation of options has attracted a good are becoming increasingly com- deal of attention from market partic- mon; and ipants and regulators. • to assess the range of valuations Market participants who deal in and risk parameters supplied by options need to ensure they have participating institutions, and to valued these products appropriately provide feedback to these insti- to measure and manage their expo- tutions. sure to credit and market risks, and calculate their profits and losses Methodology correctly for both internal manage- More than 40 institutions with major ment reporting and external statutory trading activities in London were accounting purposes. asked to participate in the survey. Regulators and supervisors are These included all the banks granted also interested in the appropriate model recognition by the Bank of valuation of option portfolios, not England, and a number of other least so they can assess whether institutions which are significant adequate amounts of capital are participants in the derivatives markets held against the risks involved. globally. It was against this background Eight products were chosen for that in December 1996 the Bank of valuation (see Table 1). Of these, England conducted its second options five were options on foreign exchange valuation survey. The earlier survey, positions, and the others were interest- conducted in February 1995, had rate related. Participants were asked yielded a number of interesting to undertake a number of different results,1 and the second survey was valuations of each product, usually

18 VALUATION OF OPTIONS

involving variations in the under- data (eg forward points, yield curves, exchange options revealed a similar lying strike price. Valuations were volatilities) and to use their own pattern to the last survey. For requested as at 4.15pm London time model to produce a valuation.3 example, the vast majority of on Wednesday 4 December 1996, While this approach makes varia- respondents who valued European using mid-market rates and assuming tions in the results more difficult to vanilla options used the Black- 4 a counterparty of good standing explain than if these data are given, Scholes model. Many respondents (AA rated) with adequate credit there are benefits in being able to also made explicit reference to 5 lines available. identify potential problems with the the Garman-Kohlhagen variation. Participants were given the input data used by the individual Several banks’ responses suggested basic economic details of each of institutions. the use of proprietary (as opposed to the transactions, and were then Survey participants’ responses market standard) analytical models, asked to collect the necessary input on the models used to value foreign but it was not clear to what extent

Table 1 PRODUCTS CHOSEN FOR VALUATION

Product 1 GBP/DEM straddle (ie a GBP/DEM call and a GBP/DEM put with the same strike) a) 10-month option at-the-money-forward b) 10-month option with a strike at GBP/DEM 2.5500 Product 2 JPY/CHF straddle a) 10-month option at-the-money-forward b) 10-month option with a strike at 95.00 Product 3 Reverse knock-in (up-and-in) barrier option on GBP/DEM a) 12-month option, strike at-the-money forward, barrier at 2.6500 b) 12-month option, strike 2.3500, barrier at 2.6500 Product 4 Double barrier (knock-out) option on GBP/DEM a) 12-month option, strike at-the-money forward, barriers at 2.4000 and 2.6500 b) 12-month option, strike 2.3500, barriers at 2.4000 and 2.6500 Product 5 Digital (binary) range option on GBP/DEM a) 12 month option, barriers at 2.4000 and 2.6500, payout if within range at expiry b) 12 month option, barriers at 2.4000 and 2.6500, payout if within barriers for entire life Product 6 European call option on US Treasury Bonds a) 12-month option on “on the run” 10-year bond, strike at mid-market clean spot price b) 12-month option on “on the run” 10-year bond, strike at clean price of 106.00 per 100.00 c) 12-month option on 9 3/8% Feb 06 bond, strike at mid-market clean spot price d) 12-month option on “on the run” 5-year bond, strike at mid-market clean spot price Product 7 Collar (Interest Rate Cap and Floor) on GBP LIBOR a) 12-month collar, 3 x 3 monthly resets, collar at 75bp above/ 25 bp below 3 month LIBOR b) as per a), but with floor set to produce zero premium product Product 8 European swaption on USD LIBOR a) 12-month receiver swaption for 10-year swap paying 6 month USD LIBOR, strike 6.7% b) 3-month receiver swaption for 10-year swap paying 6 month USD LIBOR, strike 6.7% c) 12-month receiver swaption for 5-year swap paying 6 month USD LIBOR, strike 6.25%

19 VALUATION OF OPTIONS

these approaches were simply modi- fied versions of the Black-Scholes analysis. For exotic options, the outcome was more varied. About half the respondents on the single barriers (product 3) used a standard Black- Scholes/Garman-Kohlhagen model. Others were evenly distributed among proprietary analytical models, numerical (Monte Carlo) models and binomial/trinomial tree-type models. The majority of respon- dents who valued the double barriers (product 4) used exactly the same model as they used for the single barriers. Although a smaller number of respondents valued the digital products (product 5), the pattern for these products was very similar, with around half using Black-Scholes and half using proprietary, numer- ical and/or tree models in equal proportion. Participants’ responses on the models used to value the interest- rate products also revealed a similar pattern to that observed in the 6 previous survey. Black or Black- Scholes models were universally used to value the interest-rate collar product (product 7), and were also widely used to value the swaption products (product 8). A number of other models were used in addition to Black/Black-Scholes model in the valuation of the swaptions, 7 namely Black-Derman-Toy bino- 8 mial, trinomial and Hull-White trees, while a range of models was also used for the valuation of the bond option products (product 6): one third of the respondents used

20 VALUATION OF OPTIONS

Black/Black-Scholes, one third used variation of the main results for In some survey products, partic- 9 10 Cox-Ross or Cox-Ross-Rubinstein each product, expressed in the form ipants were asked to value the same binomial models based on bond of a percentage standard deviation11 option with different exercise prices. forward yields, while the rest of the from the mean. To obtain an accurate valuation in respondents used a variety of models markets where a volatility smile is including finite difference grids and Volatility smile present, banks should use different other proprietary numerical models. The term “volatility smile” is used to volatility inputs to value options. describe a situation in which the Taking foreign exchange prod- Results implied volatility associated with ucts 1 and 2 as an example, however, A summary of the main results of options away-from-the-money differs only one third of survey respondents the survey is set out in Table 2 (see from (typically is higher than) that used different volatilities for the Box 4, Option Products, for defini- associated with an at-the-money away-from-the-money option in tions). This illustrates the degree of option. comparison to the at-the-money

Table 2 MAIN RESULTS Std Dev (%) Product No. of Responses Value Delta Gamma Vega Spot Rate Implied Volatility Product 1(a) 35 2.7 5.3 3.5 0.4 0.1 2.5 Product 1(b)39 2.4 3.9 3.7 0.7 0.1 2.5 Product 2(a) 34 7.3 9.1 8.3 0.9 0.1 7.4 Product 2(b) 35 1.9 4.3 12.1 13.8 0.1 8.1 Product 3(a) 33 5.3 2.3 7.4 2.1 0.1 3.2 Product 3(b) 33 4.6 3.2 10.7 3.8 0.1 3.2 Product 4(a) 30 66.1 19.0 32.1 27.4 0.1 3.9 Product 4(b) 30 48.8 18.1 33.9 31.0 0.1 4.1 Product 5(a) 22 4.6 21.9 16.6 15.3 0.1 1.6 Product 5(b) 26 52.6 19.8 39.1 29.8 0.1 4.2 No. of Responses Value Delta Gamma Vega Forward Bond Implied Yield Volatility Product 6(a) 18 3.7 6.9 9.9 33.0 0.3 3.8 Product 6(b) 18 5.3 8.0 8.6 34.7 0.3 4.0 Product 6(c) 18 7.4 10.5 14.1 36.9 0.2 4.1 Product 6(d) 17 7.1 9.2 13.5 48.9 0.2 3.0 No. of Responses Value Floor Implied Volatility Product 7(a) 38 10.8 Collarlet 1 38 66.1 10.2 Collarlet 2 38 20.8 6.5 Collarlet 3 38 22.5 5.7 Product 7(b) 33 0.7 No. of Responses Value Delta Gamma Vega Implied Volatility Product 8(a) 33 6.6 10.1 17.1 2.7 1.2 Product 8(b) 33 11.4 11.0 11.5 5.9 1.9 Product 8(c) 33 10.8 26.4 29.9 0.9 1.5

21 VALUATION OF OPTIONS

option. Part of the reason for this is Box 1 BASIC DEFINITIONS that volatilities for away-from-the- A call (put) option is an option that gives the buyer the right but not the obligation money positions, or less commonly to buy (sell) an underlying asset on, or sometimes before, an agreed date at an agreed traded products, may not be readily price, known as the strike or exercise price. available, and consequently many An option is at-the-money if its strike is equal to the spot market price of the under- banks may have had little choice but lying asset. At-the- money-forward means the strike is equal to the current forward to use at-the-money volatilities for market price of the underlying asset for the maturity date of the option. revaluing their positions on a daily An option is in-the-money (out-of-the-money) if, from the buyer’s perspective, its basis, regardless of whether the strike is more (less) favourable than the current spot or forward market price of the option is in or out-of-the-money. underlying asset. However, it is important in these circumstances that such institutions regularly (say, on a monthly basis) Box 2 RISK AND RISK PARAMETERS attempt a fully independent and accurate valuation, which reflects the The sensitivity of an option is the measurement of the change in its value which fact that volatilities away-from-the- results from a change in the price of the underlying asset, in the volatility of the money are not necessarily identical underlying asset, or in any other factor determining the value of the option. Option to those at-the-money, and ensures sensitivities are also sometimes described as risk parameters or — because they any valuation errors are not exces- are named after Greek letters — Greeks. sive. Delta is the sensitivity which describes the rate of change of an option’s value with Consistent with the view that respect to a change in the price of the underlying. Gamma describes the rate of volatility smiles are less prevalent change of an option’s delta with respect to a change in the price of the underlying in interest-rate markets, there was asset. Vega gives the rate of change of an option’s value with respect to a change in less evidence of different volatilities the volatility of the underlying asset. being used for away-from-the- money options in the interest-rate products. In bond options (product 6) Chart 1 US TREASURY BOND OPTION there was no clear consensus on the GAMMA AND VEGA 45 existence or shape of a volatility 650 smile. 40 In the swaptions (product 8), 550 the product specifications were not Gamma (LHS) 35 set up to enable analysis of the

450 volatility smile. The survey results 30 did, however, show reasonably

350 $000's systematic evidence of a term struc- $000's 25 ture of volatilities being used, as all 33 respondents to these products 250 20 Vega (RHS) used a higher volatility for the one- year option on a 10-year swap 150 15 (product 8(a)) than for the three- month option on a 10-year swap 50 10 Responses (product 8(b)).

22 VALUATION OF OPTIONS

“The Greeks” difference appears to stem from ucts a bimodal distribution was The key means of managing option the complexities of the particular apparent (see Chart 1). It is likely portfolios is by way of sensitivities product involved: the pay-off func- that this pattern resulted from some (eg delta, gamma and vega; see tion contains discontinuities which respondents’ definition of vega as Box 2). By using these measures can cause sensitivity measures to the change in option value arising to guide hedging strategies, banks change rapidly over a small range of from a change in the bond yield attempt to limit their risks to accept- inputs. volatility rather than from a change able levels. However, the survey Turning to the interest-rate in bond price volatility as requested highlighted quite stark differences products, the measurement of sensi- in the product specification. It is in estimated Greek sensitivities tivities in the sterling interest rate also possible, however, that the between respondents. collar (product 7) was based on a pattern was related to the model In the foreign exchange prod- wide range of methodologies. This being used. Either way it is worth ucts, further discussion with some made analysis difficult, because it noting that the high degree of vari- of the participants revealed that did not prove possible to convert ation in vega exposures, represented these differences reflected the them into standard form. by a reported standard deviation in methods used to calculate the sensi- Approaches to the measure- excess of 30 per cent for all four tivities. Some participants had ment of sensitivities in the swaption bond option products, is probably calculated their sensitivities using product (product 8) were more excessive. the partial derivatives of the option standardised, although for some All these findings reaffirm the pricing equation. Others, however, banks — which quoted delta as a need for precision and care in the calculated their sensitivity measures percentage of the hedging swap — definition of Greek sensitivities, by revaluing the option fully for a it was necessary to translate their because although the Greeks are given shift in the relevant para- reported delta and gamma into often assumed to have universal meter, eg valuing an option using a “sensitivity equivalents” to enable definitions there can be legitimate volatility of 7 per cent, and then comparison with the majority of alternative definitions which do not valuing the same option again using respondents. This translation did necessarily accord with common a volatility of 8 per cent, with the introduce some estimation error market practice. This requires that change in value taken to be the vega into the summary statistics for delta banks take great care with their sensitivity. Particularly for the more and gamma on all the swaption measurement systems, understand exotic products, these two methods products. each method clearly, and analyse produced outcomes which diverged An interesting result emerged carefully the extent to which hedging significantly. in the calculation of vega numbers positions will protect against both In product 4, for example, a in the bond option products small and large movements in the number of banks used the “revalu- (product 6) where in all four prod- underlying instrument. ation” method to calculate their vega exposures, while another group Table 3 LIQUID AND ILLIQUID VARIANTS of banks used partial derivatives. Within each group the results were Std Deviation Value Delta Gamma Vega broadly consistent, but the group Product 1(a) 2.7% 5.3% 3.5% 0.4% using partial derivatives tended to Product 2(a) 7.3% 9.1% 8.3% 0.9% report a figure about one third Product 6(a) 3.7% 6.9% 9.9% 33.0% higher than the banks using the Product 6(c) 7.4% 10.5% 14.1% 36.9% revaluation method. The size of this

23 VALUATION OF OPTIONS

Chart 2 FX DOUBLE BARRIER OPTION: It is also important, especially 200 STANDARD v PROPRIETARY MODELS where models are being used to 180 generate capital requirements, that Standard Mean-67 160 supervisors understand the implica- Range-101 tions of each particular method and 140 ensure that each bank is using a

120 Proprietary single method consistently over Mean-92 time. lue £000's 100 Range-157 Va 80 Appropriateness of models

60 One interesting feature of the latest survey was a much higher response 40 rate than in the first survey on the 20 more exotic products (in particular

0 products 3, 4 and 5). This reflects Responses the rapid growth in these products Chart 3 FX NO-TOUCH DIGITAL OPTION: in the past year or so, when rela- STANDARD v PROPRIETARY MODELS tively low volatility in foreign 300 exchange markets has led market participants to seek new and inno-

250 vative structures appropriate to such Standard Mean-77 an environment. Range-153 200 As a result, barrier and digital alue £000’s V option structures have become

150 Proprietary increasingly commonplace. Less Mean-113 Range-240 than 20 respondents to the 1995 survey valued the barrier option, 100 while 33 participants attempted in 1996. Similarly, the double 50 barrier and digital options included in the latest survey were valued by 0 Responses 30 and 22 institutions, respectively, suggesting these markets are becom- Box 3 VOLATILITY DEFINITIONS ing increasingly competitive. One distinction to emerge in Volatility measures variability in price of an underlying asset, represented by the the survey results for these exotic annualised standard deviation of the natural log of the ratio of two successive prices. products was between those banks Implied volatility is the value of volatility which is implied by a given option price. that used standard industry soft- The volatility smile depicts the graph of the implied volatility of an option against ware, and those that had developed different strike prices for a given maturity, often giving rise to a smile-shaped curve. their own proprietary valuation The term structure of volatility is the curve describing the implied volatility of models. at-the-money options with different maturities. Charts 2 & 3 highlight this issue. A large number of partici-

24 VALUATION OF OPTIONS

Box 4 OPTION PRODUCTS

Whereas a vanilla option is the broad term used to describe a standard option, an exotic option is an option that is non-standard in terms of its pay-off, its underlying asset, exercise price or expiry conditions. Some types of “exotic” — for example some foreign exchange barrier options — are now so widely traded that they are regarded by the market as “vanilla” options, but for the purposes of this article any option with a non-standard pay-off function, eg a barrier option, is defined as “exotic”. Buying (selling) a straddle involves the simultaneous purchase (sale) of a put and a call with the same strike and same maturity. A barrier option describes the broad class of option which is activated (knocked in) or terminated (knocked out) only when the rele- vant spot price reaches an agreed trigger level or “barrier” between inception of the option and maturity. A reverse barrier describes a barrier option whose intrinsic value increases as the spot price moves towards the barrier level. A double barrier option is a barrier option with two barriers, one either side of the current spot rate. A digital or binary option describes the broad class of option which has either a fixed pay-off or a zero pay-off depending on particular pre-agreed conditions. The basic at-maturity digital has a fixed or zero pay-off depending on the position of spot vis-à- vis a pre-specified trigger at maturity. Common variants are the one-touch, in which pay-off occurs if spot reaches the trigger at any time between inception and maturity, and the no-touch, in which pay-off occurs only if spot does not reach the trigger between incep- tion and maturity. A cap guarantees that the interest rate on a particular investment at any given time will be the lesser of the prevailing interest rate and the cap rate. A floor places a lower limit on the interest rate that will be applied. A collar is a combination of a cap and a floor, and specifies both an upper and lower limit for the rates that will be applied. Collars may be applied for a number of interest periods: a collar for a single period is called a collarlet. A swaption is an option on an interest rate swap, which gives the buyer the right to enter into a pre-specified interest rate swap at a pre-specified time and according to given terms and conditions.

pants used standard software pack- “correct” values against the crite- banks to respond to product 5(a), on ages for these products, and produced rion that model values should be the basis that this product was not a broadly similar group of results. reasonably representative of actual handled by standard pricing soft- However, institutions using their prices in the market. This may be ware. However, the product was one own internally-developed models because the models in question do that could be constructed using a produced a more disparate range of not take sufficient account of the combination of other options which responses. (two-dimensional) volatility smile, the software could handle. This Which values are “correct” is, or the (three dimensional) volatility suggested a reluctance to rely on of course, a matter of opinion.12 It is surface.14 knowledge and understanding of the widely known that the values One related concern for the products involved, and an undue produced by some standard pack- Bank which emerged from the reliance on computer models to ages, in their current form, can differ survey was the apparent “black provide the answers. considerably from the prices at which box” mentality that seemed to exist This analysis raises a number of deals are conducted in the market.13 in some institutions, ie the accep- important issues for institutions But, it also appears from the tance of the output of option pricing trading more exotic products. These disparate nature of some of the models without any intuitive feel include the need to ensure that staff survey results that a large number of for its reasonableness. in the back office and other support models — including some propri- One example of this mentality functions have the ability to use etary models — are not producing was the reluctance by a number of and understand more sophisticated

25 VALUATION OF OPTIONS

models (which are often developed Table 3 summarises the variation “outliers” compared with other and located in the front office) as between liquid and illiquid variants respondents to the survey. These well as the “standard” models with of the two products. In both cases banks fell into a number of cate- which they are likely to be more this appeared to be the result of a gories: familiar; and the need to ensure that much reduced consensus on the •Afew banks were infrequent valuations which are used for risk appropriate volatility input for the participants in the markets in control, management reporting and particular option. which some of the products were statutory accounting purposes are As a general observation it is traded, were not closely in touch both accurate and consistent with worth noting that the option on with those markets, and conse- the front office valuations of these quently had used volatilities products. which were out of line with those used by the majority of Liquidity The survey respondents. Among foreign exchange options, This raises issues about the extent product 2 was chosen to enable has improved of coverage of CAD model analysis of the impact, if any, the recognition for the banks in ques- liquidity of the underlying instru- supervisors’ tion (ie whether the coverage of ment had on valuation (one possible model recognition should be implication of lower liquidity being awareness of redefined to include only a sel- an increased difficulty in collecting ected number of liquid currency accurate inputs for option models). some of the pairs). The Swiss franc/Japanese yen was • One bank, throughout the range chosen as being significantly less more complex of survey products, calculated traded than sterling/D-mark, even its vega exposure using a though both the yen and the Swiss option products different method to other survey franc are well traded currencies in participants (and different to that their own right. requested in the survey specifica- Among the interest-rate prod- the illiquid bond (product 6(c)) tions). As a consequence, the ucts side, the specification of the displayed higher variation in value, bank’s reported vega exposures bond option products (product 6) delta and gamma than any of the were different to those of all also enabled some analysis of the other three bond options in the other participants by an order of impact of the liquidity of the under- survey. magnitude. lying instrument on the option price. The Bank of England’s follow- The underlying bond selected for Supervisory lessons up raised questions over the product 6(a) was the current “on the The findings of the survey are bank’s CAD capital calculations run” 10-year bond issued by the US expected to be of relevance both to for vega risk. Discussions on Treasury, whereas product 6(c) was the day-to-day supervision of indi- this subject with the bank in a much more illiquid 10-year bond. vidual banks and to broader policy. question are continuing. Consistent with the a priori • In another case, one bank’s assumption in both cases, banks Individual banks approach to the valuation of the appear to have had more difficulty The Bank of England has already bond option product (product 6) valuing the options with less questioned directly those banks was flawed in two respects: liquidity in the underlying asset. whose responses were identifiably the approach was excessively

26 VALUATION OF OPTIONS

simplistic (in that the bank had close to the average. For these banks, • an inability to answer follow-up used the same price volatility for the Bank of England drew some questions, usually due to a lack all four variants of the product); comfort in cases where the results of documentation, often com- and the volatility used was not turned out to be more consistent pounded by staff turnover. consistent with those used by than expected, although clearly the • an inability to replicate initial other respondents. Again in this results of the survey are not conclu- submissions, or to find details of case the terms of model recog- sive evidence of prior problems the input data used. nition for the bank in question having been fully corrected. Conse- These problems resulted in are being reviewed. quently the Bank of England has many of the respondents’values and • One bank’s initial valuation of continued to keep a close watching risk parameters having to be re- the sterling collar (product 7) brief on all banks which fall into estimated, or where follow-up reval- was clearly out of line with those this category. uations proved impossible, excluded provided by the other survey completely from the survey output. participants, although neither The Bank is in the process of the input data used, nor the risk reviewing the appropriateness of parameters resulting from the Technical and model recognition for banks where valuation, were similarly out of problems of this kind became line. The bank was not able practical issues apparent. to identify the reasons for the apparent mis-valuation, and emerged from Broader implications hence the result was allowed to A number of the technical and prac- stand. the results ... tical issues which emerged from the The Bank of England questioned results may have broader implica- the risk-management staff invol- particularly tions for future policy, particularly ved, and the reasons for the with respect to model recognition. problem have been identified. with respect One point which was rein- Again, the implications for the to model forced by the survey results is the bank’s model recognition are need for banks’ procedures to make being assessed. r ecognition allowance for uncertainties in the For every product there were a valuation of option products — for large number of consistent results example, arising from the use (or which fell within a relatively narrow Unfortunately, a few respon- non-use) of volatility smiles and range. In some individual cases, dents took a less professional and surfaces — in the same way as however, the Bank of England had thorough approach to the completion allowance is made for the spread prior information (for example, of the survey than the Bank of between bid and offer rates, or the arising from Section 39 reports, England expected. This manifested potential difficulty in closing out from Traded Markets Team visits, itself in a number of ways: positions in illiquid markets. or from specific losses in particular •allocation of the survey response Good practice would be to set option portfolios) which suggested to staff who had insufficient aside reserves explicitly to cover a priori that these banks had prob- knowledge of the products, the such uncertainties, and the Bank is lems with their pricing or risk valuation methods, and/or the taking the opportunity to discuss management models and conse- risk parameters associated with with banks both their reserving poli- quently their results might not be those products. cies and the level of their reserves

27 VALUATION OF OPTIONS

against model-related uncertainties resourced risk management teams visory staff more aware of some of of this kind. Another broad policy — said they found the results of the the practical issues associated with issue which might be influenced by survey extremely interesting and the valuation and risk management the findings of the Bank’s survey beneficial, and were very construc- of the more complex option prod- is the framework for including tive in both their approach to the ucts traded in the market. options risk within a Value at Risk- completion of the survey responses A number of technical areas type approach for the purposes and their comments on the results. worthy of further research have of calculating market-risk capital From the Bank’s perspective, been identified. requirements under the forthcoming the three specific objectives of the The Bank of England also Basle and CAD II market-risk survey were clearly achieved. A hopes that the results of the survey regimes. An issue for banks building number of issues arising from the will be of broad interest to the VaR models is their ability to survey are of direct relevance either market as a whole, and in particular capture the non-linearities and to particular institutions or to the that the institutions which partici- discontinuities associated with the development of future policy, and pated obtained useful information more common option products. a number of these are already being from the feedback they received on A final example is the issue incorporated into the Bank’s super- their individual positions compared concerning the measurement of vega visory work. Furthermore the with other responses gained in the exposure. The findings of the survey has made the Bank’s super- survey.■ survey reinforce the suggestion that where a bank’s capital requirement NOTES for gamma and vega exposure on 1 See Bank of England Quarterly Bulletin, November 1995, “The pricing of over-the-counter options”, exotic products is based on its pp 375-381. 2Model recognition under the CAD allows banks to determine their capital requirements for market models output, that capital require- risk using their own internal models, rather than a more conservative, standard methodology. ment might be more appropriately 3 This differs from the previous survey, where these data were also given for some products. 4 Black, F. and M. Scholes (1973), “The Pricing of Options and Corporate Liabilities”, Journal of calculated using full revaluation Political Economy, 81, May/June, pp 637-654. rather than using the instantaneous 5 Garman, M. and S Kohlhagen (1983), “Foreign Currency Option Values”, Journal of International Money and Finance, 2, December, pp 231-237. Greek sensitivities derived using a 6Black, F. (1976), “The Pricing of Commodity Contracts”, Journal of Financial Economics, 3, March, partial differential equation. Again pp 167-179. 7Black, F., Derman, E. and W. Toy (1990), “A One Factor Model of Interest Rates and its Application the Bank is considering the policy to Treasury Bond Options”, Financial Analysts Journal, January-February, pp 33-39. implications of these findings 8 Hull, J. and A. White (1987), “The Pricing of Options on Assets with Stochastic Volatilities”, Journal of Finance, 42, June, pp 281-300. 9 Cox, J. and S. Ross (1976), “The Valuation of Options for Alternative Stochastic Processes”, Journal Conclusions of Financial Economics, 3, pp 3-54. The Bank received mixed feedback 10 Cox, J., Ross, S. and M.Rubinstein (1979), “Option Pricing: A Simplified Approach”, Journal of Financial Economics, 7, October, pp 229-263. from survey participants as to the 11 Standard deviation is a statistical measure of the extent to which a set of numbers varies from its costs and benefits of their participa- mean, expressed in a standardised form. It is defined as the square root of variance. As an example, the population of responses for product 1(a) value showed a mean $1.34m, variance $1.32bn and tion in it. Some banks — notably standard deviation $36,000 (2.7% of mean). smaller banks and those with less 12 It is also worth noting that a failure to value a product accurately can also distort the sensitivity measures for that product and hence make it difficult to hedge a position accurately. resources devoted to their risk 13 So far as industry standard software is concerned, these problems are now being addressed by the management and risk control func- software suppliers, who are now developing improved valuation models for products of this kind. Whether this will reduce the variability in values observed in this survey remains to be seen. tions — found it time-consuming 14 The volatility surface usually refers to a three dimensional graph of the implied volatility of an option: and burdensome. Others — particu- in foreign exchange options the three-dimensions are usually volatility, time to option maturity and strike; in interest rate options the three dimensions are usually volatility, time to option maturity larly the highly professional market and time to maturity of the underlying instrument. participants and those with better-

28 THE GENESIS OF REGULATION be authorised to accept deposits or undertake banking business in the UK. There were disclosure require- By Victoria Robb, Bank of England ments (Protection of Depositors Act On 20 May 1997 the Chancellor announced proposals for a fundamental 1963) on those institutions that adver- change in the structure of financial services regulation in the UK, with the tised for deposits, including the creation of one large regulator spanning most (if not the whole) of the obligation to include in their accounts industry. In essence, the regulatory map is to be completely re-written. prescribed information, which were This article looks at the key changes over the past which have formed the examined by the Board of Trade. map as it now stands and at the proposals for the new arrangements. However, banks and discount houses were exempt from these As in other areas of the financial requirements. services, a number of factors have The Bank of England Act 1946 influenced the evolution of banking gave the Bank powers to “request supervision in the UK. These factors information from and make recom- have been both industry-specific mendations to bankers” and, if and general and have included authorised by the Treasury, to “issue economic developments, advances directions to any banker for the in technology and communications, purpose of securing that effect is and globalisation. There have also given to any such request or recom- been moves to increase the protec- mendation”. No such directions tion of depositors and investors. were ever issued. The supervisory regime has The Bank’s main concern was also been altered after collapses of, with those institutions with which it and problems with, major and had a counterparty relationship in its minor institutions. These develop- dealings in the bill market — ie the ments have been followed by accepting houses (the merchant increased formalisation of the Bank banks whose acceptances the Bank of England’s supervisory role. was prepared to discount) and the International developments have discount houses (through which the also played an important part in Bank provided liquidity to the UK shaping the regulatory regime for banking system as a whole). These banking, in particular the UK’s institutions were required to attend membership of the European regular meetings with the Bank: Community and the harmonisation supervision was undertaken by a of the regulatory approach that small number of staff in the Bank’s accompanied this. (The First Banking discount office. Co-ordination Directive was intro- The 1960s was a period of duced in 1977.) major structural change in the UK Before the introduction of the banking system. The favourable first Banking Act in 1979 there was economic conditions of the 1950s no statutory requirement that a bank and early 1960s, together with the or similar deposit-taking institution expansion of the consumer credit

29 GENESIS OF REGULATION

industry, encouraged the emergence had mismatched the maturities of own funds and funds from the of a secondary banking sector. Until their assets and liabilities. clearing banks. Competition and Credit Control The first signs of problems in Lessons drawn from the crisis was introduced in 1971, these insti- the financial system emerged in included the need for greater super- tutions were outside the credit November 1973 when London and vision, particularly of the secondary control regime which was applied to County Securities found itself in banks; a tighter definition of what the banking system. liquidity difficulties, being unable to constituted a “bank”; and greater A significant development at renew deposits taken through the deposit protection. It was also this time was the rapid growth of money markets. Interest rates had recognised that the system of new and important markets — in reached double figures to combat multiple bank recognitions was a particular the wholesale money the rapid growth in the money source of confusion in the market- market, the inter-bank market and supply between 1971 and 1973, place and led to gaps in supervision. the CD (certificates of deposit) which put pressure on the fringe During 1974, the Bank identi- markets in sterling and dollars. banks’ profit margins. fied most of the large companies Fringe financial institutions Soon a number of these banks registered in the UK which held were increasingly able to attract were in difficulties. The Bank judged sizeable deposits from the public. large amounts of wholesale deposits that, for the maintenance of systemic These companies were invited to (mainly at short maturities) and stability, a group of fringe (secon- submit themselves to voluntary expand their balance sheets, without dary) banks that had liquidity (as supervision by the Bank. By 1975 having established customer rela- opposed to solvency) difficulties all banks and other deposit-taking tionships. Risks arose, both interest should be rescued. Hence the launch companies of significance had rate and funding, where these banks of the “Lifeboat” using the Bank’s agreed to be subject to prudential supervision by the Bank. Box 1 STATUS OF A BANKING INSTITUTION (PRE-1979) A new department in the Bank The status of a banking institution was based in practice on what became known as was set up in 1974 with more staff. “ladder of recognitions”, depending on its position under various statutes. In 1975, the government announced 1 Authorised status under the Exchange Control Act 1947 — the Bank and the that the enhanced system of super- Treasury were responsible for establishing a list of banks authorised to deal in vision that the Bank of England had foreign exchange. This was regarded as the highest accolade a bank could attain. put in place would be confirmed in 2 Companies Act 1948 exemption from disclosure of certain information in statute. AWhite Paper was published company accounts. in 1976. It acknowledged the need 3 Exemptions from the Protection of Depositors Act 1963 restrictions on adver- to preserve the Bank’s existing tising for deposits. approach to supervising the primary 4 Companies Act 1967, section 123 certificate — issued by the Board of Trade to banks — an approach which could institutions it considered to be bona fide carrying on the business of banking. This be adapted to the changing circum- certificate secured exemption from the Moneylenders Acts. stances of the banks and the markets. 5 Recognition under Income and Corporation Taxes Act 1970 — granted by the Several other factors also led to Inland Revenue to bona fide banking businesses allowing them to pay and receive a review of the scope of banking interest gross of tax. supervision. The UK entered then Some of the “lesser” recognitions only required decisions about the type of business European Economic Community in carried on (eg section 123 certificates) rather than fitness and properness and sound- 1973, and played a leading role in ness of the institution. the discussions on harmonising banking regulation. These resulted

30 GENESIS OF REGULATION

in the First Banking Co-ordination This found that JMB’s status as Directive (1977) which introduced a “recognised bank” had been a the requirement that there should factor in the supervisors’ delay in be a prior authorisation procedure becoming aware of, and reacting for credit institutions, along with to, its growing problems. It recom- minimum criteria for the authorisa- mended that the two-tier system of tion of a credit institution. authorisation be replaced by a single The number of overseas banks authorisation to take deposits, and that opened subsidiaries or branches that all the powers given to the Bank in the UK rose sharply during the should apply to every authorised 1970s; and with the development institutions. of the wholesale market, banks had A White Paper on Banking become far more interdependent. In Supervision, released in December response to these changes and the 1985, proposed new legislation need to implement the EC First which would provide a Board of Banking Co-ordination Directive, Banking Supervision to assist the the UK introduced the first Banking Governor in his banking supervision Act. Robin Leigh-Pemberton (now responsibilities. It would replace the Lord Kingsdown). Recommended The 1979 Banking Act did not that the two-tier system of two-tier system of regulation and alter fundamentally the previous authorisation be replaced encourage increased co-operation supervisory arrangements, but it did between supervisors and auditors. introduce a system of legal powers It is worth noting that the paper and sanctions available to the Bank reported the government had consid- to underpin its supervision of the Johnson Matthey Bankers (JMB) in ered establishing a specialist banking banking sector. It also introduced a 1984 — precipitated by two large inspectorate, but concluded that it two-tier system of authorisation exposures that went bad, although would be more effective to strengthen which distinguished recognised the underlying problem was one of the Bank’s existing statutory powers banks from licensed institutions: poor systems and controls — led to and resources. this was based on track-record, another review of the regulatory The paper also set out thoughts standing and range of activities. regime for banks. on the role and aims of the banking Recognised banks were not subject Johnson Matthey Bankers was supervisor. The supervisor was “to to such a wide range of statutory authorised under the 1979 Banking ensure that the bank is managed in requirements as licensed deposit- Act as a “recognised bank”. Its such a way as not to put at undue takers. The first depositors’protection collapse highlighted the problems risk the interest of depositors, with scheme was also introduced. of making a distinction between that institution or more generally”. By the early 1980s banking recognised banks and licensed The result was a new Banking business was no longer just about deposit-takers. In December 1984, Act. The 1987 Banking Act did not accepting deposits and making loan the Chancellor of the Exchequer alter the general approach to regu- advances. Market and technical and the Governor of the Bank of lation which had been set out in the innovations were leading banks to England agreed that a committee 1979 Act (apart from removing the use more complex instruments to should be established, the Leigh- two tiers of authorisation). The manage their balance sheets. These Pemberton Committee, to consider criteria for authorisation were set developments, and the collapse of the system for supervising banks. out in general terms. These allowed

31 GENESIS OF REGULATION

the Bank flexibility in determining was, in fact, a watershed, with a members chose not to follow the what was appropriate for each insti- move towards a much a sharper advice of the independent members. tution. The Bank’s powers to focus on the adequacy of banks’ Since the 1987 Act, the Bank’s investigate and to seek information systems and controls. approach to supervision has been were enhanced substantially, as One of the main purposes of further refined as a result of the were the notification requirements establishing a Board of Banking high-profile closure of the Bank of applicable to authorised institutions. Supervision was to bring outside Credit and Commerce International The Bank’s supervision was expertise to bear on the Bank’s (BCCI) in 1991, and more recently, intensified. It continued to be based supervisory decisions. The inde- the problems at Barings in 1995. on regular discussions with manage- pendent members of the Board (a These cases reflected further changes ment on the information contained majority) had a duty to give advice taking place to the form and structure in the prudential forms, but the Bank to the ex-officio members on super- of banking business. also sent out teams to review visory matters arising out of the Following the “Big Bang” in authorised institutions’ systems and exercise by the Bank of its functions 1986, most of the leading Stock controls and obtain more detailed under the Banking Act. The Board Exchange member firms were information on their business. was required to prepare an annual bought by UK merchant or clearing The Bank could, under section report, to be included in the report banks, or by overseas commercial 39 of the Act, require an institution made by the Bank to the Chancellor or investment banks. This started to commission a report by accoun- on its activities under the Banking Act the trend towards the formation of tants on its systems and controls and each financial year. The Chancellor financial conglomerates. Increas- on areas of business of particular also had to be informed of circum- ingly UK banks sought to have an concern to the Bank. The 1987 Act stances in which the ex-officio international presence to compete

Box 2 KEY DEVELOPMENTS IN BANKING SUPERVISION

Pre-1979 Banking Act: 1979 Banking Act: 1987 Banking Act: Legislative Informal regime. Bank had some Introduced prior authorisation A Board of Banking Framework powers under the Bank of England procedure and minimum criteria for Supervision was established Act 1946 to request information authorisation. Also two-tier system and the two-tier system of and make recommendations to of authorisation; recognised banks authorisation abolished. bankers. and licensed institutions. Bank’s information and investigation powers were strengthened. Key Industry Emergence of a secondary banking Banking business becoming more Increasing globalisation of Developments sector and the fringe banks crisis of complex as a result of market and financial institutions and the 1973/4. Those banks holding technical innovation. Collapse of formation of financial sizeable deposits from the public JMB in 1984 led to review of conglomerates. Closure of invited to submit themselves to supervisory regime. Government BCCI (international bank voluntary supervision by the Bank. White Paper published in 1985. with complex structure) in 1991 and collapse of Barings in 1995 triggered further reviews of supervision.

32 GENESIS OF REGULATION

with US and Japanese banking the gateways for sharing confiden- firms. The development of the tial supervisory information and derivatives markets provided the required auditors to disclose to super- banks with new ways to manage visors any material concerns about a their risks. bank that affected its continuing authorisation. BCCI and new challenges The collapse of Barings, The international nature of BCCI’s followed large losses on futures and business, the opacity of its group options trading by its Singapore structure and the history of fraud subsidiary. This led to a review by and false accounting that was the Board of Banking Supervision, uncovered, raised new supervisory at the instigation of the Chancellor challenges both domestically and of the Exchequer in March 1995, into internationally. the way in which the Bank super- Big Bang After BCCI had been closed in vised banking groups, particularly 1991, Sir Thomas Bingham was those which contained significant started the appointed by the Chancellor of the non-banking businesses. trend towards Exchequer and the Governor of the The Board recommended that Bank of England to inquire into the the Bank should seek a better under- the formation supervision of BCCI under the standing of the risks, and their Banking Act. He concluded that management and control, in any of financial whilst there was no need for a group which contained an autho- radical change in the basic system rised bank. It also recommended that conglomerates: of supervision, the Bank should there should be an independent be more alert to suspect banks quality assurance review of the in the UK these and evidence of fraud, and that Bank’s supervision of banks. banking structures which could In October 1995, the Bank were dominated potentially stop supervisors from appointed Arthur Andersen and Co having a clear understanding of an to look at the effectiveness of its by banking institution’s business should not be supervision activities within the institutions permitted. current legislative framework. Arthur In response, the Bank set up a Andersen reported in July 1996 and new Special Investigations Unit to the Bank decided to implement all pursue any indications it received of the proposals in its report. A key fraud or criminal activity. In the EU proposal was the adoption of a risk- the problems raised by BCCI were assessment process (RATE) which addressed by the Post-BCCI Direct- would focus supervisory effort on ive (1995). This required supervisors higher-risk firms. to refuse applications from, or to From its inception, banking revoke authorisations of, those insti- supervision focused on the pruden- tutions whose structures or close tial supervision of banks and did not links with others prevented effective encompass the conduct of banking supervision. The directive widened business. The regulation of financial

33 GENESIS OF REGULATION

services, on the other hand, is more tions and improper sales techniques, complex in nature, as it encom- or exploiting positions of trust — passes prudential requirements for were prevalent at the time. The non-bank and non-insurance invest- second committee, headed by Sir ment businesses, and supervisors Alan Anderson, examined fixed Share pushing typically focused more on business trusts (unit trusts as they are known conduct. today). and share Before the Financial Services The key provision in the 1939 Act was introduced in 1986 there Act was the introduction of licensing hawking — was no legislation in the UK which requirements, which covered indi- regulated comprehensively all areas viduals and firms that dealt in advertising of investment business. The legisla- securities. Securities included shares, tion that did exist had been enacted debentures and units in a unit trust investments for piecemeal to deal with particular scheme. However, large parts of the investment activities, or types of industry which were members of sale using firm that were seen to warrant regu- bodies such as the Stock Exchange, unrealistic lation. were exempt, which created a largely Financial services developed in self-regulatory regime. claims, trying to the UK in parallel with the City of The Board of Trade (which London. Those engaged in the issued licences) was also able to sell investments financial services business made up declare persons and firms exempt a small homogeneous community from the requirement to hold a to individuals until the Big Bang occurred in 1986 licence, so long as securities dealing (although the Eurobond market was did not form the main business using false by then an internationally important activity of that institution and a market for London and based in the greater part of their securities activ- r epresentations the banks). Regulation was largely ities were conducted through the in response to a number of scandals agency of members of the Stock and improper and abuses. Exchange. sales techniques, The first major piece of legisla- The Board of Trade would grant tion to try to regulate investment exemptions only to those institu- or exploiting activity was the Prevention of Fraud tions that were deemed to be of (Investments) Act 1939. It was sound financial and business repu- positions of enacted as a result of the reports of two tation. Thus merchant and clearing government-appointed committees. banks, insurance companies and trust — were The first committee, headed by investment trusts were the main Sir Archibald Bodwin, looked at institutions to gain exempt status. prevalent “operations known as share pushing The 1939 Act also gave the Board and share hawking”. These prac- of Trade powers to issue conduct of tices — advertising investments for business rules. The Prevention of sale or subscription using unrealistic Fraud (Investments) Act 1958 was claims, trying to sell investments to virtually a re-enactment of the 1939 individuals using false representa- Act — the only addition was the

34 GENESIS OF REGULATION

Box 3 THE EVOLUTION OF THE INSTITUTIONAL STRUCTURE FOR FINANCIAL SERVICES REGULATION — GOWER PROPOSALS AND BEYOND

Government/self-standing body Self-regulating organisations Gower Department of Trade responsible for policy Four recognised self-regulatory agencies: Discussion and overall supervision under a new Secu- The Public Issues and Take-over Agency responsible for the Document rities Act. Also responsible for the scrutiny of all prospectuses, including primary and secondary 1982 recognition and supervision of self-regula- issues and takeovers. tory agencies. The Stock Exchange which would continue to be the recognised manager of the activities of member firms, except insurance broking and unit trust and investment management. An agency/association with responsibility for over-the-counter markets, dealings off The Stock Exchange and investment management and advice. Unit Trust Agency responsible for all unit trusts and mutual funds. Gower The governmental role could be left to the Based as far as possible on existing associations. Those which Report DTI. If this would involve a substantial might be expected to qualify for recognition included The Part I volume of day-to-day supervision, a self- Stock Exchange, National Association of Security Dealers and 1984 standing Commission answerable to the Investment Managers, a Unit Trust Agency, the Association of Secretary of State should be established. Futures Dealers and Brokers, Lloyds and the certifying body for life assurance intermediaries (being explored by the Institute of Insurance Consultants and the Life Insurance Association). A number of other professional associations (lawyers, invest- ment analysts, pension fund consultants, accountants, actuaries) would prefer members to register through their membership. Government Legislation to confer regulatory powers on Creation of self-regulating organisations which were primarily White the Secretary of State which he could dele- function-based. Paper gate to a Body(ies) that recognise SROs and 1985 have authority over their rules and practices. Proposed two bodies: the Securities and Investments Board, (regulation of securities and investments), and the Marketing of Investments Board, (marketing of pre-pack- aged investments). Financial One statutory Body, (Securities and Invest- Self-Regulating Organisations recognised by the SIB (five Services ments Board) to which the Secretary of State initially). Also Recognised Professional Bodies (such as Act for Trade could delegate powers under the accountants and lawyers) Recognised Investment Exchanges 1986 Act. Responsible for oversight of the (such as the Stock Exchange) and Recognised Clearing Houses. Recognised Bodies and the FSA regime more generally.

35 GENESIS OF REGULATION

power given to the Board of Trade to banks and insurance companies. In 1977 a committee headed by to appoint an inspector to examine In the event, the proposed changes Harold Wilson (later Sir Harold) the administration of a unit trust were delayed because of a change in was appointed to enquire into the scheme. government. role and functioning of financial In the 1960s and 1970s there In the 1970s, two new supervi- institutions and consider what , if was a large expansion in the number sory-type bodies were established: any, changes were required in the of participants in the financial the Joint Review Body (in 1977), arrangements to supervise these services market, of which only a which was responsible for the general institutions. small number were subject to the oversight of all aspects of securities The committee’s report, published Prevention of Fraud (Investments) market supervision and was made in 1981, observed that at that time Act 1958. up from senior officials of the Bank there was no single authority with In 1974 the Department of Trade of England and the Board of Trade; clearly defined responsibility for the and Industry (DTI) conducted a review and the Council for the Securities overall regulation of the financial of the adequacy of regulation of the Industry which was responsible for system, statutory and non-statutory. securities market. After this review the non-statutory aspects of super- It proposed that a single review the government proposed to amend vision of the securities markets not body should keep the regulation of the Act to tighten up the system for covered by the Stock Exchange. (Its all parts of the financial system granting licences, to reduce the number chairman and deputy chairman were under regular review. This body of exemptions and to confine them appointed by the Bank of England.) would not be involved in day-to-day

Box 4 THE UK REGULATORY STRUCTURE

Government HM Treasury Department

Supervisor Bank of England Securities and Investments Board (SIB) Building Societies Commission

Self-Regulating Recognised Investment Recognised Building Organisations Exchanges “RIEs” & Professional Societies SFA IMRO PIA Clearing Houses “RCHs” Bodies

Money Market Securities Fund Financial Institution Banks* Institutions Firms Managers Advisers

Banking Act Financial Services Building Societies Legislation 1987 Act 1986 Act 1996

* Also regulated under the Financial Services Act for their investment business activities.

36 GENESIS OF REGULATION

regulation (this responsibility would City specialist firms began to remain with existing bodies) nor in merge forming integrated finan- the formation of legislation, which cial services businesses. The would remain the responsibility anticipation that the Stock of government departments and Exchange would eventually allow Parliament. dual capacity trading acceler- During the late 1970s and early ated this process. 1980s pressure for change in the •Technology was transforming the regulatory system grew. There were way business was conducted: several reasons: market information could be • The securities industry had transmitted world-wide round the become international and the clock. lifting of exchange control in • The number of people directly 1979 had made it easier for UK investing in shares fell substan- securities to be traded abroad tially, in favour of investments and for foreign shares and funds such as unit and investment to be traded in London. trusts. The business of managing • In response to competition from and advising on investments

overseas counterparts, traditional grew in importance. Self-regulation ... involves • The development of new prod- something more than trusting ucts such as commodity futures people to behave themselves. Professor Jim Gower. Review of funds led to gaps in coverage, as Investor Protection, A Discussion Department of Trade & Industry such funds were not classed as Document (HMSO January 1982) securities under the Prevention of Fraud (Investments) Act. Early resistance by the industry to a move away from the largely Early resistance self-regulatory system was diluted by several scandals. Two of the by the industry Insurance more prominent were the collapse Companies of Norton Warburg in February to a move away 1981 and the investigation into the behaviour of senior executives at from the largely Halliday Simpson between 1978 and 1983. self-regulatory Norton Warburg, an investment management firm, misused clients’ system was money and its failure led to losses of Insurance large sums by individual investors. diluted by Companies Act 1982 Halliday Simpson, a regional stock- several scandals broker, abused its position as broker by buying shares for resale to clients for its own profit.

37 GENESIS OF REGULATION

These cases reinforced concern for the securities and investment that the current regime for investor industry. protection was inadequate. Of specific concern were the The government realised that marketing methods (in particular to achieve its goal of increased “cold-calling”) used to sell insurance public participation in the market, products that were, to all intents and public confidence in the security of purposes, investments. investments was vital, and this The report proposed that “Invest- could be achieved only by a visible ments” and “Investment Business” improvement in the regulatory should be defined widely, and should Regulation must system. include contracts for commodity, In July 1981 Professor L C B financial futures and options and life stimulate (Jim) Gower was commissioned by policies, that had not fallen under the the Secretary of State for Trade to scope of the Private Finance Initiative competition and review investor protection. His Act. It also suggested that a system innovation ... review was to consider the statutory be put in place for “recognising” protection now required by private investment exchanges (in addition the system and business investors in securities to self-regulating organisations). and other property, including invest- The main force of the report must inspire ors in unit trusts and investment was directed at improving protec- trusts; the need for statutory control tion for the retail investor, with a confidence in of dealers in securities, investment carve-out for professional investors consultants and investment mana- proposed for some of the provisions issuers and gers; and to advise on the need for of a new regulatory regime. new legislation. The government then consulted investors ... His initial output was a discus- interested parties. In May 1984, the sion document which considered Governor of the Bank of England and have the the shortcomings of the existing appointed an advisory group of r esilience not system and put forward proposals senior City figures under Sir Martin for comment on a possible new Jacomb to advise on a structure of to be over-run regime (see Box 4 The UK Regula- self-regulatory groupings that would tory Structure). He believed that most appropriately cover all types of by events regulation would operate best within securities activity and would gain a statutory framework, with day- sufficient support from practitioners to-day regulation undertaken by to be capable of implementation as self-regulatory agencies and the early as possible. government exercising a residual, Sir Martin Jacomb’s advisory supervisory role. group recommended concentrating This was followed in January on an institutional structure of regu- 1984 by first part of his report, lation, under a practitioner-based “Review of Investor Protection”. system. This set out recommendations for a Shortly after the appointment of comprehensive regulatory regime Sir Martin’s group, the Parliamentary

38 GENESIS OF REGULATION

Under Secretary of State for Nearly all firms carrying out Corporate and Consumer Affairs investment business in the UK had invited a group, chaired by Marshall to seek membership of an SRO and Field (the chairman of the Life abide by its rules, or be directly Offices Association) to advise on regulated by the SIB. An exception the prospects for practitioner-based was professional firms such as regulation of the marketing of life solicitors where investment busi- assurance and unit trusts. ness was ancillary to their main The recommendations of these activities: they could be supervised advisory groups were important in by their professional association. shaping the government’s proposals The Bank of for legislation detailed in its White Criticism Paper published in January 1985. It The Financial Services Act regime England Bill stated the government’s objectives has always been the subject of as being: efficiency, competitive- debate and criticism. In the early provided a ness, confidence and flexibility. years criticism concentrated on the window of “Regulation must stimulate compe- legalistic nature of the regime and tition and innovation ... the system the plethora of rules that firms had opportunity must inspire confidence in issuers to follow. The second chairman of and investors ... must be clear the SIB rationalised the rules by to transfer enough to guide but not cramp issuing 10 statements of principle structural and other change ... and backed up by a set of “core rules” banking have the resilience not to be over- established by the SIB, leaving the run by events.” SROs to produce their own tailored supervision Market forces, facilitated by set of detailed requirements suited adequate disclosure of information to their particular membership. to a new and to customers and a competitive More recently, the Maxwell industry, were seen as providing the affair and the mis-selling of home strengthened best means for the industry to meet income plans and personal pensions, Securities and the needs of its customers. led to a fundamental rethink of the The Financial Services Bill was self-regulatory system under the Act. Investments published in December 1985 and Following the Maxwell affair, it concluded its progress through the new chairman of the SIB, Sir Board Parliament in 1986. It provided, for Andrew Large, was asked by the the first time, a comprehensive Chancellor of the Exchequer to regulatory framework for invest- review the way the SIB carried out ment business carried out in the UK. its regulatory responsibilities under It retained practitioner involvement the Act, with particular reference to using self-regulating organisations its oversight of the recognised (SROs), but with oversight by a bodies (the recognised self-regu- statutory body, the Securities and lating organisations, professional Investments Board (SIB) giving a bodies and investment exchanges) two-tier structure. for which it was responsible.

39 GENESIS OF REGULATION

His report, “Financial Services sured, prompted a reorganisation Regulation: Making the Two-Tier at the DTI and an increase in the System Work”, was published in number of staff directly involved in May 1993. He concluded that “SIB supervision. Further powers were and the regulators need to improve also given to the DTI in the Insurance the system’s effectiveness and Companies Amendment Act 1973. confidence in it, by a more active In 1974 a new Insurance Companies and formal exercise of SIB’s lead- Act came into force which set down ership and by achieving greater a stricter set of criteria for insurers transparency for the system”. to satisfy before they were granted However, he warned “without authorisation by the Department of the changes in attitude and approach Trade and also consolidated previous ... recommended in this report [I] do legislation. not feel confident that the two-tier Pressure increased to introduce system will, in fact, deliver the new legislation to protect policy- necessary results ... This could lead holders following the failure of to a cry for even more radical Nation Life in 1974. The govern- change, including perhaps for a ment responded by introducing a fully statutory unitary system”. compensation scheme for policy- holders through the Policyholders Insurance companies Protection Act 1975. The legislation The first key legislation affecting currently governing insurance compa- insurance companies was the Life nies is the Insurance Companies Act Assurance Companies Act 1870 1982. which was introduced in response to While the DTI is responsible for a number of insurance companies supervising UK insurance compa- defaulting, culminating in the collapse nies, it relies on the Government of the Albert Life Assurance Company. Actuary’s Department for the tech- It imposed a deposit requirement of nical, actuarial review of the annual £20,000 for insurers engaged in life returns of long-term companies. business. The collapse of Fire Auto Like banking supervision, insur- and Marine in 1966 left many poli- ance supervision is focused on the cyholders unprotected. To rectify prudential supervision of compa- this sort of problem, the Companies nies, not business conduct. Sales Act 1967 introduced a system of practices for life assurance products prior authorisation of insurance were eventually covered by the companies: this included assessing Financial Services Act. whether the companies’officers and controllers were fit and proper and Building societies Sir Andrew Large: “Without the it gave the DTI intervention powers. changes in attitude and approach ... In contrast to the history of regula- The collapse of the Vehicle and recommended in this report [I] do tion of banking, financial services not feel confident that the two-tier General in 1971, which left nearly system will, in fact, deliver the and insurance, successive changes one million policyholders unin- necessary results ...” to the regulatory regime for building

40 GENESIS OF REGULATION

societies have been aimed at progres- in principle for bringing the regula- sively relaxing what was a relatively The regulators tion of banking, insurance and prescriptive regime which governed have embarked securities under one roof. societies since the first significant The Bank of England Bill, which building societies legislation of on an intensive gives the Bank operational indepen- 1874 (in effect building societies dence for monetary policy, provided were limited to offering mortgages work programme the opportunity to transfer banking funded by savings accounts). supervision to a new and strength- The 1986 Building Societies to ensure that ened SIB. The changes to the Financial Act and subsequent amendments in Services Act that are necessary to 1988, enabled building societies, for the new give the SIB direct responsibility for the first time, to offer such products the regulatory regime under the Act as current accounts and life assur- arrangements (by merging all of the recognised ance in competition with the banks. bodies with the SIB) would come The most recent legislation, the will operate as into force in the next session of Building Societies Act 1996, removed smoothly and Parliament. The Bank will remain most of the remaining restrictions on responsible for the overall stability of building societies’ activities. effectively as the financial system. The DTI subsequently annou- The future possible nced (on 23 July) that responsibility The Labour Party made clear before for the regulation and supervision of the general election that it intended insurance companies would transfer to reform City regulation. The to the Treasury. The final changes current structure, it said, was both necessary to create a single, enhanced costly and bureaucratic, with too financial services regulatory body many layers. were announced on the same day by It proposed a move from self- the Treasury: insurance supervision, regulation to direct regulation by together with all the functions merging the self-regulating organ- currently carried out by the Building isations under the Financial Services Societies Commission, the Friendly Act — the PIA, SFA and IMRO with Societies Commission and the Central the SIB. The aim would be to reduce Office of the Registry of Friendly the layers of regulation, clarify Societies, would be brought under responsibility, end tensions between the remit of the new regulatory regulators and enable savings by body. The provisions would be reducing duplication and overheads. included in the Financial Regulatory On 20 May the Chancellor of Reform Bill, to be published for the Exchequer announced reforms consultation in summer 1998. that were more radical than had The regulators have embarked been expected. He argued that as on an intensive work programme to distinctions between financial insti- ensure that the new arrangements tutions were becoming increasingly will operate as smoothly and effec- blurred, so there was a strong case tively as possible.■

41 THE PRE-COMMITMENT APPROACH TO derivatives that are held for short- term trading purposes. In 1993, the Basle Supervisory Committee3 SETTING CAPITAL REQUIREMENTS (Basle) had proposed a similar method to setting risk-based require- By Patricia Jackson and Simone Varotto, Bank of England ments for trading books, known as Arupratan Daripa, Birkbeck College the Standardised Approach. However, Basle has now agreed to offer an This article looks at the Pre-Commitment Approach (PCA) to setting alternative regime with capital capital requirements for the securities, foreign exchange and commodi- requirements based on the internal ties trading books of banks, which has been developed by economists in VaR models of the firms. the US Federal Reserve Board (Kupiec and O’Brien1) and put forward as The Basle VaR method for a possible alternative to the Basle Standard and Value at Risk (VaR) setting capital requirements for methodologies. Under the PCA, banks themselves would determine how trading books was developed largely much capital they needed to back their trading book and would be in response to a concern on the part penalised if losses exceeded this level. of large diversified players that they would, under the Standardised We look at how the approach would Approach, be carrying dispropor- work, the reasons for its develop- tionally more capital than specialist ment and compare the approach to players in relation to the actual the Value at Risk methodology risk of loss. The Basle Standard under which the banks’VaR models Approach (and CAD1) delivers are used to set capital requirements, capital which covers about 99 per subject to parameters set by the cent of price moves over a two week regulator. Drawing on the analysis period for each market (eg UK in a Bank of England discussion interest rate risk). But if a firm has paper by Daripa and Varotto,2 this risks in a number of markets (say work focuses, in particular, on US interest rate risk, UK equity risk, whether the incentive effects regarding US equity risk) all the separate risk-taking would be different under capital figures for each type of risk the pre-commitment as opposed to are simply added together to give the Value at Risk methodology for the overall requirement, delivering setting capital requirements. what can be an excessively large cushion of capital relative to the VaR and Basle Standard risks being run because the benefits In January 1996, the European of diversification across markets are Union Capital Adequacy Directive not taken into account. (CAD1) laid down rules setting It is almost impossible to take risk-based capital requirements for diversification across different types the trading books of banks and of risk properly into account using securities houses. The trading book rule-based capital requirements as consists of securities, foreign exch- this would necessitate taking into ange, commodities positions and account all of the cross correlations

42 CAPITAL REQUIREMENTS

in returns on the different expo- would lead to an increase in the sures. The VaR models, which some capital requirement based on the banks had developed to give senior output of the model. This would be management a consistent measure achieved by increasing the over- of overall portfolio risk, provided a riding multiplier which is used to means of achieving this. The convert the output from the models models calculate the capital neces- into a capital requirement. sary to cover losses that, given the actual portfolio composition, could Possible drawbacks occur with a pre-determined frequency Although the adoption of VaR, as over a set time horizon. They take a methodology for setting capital into account past correlation between requirements for trading books, the returns on the various assets in enables the diversification element to the portfolio as well as the volatility be captured in a flexible and rational of the individual returns. Each day way, some concerns have been The Bank for International the firms would run their VaR expressed about the approach. One Settlements, Basle, Switzerland. In 1993, the Basle Supervisory model and compute the expected of these concerns is that the setting Committee had proposed loss on their current portfolio. This, of the parameters in the models by a similar approach to setting risk-based requirements for trading with additional safeguards set by the regulators is too intrusive — it books, known as the Standardised the regulators, would form the basis would be better for the firms to be Approach. for the capital requirement over the permitted to decide on an appro- However, Basle has now agreed to offer an alternative 4 following 24 hours . priate model given their activities. regime with capital Basle proposes that the confi- For example, the fixed parame- requirements based on the internal VaR models dence interval for the model should ters set by the requirements, such as of the firms. be set at 99 per cent (ie it would the 10 day holding period, make no calculate the loss which would not allowance for the relative sophisti- be exceeded on more than 1 per cent cation and placing power of the of occasions) and that an ex-post firms which would affect the likeli- model validation — backtesting — hood of losses. One firm might should be carried out to try to check reasonably expect to trade out of a whether the models do in fact position in a few hours: for another such rare occurrences to allow for deliver this. it might be a matter of days. sensible forecasts. Basle lays down a number of Whether such differences would By using Monte Carlo simu- other elements which help to make also apply in stress conditions is, lations, Kupiec (1995)5 shows that the result conservative. Losses are however, more open to debate. in some cases, backtesting would estimated for a 10 day holding period The accuracy of VaR models be subject to substantial errors and an overriding multiplier of 3 is depends on the correct estimation of even in samples as large as 10 used to convert the VaR amount into the expected frequency of rare years’ of daily data. Although, a capital requirement, which in events, that is, losses in the portfolio Jackson et al (1997)6, in an empir- effect delivers a confidence interval exceeding a given value. The problem ical analysis based on real trading far in excess of 99 per cent. Too is that, even though the observation book data, show that backtesting many exceptions (ie dates on which period might be very long, there might actually be effective in the model under-forecast losses) would not be a sufficient number of identifying inaccurate models.

43 CAPITAL REQUIREMENTS

Box 1 PCA AND VAR WITHOUT AGENCY PROBLEMS

Obviously, even under PCA, banks should run a VaR-like As a consequence, under PCA, if actual cumulative losses, at model to forecast the future behaviour of their portfolio returns. any given time in the observation period, go beyond k, a This would help in estimating the probability of breaching the penalty f will be charged. Instead, under VaR, the portfolio pre-committed capital and ultimately incurring a penalty. volatility will be upper bounded or equivalently the probability So, even though the regulator exercised ex-post control, the of an overall increase of future losses above k will not be problem that the bank would face, when optimising its port- allowed to top p (p is set by the regulator). folio choice, reduces to an ex-ante forecast of the pattern of Indeed, it is indifferent if the regulator sets p or σ because, future trading book yield. given µ, there is a one-to-one relationship between the two. Intuitively, under PCA and in the absence of agency problems, σ: VaR volatility upper bound supervisors would only need to set an appropriate penalty schedule in order to obtain the same outcome as would be produced by a VaR regulatory framework. σ To describe this point, let us introduce some simplifying hypotheses. Suppose that, by using historical data, it is possible to forecast the future distribution of portfolio returns correctly. This means that, even though uncertainty still remains, a port- - ∝ folio’s risk and expected yield can be correctly measured in -k o net returns advance. It is further assumed, for illustrative purposes only, When a bank selects its trading book it can choose among a that the distribution function is normal with mean µ and stan- number of combinations of expected return µ and volatility σ. dard deviation9 σ. In a world where investors choose portfolios only according to In what follows, we compare VaR and PCA over a given time the mean and variance of the returns, there exists a feasible horizon, say a quarter, in which the trading book capital k is set of portfolios which remunerates higher risk with higher assumed to be constant (k = k ). returns10. The set is defined efficient frontier.

Pre-commitment of capital it needed to back the sense that banks, by a self-assess- To address these concerns the pre- trading book and would be penalised ment of the efficacy of the internal commitment approach has been if it under-estimated. risk management system, would suggested as an alternative to VaR. The firm would almost certainly decide whether to be more or less Under the PCA, a bank would need to have a VaR model to assess conservative with respect to the specify how much capital it needed how large a portfolio it could run output of their internal models. As a to back its trading book over a day-by-day given the capital amount consequence, Kupiec and O’Brien period, say the next quarter, and to which it was committed. But the (1995) believe that the threat of would be penalised, for example supervisors would not have to recog- penalties would encourage the adop- by a fine, if cumulative losses nise or attempt to check the accuracy tion of more comprehensive risk exceeded this figure at any time in of the model, nor would they have management procedures to take the quarter. In other words, rather to lay down parameters in advance account of risk sources such as than meeting the capital require- for the model such as the 99 per cent operational and legal risk. These ments set by the regulator, the firm confidence interval. This would risks could not be explicitly incor- would itself decide on the amount constitute an enhancement in the porated in ex-ante risk-based capital

44 CAPITAL REQUIREMENTS

µ set of feasible portfolios the regulator to achieve a pre-committed level of capital consis- efficient frontier tent with that produced by a VaR framework11. It follows that shareholders will accept more risk only for higher yields. Banks will then allocate their resources by selecting portfolios lying in the efficient frontier. Under VaR, as capital is fixed, the level of volatility and expected return are upper bounded. Under PCA, σ can be increased without limit s allowing higher returns while not changing the cost of capital. The combined effect of regulation and the cost of capital make Penalties establish a link between the capital and risk. The banks’owners risk averse. In fact, regulation establishes a link link consists of an economic cost which increases as the between level of capital and portfolio riskiness via a priori capital declines and decreases as the trading book risk goes limits to the variance (VaR) or ex-post fines (PCA). up. By using an appropriate fine schedule, it would be It can reasonably be assumed that banks also select trading possible to force in a PCA world the same ex-ante level of σ portfolios lying within the efficient frontier. In fact, regulation as in VaR. places a link between risk and capital level. It turns out that At -k , q>p σ : VaR volatility upper bound bank owners accept higher risk only if they can earn higher f1: Penalty VaR equivalent

returns as a compensation for the cost of injecting more capital f1 (VaR) or the potential cost due to the higher likelihood of being σ

penalised (PCA). The following picture can help summarise f2 this point. q Interestingly, it can be proved that if the bank owner can fully control the manager — that is there are no agency problems p -k o ∝1 ∝2 net returns in the bank — there exists a penalty schedule (f1) that permits approaches (VaR and Standard Appr- threat because losses could be close penalty could be implemented either oach) as it would be difficult to to or above the level of proprietary via the imposition of a compulsory implement a general rule for their funds. Having nothing to lose, share- capital level unrelated to the next identification and measurement. holders might be tempted to gamble period of pre-commitment, or However, the PCA has only for resurrection by taking huge through an “add-on” to future pre- been proposed for those banks for risks. commitments. The former alternative which trading risk is not the domi- could simply mean a reversion to nant risk. They would therefore Penalty structure the standard rule-based approach — have a large cushion of regulatory Finding an appropriate penalty in effect, a temporary abandonment capital generated by the banking mechanism is not a straightforward of pre-commitment for the partic- book, with the trading book require- task. Kupiec and O’Brien (1995) ular firm. If the penalty was simply ments very much a subsidiary suggest either capital charge penal- an add-on to the pre-committed element. For other banks for which ties or monetary penalties.7 amount, the bank could nullify the trading was the determinant risk, The application of increased penalty either by reducing each penalties would not represent a future capital requirements as a period’s pre-commitment to main-

45 CAPITAL REQUIREMENTS

tain the same relationship between Disclosing a breach pre-committed capital (after the It is proposed that, under PCA, any penalty) and risk, or by increasing the breaches and penalties would be amount of risk taken in the trading disclosed to increase market disci- book for a given pre-committed pline. The paper by Daripa and capital. Varotto explores the effect of public Therefore, although additional penalties for breaches on conserva- capital charges intuitively appeal tive banks. When the breach of because, on the face of it, they pre-committed capital is publicly would require those banks that have announced, it could have a serious had a breach to restore an adequate impact on the reputation of the capital cushion, in practice the bank, which might lead to a lower When the banks may be able to negate the future profitability. Clearly reputa- effect. tional effects would affect the whole breach of Fines would be even more prob- bank and not simply the trading lematic as they would weaken an arm. This could well (where there is pre-committed already compromised financial situ- less than full deposit insurance) capital is ation (with further complications increase the cost of funding across potentially arising from adverse the whole bank, which would reduce publicly public reaction) without providing the return on the trading portfolio, debt holders with any additional and also on the banking book. This announced, it guarantees. is because a PCA breach could be Another issue, common to both taken as an indication of generally could have penalty structures, is that in the case poor systems and controls. Concern of a breach, penalties should not be about reputational damage reflects a serious applied if the bank had in fact esti- the fact that: mated capital conservatively but • the extent cannot be predicted. impact on the had been subject to extreme market • it is likely to depend on the conditions that could not have been contingent economic situation, r eputation of anticipated. with most severe consequences the bank However, to make the policy in periods of market instability credible, supervisors would have to — ie the periods when such state clearly that the penalties would effects are least welcomed by be waived only when a violation of regulator and regulatees. the commitment was caused by truly An important point to note is extraordinary market events, result- that there could be asymmetry ing in widespread disruption. between gains and losses. Under the Nonetheless, drawing a distinc- PCA, the bank would have a large tion between periods of this extreme banking book relative to the trading kind and others for which a relax- book. While a profit on the trading ation of the rule should not have book would not be very important been allowed, would not be straight- for the overall return of the bank, a forward. loss on the trading book, in excess

46 CAPITAL REQUIREMENTS

of the pre-committed capital, could PCA compared with VaR have a significant effect on the A key difference between VaR and entire bank through the funding pre-commitment is that the former costs. This could lead banks to provides a relatively hard link marginalise their trading books between the exposures which a firm even further, and also to over- is running and the capital it must commit trading book capital. hold. A firm cannot take an expo- The US Federal Reserve Board sure in excess of the capital set set in train a pilot project in October aside, because this can be audited 1996, to look at the effect of using and the requirements must be met at pre-commitment. Ten banks — a all times. In contrast, with pre- mixture of US and foreign banks — commitment, the firm must take a agreed to pre-commit to the capital judgement about how much capital Shareholders they would need to back their it needs, reflecting expectations trading book for four measurement about the kind of positions it intends may find it periods. to undertake in view of predicted The reports of the results for the market changes and the speed with difficult to first period highlight the problem of which positions can be sold. excessive capital commitment for For example, it could choose to influence high franchise value firms. The run a very large exposure relative to decisions other president of the New York Clearing the capital set aside, far in excess House Association said at the end of of the amount which would be than by voting the first period8 that none of the possible under the VaR approach, banks tested had needed more on the expectation that it would be with their feet capital than they had set aside, but sold very quickly. they had taken “a very conservative Another difference between the and selling their approach ... They did not want their two approaches is that, under VaR peer group or their primary regu- the capital requirements can be set shares if they lator to know that they went through by the regulator at a level which the capital ... They took it very seri- would reflect the social cost of are dissatisfied ously ... They saw it as a threat to failure. In the case of default of a their reputation”. sizeable firm, the social costs could with the returns In sum, banks with large be higher than the private costs of banking books are the best candi- failure to the shareholders, because dates (from the point of view of of the magnitude of the domino social optimality) to bear greater effect that the failure of a big player trading book risks. Yet these are can trigger. Under PCA, as the firm precisely the banks that might itself decides on the amount of reduce their trading books and capital to set aside, the considera- commit to excessive trading book tion of only private costs may lead capital, if the violation of pre- to under-capitalisation. committed capital was publicly An interesting question is disclosed. whether pre-commitment would

47 CAPITAL REQUIREMENTS

Box 2 PCA AND VAR WITH AGENCY PROBLEMS

Let us consider the following example: the bank owner and if an investor is willing to accept higher risk, there are oppor- manager are risk neutral. However, the bank owner offers a tunities of getting higher expected returns. So increasing r has risk-sensitive compensation scheme which induces the the effect of reducing σ and, at the same time, the expected manager to behave according to any desired degree of risk yield of the bank's trading book. aversion. The level of induced risk sensitivity is measured by Suppose12 µ=0.5σ , then, in the event of a full protection the parameter r. According to the penalty structure set by the strategy, the type 2 manager should produce a variance equal regulator, the principal’s optimum choice of σ turns out to be, to 1, hence r would have to be set equal to 9. Full protection σ * say, p =1. In the market, there are two types of managers. This will then give an expected return of 0.28, while in the normal differentiation comes from the consideration that managers can case, for expected σ equal to 1, µ would be 0.5. be differently concerned about their reputation. Daripa and Varotto 1997 show that if the manager were interested only in PCA remuneration the owner could write a contract that, under PCA, OWNER’S EXPECTED AND ACTUAL PAY-OFF would reduce the risk of overtrading to VaR levels. However, the manager might well be attracted by non-pecuniary benefits Expected Actual such as maintaining star performer status. The importance of pay-off pay-off these benefits varies among managers, thus producing different type 1 type 2 receptiveness to contractual agreements. Then, manager manager σ σ σ σ σ • type 1 is more receptive to the risk limiting incentives µ E p µ p= 1 µ p= 2 incorporated in the contract. Her optimum choice of Owner maximises portfolio volatility will be equal to: pay-off 0.5 1 0.1 0.2 0.9 1.8 σ 1 r=5 (E p=1) σ * 1 (r)= r Owner takes full risk protection 0.28 0.56 0.06 0.11 0.5 1 σ * • type 2 is less receptive. By maximising her pay-off func- r=9 ( p ≤1) tion subject to the same contractual constraint, the optimum trading book volatility is given by: The efficient frontier is assumed to be µ=0.5σ 9 σ * 2 (r)= r In sum, VaR guarantees full protection because the level of σ is directly constrained by the capital in place. So As the bank owner has not a priori information he will choose under VaR it is possible to achieve an ex-ante volatility σ σ an r such that the expected portfolio volatility (E p) is equal to the bank owner’s optimum choice p=1 and a exactly 1. In the optimum contract r will then be equal to 5. return of 0.5. σ * σ * Ex post, if the manager type is 1 then 1 (r)< p , while if it Under PCA the same constraint on the volatility level can σ * σ * is 2, 2 (r)> p . So, if the principal maximises his pay-off only be obtained at the cost of a lower return: 0.28. In this function there will be a fair probability that the risk taken on circumstance, bank owners could choose to give priority will be very high and, probably, unsuitable for the level of to the portfolio returns and relax the initial risk limit. trading book capital set aside. To overcome the problem, the It turns out that under PCA the extent of the agency principal might decide to be more conservative and increase problem may be a determinant of the risk-taking policy in the level of induced risk aversion r. In a mean-variance world, the bank.

48 CAPITAL REQUIREMENTS

create the same incentives for firms agencies, in the attempt to over- are taken by managers with limited in terms of risk taking as VaR. Most come the difficulties in validating liability while, under PCA, the of the analysis of this question has internal models and detecting banks’ owner would have to suffer the assumed that the shareholders (who risk-management efficacy, set a losses and pay the penalty in the would bear the effect of any penal- high multiplying factor. case of a breach. Indeed, market ties) take the decisions about the However, for most large banks pressure provides constraints in the exposures being run. In fact, for a potential agency problem does design of a pay structure for large banks with diffuse share- exist. The paper by Daripa and top management or “high-flying” holding, this is not the case. The traders. At that level, the possibility managers take the decisions about of being sacked may not represent the exposures and the shareholders a real penalty as individuals may may find it difficult to influence well be re-hired by another firm. these decisions other than by voting In theory, if the principal knew with their feet and selling their his agent’s preferences, the risk shares, after the event, if they are It seems that constraint that the imposition of dissatisfied with the returns. This PCA fines generates could simply creates what is known in the finance there is not be transferred to the manager — literature as a principal/agent prob- for instance by a risk-sensitive lem. The shareholders, as principal, yet a completely compensation scheme. Under these bear the risk of the penalties whereas circumstances, the PCA rule would the managers — the agents — who fault-free pervade the bank decision-making may well bear much less risk, set solution as to process and achieve the target regu- the trading book exposures. latory risk limit. But this may not be The paper by Daripa and how to impose the result if there is an information Varotto explores the effect of this gap (asymmetric information) bet- principal-agent problem on the minimum trading ween principal and agent. Where incentives for the firms posed by the there is asymmetric information pre-commitment approach. The book capital shareholders could tailor a compen- analysis shows that if the share- sation scheme, which on average holders were able to determine the r equirements delivered the desired level of risk- exposures and therefore there was taking but which would leave scope no agency problem, the incentive for some PCA breaches. Alternati- effects of VaR and PCA would vely, they may be more conser- probably not differ substantially. In vative and penalise risky trading fact, if this were the case, in strategies to an extent that in no either approach the regulator could Varotto shows that the shareholders circumstance would risk be high successfully bind banks to imple- cannot design contracts for the enough to generate a significant ment portfolio strategies consistent managers which will align the probability of breach. with their capital cushion (see objectives of both parties. This Even if the last option could Portfolio Strategies, Box 1). Indeed, is because the managers cannot guarantee full protection against the PCA could be more efficient. usually be fined (ie paid negative regulatory penalties (breaches occur This might be the case if, for salaries) in the event of losses. Thus with negligible probability) its instance, under VaR supervisory decisions about trading book risk implementation would compromise

49 CAPITAL REQUIREMENTS

the profitability of the firm. There shareholders to effect a transfer of There is not yet a completely would effectively be no way in regulatory constraints to the deci- fault-free solution on how to impose which the shareholders could ensure sion makers in the firm. PCA would minimum trading book capital that the managers would not take then fail to guarantee that a volatility requirements. VaR models represent excessive risk while maintaining upper bound, relative to capital’ is an advance on the Standard Approach profitability (see Box 2). actually in place. but have a number of drawbacks. In Hence, it is likely that share- In contrast, as VaR sets rules for particular, the fixed parameters set holders would opt for a form of generating the capital requirements by the regulator do not take into compensation that would not guar- it would provide the same risk account the different circumstances antee full protection against breaches limitations in all circumstances and of different banks. or losses, leaving the firm open to would therefore not be affected by PCA represents a way of excessive risk-taking relative to any agency problems within the putting the onus on the firms to capital. For example, if the manager bank. decide how much capital they need, is more concerned with maintaining The only other option would be given the penalty structure. This his/her star performer status than for PCA to set penalties — not for represents an important shift in his/her monetary compensation, then the firm (and therefore for the share- paradigm with interesting implica- PCA will fail to guarantee adequate holders) but for the managers. It is tions. At present, it is very difficult protection against high-risk trading not clear, however, that it would be to have a complete understanding of strategies. possible to do this. Managers are PCA’s potential impact on firms’ In conclusion, as PCA hinges penalised in various regimes for risk taking. Any development in this upon the way in which the owner failure to supervise subordinates or approach would need to deal with can influence managerial choices, if a lax control environment but to some key issues, namely: there is asymmetric information exact penalties from individuals for • The agency problem within banks about trader/management prefer- underestimating capital might exceed caused by information asymme- ences, it may not be possible for the the bounds of natural justice. tries between ownership and management. • The difficulty of imposing capital NOTES levels consistent with public costs 1Kupiec and O’Brien (1995) “A Pre-Commitment Approach to Capital Requirements for the Market Risk”. Federal Reserve Board. of failure. 2Daripa A and Varotto S. (1997) “Agency Incentives and Reputational Distortions, a Comparison • The development of an effective of the Effectiveness of Value-at-Risk and Pre-Commitment in Regulating Market Risk”. Bank of penalty regime. England Divisional Paper Series. 3 Basle Committee on Banking Supervision (1993) “The Supervisory Treatment of Market Risks”, An optimal solution could well April, Bank for International Settlements, Basle, Switzerland. be a framework which integrated 4 See Jackson (1995) “Risk measurement and capital requirement for banks”, Bank of England Quarterly Bulletin, May. the most attractive features of the 5 See Kupiec (1995) “Techniques for verifying the accuracy of risk measurement models”. The Journal two methodologies (VaR and PCA). of Derivatives. 6 See Jackson, Perraudin, Maude (1997) “Bank Capital and Value at Risk”, Journal of Derivatives, It might be possible to combine both Spring. to deliver regulation that imposes a 7 They also propose to combine monetary and capital charges in a mixed penalty structure. level of capital consistent with the 8 American Banker 3/3/97. 9 Returns standard deviation is their average dispersion around the expected value. In the remainder riskiness of the positions taken of the paper standard deviation will be indifferently referred to as volatility and risk. Standard devi- while at the same time being flex- ation is defined as the square root of variance. 10 See Markovitz, H. (1952) “Portfolio Selection”, Journal of Finance, 7: 77-91. ible — ie permitting relatively lower 11 For a discussion of this issue see Daripa and Varotto 1997. requirements for firms with greater 12 This is the aforementioned efficient frontier. placing power and market expertise.■

50 their ability to look after their own REFLECTIONS ON FINANCIAL interests. Conduct of business regu- lation aims at business functions REGULATION and how (solvent) firms conduct business with their customers. Again, By Charles Goodhart and Philipp Hartmann, London School of Economics this mainly concerns the retail David Llewellyn, Loughborough University sector, while the wholesale sector may be largely governed by codes Many countries have experienced significant banking sector problems in of conduct among the professionals the past 15 years. The main causes have been those that have tradition- themselves. ally attended commercial banking problems — poor credit control, connected lending, insufficient liquidity and capital, and poor internal Systemic issues governance. The outcome has been worse than in any similar period since The systemic, though not the pruden- the great depression of the 1930s. In most countries, especially perhaps tial, dimension to regulation occurs the emerging and transitional countries, there is a need for improved when the social costs of the failure external supervision to reinforce internal controls. of a financial institution (particu- larly a bank) exceed the private Given the scale, nature and costs costs and such potential social costs of the financial and banking prob- are not incorporated in the decision- lems around much of the world, making of the firm. what is the rationale for financial However, systemic issues do regulation? not relate to all institutions. The key In contrast to, for example, point is that banks are subject to public utility regulation, financial potentially contagious deposit runs regulation is less about the control which can cause otherwise solvent of monopoly power and more about banks to become insolvent. This is maintaining systemic stability (a primarily because a large proportion question of risk allocation) and of the assets, whose value is based consumer protection. While clearing on inside information, are not easily and settlements systems for finan- marketable, while a big part of the cial transactions incorporate network liabilities (usually deposits with a effects with certain natural monopoly fixed guaranteed value) can be properties, most of the financial withdrawn almost instantaneously. industry is operating in extremely Moreover, banks are also special competitive conditions. because of their pivotal position in We divide customer protection clearing and payments systems, into two aspects: prudential regula- which provide the market infra- tion and the conduct of business structure through which systemic regulation. Prudential regulation shocks can be transmitted. aims at avoiding institutions’ fail- Some analysts challenge the ures with adverse consequences for basic premise that banks are prone individual retail depositors, whose to runs, or argue that improvements informational disadvantages lessen in payments systems technology,

51 REFLECTIONS ON REGULATION

such as netting arrangements and insurance companies in industrial shift many risks off balance sheet, real-time gross settlement (RTGS) countries, have transformed the task which makes the use of traditional, can deal effectively with any systemic of external financial regulation. occasional accounting data of less hazards. Another dimension of greater insti- use as a reliable guide to the banks’ However, there is a particular tutional complexity is the geographic state of health. This process is likely dimension to the debate that is worth diversification of major banks, which to continue in the future with the noting: the risk versus the serious- are now active in all the main finan- further evolution of new markets, ness of the issue. The probability of cial centres and markets worldwide. such as that for credit derivatives. the failure of a single bank inducing Credit derivatives make default a systemic problem may be low but, risk tradable and might therefore if such systemic failure were to New and render the traditional regulatory occur, it would be very serious and distinction between banks’ trading the cost would be high. Thus, regu- increasingly and banking books obsolete. lation to prevent systemic problems While Value-at-Risk (VaR) mod- may be viewed as an insurance sophisticated els for market risks have already premium against a low-probability found their way into the 1996 occurrence. financial amendment for market risks of the In general, the requirement for Basle Capital Accord, more sophis- systemic regulation is less evident instruments and ticated quantitative risk management for securities firms or insurance techniques of models for credit risk might well companies than for banks. They do have to be considered by regulators not play a comparable role in the risk management in coming years, if market partici- payments system, their assets are pants succeed in overcoming the less opaque and largely marketable make the task lack of comprehensive data on and their funding is not based on defaults and loan recovery rates. cash deposits, which can be quickly of external For all these reasons external withdrawn. Even so, securities and regulation, both in its regulatory insurance firms may under some r egulation and mode of seeking to lay down, ex circumstances pose systemic threats. cathedra, common rules and ratios One obvious case is the potential supervision that all banks should follow, and in impact on banks when securities a supervisory-monitoring mode of operations represent part of the much more checking whether banks are com- banking group. Even if a firewall difficult plying with such rules, is becoming arrangement establishes a legal inde- less effective and less feasible. pendence, the default of a bank’s The Barings’ failure produced subsidiary would inevitably damage Finally, new and increasingly a standard reaction among (less its own credit standing. sophisticated financial instruments informed) observers that supervi- and techniques of risk management sion needed to be increased and Need for reform make the task of external regulation improved. However, among the The degradation of the dividing and supervision of financial institu- more perspicacious commentators, lines between commercial and invest- tions much more difficult. most of the supervisors themselves ment banks and, to a lesser extent, For example, securitisation tech- and certainly among most practi- the coming together of banks and niques and derivatives instruments tioners, the lesson drawn was that

52 REFLECTIONS ON REGULATION

the external regulation required to prevent such cases of fraud (when it had escaped internal management) would have had to be so pervasive, so intrusive and so expensive as to be practically impossible. The failure arose from poor internal monitoring and supervision, not from failure of regulation. What went wrong was lack of internal control systems, failure of manage- ment to monitor its staff and failure The role of supervisor or compliance officer is not generally a glamorous one of the bank to enforce its own rules. when compared with a “star” trader. Indeed, we would argue that the In the words of Professor Goodhart: “As Gilbert and Sullivan wrote about the lot of a policeman, the position of a regulator is not a happy one. He/she can most so-called regulatory failures hope at best to be ignored, at worst to be reviled.” are actually internal supervisory failures of monitoring and supervi- the focus and the needs of the regu- it might lead the way to more far- sion, rather than the absence of latory authorities of these countries, reaching reforms in the longer term. prescriptive rules. which need to better enforce the old, It stipulates the conclusion of There is really no alternative to traditional lessons of credit control an incentive contract between the placing the primary responsibility and properly diversified and arms- regulator and the regulated. At the for risk control on the shoulders of length lending, and the authorities beginning of a quarter, half-year or internal management, and on their in more industrialised countries, similar period, a bank would have to auditors. The internal managerial which are beginning to experiment pre-commit to a maximum cumula- risk-control mechanisms of those with new regulatory approaches, tive trading loss over the regulatory banks in industrial countries with such as the reliance on banks’ portfolio holding period, which is large trading books should be internal VaR models. 10 business days. This pre-commit- strengthened, and the nature and ment amount would be equal to functions of external regulation Incentive-compatible the minimum capital requirement; should be recast, with more emphasis The basic underpinnings of such a hence there are disincentives against on establishing incentives/sanctions new philosophy are, maybe, best over-commitment. Should the bank’s and less on prescriptive rule-setting. illustrated with the example of the trading loss (which, in the case of However, we also note that it is pre-commitment approach, which market risk, is relatively easily to not a question of internal versus has recently been developed by verify) exceed the pre-commitment external regulation. The recommen- researchers of the US Federal Reserve at any time during the quarter/half- dation is rather about the balance Board and has received strong year, the regulator would impose between the two and we would backing by the chairman, Alan penalties. Hence there are also disin- consider it wrong to polarise the Greenspan. centives against under-commitment. issue. This proposal aims, in the first With an adequate and credible penalty We would not apply the same place, at the reform of minimum structure it is in the banks’own self- recommendation to emerging and capital requirements for market risks interest to maintain the optimal transitional countries. This leads to (such as fluctuations in interest rates, regulatory capital and ensure satis- something of a dichotomy between share prices or exchange rates), but factory internal controls.

53 REFLECTIONS ON REGULATION

Pre-commitment has the advan- ness in achieving lower systemic tage of focusing on the outcomes risk and more effective client protec- rather than the whole process of tion without causing any distortionary market-risk management, keeping side-effects, depends to an impor- the regulatory intrusion into bank tant extent on market expectations management and the costs of regu- about the ex ante credibility of regu- lation relatively low. latory intervention. In particular, it relieves regula- Rules and regulations, which Pre-commitment tors from the burden of deciding were perceived to be optimal ex ante, about the quality of banks’ internal may be time-inconsistent. There focuses on the market risk management (Value-at- may, for instance, be many reasons Risk) models, which they will have why the regulatory authorities may outcomes rather to shoulder with the full implemen- choose not to close a bank in tation of the recent amendment of evident difficulty. That does not than the whole the Basle Capital Accord for market mean that regulators or other public process of risks. officials are dishonest or act with It also avoids the potentially guile, intentionally misleading the market-risk distortionary effects on banks’port- public. folio choices by simple “one size However, they may be exposed management, fits all” capital adequacy ratios, to pressures from different interest which often contain imprecise risk- groups in the political or financial which keeps weightings applied uniformly to all sector to deviate from the ex ante institutions’ positions in different optimal responses to financial diffi- the regulatory financial instruments. culty. In particular, incentives for However, a debate revolves forbearance — to delay intervention intrusion into around the character and size of and disclosure — can be strong. penalties. Most US banks welcomed If the political cost of prompt management of pre-commitment in general, but are and rigorous action is high, then a bank and the strongly opposed to monetary penal- it might be easy to avoid early ties. One argument brought forward disclosure, use the lender of last costs of is that they would hit banks when resort to provide any required they were already down. Non-mone- liquidity to an individual bank and r egulation tary penalties could be formulated hope that through some condition- as restrictions on trading activities ality of that liquidity provision it r elatively low or temporary increases of the finds its “way back” to solvency minimum capital requirement. We and profitability. feel that incentive-oriented regula- The problem is that financial tion is an area in which increased markets are, of course, aware of academic research could make a time-inconsistency and credibility substantial contribution to the regu- problems, in particular with respect latory debate and future reforms. to large financial institutions which The incentive compatibility of might fall under the “too-big-to-fail financial regulation, ie its effective- doctrine”.

54 REFLECTIONS ON REGULATION

Expectations about forbearance ities as capital diminishes. Beyond and public bailouts imply a reduc- that, the fact that each, and every, tion of incentives to set up better dividing line (say the 8 per cent internal controls. This itself increases minimum level of the Basle Capital the likelihood of financial distress. Accord about the regulation of From an international perspec- credit risk) is arbitrary, makes it tive, there are too many examples of undesirable to put too much weight unsuccessful forbearance — it is our on any one such number. It is view that these provide a presump- preferable to have a series of such tion against it. Hence, the question dividing lines, with the effect of Expectations arises as to whether there are possible going through any one of them rela- solutions to overcome such time- tively minor, but the cumulative about inconsistency in financial regulation. effect large. In this respect, the prin- One approach to this is pre- ciple of the graduated sequence of forbearance commitment on the regulatory side. response in FDICIA seems right. Formulating rules for regulatory There is a good general argu- and public reactions in the form of an incentive ment that pre-commitment and rules contract is relatively rare. However, are normally advantageous. But if bailouts imply there are examples of rules-based the external regulator feels very a reduction regulation. strongly about a particular case, it The most far-reaching and best should be allowed to override the of incentives known is the ladder of graduated pre-commitment. responses to declining capital ratios That, however, leaves a problem. to set up fixed in the 1991 United States Both the justification for, and perhaps Federal Deposit Insurance Improve- even the existence of, such an better internal ment Act. (In response to current override in this matter may, for banking difficulties in Japan, a obvious reasons, need to be kept controls — this corrective action study group, set up confidential. by the Ministry of Finance, has How then can one prevent the itself increases recently published an interim report external regulator from effectively on a proposal for similar regulations just exercising discretionary control? the likelihood in that country.) Our answer is that the report of financial The greater a bank’s capital, the and justification would have to be less likelihood there is of insolvency made to an “independent” over- distress for a given expected risk. But there seeing body, with full publication cannot be a single point value above after an appropriate time-lag. which the financial institution is safe, and below which it must be Proportionality closed. Clearly, danger increases One of the problems of regulation gradually as capital diminishes. is that it is usually not supplied That fact by itself implies that through a market mechanism, and there should be a graduated series of hence consumers are unable either responses from the external author- to signal what is required or to indi-

55 REFLECTIONS ON REGULATION

cate how much they are prepared to supervisors are clear: to try to avoid, pay for the benefits of regulation. or to avoid taking responsibility for The absence of a market in regula- market failures, while at the same tory services creates two hazards: time ensuring that the costs of such there is a major loss of information (additional) regulation and supervi- about what extent of regulation the sion are fully met by others, usually consumer demands and how much by passing them on to the private the consumer is prepared to pay. sector. In general, the public is blithely One of the Of course, the financial inter- unaware of the cost of the regula- mediaries themselves will pass on tory system. The financial cost of problems of most of the costs of such interven- intermediation affects the size of the tion in the form of higher spreads. spread between a bank’s deposit and r egulation is Even if there is competitive neutrality lending interest rates. The cost of between institutions, regulated insti- complying with regulation and super- that it is usually tutions may still lose business to the vision amounts to no more than a not supplied capital markets. Put another way, few basis points, widely diffused regulation of financial intermedi- over all the customers of financial through a market aries may be seen as an implicit intermediaries. Customers are usually subsidy to primary capital markets. unaware that the regulatory regime mechanism, and This is not necessarily unwarranted: imposes any costs upon them at all. indeed, it will be justified if the This creates the impression that hence consumers regulation is efficient. It is of no regulation is a “free good” — a concern if banks lose business to problem in that it is likely to create are unable either markets because regulation forces excess regulation, in particular with them to pay the full costs of their respect to regulatory institutions to signal what is activity, including potential exter- which have their own self-interest nalities such as systemic costs. and reputation at stake, so that the r equired or This does, however, cause prob- burden (costs at the margin) of regu- indicate how lems both for regulators and for the latory compliance can exceed the legislators, who otherwise normally benefits. much they are treat such regulation as a free good. On the other hand, the costs to If the authorities lose market share, the general public of a financial prepared to pay and financial institutions emigrate crisis, in which people lose money to another jurisdiction, regulators as a consequence of the behaviour for its benefits will find both their potential tax of an intermediary, are felt directly. base and their power base shrinking. Under such circumstances, the result So, the natural tendency within each of such losses is inevitably to call regulatory regime to over-regulate for more, improved, better regula- is held in check by fears of losing tion and supervision. market share. Regulators naturally In addition, and under the influ- deplore this potential erosion of ence of such public pressure, the power and denigrate it with the term incentives for the regulators and “regulatory dumping” or “the rush

56 REFLECTIONS ON REGULATION

to the bottom”, implying that the raises the question of what (cred- end result of such competition will ible) sanctions, if any, could be be grossly insufficient public regu- imposed by self-regulating practi- lation. tioners without running into such It is, however, rarely demon- issues as whether self-regulating strated that an informed public practitioners’ clubs might them- and practitioners necessarily prefer selves become anti-competitive and to operate in “insufficiently” regu- even detrimental to the wider public lated environments. Competition Regulatory interest. between regimes does not neces- A problem in trying to move sarily result in a complete absence r egime towards proportionality between of regulation. Indeed, if there is a costs and benefits of financial regu- consumer demand for the services competition lation is that it is so hard to measure of regulation, then it is quite likely either. This comes, theoretically, that less-regulated centres would does not result within the purview of cost/benefit lose business and this could call in an absence analysis, but these are notoriously forth an industry demand for tighter difficult to complete successfully in regulation. of regulation — practice, not least in the financial In the meantime, one of the sector. This is primarily because it is greatest concerns of regulators is if there is difficult to quantify the benefits of how to deal with the growing such regulation and supervision. globalisation of wholesale financial consumer At least the costs are fairly intermediation, and prospectively of concrete. However, a problem in such retail financial services as well. demand then calculations is distinguishing between The previous argument about the activities that financial interme- the tendency for excessive regulation less regulated diaries would have undertaken for should make one hesitate before their own risk control and internal advocating a global regulatory frame- centres would governance, and the additional work work, even if that were practically lose business, that the regulatory regime imposes. feasible. Given differences between That said, much more could, national structures, history, culture, raising industry and needs to be done, to attempt to interests and viewpoints, any inter- identify and quantify such costs. national harmonisation would be a demand for How do these relate, for instance, to compromise; and like most compro- the average spread? If translated mises, it is likely to be unsatisfactory tighter rules into terms of basis points, how (from the outset) and simultane- much is the average user of finan- ously difficult to amend in the light cial services paying for the regulatory of changing circumstances. regime? This suggests that the formula- Currently there is no apprecia- tion and application of global tion that the regulatory regime “standards” might be better allo- entails costs either for the user of cated to private sector institutions to such services, for the equity owners impose upon themselves. This of such intermediaries, or for the

57 REFLECTIONS ON REGULATION

taxpayer. Legislatures and supervi- costs, systemic costs of failure) can sors press ahead with proposals be corrected through regulation and having little knowledge, considera- supervision, the magnitude of the tion or discussion of what these may benefits could be very high even if cost. Commissioned studies, both at they are not easily quantifiable. Nor national and international level, is it the case that the benefits of the (true) costs of the various will accrue to consumers only as regulatory regimes would be most regulated firms can also gain in a welcome and could contribute to regime of effective and efficient Where making regulatory regimes more regulation. efficient. There are benefits from regula- supervisors While measuring costs is diffi- tion, and there is an evident consumer cult, measuring benefits is even demand for them. But, regulation is have managed harder. The main purpose of regula- not a free good, and the costs are tory regimes is to prevent events varied and can be substantial. One to defuse a happening; for example, systemic of the most difficult exercises in the potentially collapses or opportunistic mistreat- regulatory process is to make the ment of clients. So success involves ultimate cost-benefit judgement: do dangerous the absence of such “bad” events. the benefits derived from regulation But of course we do not know exceed the costs? situation, there whether the absence of such bad Complex judgements have to events is due exclusively to the be made because it is in practice will usually be regulatory regime. Moreover, even extremely difficult to be precise where the supervisors have managed about costs and benefits. There are constraints on to defuse a potentially dangerous serious methodological problems situation, there will usually be involved in conducting a cost-benefit their ability constraints on their ability and will- analysis of financial regulation: ingness to reveal their successes • The difficulty of placing a precise and willingness publicly. value on the consumer-welfare to reveal their While the benefits of regulation benefit of correcting any market in general, and specific elements of failure. successes a regulatory regime, may be diffi- • Identifying the value placed on cult to measure accurately, this does information by consumers that publicly not mean that they are not real. The is relevant to them. ultimate rationale of regulation and • The extent to which a probability supervision is to correct various of failure is reduced. forms of market imperfections and •The problem of first defining, failure, including externalities such then measuring and locating the as the social costs of failure incremental costs of regulation exceeding the private costs. on firms. If significant market imperfec- • Determining what affect regula- tions (such as those associated with tion has on the competitive asymmetric information, agency conditions in the industry.

58 REFLECTIONS ON REGULATION

How then are the benefits of tion is to try to give clients of finan- regulation to be measured? One cial services a choice between approach, which is commonly used institutions subject to more, or less, to assess the perceived value of non- regulation. One example is the market services, is to use survey “narrow bank” proposal. Under this techniques, asking consumers how scheme, a bank, or even separated much they are willing to pay for parts of the same bank, can elect to protection against losses from finan- be made “safe” by being required to cial failures or mistreatment. But hold only a limited set of highly once again, the insurance nature of liquid, almost capital-certain assets, the regulatory service makes it more as counterparts to their deposits. difficult for anyone to assess its This “safe” part of the bank could value. Thus, after a crash of some be granted 100 per cent deposit kind, the ex post value attached to a insurance. strong regulatory regime may well There are, however, problems be a multiple of the previous ex ante with this approach. First, it does not value. deal fully with differential informa- Alan Greenspan, the chairman of However, practitioners, and their tion between the suppliers and the US Federal Reserve Board. He firms, have a more informed under- demanders of financial services. It has backed the pre-commitment approach, which has recently been standing of the costs and benefits to may not be possible for consumers developed by researchers of the themselves of the regulatory regime. to make judgements between so- Board Once again, however, there is likely called safe and less safe financial to be a bias in their responses to any institutions. Also, it may not always into the higher risk/higher return such survey, as the social benefits be possible to require consumers to entities, thereby driving the relative from preventing systemic collapses accept the consequences of their price of risk assets up and further are likely to considerably exceed judgements. reinforcing the boom. When the their private benefits. Even so, a It simply will not be politically boom breaks, there will be an equiv- relative ranking by practitioners of feasible, at least in the aftermath of alent flood back into the safe entity. which aspects of the regulatory financial disasters, to leave Aunt The flow of funds from riskier to regime are perceived as better value, Agatha to the mercy of caveat safer entities would then force a relative to their cost, and which emptor, should she, particularly risky bank to have to have to liqui- were not, could be helpful. Such during boom and “bubble” condi- date their (risky) assets, thereby surveys have their limitations; but tions, choose to invest in “risky” reinforcing the severity of the given the paucity of information on activities. slump. the cost/benefit balance, and the In addition, the division of generalised complaints of practi- banks into low risk/low return and Institutional structure tioners about the lack of due higher risk/higher return entities The structure of supervisory bodies proportionality, it would seem that may have the undesired side-effect and their internal organisation has more could usefully be done in this of exacerbating macro-economic recently come to the forefront of the field. volatility. During booms, when confi- debate about financial regulation, A more ambitious proposal dence is high and the prospects not only in the UK (which has opted than consumer surveys to achieve of failure correspondingly low, in favour of a single mega-regulator, proportionality in financial regula- investors are likely to shift en masse probably embracing the whole UK

59 REFLECTIONS ON REGULATION

financial services industry in the We believe that, in practice a conduct of business agencies may future), but also in Australia and matrix approach is needed when be the most effective and efficient other countries. We should first financial institutions conduct a wide approach. emphasise that institutional struc- range of business. In this case, insti- Overall, we emphasise that the ture does matter, because the way tutions need to be regulated on both design of regulatory structures, or that the dividing lines between and the internal structure of a mega- within institutions are drawn can regulator, has to be focused around influence the effectiveness and costs the objectives of regulation, because of regulation. the ultimate criterion for devising However, in our view it is not the optimal structure should be an issue of primary importance for the effectiveness and efficiency of effective regulation. Although of regulation in meeting its basic objec- course not entirely independent, the tives. way in which regulation and super- An alternative school of thought vision is conducted (see, for Institutions need to the concept of a mega-regulator example, the discussion of incentive argues that regulatory agencies are issues addressed earlier) has to be at to be regulated most effective and efficient, but also centre stage. In particular, given the more transparent and accountable, great variety of different financial on both a when they have clearly defined, and institutions and instruments and precisely delineated, objectives and their links with each other, there is functional when their mandate is clear and not one single, right institutional precise. structure which can be advised. (conduct of Moreover, there will be times There are different approaches which when the objectives of regulation all have their merits and disadvan- business) and are in conflict and one of the issues tages. to consider is which structure is There are three broad structural institutional most efficient at resolving conflicts. approaches which may be distin- (prudential) In a single agency, for example, all guished by institution, function, and conflicts are internalised. One merit objective. basis of focusing institutional structure In the case of an institutional upon objectives is that it requires focus, regulation is directed at finan- significant conflicts between such cial institutions irrespective of the different objectives to be resolved at mix of business undertaken. This is the political level. particularly appropriate when consid- Additionally, making different ering prudential issues which must agencies responsible for wholesale necessarily focus on institutions. It and retail business (even though is, after all, institutions and not the distinction is fuzzy at the functions that become insolvent. margin) may reduce the danger that Functional regulation has the busi- a functional (for conduct of business specific regulations are extended ness undertaken by institutions as its purposes) and an institutional basis beyond the sphere for which they focus, irrespective of which institu- (the prudential dimension), and that are appropriate. The nature of market tions are involved. having specialist prudential and imperfections and failures is different

60 REFLECTIONS ON REGULATION

in the two areas and the rationale, exception is the monetary authority The second issue is possible approach and conduct of regulation in Singapore. This is not accidental. conflicts of interest. This is often needs to be different. If the central bank has independent advanced by academic economists One of the central objectives of powers to set the interest rate, the as the main argument against central regulation, generally applicable to combination of a widespread regu- bank participation in regulation, in wholesale markets, is to maintain the belief that a central bank with systemic stability. In the absence of responsibility for preventing systemic a mega-regulator this should be the risk is more likely to loosen mone- clear focus for a specialist agency. tary policy on occasions of difficulty. As already noted, there is also a We see no reason why assistance to requirement for prudential regulation a bank in difficulties need affect the even where no systemic considera- aggregate provision of reserves or tions arise. This can also be addressed level of interest rates. Any lender of by a separate agency in charge of last resort assistance can be offset in securities firms, insurance compa- Combining a the aggregate by open-market oper- nies and other non-bank institutions ations. In our previous empirical where continued solvency is a regu- widespread and historical studies of this issue, latory issue. we have come across few attested As the retail conduct of busi- r egulatory cases where the concern of a central ness regulation raises fundamentally bank for the solvency of its banks different issues and requires a totally function with has been a major factor in an exces- different approach to that for whole- sively expansionary monetary policy. sale business, those who advocate monetary control The bottom line in our view specialist agencies would argue in is that banking realities will force favour of a dedicated conduct of might give there to be considerable co-ordina- business regulator for retail business. tion and interaction between the top However, exchanges, and perhaps excessive power officials dealing with monetary wholesale conduct of business issues, to unelected macro policy and those in charge of could be largely self-regulated, except bank regulation. for competition-policy issues related officials The question of whether the to market access. banking supervisory body is formally When each country has decided within, or outside, the central bank on its optimal regulatory structure, is then essentially a subsidiary issue, the next question is the role of the depending on perceptions of the central bank within that. appropriate locus of power and The first issue we address is responsibility. These perceptions that of power. Altogether we found will vary depending on accidents of seven countries (notably Denmark, history and culture. There is no Norway and Sweden) that have put single best approach under all in place a single, all-embracing latory function with monetary circumstances, as is evidenced by the financial regulatory authority, all control might be thought to place variety of regulatory structures in but one have made this separate excessive powers in the hands of different countries and the lack of from the central bank; the sole unelected officials. tendency towards a single model.■

61 THE UK MARKET FOR HIGH-YIELD DEBT However, the UK and other European companies have issued in the US. Such “Yankee” high- By Jeremy Leake, Alex Crowe and Rupert Watson, Bank of England yield issuance by European based companies amounted to about $3bn There has been an active market in high-yield bonds (“junk” bonds) in 1996 and, since 1993, the market in the US for some years. Yet, in spite of apparent attractions to has been used extensively by media, UK issuers, there is no effective market in the UK. This article looks at cable communications and telecom- the role which high-yield debt can play and examines the development munications companies (see Chart 1). of the sterling market. Swapping the proceeds of a US dollar issue into the borrower’s There has been an active market in domestic currency can be costly, so the US for high-yield bonds — this recourse to a foreign market sometimes referred to as “junk” suggests strongly that there are bonds — for some years. These are deficiencies in the UK market. sub-investment grade bonds which Certainly most UK companies which have a credit rating below Standard have issued in the US agree that a and Poor’s BBB- rating or below sterling issue in the UK, were it Moody’s Baa3 rating. The bonds possible, would have been prefer- offer investors the chance of a able. Indeed, the lack of an effective higher prospective return than less UK market may be restricting UK risky investment grade bonds. Most firms’ access to capital. are issued by US companies in The implication — that there the US market and are held by may be benefits to be gained from American investors. a deeper and more active sterling Total issuance in the US market high-yield market attractive to UK was about $60bn in 1996 — a investors — prompts the search for typical issue is about $100m to explanations as to why it does not $150m in size, has an intermediate yet exist. This article examines the seniority ranking, and a maturity of question from the issuer and about seven years. Spreads over US investor perspectives, and assesses Treasury bonds — in effect, the various factors which have been additional return offered in exchange suggested as potential obstacles to for the added risk — ranged from the sterling market’s development. 150 basis points for BB+ rated issues to nearly 500 bp for B- rated The issuer perspective issues. High-yield debt is typically of long In the UK, as in the rest of maturity, unsecured (or less well Europe, bonds of this type are rare: secured than typical bank loans) according to one estimate, total ster- with more flexible — generally ling high-yield issuance was £425m non-financial — covenants and is in 1995 and £745m in 1996 — all of repaid in full on maturity, rather which were unrated (see Table 1). than amortised over its life. So it

62 HIGH-YIELD DEBT

Table 1 UK STERLING HIGH YIELD ISSUES 1995-97

Issuer Date Amount Maturity Coupon Initial spread over Gilts Eco-Bat Technologies 20/06/97 £65 m 2007 9.125% 200 bp Castle Transmission 14/05/97 £125 m 2007 9.000% 195 bp CGL Rail 15/10/96 £165 m 2012 9.375% 130 bp Fitzwilton Finance (UK) 11/09/96 £80 m 2006 9.750% 175 bp Daily Mail & General Trust 14/03/96 £100 m 2021 10.00% 145 bp First Hydro Finance 05/01/96 £400 m 2021 9.000% 115 bp Computacenter 21/11/95 £50 m 2002 10.000% 225 bp Independent Newspapers 16/06/95 £75 m 2005 9.250% 125 bp Slough Estates 11/04/95 £100 m 2017 10.000% 160 bp DeBeers Centenary 06/03/95 £100 m 2020 9.750% 165 bp Daily Mail & General Trust 25/01/95 £100 m 2005 9.750% 76 bp

Source: Bankers Trust NB: All issues were unrated affords financing of a kind and with ment of middle-market UK firms rity in their initial phases. In PFI a cash-flow profile not readily avail- is becoming increasingly sophis- projects to date, contractors have able from other sources and is likely ticated and such firms now want supplied much of the equity, but for that reason to be attractive to a to use more flexible financial there is a limit to how much range of issuers. On the basis of US instruments, such as bonds, which their balance sheets can bear and — and limited UK — experience, were previously available only a small number have tapped the these include: to larger corporates. bond markets. The PFI may even- • High-growth (perhaps high- tech) • Large management buy-outs: tually become a source of higher mid-sized companies: these firms for similar reasons, high-yield yielding bonds. often have limited security to bonds are likely only to be There is, however, a price to pay by offer and limited or negative attractive for larger buy-outs — issuers for the advantages of high- current cash flow, so high-yield but these are accounting for an yield bonds over more conventional debt may provide a more ready increasing proportion of the sources of funding: they can be source of development capital MBO market. expensive. They are usually unse- for them than banks. • Mergers and Acquisitions: M&As cured and subordinated to bank Such paper is unlikely to be have been a major application of debt, so they typically have an useful to smaller companies and high-yield debt in the US and interest rate at least 100 bp higher start-ups because the issuance could play a similar role in the than bank debt for a given issuer. of bonds can be complex and UK. Gearing in UK mergers and Issuance costs are also relatively costly. The consensus among acquisitions tends, however, to high. Investment bank fees are issuers and intermediaries in the be lower than in the US. typically 3 to 4.5 per cent, and legal US is that issues should be at • Project finance and the PFI fees and other administrative costs least $50m, although the preferred (Private Finance Initiative): can add another 80bp — so they size would be more than $100m. projects often show negative are not a cheap substitute for bank However, the financial manage- cash flows and offer little secu- debt.

63 HIGH-YIELD DEBT

Also, US experience suggests be invested in such paper in the UK: high returns and capital growth, for that issuance can be an uncertain institutional assets in the UK which they look to equities, and process. This is partly because, in exceeded £1.3 trillion at the end of certainty of returns, for which they the US at least, legally binding 1996 — 20 per cent of which were typically rely on investment-grade underwriting is not commonly bonds (including gilts, foreign debt. As a result, UK fund managers available. But it is also apparent that government bonds and corporate generally have little expertise in investor appetite for paper can bonds). Chart 2 shows the bond assessing the credit of corporate change quickly. Several UK issuers holdings by different classes of bonds. in the US market have suddenly financial institutions in the UK. The decision to invest in found themselves unable to place UK investors have so far equities will be underpinned by issues on the terms they expected, shown little appetite for high-yield extensive research into company having incurred substantial issue debt. There seem to be two kinds performance and prospects, but there costs. Zero coupon bonds are espe- of potential obstacles: investors’ is little perceived need, and there- cially susceptible to such problems. ability and willingness to evaluate fore no established expertise, to and then manage them; and restric- undertake a detailed evaluation of The investor perspective tions — administrative or statutory the likelihood and cost of issuer High-yield debt is essentially a — on investment powers. default. The lack of a significant wholesale market instrument, so it high-yield debt market in the UK is is principally aimed at institutional Evaluation and realisation itself a deterrent to investment in it, investors. There are very substantial For most UK institutional investors, as it limits the scope for benefiting pools of funds which could therefore the key portfolio decision is between from credit assessment expertise.

Chart 1 1996 NEW ISSUE ACTIVITY IN THE YANKEE HIGH-YIELD BOND MARKET

Utilities 5% Other 3% Retail 2% Telecommunications 27% Consumer Products 2%

Technology 3%

Finance 5%

Transportation 3%

Sovereign 4%

Energy 4%

Service 5%

Manufacturing 5% Media 21%

Forest Products 11% Source: SBC Warburg

64 HIGH-YIELD DEBT

Chart 2 BOND HOLDINGS BY DIFFERENT INSTITUTIONS

Pension Funds

Long-Term Insurance Funds

Insurance Companies other than Long-Term Funds Foreign Corporation Foreign Government UK Corporation Investment Trusts Gilts

Unit Trusts

£bn 0 20 40 60 80 100 120 140

Source: Financial Statistics

UK investors’ attitudes are, in part of investment grade debt. An active constrained in this respect. For at least, a product of the market envi- post-default secondary market can, example, many managers of pension ronment here, but they could change at least to some extent, substitute for funds lack trustee approval to invest if the incentive were sufficient. workout expertise, as it provides a in such assets: trustees tend not to Prudent investment in high- mechanism whereby defaulted debt allow investment outside the main- yield debt requires a careful review can be transferred to those willing stream of domestic and overseas of the debt-servicing capacity of the and able to engage in the workout equities, investment grade bonds issuer, a thorough understanding of process. But there is still a need for and property. There is, however, no the ranking, the likely treatment of expert participation, if the exit route statutory obstacle to them doing so. the debt in an insolvency and an provided by such a secondary market Many pension fund managers assessment of the recovery potential is to exist at all and to give assur- are allowed to invest in the equity of in the event of issuer default. This ance that it will reliably deliver fair companies whose high-yield debt is because a significant factor in value. they would be unable to hold, even obtaining attractive ex-post returns though equity is riskier than debt. is not just the high coupon payments Investment powers This suggests that, if trustees were but also the investors’ ability to The other key aspect from the insti- more familiar with the instrument recover value in workouts. This can tutional investor’s perspective is and satisfied that fund managers call for expertise rarely required of their formal capacity to invest in had the capacity to realise its poten- either the equity investor or (because sub-investment grade paper. Many tial returns, this constraint on default is a remote event) the holder institutional investors in the UK are investment could be relaxed.

65 HIGH-YIELD DEBT

A significant number of unit Regulations on issuance issues to take place without the trusts, unit-linked insurance funds In the UK, the framework for secu- issuer contravening the Banking Act. and investment trusts are also rities issuance does not differentiate Companies proposing to issue constrained from investing in high- between the treatment of high-yield debt securities may also have to yield debt. debt securities and that of other debt consider the Bank’s guidelines on In these cases, the restriction securities: any issuer of debt secu- the lead management of sterling originates from their fund descrip- rities must comply with the Banking capital market issues. Firms which tion and is presumably the result of are incorporated elsewhere in the the actual or perceived lack of EEA may lead manage such issues interest among the retail investors Life funds if permitted to do so by their home at whom the funds are targeted. supervisors, either in their home Some unit trusts do undertake seem to be country or elsewhere in the EEA if such investment: in particular, the this is allowed under the passport. “UK gilt and fixed interest” sector unrestricted Firms incorporated in the UK or invests in corporate bonds and in their choice outside the EEA may also lead some of these may be high-yield. manage issues in London if they However, many of these funds are of investments have satisfied the Bank of England marketed as low-risk vehicles and as to their competence and experi- as such they do not invest heavily — directors ence. in high-yield paper. Tax issues Life funds seem to be unre- usually set broad There are significant structural stricted in their choice of differences between the US and UK investments: they are of course guidelines as to tax systems which may affect the subject to solvency tests by the servicing costs of debt and equity, DTI, but these seem neutral the maximum and so contribute to the success of between investment grade and the US high-yield market. The fact other debt securities. Directors that may be that in both countries the return on usually set broad guidelines as to invested by debt (whether interest or discount) the maximum that may be invested is deductible from corporation tax by a fund in particular credit quality a fund in profits, whereas dividends are not, ranges. It appears, though, that gives debt finance an advantage these limits are rarely reached. particular credit over equity — encouraging higher gearing. Nonetheless, the UK system, Structural impediments quality ranges unlike that in the US, partially A range of structural factors have imputes corporation tax to share- been cited as obstacles to the Act 1987 as well as with the general holders and makes an indexation development of this market, perhaps law on prospectuses, listing and allowance for capital gains on equi- because they are respects in which advertising. The Banking Act can ties. So (until the recent Budget the UK and US environments differ. restrict issuance if this involves the restricted dividend tax credits) the The most common factors cited acceptance of deposits as defined in UK system provided more flexi- include differences in issuance the Act. But even here the Exempt bility and created less incentive for regulations, insolvency procedures Transaction Regulations 1997, allow issuers to prefer debt over equity and tax regulations. a wide range of corporate debt finance.

66 HIGH-YIELD DEBT

When a UK company does tax authorities and employees) and bankruptcy court by an interested decide to raise debt, its interest costs secured creditors (usually banks) party, such as a creditor, a share- tend to be lower when borrowing for repayment. Secured lenders are holder or a director of a business in from a UK bank than when issuing afforded greater privileges in financial trouble. Its objective is to domestic debt securities. This is Administrative Receivership than try to find terms to preserve a busi- because interest can be paid gross to in Chapter 11 (respectively the most ness that are acceptable to a majority a bank — avoiding the need to commonly used frameworks for of creditors. Chapter 11 also gives deduct withholding tax from interest explicit rights to any interested party payments and remit it to the tax to be consulted and to appeal to the authorities. Since withholding tax The prospect courts should they believe that their does not apply to the discount secu- interests are being overridden. rities, however, there is an incentive of a prolonged Differences between US and UK to issue zero coupons instead of insolvency procedures may not, interest bearing debt, if there is period of low however, be as marked in practice as investor demand for it. inflation may they would appear on paper. Most However, despite these differ- receivers make every reasonable ences, the general opinion among increase effort to preserve the business, and issuers and investors seems to be this frequently involves negotiating that the UK tax regime has not in institutions’ with different classes of creditor to practice presented an obstacle to gather their views about partici- high-yield issuance. appetite for pating in a financial reconstruction. Insolvency However, the ranking of different Of potentially greater significance, credit exposure types of creditor is much the same in given the default risk associated both countries: security rights, in with high-yield debt and the impor- as a source of particular, are respected. But there is tance of recovery rates, is the effect a greater difference between the rank- of insolvency law and informal higher returns ings of different creditors’ claims in procedures for restructuring the — similarly, the Anglo-Saxon countries and in capital base of a company in finan- continental Europe where obligations cial difficulty. It has been argued EMU may to employees are given relatively that UK insolvency law provides a more weight. disincentive to high-yield issuance stimulate A further practical issue is the compared with the US, because it liquidity of high-yield debt that is in gives less power to unsecured cred- investor demand default. There is an active secondary itors to influence the outcome of an market in distressed debt in the US insolvency. corporate insolvency in the UK and which enables holders of defaulted Most issues of high-yield debt the US). Receivership allows a high-yield bonds to realise their are unsecured or secured by a secured creditor to appoint a receiver value and thereby transfer responsi- second charge over certain of a to recover his loans by realising his bility for deciding the terms of a borrower’s assets. In the event of security. The courts are not involved capital restructuring to specialist a borrower becoming insolvent, in the appointment. intermediaries. holders of high-yield debt will rank Chapter 11, by contrast, is initi- Amarket in distressed debt has behind preferred creditors (such as ated by an application to the US emerged in the UK in the past five

67 HIGH-YIELD DEBT

years, and it has played a role in to favour bond investment generally a number of recent extra-statutory rather than high-yield instruments in workouts for companies which had particular, there could nevertheless encountered financial problems. be benefits to the high-yield market. But some argue that information Similarly, the abolition of repayable is typically not as widely available tax credits announced in the Budget in UK workouts as is in the US, is likely to favour investment in because much of the negotiation bonds by pension funds. takes place between a company’s Of course, the high-yield market bankers. To the extent that that is by its very nature brings risks as true, it will tend to obstruct reliable Some argue that well as benefits. Issues may, for price formation in the secondary example, be mis-priced so ex-ante market, and so lower its liquidity. information spreads do not compensate investors for the default risk they are taking. Outlook for the UK is typically not This could be exacerbated in the A number of factors are working to as widely early stages of the market by low change UK investor attitudes to liquidity. Certainly, issuing houses credit risk, which could provide available in UK and investors in this paper would impetus to the high-yield market need to develop the credit skills to here. It is often remarked that low workouts as in price and assess this type of security interest rates prompt the search for properly. higher yields. The prospect of a the US, because So there are potential risks to prolonged period of low inflation financial intermediaries both in may increase institutions’ appetite much of the holding high-yield bonds and in for credit exposure as a source of underwriting them — either at issue higher returns. negotiation or (by committing to provide Similarly, EMU may stimulate liquidity to the market) in the investor demand: it will remove takes place secondary market. If the secondary opportunities to earn returns through between a market does not function efficiently, exposure to currency risk; and a it may be difficult to value holdings consolidated Euro market would company’s of this kind of paper — or to assess provide greater scope to diversify the mark-down which might be credit risk. It might also create bankers associated with a rapid sale. greater secondary market liquidity. Nevertheless, high-yield debt Separately, maturing pension would plainly add to the range of funds are widely expected to increase financing options available to UK their bond holdings. Also various companies and, if such a market can legislative changes, such as the develop here, it would eliminate the Pensions Act which came into force hedging costs and other uncertain- in April 1997, contain a number of ties created by issuing such paper in features that could impact upon the the US market — where UK investment strategy of pension companies inevitably also suffer funds. Though on balance these tend from poorer name recognition.■

68 After a lengthy dialogue with REGULATION OF CUSTODY IN THE UK the industry, investors and other regulators, including the Bank of By Liz Murrall, the Securities and Investments Board England, we consulted on proposals for regulating custody that have Following a review by the Securities and Investments Board of the now resulted in: regulation of custody, and enactment of the necessary legislation, anyone who • An amendment to the Financial provides custody services in the UK must, from 1 June this year, be authorised Services Act (FSA) which, from under the Financial Services Act. The SIB review was prompted by a variety 1 June 1997, means that any of factors, including the Maxwell debacle, implications of EC Directives and person that provides custody developments in both the domestic and international securities markets. needs to be authorised to do so. • New standards for the regula- In developing proposals for a regu- tion of custody. These indicate latory framework, the Securities and what the SIB will regard as Investments Board (SIB) wanted to “adequate” investor protection cover the main risks associated with in the area of custody. The stan- custody. These are: dards set out the safeguards • Theft or other loss of investors’ which the SIB expects each of assets resulting from manage- the regulators under the FSA to ment or staff having unautho- implement. The standards do rised access to investments or not relate to authorised firms overriding control procedures. directly. Each FSA regulator had • Misuse of investors’assets where to give effect to the standards the custodian’s own investments through their own rule books. are mixed with the clients’, so that the latter are at risk of being Authorisation of custody lost or misused. Under the FSA, unless specifically • Legal uncertainty where there is exempt, any firm which conducts inadequate legal documentation investment business in the UK is of the custodian's responsibility required to be authorised and is to its clients. (If roles are clearly regulated for its investment activi- defined then fraud is less likely ties. Custody was originally excluded to flourish and the confusion from the scope of the FSA. However, and cost of sorting matters out as custody tends to involve associ- can be reduced.) ated activities which are authorisable • Inadequate record-keeping so under the FSA, most providers of that it is not clear who owns these services were in any event which investments. authorised. The FSA regulators • Bad administration such as the monitored a firm’s custody operation failure to account for income when the investments were held in from investments or to handle connection with the firm’s invest- corporate actions, such as proxy ment business. To a certain extent, voting and exercising rights. the FSA regulators also monitored

69 REGULATION OF CUSTODY

in an indirect way custody services authorisable levelled the regulatory believed that making custody provided to authorised firms by playing field and ensured that this authorisable would ensure that third parties. The SIB considered third-party custody was regulated investors in these circumstances that there was a case for the FSA directly. were similarly protected. It would being amended to make custody an also mean that a private investor authorisable activity in its own could claim on the Investors right. It published a consultative Compensation Scheme in the event paper in August 1995 in which it of default by an authorised custodian. put forward a three-fold argument. First, whereas a person who Industry opinion took deposits was required to be Arguments for These arguments for making custody authorised under the Banking Act authorisable found support among and was supervised by the Bank of making custody most of the respondents that were England, anyone could provide consulted. custody services and hold assets authorisable A SIB analysis showed that without needing to be authorised or found wide- the additional costs associated with supervised. It was not felt that the making custody authorisable would services of custodians were inherently spread support be small and that the cost-benefit hazardous — custodians usually have balance was favourable. In partic- sound management and control and analysis ular, as most custodians were their operations well — but the already authorised by virtue of value of assets held in custody was showed that the their other activities, few firms such that the loss involved in a were likely to need authorisation for disaster could be vast (the top seven additional costs the first time. (This analysis has custodians in the UK hold some been supported by events — only £700bn of domestic assets in would be small six firms have become authorised custody). exclusively for custody business Inevitably, the collapse of and that the since the legislation was changed Barings heightened concern about cost-benefit and custody was made an authoris- the risks associated with custody. able activity.) The assets that Barings held in balance was Furthermore, as an FSA-autho- custody were not lost but for some rised firm’s custody operation was time they were frozen and could not favourable already monitored by the FSA regu- be returned to investors. lators when it related to the Second, although investments investments of its investment busi- in custody frequently fell to be ness customers, the main increase regulated, there was a gap in circum- in costs would relate only to the stances where a firm, whether third-party custody of such firms. authorised by the FSA or not, As these firms typically manage provided stand-alone or third-party The third argument was that a their investment business and third- custody, and investments were held gap existed where the investor, party custody operations in a other than in connection with its rather than an authorised firm, similar way, these additional costs investment business. Making custody appointed the custodian. The SIB were likely to be small.

70 REGULATION OF CUSTODY

Following further consultation cated that a firm would be safe- intangible assets (such as share by the Treasury on the details of the guarding where it: certificates); or amendment to the FSA, the neces- • has physical possession of • protects the integrity of intan- sary change to the legislation was tangible assets (such as gold); or gible assets not evidenced by a approved by Parliament and, on • has physical possession of docu- physical document (such as 1 June 1997, it became a criminal ments which are evidence of shares in CREST). offence to carry on “custody” in the UK without being an authorised or exempted person. Box 1 CUSTODY AS AN AUTHORISABLE ACTIVITY Paragraph 13A of Schedule 1 of the Act reads as follows: Authorised activities Custody of investments What precisely is meant by “custody” (1) Safeguarding and administering or arranging for the safeguarding and administra- and what activities are authorisable? tion of assets belonging to another where: To help firms decide whether or (a) those assets consist of or include investments; or not they needed to be authorised (b) the arrangements for their safeguarding and administration are such that those for their custody business, the SIB assets may consist of or include investments and the arrangements have at any issued guidance (The Securities time been held out as being arrangements under which investments would be and Investments Board, Guidance safeguarded and administered. Release 5/97, June 1997, Custody (2) Offering or agreeing to safeguard and administer, or to arrange for the safeguarding of Investments under the Financial and administration of, assets belonging to another where the circumstances fall Services Act 1986) on our interpre- within sub-paragraph (1) (a) or (b) above. tation of the amendment to Notes: Schedule 1 of the FSA. In the SIB’s opinion, a firm will be “safeguarding” assets consisting of or including In summary, custody is now investments in cases where the firm: covered by the FSA where the (a) has physical possession of tangible assets (such as gold); or assets consist of or include, or may (b) has physical possession of documents evidencing intangible assets (such as share consist of or include, investments certificates), or (see Box 1). A firm is carrying on (c) protects the integrity of intangible assets not evidenced by a physical document (such custody if it holds a portfolio of as shares in CREST). assets which include paintings, cash In the SIB’s view, “administration” will include any one or more of the following or land provided the portfolio also services provided by a firm for its customers, though this is not meant to be an exhaus- contains, or may from time to time tive list: contain, investments as defined by (a) maintaining accounts with clearing houses; the FSA. (b) settling transactions in investments; The new legislation defines cust- (c) operating through depositories or sub-custodians in the United Kingdom or else- ody as consisting of the safeguarding where; of assets combined with their admin- (d) operating nominee accounts, including pooled accounts, which identify each istration. customer’s assets in a ledger; The Treasury did not seek to tie (e) cash processing associated with customers’ assets; down the definitions of either (f) collecting and dealing with dividends and other income associated with the assets; safeguarding or administration in (g) carrying out corporate accounts such as proxy voting (including exercising rights the legislation. The Securities and conferred by an investment on behalf of the beneficial owner). Investments Board’s guidance indi-

71 REGULATION OF CUSTODY

We also give some examples of The legislation covers not only activities which we understand will the provision of custody, but also be included within “administra- the arranging of custody. Firms may tion”, such as: settling transactions, not avoid authorisation by struc- operating nominee accounts, cash turing their business so that one processing, collecting and dealing entity is responsible for the safe- with dividends and appointing sub- guarding of investments and another custodians. for their administration. The SIB Firms do not need to be believes that whenever assets are providing all these administration safeguarded and administered, Firms may services to be carrying on custody someone will always be responsible business. A firm could be carrying for arranging the safeguarding and not avoid on custody when, in safeguarding administration. Consequently, the assets, it performs any aspect of firms which only carry out adminis- authorisation administration. tration and arrange for their nominee by structuring The two activities of safe- to do the safeguarding will be guarding and administration must carrying on a custody business. their business so be combined — but, pure safe- keeping operations (such as the Regulation of custody that one entity holding of jewellery, wills, title The standards for custody set out deeds or share certificates in locked the safeguards the SIB expects each is responsible boxes), where there is no admi- of the regulators under the FSA to nistration, are not required to be implement and to ensure that their for the authorised. Likewise, a solicitor authorised firms comply. The stan- who merely holds share certificates dards address the risks identified safeguarding of for an investor in a safe does not with custody by reference to: need authorisation for this activity. responsibilities, segregation, pro- investments The legislation contains a tection and identification and recon- and another number of specific exclusions. ciliation. Nominees as registered holders of Responsibilities for their investments, for example, will not Key responsibilities must be agreed be carrying on custody business in writing between the parties to the administration where an authorised firm which is custody arrangements, ie the firm permitted to carry on that business providing custody and its customer. takes responsibility for them. Foreign (For example, it must be clear who exchange business and transitory will be responsible for appointing safekeeping (such as where inde- sub-custodians, for reviewing the pendent financial advisers receive latters’ performance and for losses policies from product providers to of customers’ investments.) pass on to investors) are also When investments are held in excluded. Furthermore, holding a nominee company, the authorised investments as security against firm must always accept responsi- loans is not custody. bility for its nominee. An investor

72 REGULATION OF CUSTODY

would, therefore, have a right of indirect controls over the selection action against the firm for any and appointment of global custo- losses arising as a result of its dians are of obvious importance. nominee’s behaviour. If the firm These indirect controls are were then to default on its liabilities, achieved through three main means. a private investor could expect • First, a firm must exercise due to be able to make a claim on the skill, care and diligence when Investors’ Compensation Scheme. selecting and appointing a third- Segregation and operational party custodian. safeguards • Second, the types of institution The standards aim to improve oper- that can act as a custodian are ational safeguards by preventing set out — the custodian must be Enhancing customers’ investments from being an institution that is regulated or mixed with those of the firm and of supervised or subject to some r egulatory other customers. This decreases the form of independent scrutiny. standards likelihood of fraud and misuse. • Third, the custodian needs to Protection against loss agree to contract to provide a should assist Custodians must have arrangements particular level of service. For in place to ensure that customers’ example, when a third-party in discouraging assets are securely held. They custodian is used, it is important should ensure that they act only on that it has adequate procedures fraud in instructions given in accordance in relation to its custody opera- with what has been agreed with the tion. handling customer. Identification and reconciliation Conclusion investments Custodians must maintain records The new legislation provides a to identify the assets and entitle- more secure regulatory framework and increase ments of each of their customers. for dealing with third-party or the likelihood Where assets are held as collateral, stand-alone custody and also with these must be identifiable and new, entirely unregulated firms of its early customers should receive a record arriving in the UK. Furthermore, of their investments at appropriate enhancing regulatory standards detection intervals. should assist in discouraging fraud The standards apply not only in the handling of investments and where firms provide custody direct increase the likelihood of its early but also where they delegate custody detection. to third-parties. Where custody is Firms should generally be able delegated, and where the third-party to meet these standards since most custodian is carrying on business custodians were already authorised outside the UK and thus outside under the FSA for their investment he reach of UK regulation, the stan- business and the standards for dards rely on indirect controls. custody codify what the better firms Since custody is a global business, were already doing in practice. ■

73 General of Fair Trading. Unlike the MARKET MANIPULATION Office of Fair Trading, however, whose interests in this area are By Guy Sears, the Securities and Investments Board principally economic, the Financial When addressing market manipulation what matters is the appropriate Services Act regulators do not always regulatory response. Some responses are designed to preserve and protect, act solely because of considerations on a real-time basis, the integrity of the market; others to of economic efficiency. identify and discipline any market participants whose lack of integrity The requirement for firms to damages that market. act with integrity carries with it considerations of moral account- Manipulation damages markets. It ability. Regulatory responses which distorts prices, which may lead seek to provide proper protection directly to economic inefficiencies for investors and uphold the integrity and resource mis-allocation; it of markets, at times address defects allows transactions to take place at in the trading environment and at prices which do not reflect a true other times judge whether anyone is balance between supply and demand, to blame. so some market participants benefit Under the current system not unfairly at the expense of others; all those who use markets, let alone and it may reduce confidence in participate in them, are subject to both the market affected and in the same sets of standards or regu- markets generally, and so discourage lations. This can therefore limit the their use and indirectly damage the effectiveness of regulatory responses economy. which seek to protect the environ- Some manipulative events on ment; however, careful consideration their own may not damage the of the design of contracts and the use market environment. If they became of market surveillance have a role in widespread, however, these activi- stopping manipulation before it ties could damage the environment. starts. Therefore, regulatory responses must include disciplinary measures that Markets’ functions are designed to deter repetition Securities and futures markets are in the interests of preserving the places where people come together integrity of the market. In this way to trade investments. The trade in there are regulatory responses to some investments is driven by a manipulation which distinguish desire to buy something cheap and between damaging situations and sell it at a profit. Other trading may abusive participants. be driven by the wish to redistribute This article focuses on the risk in an efficient manner. manipulation of financial markets. Reliable pricing is important — Manipulation of markets is a more it inspires user-confidence. Markets general issue and in other contexts will exist where potential buyers falls within the remit of the Director and sellers of a product compete in

74 MARKET MANIPULATION

an environment that provides them movements) depend on the relia- with access to information about the bility, and the perceived reliability, product. of the prices used. There are two types of needs Markets should be a focus of that markets can answer. The first is reliable information about a product. to provide an efficient means of The trading activities of participants determining the price of the product should be a reliable indication of the traded — it may be a share, a bond value placed on the product. If the or the price of an underlying information received by the market commodity. For example, the price is not reliable or the trading activi- If the of copper is fixed by reference to ties of participants are thought to be the price at which it is trading on the artificial, then confidence in the information London Metal Exchange. price and the market will diminish. Markets also enable risk to be Assets will be mis-priced, some r eceived by transferred. This is the second need participants will stay out of the that a market answers. Those who market altogether and others will the market is are averse to risk may transfer it to spend more money conducting their not reliable those who are tolerant of it. own enquiries and double checking Consider a manufacturer of catalytic the validity of statements made. or the trading converters for cars. The manufac- The volatility of prices may turer wants to know the profit from increase as rumour, counter-rumour activities of each converter. If a producer of and fear dominate the price-forma- converters is to supply a car manu- tion process. participants facturer at a fixed price over the As volatility and prices increase, next year, it must try to ensure that the producer of catalytic converters are thought to the price of the palladium or plat- would find it more expensive to inum used in the converters will not transfer the risk of price movements be artificial, vary. The futures market enables the to the market. Since the changes in producer to hedge the risk of price price movements are increased, the then confidence movements by buying certain futures producer would have to pay more to in the price and contracts. guard against them. At its simplest, in return for an the market will agreed fixed premium the producer Markets and exchanges is no longer exposed to a price The label “market” is often used diminish movement which could wipe out his interchangeably with “exchange”. profits. Someone in the market who They are not synonymous but, for assesses the risk differently, has convenience, this article uses the indemnified him against such term “exchanges” to refer to the movements. organised and supervised environ- Both functions (to provide an ment in which certain markets have efficient means of determining the their principal existence, and it price level of products and to concentrates on those markets rather transfer efficiently the risk of price than markets generally.

75 MARKET MANIPULATION

Regulators are concerned about is in practice decided not only by the general economic damage that considerations of economic effi- manipulation causes through the ciency but also those of ethics. mis-pricing of assets and the mis- Norman Barry1 said of insider allocation of resources. They also dealing: “The real controversy is aim to protect investors. These not about the need for some utility- complementary concerns have led independent moral rules..., but to two approaches to market manip- about their precise content and the ulation. The first seeks to address domain over which they are The rules and the interests of the environment, the enforceable.” second to address the conduct of In considering what is and what practices of the entities within the environment. is not acceptable market conduct in The nature of regulation and the the UK, care should be taken to exchange must structure of the Financial Services distinguish between regulatory Act 1986 recognises these concerns responses which seek to maintain ensure that and approaches and their use in orderly markets and reduce econ- business analysing and evaluating the nature omic damage, and those which of activities. For example, invest- involve an assessment of the moral conducted by ment exchanges recognised under culpability of the participants. Section 36 of the Financial Services Where manipulation is alleged means of its Act 1986 must satisfy the require- to have occurred, proving intent can ments of Schedule 4. One of the be extremely difficult. Inferences facilities is obligations under Schedule 4 is may be drawn from the surrounding found in paragraph 2 (1) under the circumstances but the stigma of conducted in an heading Safeguards for Investors manipulation ensures that the accused and states: “The rules and practices will criticise a system that judges on orderly manner of the exchange must ensure that inferences that are really leaps of business conducted by means of its faith. and so as to facilities is conducted in an orderly Often such disputation is not afford proper manner and so as to afford proper the primary concern of regulators. protection to investors.” Steps may be taken, for example, to protection to There is a proper market require- ensure price or position limits are ment in paragraph 2 (2) and both of imposed in order to re-balance the investors these are supported by paragraph 5 environment. These steps are needed under the heading, Promotion and whether or not anyone is to blame. Maintenance of Standards, which (If a traffic jam is the result of an commences: “The exchange must accident caused by a dangerous be able and willing to promote and driver, it is right that dangerous maintain high standards of integrity driving be deterred through the and fair dealing in the carrying on of driver being held to blame.) Markets investment business ...” may be disorderly merely because What constitutes market manip- of the diverse competitive interests ulation and the appropriate response, of human beings.

76 MARKET MANIPULATION

Manipulative activities Many strategies involve both or its participants from one or two One classification (see Box 1) divides information-based and trade-based of these trades will usually be unwarranted activities into two manipulation. The crudest forms of extremely low. But the preservation classes: information-based abuses “boiler-room” operations involve of orderly markets and the proper and trade-based manipulations. pressure-selling using misleading protection of investors would be put Other types of market abuse are information. The manipulators may at risk were such practices to insider trading and front-running. be trading among themselves to become widespread. Information-based abuses are enhance the appearance of interest Quite apart from any economic those in which the manipulator in the share and support its price. damage, most market participants creates a false market by dissemi- Any damage to an organised market expect their peers to act with nating false or misleading information. Disseminating such information Box 1 MANIPULATIVE MARKET CONDUCT — EXAMPLES via, for example, internet discussion groups, to “talk up” share prices, is (A) Trade-Based Manipulations universally considered manipulative. (1) Generalised (with no specific time window) Trade-based manipulation is Actions intended to create a shortage of stock: characterised by the manipulator • corners leading others to believe that trades • squeezes are or should be occurring in certain Actions intended to create an impression of false activity for example to ramp: volumes or at certain prices and to • wash trades treat that information as if it was a • matched orders bona fide reflection of the needs and • support operations beliefs of market participants. If a • firms buying own shares (or using concert parties) to raise price or person conspires with another to prevent it falling carry out 20 or 30 trades at ever- (2) Time-Specific (manipulations within a specific time frame, eg related to a increasing prices and everyone can contract): see the activity, it is likely that an • moving share price on a given day to a trigger bonus clause in remuneration innocent party may buy from the contracts first at or around that price. • trading specifically to interfere with the price or settlement of derivative There are restrictions to the contracts creation of false or misleading • trading to influence the particular spot price for a stock that had been agreed appearances of active trading on all as determining the value of a transaction the world’s exchanges. In some • marking the close (changing prices at market close). cases the restrictions are particu- larised. Trades in which there is no (B) Information-Related Manipulations real change in beneficial ownership, (1) False or Misleading Information: where prices are pre-arranged and • making untrue statements of material facts not obtained through the price- • spreading false rumours to induce buying or selling by others discovery processes, or where private (2) Failure to release information : agreements are reached to indem- • non-disclosure of material facts (including material interests) nify buyers against loss, such that they do not take on any market risk, This classification was developed by Francis McGee in the Bank’s Regulatory Policy Division. are commonly seen as wrongs.

77 MARKET MANIPULATION

integrity. A false trade will be seen released it drops, not back to the and reacted to as morally wrong. equilibrium price, but a little below From any view, therefore, they are that price, before it returns to the manipulative. equilibrium level. This dip below Classifying another trade-based the line is called, “burying the activity as manipulative — an activity corpse”. It is caused because supply which results in a squeeze or corner briefly exceeds demand as the — generates much debate. The terms It is difficult efforts of those who needed supplies squeeze and corner are used almost diverted the commodity from other interchangeably, although this is to argue that the users into the delivery mechanisms technically incorrect. Since in prac- of the market. tice the one rarely comes without intention of a Should those who participate in the other, I shall use the term the squeeze be seen as morally squeeze. person creating culpable? To what extent, if any, Before characterising a squeeze a corner or do squeezes damage the efficiency as “manipulative”, consideration of markets? must be given to its potential for squeeze is The answers to these questions economic damage and its impact are important because there are from a utility-independent ethical to induce or different regulatory responses to stance. Only then can the appro- situations which damage market priate regulatory response in a inhibit trades, efficiency and to those which are particular market be decided upon. moral wrongs. These can be seen in A squeeze happens when the given shorts will the provisions of statute, regulation owner of a huge amount of a and guidance. commodity goes into a futures already have market and buys contracts requiring Legislative environment people to deliver that commodity at contractually Section 47 FSA 1986 includes the a specified price in the future. Those committed to nearest equivalents to criminal who are bound to deliver discover offences of manipulation. One is the they may not be able to. Their trades with dissemination of false or misleading efforts to obtain the commodity information about investments to drive up its price. The holder of the the longs induce others (broadly) to buy or commodity stock then slowly leaks sell or exercise or refrain from exer- it out at the new high price to those irrespective of cising rights connected with the who have to sell it back at the lower investment. price which had been agreed on the the manipulation The other offence is creating a futures market some months before. false or misleading impression of During a squeeze, a graph of the market in, or price or value of, the price against time generally has investments when done for the its own characteristics — similar to same improper purpose. the representation of a heart beat. Criminal offences under Section There is a spike as the price climbs 47 require certain events and the rapidly and then as the pressure is existence of specified intentions or

78 MARKET MANIPULATION

purposes on the part on the wrong- should observe high standards of doer. The range of wrongful market conduct.” purposes, while certainly not narrow, There is no specific require- has been argued by others to be ment under the FSA for firms to insufficiently wide. In his December become authorised persons by 1996 paper, Manipulations of An essential reason of their being members of an Metals Futures: Lessons from organised market (whether or not it Sumitomo, Professor Christopher characteristic is an exchange). Gilbert2 writes that it is difficult to The need for authorisation argue that the intention of a person of exchanges depends on whether or not a firm is creating a corner or squeeze is to conducting investment business, induce or inhibit trades, given shorts is that the and many firms whose only busi- will already have been contractually ness involves trading their own committed to trades with the longs parties to funds, will neither be nor need to be irrespective of the manipulation. authorised persons. Indeed, in the Whilst I do not propose to enter the transaction commodity markets, even those into a debate about Section 47, it are not entitled which deal for clients may not be cannot be denied that it stands out, authorised persons because they are in contrast with the provisions to retain not dealing in investments but the mentioned below, because of its underlying commodity, for example application to any person and ownership in copper or cocoa itself. regardless of their status as a For this reason, in any organ- licensed or recognised entity under of certain ised market a number of the users the Financial Services Act. will not be subject to the SIB information Principles in carrying on their busi- SIB principles ness. This limits the effectiveness of In discussions about market manip- about price and the statements of what constitutes ulation, reference is often made to high standards of market conduct, the “SIB Principles”, in particular volume — the and at times a stricter approach will Principle 3. exchange be applied to the activities of regu- The 10 Principles, were promul- lated firms than to others. gated by the SIB in 1990. They are appropriates Broadly, the requirement for intended to form a universal state- authorisation exists where activities ment of the standards expected and such knowledge involve direct transactions with they apply directly to the conduct of investors. In this way, investor investment business and to the and makes it protection issues are adequately financial standing of all authorised addressed in relation to entity-based persons. available to all risks (those arising from the rela- Principle 1, Integrity, states: “A tionship a firm has with its firm should observe high standards customer). of integrity and fair dealing.” More difficult is the environ- Principle 3, which deals with market ment-based risk (where the activities practice, commences: “A firm of someone else drive prices away

79 MARKET MANIPULATION

from a true reflection of supply and conduct as required under Principle the result of improper trades may be demand estimations and predic- 3, which, as explained, will only to call into question the recognition tions). The investor’s broker may apply to those who are authorised requirement for the proper protec- get the best traded price, but it may persons. tion of investors or the maintenance not be as good a price as it would This further exemplifies the of a proper market, for example by have obtained had the manipulative current legislative systems which giving rise to misleading impres- activities not been present. imposes the law on everyone — the sions as to the value of a particular There is therefore a need to SIB Principles on the authorised contract. The guidance stated that in improve the environment. Organised firms, whether exchange members such an event, it would be appro- markets, whether involving anony- priate for the exchange to take mous trading or not, are more than action to end the improper market just the aggregation of a number of situation. bilateral contracts. Buyers and There will always The first approach (relating the sellers are not merely brought appropriateness of the activity to together, nor is information. The be a temptation Principle 3) addresses the actions of information that is generated by the firm and its effect on others, and their trading is essential to the price for those who through disciplinary process seeks discovery process. to deter activities that are considered Were I to sell you my motor- would like the unacceptable. The second approach cycle, the price would be a private (which relates to the appropriate- matter between us. But an essential financial futures ness of the activity to proper characteristic of exchanges is that contract to be markets) addresses the effect on the the parties to the transaction are not trading environment and, through entitled to retain ownership of more valuable an intervention process, seeks to certain information about price and protect that environment. volume. The exchange appropriates to try to alter the such knowledge and makes it avail- Prevention able to all. Exchanges have rules price of the Leaving the problems of how which govern the activities of all authorities can effectively regulate exchange members whether or not underlying market users (that is those whose they are authorised persons. activities can pollute the market) if An exchange member’s activi- index they are not exchange members or ties can be scrutinised and subjected authorised persons, this final to exchange rules designed to section deals with the exchange promote efficiency and for investor environment, design of contracts protection. The guidance release of or not (the latter would include fund and disciplinary processes. 1993 issued by the SIB, Proper managers) and the exchange rules Regulatory authorities around Trades in Relation to on-exchange which apply to all, but only, exch- the world have learnt how impor- Derivatives, addresses exchange ange members, whether authorised tant are issues of contract design, participants and those who deal firms or not. market surveillance and market directly or indirectly through them. The guidance also related to the sanctions in international markets. It shows how firms may fail to statutory recognition requirements Where prices of futures contracts observe high standards of market for exchanges themselves. It noted are set by reference to some other

80 MARKET MANIPULATION

financial instrument or index at a - complements but is not a market situation4 and, if necessary, certain time on a certain date, there substitute for an appropriate to punish abusive conduct. will always be a temptation for market surveillance prog- One area which has received those who would like the financial ramme. less publicity is the design of exch- futures contract to be more valuable • That an active and effective ange disciplinary processes. This to try to alter the price of the under- market surveillance programme can no longer be considered in rela- lying index. Contracts have been by the market regulatory auth- tion to a particular exchange as altered to make it more expensive ority: certain instruments are traded in and risky to try to do that, for - is essential to ensure that more than one exchange; they are example, by altering the manner and commodity futures markets also traded over-the-counter as well timing of the calculation of the operate in a fair and orderly as in cash markets. So those respon- index. To hold the price for two manner; sible for disciplinary procedures minutes may cost a lot of money on - should be designed to detect, must be informed of the treatment some markets; to hold that price for to prevent, to take corrective of faults and abuses in these envi- an hour may be impossible. action with respect to, and to ronments. But in commodities, the char- punish abusive conduct ... Even if rules are, on the face of acteristics are different — particularly • That regulatory measures which it, identical, if different exchanges when supply is limited, subject to facilitate the identification of have different enforcement policies, relatively high production, trans- large exposures should be devel- or if the level of fines for default port, storage or delivery costs, or oped. These measures may differs between exchanges, one subject to seasonal shortages or involve access to information exchange may become more attrac- long production lead-times. relating to persons holding or tive to the manipulators than another. The London Communiqué,3 controlling such large exposures International agreement on these recognised both the distinction and their related derivatives, matters would be complex but it is between situations which damage over-the-counter and cash market hoped that, at least within the UK, the market and those which involve, positions. These measures may further thought can be given to this in terms of the Communiqué, a also involve access to informa- area. necessity to punish abusive conduct. tion on deliveries. The Market Conduct Group In matters of contract design The authorities also agreed to will consider many of these matters and the market environment, discu- promote those matters concerning with a view to publishing a discus- ssed in the Communiqué, under the contact design and the adoption of sion document to open debate heading Market Surveillance, the rules and procedures that authorise among market participants and the authorities reached the following a market authority to intervene in a public generally.■ points of consensus: • That the proper design of com- NOTES modity contracts: 1 Insider Dealing — An Exploration into the Existing Law and Practice by Professor Norman Barry, - not only enhances their published by Foundation for Business Responsibilities issues paper No 1 June 1996. 2Professor of Applied Econometrics, Queen Mary and Westfield College, University of London. economic utility but is also a 3On supervision of commodities futures markets issued by representatives of regulatory authorities of critical aspect of market 17 countries responsible for supervising commodities futures markets which met in London on 25-26 November. integrity in that proper design 4 An article which questions the efficiency of interventions generally (though not about the London reduces their susceptibility Communiqué) is “Squeezes, Corpses, and the Anti-Manipulation Provisions of the Commodity to market abuses, including Exchange Act”, by Craig Pirrong in CATO Regulation (http\\www.cato.org/pubs/regulation). The article also contains a more detailed description of market conditions in a squeeze. manipulation;

81 These developments prompted GLOBAL INSTITUTIONS, NATIONAL the Group of Thirty to commission a study of the management and SUPERVISION AND SYSTEMIC RISK supervision of global financial insti- tutions and the potential for systemic By John Heimann, Merrill Lynch risk. The Group had done substan- Lord Alexander of Weedon, NatWest Group tial work aimed at reducing risk in securities and derivatives markets, A study by the Group of Thirty — a Washington-based body of first in its 1989 report Clearance academics, bankers and government officials — into the management and Settlements in World Securities and supervision of global financial institutions and the potential for Markets and then in Derivatives: systemic risk, found that an industry initiative is needed to promote Practices and Principles in 1993. consistently high standards of risk management and control for global However, neither study focused institutions. on the safety and stability of the financial system as a whole. It was The threat of serious disruption to consider more fully the implica- to the international financial system tions of the emergence of global may be small, but it is nonetheless a institutions and markets that, two serious concern. Rapid changes years ago, we agreed to co-chair a occurring in the international finan- new study group. We invited a cial system have resulted in new distinguished group of financial sources of, and transmission mech- executives, senior supervisors and anisms for, systemic shocks. The experts from the law and academia institutions active in international to take part. markets are becoming larger and It should be noted that our more complex. A rapidly-growing deliberations took place against a volume of transactions and an background that was far from calm. expanding array of new products Barings collapsed under the weight are moving across borders at ever of losses by a rogue trader in 1995. faster speeds. Many other financial institutions New entrants to the system, have suffered trading losses arising often from outside the G-10 coun- from a variety of sources in the tries, are less well known to the intervening years. The largest cases international financial community involved accumulated losses from and may be weakly supervised or not bond and commodities trading that supervised at all. Indeed, the global had been successfully hidden by operations of major financial insti- traders. Other cases involved lesser tutions and markets have outgrown failures of internal controls and risk the national accounting, legal and modelling and monitoring. Taken supervisory systems on which the together, they generated a high level safety and soundness of individual of official concern. institutions and the financial system Most notably, the safety of the rely. international financial system and

82 GLOBAL INSTITUTIONS

the need to strengthen supervision Sixty-six institutions, about 75 party risk management was based have become regular topics at G-7 per cent of the sample, responded to on very thin information. There was Economic Summits and featured in the survey. The respondents (Box 1) almost unanimous support for action the last three summit communiqués. represented a reasonable cross- by industry and further steps by The Basle Committee and section, not of the financial services supervisors to strengthen the inter- IOSCO have redoubled their efforts industry overall, but of global finan- national financial system. to improve international supervision cial institutions. Confident that our concerns and co-ordination and have expanded Many survey questions were were justified, the study group set to their co-operation to include insur- general and impressionistic; and the work. Our focus was to be on ance in the pursuit of enhanced answers lent themselves to a range shocks that could cause interna- supervision of financial conglomer- of interpretations. Therefore, the tional systemic risk, whether arising ates, now formalised in the “Joint study group conducted follow-up from the spillover of domestic Forum”. interviews with a dozen institutions shocks or disruption of institutions, The Barings collapse triggered in Europe, the US and Japan. markets or the clearing and settle- the most specific action. Beginning ments mechanisms which now link with the Windsor Declaration in Serious disruption domestic financial systems. May 1995, supervisors of the world’s While the survey yielded a wide Looking back through recent major futures and options markets range of information, several results history, the failure of Bankhaus have expanded their co-operation were particularly noteworthy. First, Herstatt in 1974, the LDC debt and information sharing, especially respondents not only agreed that a crisis of the 1980s, the Mexican with regard to large exposures. New serious disruption of the interna- peso crisis of 1995 and others repre- agreements have been signed among tional financial system was possible, sented shocks to the international exchanges and among their super- but rated the likelihood at one in financial system. Yet none brought visors to strengthen markets and five over the next five years. down the system and each was market infrastructure. At the same time, respondents successfully resolved with the aid of While these steps have offered generally doubted that such a shock some sort of official action, ranging some reassurance, the study group would threaten widespread disrup- from official financial support to remained convinced that more was tion of the financial system because collaboration among central banks needed than the enhanced co-oper- of what they perceived to be the and market participants to prevent ation embodied in these various vitality of the major institutions and gridlock in foreign exchange markets. supervisory initiatives. the ability of central bankers and However, the growth in size, To assess if these concerns supervisors to limit damage and velocity and complexity of interna- were widely shared among financial calm markets. tional transactions, and the higher institutions, the group sought the Despite this expressed confi- concentration of trading activity in assistance of KPMG Peat Marwick dence, the firms had concerns about a relatively small number of institu- both in designing and conducting a the quality of risk management in tions that play a leading role survey of global financial institu- areas such as internal control, risk- in multiple markets, suggest that tions. A questionnaire was prepared measurement systems, legal risk, regulators will find it increasingly and sent to 90 financial institutions, the efficacy of risk-reduction tech- difficult to improvise effective crisis- most of them large global banks but niques such as collateral, and the management in the event of a shock including a selection of large invest- safety of clearance and settlement occurring. The threshold of concern ment banks, insurance firms and systems. They acknowledged that for the study group was a shock that non-bank financial firms. their confidence regarding counter- would not only threaten a major

83 GLOBAL INSTITUTIONS

Box 1 SURVEY RESPONSES

Australia: Italy: Switzerland: Australian Mutual Provident Society Banca Nazionale del Lavoro Swiss Bank Corporation ANZ Banking Group Banca di Roma Union Bank of Switzerland Commonwealth Bank of Cariplo Australia Credito Italiano United Kingdom: National Australia Bank Eptasim Abbey National Westpac Banking Corp Istituto Bancario San Paolo di Torino Barclays Bank HSBC Holdings Belgium: Japan: Lloyds TSB Group Générale Banque Bank of Tokyo-Mitsubishi NatWest Group Dai-Ichi Kangyo Bank Schroders plc Canada: Daiwa Securities Standard Chartered Bank of Montreal Fuji Bank Bank of Nova Scotia Industrial Bank of Japan United States: Canadian Imperial Bank of Commerce Meiji Life Insurance American International Group Royal Bank of Canada Mitsui Marine and Fire Bank of America Nikko Securities Bankers Trust France: Nippon Life Chase Manhattan Banque Indosuez Nomura Securities Citibank Banque National de Paris Sakura Bank Depository Trust Company Banque Paribas Sanwa Bank First Chicago Crédit Lyonnais Sumitomo Bank General Reinsurance Crédit Agricole Yamaichi Securities Goldman Sachs Yasuda Fire and Marine Lehman Brothers Germany: Metropolitan Life Commerzbank Netherlands: Morgan Stanley Deutsche Bank ABN-AMRO Republic Bank Dresdner Bank Fortis Group Salomon Brothers

Hong Kong: Spain: Bank of East Asia Banco Santander Total: 66

financial institution, but could study group agreed that such a situ- — defined as large, internationally cascade through the international ation could not be ruled out. active commercial banks, which are financial system threatening addi- One reason for this is the emer- major participants in large-value tional major institutions and, in turn, gence of large integrated financial payment systems, along with the the financial infrastructure of the firms with corporate structures and largest investment banks which are entire international system itself. finances of extreme complexity and key participants in the clearing and While past shocks and crises have global scope. The study group settlement systems for globally- not risen to this level, those in the focused on a set of core institutions traded securities.

84 GLOBAL INSTITUTIONS

Core institutions do not include and regulations of each. The sudden place may increase the risk of a large insurance companies or large collapse of a large, internationally misjudgement. finance companies, even those that active participant in a payments are highly active in international system would, for example, cause Market discipline markets. Although these institutions unexpected difficulties for all of its The study group concluded that are important by virtue of their size, counterparties if agreed netting confronting these challenges succ- they present substantially less risk arrangements prove not to be legally essfully will require improvements to the system than the failure of the binding — even a netted claim may in management, market discipline core institutions of which they are not be collectible in an emergency. and supervision. Large, multi-juris- customers. On the positive side, improve- dictional, financial firms which have Core institutions tend to be well ments in technology are making complex corporate structures and capitalised and have their headquar- management information and control finances require sophisticated risk ters in well-supervised jurisdictions. systems and analytic models, which management on a global basis. While it makes sense for interna- help institutions manage risks, a lot Counterparties wishing to deal with tional financial activity to be more effective. Positive develop- them and supervisors charged with concentrated in a small number of ments include: their oversight need to adopt a simi- capable firms, this concentration is • Continual improvements in larly global view of these firms’ worrying given the complexity of measuring credit and market risk. operations and finances. National their interconnections. These insti- • Enhanced diversification of port- systems of accounting, reporting tutions tend to be each other’s folios across markets. and supervision, however, fall short largest counterparties, have exten- • Expanded use of netting and of this objective. The legal structure sive dealings with many of the same collateral. upon which national regulators customers and are members of the • Greater disclosure of off balance focus is no longer relevant to overall same clearing houses and exchanges. sheet risk. control of risk. Indeed, undue focus Direct and indirect risk expo- • Substantial increases in equity on legal structure could diminish the sures within this group are so capital of many major financial effectiveness of overall risk-control complicated and opaque, and change institutions. in a global group. so rapidly, that it is virtually impos- • Financial sector consolidation. Cognisant of this challenge, sible to monitor them in anything • The growth of securitisation. financial institutions have devoted like real-time. Accounting and disclo- Despite these improvements, substantial resources to improving sure practices have not begun to substantial uncertainty remains over risk management and global controls. keep pace; risk exposures can build the level and direction of risk in the National supervisors have likewise up undetected by existing moni- system and the effectiveness of focused their attention on risk toring systems. In a crisis, both peer measures to control it. Given the management, including active board institutions and regulators may feel speed with which market partici- and management oversight; the they have too little information pants can react to events anywhere capacity to measure, monitor and about the condition of a faltering in the world, reaction times in the control risks by activity; and the institution and insufficient time to event of a shock are virtually instan- adequacy of internal controls. assess this complex information to taneous. Managers and regulators The many international supervi- warrant taking action. have very little time to analyse the sory initiatives that are under way to An institution active in scores problem, formulate and implement expand co-operation and improve of jurisdictions is also subject to the a response. The sheer velocity with co-ordination are signs of progress. vagaries and interactions of the laws which international transactions take But there are limits to what can be

85 GLOBAL INSTITUTIONS

Box 2 MEMBERS OF THE STUDY GROUP

Co-Chairs: Rt Hon Lord Alexander of Weedon QC Mr John G Heimann Chairman Chairman NatWest Group Global Financial Institutions Merrill Lynch Members: Dr Ulrich Cartellieri Prof Charles Goodhart Prof Richard Herring Member of the Board of Managing Norman Sosnow Professor of Banking Julian Aresty Professor of Finance Directors and Finance The Wharton School Deutsche Bank AG London School of Economics University of Pennsylvania

Mr H Rodgin Cohen Mr Harry P Kamen Sir David Walker Banking Partner Chairman of the Board & CEO Chairman Sullivan and Cromwell Metropolitan Life Insurance Company Morgan Stanley Group (Europe) Plc

Mr Paul Collins Dr Tommaso Padoa-Schioppa Mr John Walsh Vice Chairman Chairman Executive Director Citibank NA CONSOB The Group of Thirty

Mr Tom de Swaan Mr Brian Quinn Mr Nicholas S Wilson Executive Director Chairman Legal Adviser to the Chairman De Nederlandsche Bank NV Nomura Bank International plc NatWest Group (Retired)

Mr Toyoo Gyohten Dott Mario Sarcinelli Mr Kosuke Nakahira Senior Adviser Chairman Special Adviser to the Minister of The Bank of Tokyo-Mitsubishi Ltd Banca Nazionale del Lavoro Finance Japan

Sir Andrew Large Mr Arthur Levitt, Jr Chairman Chairman Securities and Investments Board Securities and Exchange Commission

Mr William McDonough Mr Jean Peyrelevade President Chairman and Chief Executive Officer Federal Reserve Bank of New York Credit Lyonnais

achieved in this way. The group Major financial institutions A single, globally consistent, concluded that an industry initiative should lead the development of framework should apply to all global is needed to promote a consistent, global risk management principles, institutions. This framework must high standard of risk management in co-operation with supervisors start from the premise that it is the and control for global firms. and as a continuing enterprise. board and management of global

86 GLOBAL INSTITUTIONS

institutions that have the funda- likely causes of failure of a financial mental responsibility for ensuring institution: the stability of financial institutions, • Inadequate management proce- thereby limiting systemic risk. dures. Although this is not a new • Failure of internal controls. thought, assigning enhanced respon- • Actions of a rogue employee. sibility to financial institutions Second, it is patently unreason- suggests an approach that is much able to expect supervisors alone to more than the status quo. It implies keep global institutions from having that supervisors will rely to a mishaps. Even if they had the much greater extent on the institu- resources, their task cannot be to tions that they supervise, and that evaluate the quality of traders or the the institutions themselves will current daily Value-at-Risk in accept the responsibility to improve trading exotic derivative instru- the structure of, and discipline ments. The speed and complexity of imposed by, their internal control innovation in the markets, the super- functions. visors’ inevitable position “behind Furthermore, it suggests a the curve” and their real handicaps regime for global institutions that is in competing for talented staff all Lord Alexander different and more elaborate than argue for private institutions to take that imposed on smaller or less on the responsibility. geographically diversified competi- An industry framework must tors, although the responsibility of address the difficulties posed by management in the area of internal institutional complexity, market controls is no less at smaller firms. volatility and geography. Yet the There are two reasons for this greatest challenge is the excessively change in approach. First, by far the risky behaviour that is the most largest proportion of serious finan- likely underlying cause of losses to cial problems which beset financial a financial institution. Since no code organisations (whether or not they of ethics is likely to eliminate this involve systemic risk) arise from tendency, an institution’s control problems which the organisations system must at least aim to check ought to be able to control them- the excesses of human nature by selves. By way of example, the establishing an internal vigilance failures of Continental Illinois and system that will provide early Barings, and the trading losses at warning of such behaviour. Controls John Heimann institutions such as Daiwa Bank, must withstand both external shocks Morgan Grenfell and Sumitomo and internal breakdowns. Corporation could have been avoided Comprehensive and effective if these firms had had stronger controls must cover: management oversight and fully • the management structure; comprehensive internal controls. • the internal audit function; Survey respondents listed three • the risk management function;

87 GLOBAL INSTITUTIONS

• and the compliance functions, Independent audit internal controls generally and, in including legal, regulatory and Global institutions should subject particular, of internal controls over ethical review. their world-wide operations to review the preparation of applicable finan- Implementation will call for a by a single, independent, external cial statements and risk information major commitment in the areas of audit firm or group, and should on a comprehensive, global basis. management, staffing, investment agree upon more consistent and Information from the expanded in global risk-monitoring systems meaningful disclosure of financial audit should be made available to and a more sophisticated approach and risk information on a global supervisors and the public. How to many categories of risk and risk consolidated basis. best to provide this information to control. Global institutions require a supervisors should be a topic for At the same time, the study truly global audit that provides more consideration by the industry com- group is proposing that global insti- than the traditional audited financial mittee and relevant supervisors. The tutions undertake the difficult task statement. As a starting point, a extent to which supervisors rely on of devising a framework that is single external auditing firm or the reports of external auditors will appropriately aligned with supervi- group should act as the principal depend on the structure of supervi- sory requirements. auditors and report on the global sion in each country, but the The objectives and methodolo- audit of the global institution. The additional information should better gies of a firm’s management and clear goal of the board and its inform the supervisory process, supervisors do not always coincide. management should be the most even where direct examination is In some cases, they will have to be comprehensive audit possible, even the norm. reconciled. To expedite the devel- if this means changing audit firms. The industry committee should opment of a framework, the industry Nothing less should be acceptable. also agree upon a framework for and the supervisory community will Institutions should also support more consistent and meaningful have to maintain close working adoption of common accounting disclosure of financial and risk infor- contact and co-operate throughout standards internationally. A project mation to the public. This should the process. is under way in the International include business objectives, risk While the study group envis- Accounting Standards Committee appetite, approach to risk manage- ages creation of this framework as a and it should be supported. ment and the actual risk-earnings voluntary exercise, if the frame- The other key ingredient in performance. An agreed approach work comes to be identified with a improving the external audit is an for all core institutions would over- strong internal control environment, expansion of what is reviewed. The come the reluctance of individual and if strong controls are viewed audit proposed in the report would firms to publish information about as a key component of effective go beyond traditional review of a their inner workings that is not management, the markets are likely firm’s financial statements. It would matched by other firms. to provide incentives to implement assess whether the risk-management it — in the form of higher credit policies and procedures promulgated Legal standards ratings or earnings multiples. by senior management to fulfil the Assessing legal risk in international Likewise, supervisors, whose duty objectives of the industry frame- transactions should take into account it is to make sure that management work had in fact been implemented. the enforceability of contract provi- pursues strong internal controls in The independent external auditor sions, including netting, and the any case, are likely to provide direct would attest, in accordance with effectiveness of insolvency proce- regulatory incentives for the general accepted auditing standards, to the dures. To reduce such risk, countries adoption of this framework. implementation of the institution’s should strengthen legal standards.

88 GLOBAL INSTITUTIONS

Enforceability of netting, col- Achieving a system of supervi- lateral and derivatives contracts, sion that works as effectively as if and incompatibilities in national there were a global supervisor insolvency laws have concerned requires someone at the centre of global institutions and their super- the process to co-ordinate contacts visors for a number of years. among supervisors and their sharing Despite considerable attention of information. to these issues, problems remain. Bank and securities supervisors Netting is legally binding in only are now pursuing a lead co-ordi- Understanding a few countries. There is similar nator model for information-sharing cause for concern over collateral in emergencies and this model the intricacies arrangements, both as a legal matter should become part of the ongoing and as a management practice; process of supervision as well. Over of risk defined management standards are the long-term, the role of adminis- needed. trative co-ordinator may change to management Perhaps most intractable are that of a co-ordinating supervisor inconsistencies in national insol- who will take the lead in routine systems and vency laws. While these are not supervisory matters affecting a overseeing likely to be resolved soon, other global firm. work by the Group of Thirty indi- With or without a strong co- a complex firm cates that the most important ordinator at its centre, the basic considerations in achieving a speedy underpinning of international super- in a crisis will resolution are: vision should be strong capital •Effective netting. standards and a strong framework r equire a high • Ready access to good informa- for comprehensive and effective tion on an insolvent firm's management controls as described level of skills — exposures. above. The framework would apply • Legal distinction between the to the entire global institution, supervisors firm’s own and client funds. including subsidiaries and affiliates Careful evaluation of legal risk that may not be subject to supervi- will have as part of overall risk control should sion at present. Ideally, it should be to exercise provide an incentive, at the margin, based upon, if not identical to, the to negotiate contracts in safer juris- industry framework described previ- judgement dictions. Over the long term, the ously. goal must be stronger national laws However, it can only become about a firm on netting and on insolvency. the standard for global supervision National and functional super- if it is acceptable to national and visors should agree upon a lead functional supervisors, which will co-ordinator for global firms, apply require close collaboration in its a global review framework to all design. Although this will be parts of a financial group and agree complicated, it is clearly in the upon consistent reporting require- interests of supervisors and super- ments for global firms. vised institutions to do so.

89 GLOBAL INSTITUTIONS

The resulting framework could training in co-operation with industry, Going further, effective co- then serve as a consistent basis for or a system of inward and outward ordination between supervisors will evaluating all global institutions. It secondments between supervisors only come about if they recognise would serve as a general blueprint, and supervised. These steps would their mutual interdependence and within which detailed procedures bring supervisors closer to the adopt common techniques. This would be implemented by each market, but carry the potential risk might make co-ordinated, global firm. The evaluation would be or perception of regulatory capture inspections possible over time. But performed in the first instance by an and conflict of interest. even where multiple reviews of institution’s internal audit function, management controls by an inde- and would be verified by the pendent external auditor and various external auditor. The external audit supervisors continue to occur, at would be provided to supervisors in Effective least they would all be based upon a appropriate form. co-ordination consistent, mutually reinforcing This audit and evaluation would global framework. not be a substitute for supervisory between Application of the new frame- judgement, nor relieve supervisors work in the course of supervisory of their statutory duty to evaluate supervisors reviews is not, of course, the same the firms they supervise. It would as having it fully implemented by provide additional information to will only come financial institutions. The preferable supervisors on critical issues and, to approach is to provide incentives for the extent that supervisors were to about if they a financial institution to make the pursue their own evaluation of the agreed framework part of its firm, they would likely use the same r ecognise their internal culture. Supervisors can framework. offer a variety of incentives, positive Understanding the intricacies mutual and negative, to global institutions of risk management systems and interdependence in areas such as capital, differential overseeing a complex firm in a supervisory monitoring and reporting, crisis will require a high level of and adopt reduced supervisory fees, fewer skill. Supervisors will have to exer- prescriptive rules or harsher penal- cise a lot of judgement about a firm, common ties for poor performers. making assessments of risk, not The incentive for the industry only compliance with rules. They techniques to develop a global framework will be called upon to provide would be greatly increased if super- feedback to the firm. People with visors used the occasion of reaching scarce skills will have to be Co-ordination of supervisors agreement with the industry on such recruited, and they are likely to across borders and functions is standards to promulgate consistent, demand compensation closer to important in achieving a global global supervisory reporting requ- market rates than current public- view of a firm. Naming a lead irements to the maximum extent sector pay scales. co-ordinator and encouraging all possible. With nearly half of survey supervisors to use the same evalua- respondents filing in excess of 500 Regulatory capture tion framework would simplify this supervisory reports a year, and over Keeping supervisors’knowledge up process. A common capital frame- half operating in more than 25 coun- to date may require structured work now exists for banks. tries, simplification of reporting

90 GLOBAL INSTITUTIONS

would be a powerful incentive to co- are of increasing importance in the contingency planning, etc. Futures operate. global financial system. Market exchanges and related clearing Promoting adherence to a infrastructure includes deal-reporting organisations should publish infor- global control framework would, of and confirmation systems, clearing mation on their own finances, course, be only one ingredient in the houses, central depositories and market protection mechanisms, supervisors’ broader efforts to payment systems. These mecha- sources of financial support and ensure the safety of markets and of nisms are essential to the operation default procedures. Each exchange the international financial system, of financial markets and main- or clearing house should develop albeit a very important part of the a mechanism for communicating overall picture. Supervisory respon- information to market participants sibility for protection of depositors, if and when default procedures are transparency for investors, smooth Institutions must initiated. Cross-border co-ordina- functioning of markets, appropriate tion and communication among capital standards, crisis manage- accept the need exchanges, clearing houses and their ment and the full range of other for a common supervisors should be improved. duties would be unaffected. There Governments should strengthen would be no reduction of responsi- framework for laws and procedures for market bility or yielding of sovereignty. default and protection of customer Ultimately, it remains the super- assessing, assets, funds, and positions. visor’s responsibility to decide when The study group views these deficiencies in management control managing and as the essential components of a warrant corrective action, and to coherent and acceptable interna- initiate appropriate enforcement r eporting risks, tional approach. While it may be measures. Whether this is left to difficult to put them all in place, supervisory judgement, based on a and be willing to there is little argument against them. specific problem-response regime What is clear is that no initia- or on pre-commitment by the firm work with tive of this kind will work without to certain standards of performance, supervisors two key ingredients. Major finan- the threat of sanctions against cial institutions must accept the internal management must be clear. in developing need for a common framework for Adoption of a consistent manage- assessing, managing and reporting ment framework by all supervisors standards risks, and be willing to work with would offer greater opportunity for official supervisors in developing co-ordinated action across interna- standards. tional borders to remedy problems taining their integrity, financial Supervisors must continue to in global firms. stability, and their ability to with- co-operate with one another to stand shocks and recover from understand the complexities of the Guidelines disasters is essential. modern financial marketplace to Supervisors should formulate risk There are a number of areas in ensure the effectiveness and the management guidelines in organised which action should be considered. consistency of their approaches. markets and for institutions that are There may be need for overall stan- We hope that the study group part of the market infrastructure. dards at exchanges and clearing report marks the start of that Markets and market infrastructure houses regarding back-up sites, process.■

91 THE UK CREDIT CARD MARKET The two businesses are distinct, and in the UK only a few firms are active in both. Box 1 shows the By Collette Brooking, Bank of England interaction between these two busi- The credit card market in the UK has attracted interest over recent nesses. If the card issuer is a months, partly because of the strong growth it has enjoyed and also member of an international payments because of the aggressive behaviour of a number of new entrants. This system, such as MasterCard or VISA, article looks at the market and compares some of its characteristics with an additional “link” is involved those of the credit card market in the United States. between the merchant acquirer and the card issuer.1 The credit card market is unusual, This article concentrates on the so any analysis requires an under- card issuance business — the area standing of the mechanics behind of greatest change. This business the product and the market. The first earns income from both interest part of this article outlines some of and non-interest sources, although the key sources of income and costs interest income is usually the main faced by card issuers and how these source. influence the net interest margin of Non-interest income the product. The next section UK card issuers receive most of considers the recent growth in the their non-interest income from the UK card market. Observers have fees charged to customers. Annual commented on the similarity between card fees were introduced in 1990 the UK market now and the US and are becoming common. market 10 years ago, so this section However, some of the new entrants also concentrates on some recent to the UK card market have not developments in the US market. charged an annual card fee — Then some of the similarities and which could suggest that as compe- differences between the UK and US tition increases, issuers will face card markets are examined, which pressure to remove card fees. may indicate whether the UK will Card issuers also receive an follow the US trend and issues that interchange fee, a proportion of the may arise if it does so. merchant service fee that a merchant acquirer receives from a HOW CREDIT CARDS WORK retailer. A number of card issuers The credit card market consists of charge a fee to the cardholder if two very different businesses: card they use a credit card to withdraw issuance — the “consumer end”, cash — either fixed at a minimum which provides credit cards and level or a percentage of the cash bears the credit risk of the customer amount. Issuers may also charge and merchant acquiring, and the penalty fees — if cardholders go “backroom business”, which “signs over their imposed credit limit or if up” outlets to accept credit cards payments are in arrears. Penalty and carries out processing. fees are more common in the US,

92 CREDIT CARDS

Box 1 LINKS BETWEEN THE CARD HOLDER, RETAILER, MERCHANT ACQUIRER AND CARD ISSUER

Purchase of goods, value of £50 Cardholder Retailer

Merchant service fee Reimburse (on average 1.65% £50 of transaction value)

Reimburse £50 Merchant Card Issuer Acquirer Interchange fee (on average 1% of transaction value

VISA/ MasterCard

1. The card holder purchases goods, value of £50. 2. Retailer is reimbursed by the merchant acquirer for £50, but the retailer has to pay the merchant acquirer a merchant service fee for arranging the transaction. 3. The merchant acquirer is reimbursed by the card issuer, but has to pay an interchange fee — the acquirer receives the difference between the merchant service fee and interchange fee as income. 4. The card issuer will send a monthly statement to the card holder, normally stating the minimum balance to be paid.

where they play an important part in depends on a number of factors, Interest income is significantly the debt-recovery strategy and can primarily the rate of interest that affected by the proportion of cust- represent a significant source of the bank charges in relation to the omers that do not pay their balances income. cost of borrowing and the extent of in full each month and take extended Interest income any interest-free period offered. UK credit (and the average amount of This is earned as issuers charge card companies commonly offer a extended credit that they take). The customers for using credit cards as period of time between the transac- level of non-interest bearing loans, a source of revolving credit. Card tion date and the date of repayment, either through bad debts or through holders have the option of repaying during which no interest is charged those customers who do pay their only a proportion of the outstanding (up to 56 days). It is therefore balance in full, will drag down the balance and they accrue interest on possible for a customer to time the overall level of interest income — a the amount of credit that is “rolled lending and payment of the credit typical credit card lender could over” to the next statement. The card bill to maximise this interest- expect to lose about half its of level of interest income earned free period. “headline” spread2 by subsidising

93 CREDIT CARDS

Chart 1 DECOMPOSITION OF CREDIT Cost of funds CARD MARGIN 1996 In addition to the basic costs of running any company, a card issuer must have the resources to lend Net yield Cost of full payers money to card holders. Most issuers 35% 45% borrow funds and subsequently have to pay interest on that money. For simplicity, it can be assumed that the issuers base their “cost of funds” on LIBOR (London Inter- bank Offer Rate). However, this simple proxy of funds does not account for those issuers who cross- subsidise their funding or are Other Costs internally funded. It should also be (customer incentives, fees, noted that the issuer is required to connected lender liability) 4% fund the entire asset base, irrespec-

Bad debts tive of whether it is earning interest. 16% In the US securitisation is a key source of funding for specialised cardholders that always pay off their risk customers who pay off their credit card banks, with the benefits balance (see Chart 1). balance fully (but do not earn much of funding at lower capital charge It has been estimated3 that in interest income). and the transfer of credit risk to 1996, 75 per cent of customers were It has been suggested that the investors. US card issuers earn extended credit takers, compared growth in the level of extended significant income from servicing with an average of 59 per cent in the credit takers since 1990 could securitised and sold credit card early 1990s (Table 1). However, reflect the introduction of an annual receivables. After deducting interest these percentages are only industry fee by several credit card issuers. If payments to investors, credit losses estimates and the proportion of customers have to pay for cards, and other trust expenses from extended credit takers among card they are more likely to make use of finance charge collections, they issuers is likely to vary according to them as a source of revolving credit. receive the residual income, the the types of consumers targeted. However, it is questionable whether “excess spread”. In recent months, Those issuers who target high- given the size of average credit card securitisation in the US has slowed. risk customers may have a greater balances in the UK, the average This is possibly a response to rising proportion taking extended credit card fee of about £10 would be losses on securitised portfolios (albeit at a higher risk of default) significant enough to result in such which have reduced some of the than those issuers who target lower- a shift. excess spread and may also reflect

Table 1 PERCENTAGE OF INTEREST-INCURRING BALANCES 1990-96 (UK)

Year 1990 1991 1992 1993 1994 1995 1996 Percentage of balances incurring interest 59 66 67 69 73 73 75

94 CREDIT CARDS

Box 2 SECURITISATION OF CREDIT CARD RECEIVABLES

Essentially there are three key stages in a securitisation: Any residual income is the excess spread, which usually goes 1. Issue bonds; to ensure that there is no recourse to the card to the card issuer. issuer, assets are transferred to a separate entity; in the UK, this (b) Repayment of the principal by card holders: the revolving is normally through a special purpose vehicle (SPV). It is structure of credit card securitisation has been established to important to separate the interest in the trust into seller interest offset those card holders who pay off their principal early. (the institution which originally issued the cards) and investor Instead, repayments are reinvested in purchasing new assets, interest, although both parties have a claim on the income until they get towards the end of the bond’s life and redemp- generated by the card receivables. Investors’ interest can be tion occurs. protected against default by cardholders, through a form of 3. Redemption of the bonds: redemption typically occurs over a credit enhancement. A simple example is a senior/subordinated twelve month period, at the end of the revolving period. The structure, where there is a priority allocation of the cash flow: cardholders’ repayments are not reinvested, but are retained Class A (priority); Class B (subordinated, therefore has a lower to pay the investors’ principal. Redemption can either be credit rating than class A). through an accumulation method or through amortisation. 2. Servicing the bonds (the revolving period): income received (a) Accumulation: an amount of the cardholders’ repayments are from securitised credit card receivables is two-fold: placed in trust and accumulated until they are used to make a (a) Interest and fees from cardholders (trust yield): this is used to singe “bullet” payment of the investors’ principal. pay for the administration of card accounts, to cover any (b) Amortisation: a proportion of the cardholders’ repayment is defaults on cards and to pay the investor and sellers’ interest. used to pay off the investors’principal over monthly payments.

reluctance on the part of purchasers a pool of securitised assets, both sation may be used to accelerate the of asset-backed securities (ABS) to during the revolving period and payment of investor interest. This take on high-risk assets. during amortisation. This means can be viewed as being at the The US securitisation market is they differ on what exactly consti- expense of the seller, who may be more developed than the market in tutes “a recourse to the card issuer”. left with the remaining, riskier, the UK. This partly reflects the fact The Bank of England aims to accounts that were acquired prior to that UK banks have historically ensure that the structure of a secu- amortisation. been well capitalised and have had ritisation does not favour investors In future there may be an access to cheap funds, but also and that any interest (or losses) are increased appetite for UK credit reflects regulatory restrictions in the appropriately placed with both the card securitisations from those UK. Policy on the securitisation of seller and the investor. investors “searching for yield” and revolving credit was only intro- Before the 1996 update of sec- new products in which to invest. US duced in the UK in 1992 uritisation policy, the Bank of card issuers have suggested that (BSD/1992/3) and was updated in England only allowed a disaggre- there has been increasing interest in 1996 (S&S/1996/8). gated structure5, which differs from securitised card receivables, from Neither US nor UK regulators the more common US approach, an European investors, an interest allow off balance sheet treatment if aggregated structure. A concern that which may accelerate after EMU. the investors in the bonds have any the Bank of England has with this Furthermore, as competition in the recourse to the card issuers4. US and approach is that, during amortisa- credit card market increases, issuers UK regulators, however, have tion, interest from assets that were may prefer not to tie up capital in different views on the credit risk of securitised after the start of amorti- assets with low returns. If assets are

95 CREDIT CARDS

Chart 2 CREDIT CARD SPREAD AND NET INTEREST MARGIN 1978-96

20%

Unadjusted Spread 15%

As technology 10%

improves and Spread adjusted for interest-free period 5% more detailed 0% databases of 78 80 82 84 86 88 90 92 94 96 F97 A97 J97 the UK -5%

population are -10% Cost of interest-free period

developed, card -15% issuers will be moved off balance sheet, more play a vital role. As credit card able to target capital is available to take on more debt is unsecured, companies need remunerative assets — which could to consider the financial back- different be an important factor in a market ground and position of the borrower growing as fast as the credit card before agreeing to lend. Credit socio-economic market. The level of securitisation scoring is, therefore, becoming in the UK will also be influenced by increasingly common in the UK. groups more US entrants to the market which The benefits are simple — the more accurately and have more experience in moving sophisticated the credit card credit card receivables off balance company is with regard to its exploit new sheet and already take advantage of customer database, rejection rate diversification of funding. and scientific bad-debt charge, the niches in the Bad debts more likely it will offer adequate A key cost faced by credit card risk pricing. market companies is bad debts, as non- As technology improves and performing loans have an impact on more detailed databases of the UK the net-interest margin — typically population are developed, card about a fifth (Chart 1). Although issuers will be able to target the level of bad debts will be influ- different socio-economic groups enced in part by the economic more accurately and exploit new environment, the issuer’s credit niches in the market. However, controls and debt-recovery systems there is a danger that if card issuers

96 CREDIT CARDS

become too reliant on credit-scoring Customer incentive schemes were made to the regulation. models, an element of discrimina- A number of issuers in the UK Chart 1 shows how the costs tion could be introduced. compete through measures other outlined influence overall credit The level and trend of bad debts than price — non-price competi- card profitability. has become a more significant issue tion: although the level of APR in light of the recent rise in bad charged may not be particularly Net Interest Margin debts and charge-offs experienced low, additional benefits are offered A standard estimation of the net in the US. In the UK, bad debts are to customers who have a card. Not interest margin is a measure of on average about 3 per cent in the all of these benefits will be fully interest income to average earning credit card market, compared with exploited by customers but they do assets, but this does not include the the present level in the US of about represent a cost to the card issuer — cost of the interest-free period 5 per cent. If the UK follows the US typically of between 3 per cent and (which may be a significant trend, credit standards may decline 5 per cent of the overall spread. element). as competition increases. Connected lender liability Chart 2 shows the credit card Fees An additional cost faced by credit spread adjusted for the cost of the Card issuers are also liable for a card companies is the Section 75 interest-free period. In addition to range of fees, including fees due to claim — connected lender liability being affected by the proportion of banks which provide cash to their — which states that card issuers are customers taking extended credit, card holders. Fees also have to be liable in the case of customer dissat- the adjusted spread is also influ- paid by those credit card issuers that isfaction if goods are faulty. enced by interest-rate sensitivity, are eligible and decide to join Although this clause was recently particularly as credit cards are payments organisations such as reviewed by the Department of assets with “sticky” rates, which are VISA or MasterCard. Trade and Industry, no amendments funded by floating rate liabilities.

Box 3 APR AND NET INTEREST MARGIN

The Consumer Credit (total charge for credit) Regulations 1980, ison of APR cannot accurately be made between a card that has an outlined the method of calculating the rate of total charge for interest-free period and one that does not. credit. The annual percentage rate of charge, APR, should in theory The calculation of the APR also involves an assumption that allow customers to make comparisons between the cost of credit the card issuers make about the average amount they expect to be from different providers, as calculations of the APR for credit borrowed — particularly the relation between this amount and the cards should account for the interest rate and any other fees. annual fee. For example, an annual fee of £12 spread over an The main problem concerning the calculation of the APR for average amount of £1,000 will have less of an effect on the APR credit cards under these regulations is that the “interest-free” than if it were spread over an average outstanding balance of £250. period is not taken into account. Therefore the “headline” APR An additional problem that is faced in analysis of card rates assumes that interest is paid on balances immediately. Although is that few card issuers publish separate figures for interest some card issuers do not offer an interest-free period (and there- received and interest paid on their credit card business. Therefore, fore this problem does not apply), the majority of card issuers do interest elements must be calculated using published statistics on offer interest-free periods of up to 56 days. Because the cost of the values of transactions and volumes of customers taking extended interest-free period is not included in the calculation of the APR, credit. Figures may overstate the use of cards as sources of credit the “headline” APR that is often quoted may underestimate the for individual issuers because a significant proportion are full costs that card issuers face and it also means that a direct compar- payers and don’t use them for revolving credit.

97 CREDIT CARDS

card issuers and by 1996, this had Chart 3 CREDIT CARD AND NON-CREDIT CARD CONSUMER DEBT (UNSECURED AND SECURED); grown to more than 800, a ten-fold ANNUALISED QUARTERLY CHANGE 1988Q1-97Q1 increase. This growth partly reflects the overall growth in consumer

25 credit in recent years — in Other lending December 1992 annual growth of consumer credit was 1 per cent and 20 by June 1997 was more than 18 per Credit card lending cent. However, Chart 3 shows that since 1990 the growth in credit 15 cards has generally been faster than that of other forms of consumer credit. At present, credit cards 10 account for about 21 per cent of net lending of consumer credit in the

Secured lending UK (up from 19 per cent in 1991). 5 The strong growth in the UK credit card market is all the more significant because a number of 0 commentators had believed the 1988 Jan 1989 Jan 1990 Jan 1991 Jan 1992 Jan 1993 Jan 1994 Jan 1995 Jan 1996 Jan 1997 Jan market had reached maturity five or 10 years ago — this belief was The adjusted spread will reflect the APRs charged (see Box 3) and influenced by the fact that the interest rate mismatch and any therefore the net interest margin. payment card market in the UK is hedge taken to reduce it. Particularly if a card issuer intro- one of the most developed in As credit card rates tend to be duces an introductory “teaser”7 rate, Europe. However, a comparison less variable than other lending, this will place pressure on the net with the US credit card market card issuers generally perform interest rate margin in the short shows that the UK has some way to better when interest rates are falling term. go to reach the maturity of the US and the spread over the cost of market. In 1994 there were twice as funding widens. In addition, the GROWTH IN THE UK many cards per head in the US opportunity cost of funding non- In 1990, total credit card debt as in the UK and borrowing on interest carrying balances also outstanding was almost £9bn and by credit cards accounts for a larger declines. Competitive pressure on 1996 this had grown to about proportion of unsecured consumer credit card rates will influence the £16bn. In 1990 there were some 80 borrowing in the US than in the UK

Table 2 NUMBER OF MASTERCARD AND VISA CREDIT CARDS IN ISSUE 1986-96 (UK)

Year 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 MasterCard 9,846 11,370 11,454 12,128 12,294 11,554 11,169 10,351 10,944 11,702 12,930 VISA 12,121 13,106 14,318 16,485 17,552 15,257 15,307 15,119 14,807 16,225 20,064 Total 21,967 24,476 25,772 28,613 29,846 26,811 26,476 25,470 25,751 27,927 32,994

98 CREDIT CARDS

(about 40 per cent). Recent devel- adjusted pricing is vital. Targeting ation” accounts, which involve new opments in the US market may high-risk customers is viable as segments of the market, are going to indicate how the UK market will long as adequate pricing is develop. develop in the future. employed. However, in the US, Bad debts In the past few years, the Developments in the US market there has been evidence that low- US has suffered from nation-wide Market saturation In recent years income consumers have been increases in consumer credit delin- the UK has been a target market for quencies8 and bankruptcies which US credit card companies. Their have led to increases in credit card interest in the UK market may in delinquencies and net charge-offs. part reflect the fact that the UK Strong growth Since mid-1995, the rate of net credit card market is highly concen- charge-offs on US credit-card debt, trated; a relatively limited number in the UK credit for securitised and on balance sheet of providers control a significant card market is accounts, has been rising and net percentage of the market share. By charge-offs as a percentage of loans contrast, there is far lower concen- all the more are now comparable to 1992, when tration in the US and the credit card US unemployment was 7.5 per cent. market is approaching saturation. significant It is now only 5 per cent. There have Any recent growth in the US been nation-wide increases in has been based on either winning- because a consumer credit delinquencies, over customers or exploiting niche although delinquencies have not markets, which could include high- number of increased to 1992 levels. risk customers. Although “balance The increase in bad debts in the transfer” was the approach favoured commentators US has raised a concern, particu- by US card issuers in the US market, larly as it has come in a period of as competition has increased, it is no had believed strong economic growth. A recent longer as popular. There are high the market report by Salomon Brothers noted acquisition costs in trying to “poach” that the level of credit card debt customers from competitors as the had reached being written off by some of the issuer may have to pay a premium large banks in the US is higher than for each balance. The need to maturity five in the 1990-91 recession. The dangle additional “bait”, possibly a Economist (July 5-11 1997) low “teaser” rate, will have an or 10 years reported that 53 per cent of bank impact on the net interest margin, loan losses are coming from 6 per albeit in the short term. ago cent of bank loans. The most recent Targeting of “underdeveloped” figures for 1997 Q2, however, show credit consumers might involve a decline in delinquencies. reaching out to high-risk segments granted pre-approved credit lines. Some card issuers in the US are of the market in the quest for growth This undoubtedly leads to a decline tentatively suggesting that they may — those customers with a low in quality. A number of US entrants have seen the worst and that income or high debt, the elderly, or to the UK market have said that consumer debt problems may start those with a problem credit history. “balance transfer” is an appropriate to fall in the coming quarters. Obviously, the increase in risk method now, but as competition However, the results from just one means that high quality risk- increases in the UK, “second gener- quarter make it too early to assume

99 CREDIT CARDS

Credit card companies in the US are Chart 4 LOAN LOSS COMPARISON IN THE US (VISA AVERAGE) now campaigning for changes to legislation. VISA and MasterCard

6% are pressing for a move to a needs- based system, whereby those debtors receiving protection would 5% still be required to continue repaying debts in proportion to their

4% income.

THE US PATTERN 3% There are a number of similarities between the UK and the US credit card market, which could indicate 2% that the UK is following a similar pattern to that of the US. However, 1% this section also highlights some differences between the two markets, which may influence the 0% 92 Q1Q3 93 Q1 Q3 94 Q1 Q3 95 Q1 Q3 96 Q1 Q3 future behaviour in the UK market.

Similarities that the peak of bad debts has economist of VISA, about $30bn of Increase in supply passed. debt was wiped out as a result of Growth in the US credit card market The influence of external bankruptcy in 1996. This may was partly a supply-side phenom- economic conditions, both national partly reflect changing attitudes enon, as more lenders, attracted by and local, on delinquencies has towards individuals declaring them- the potential for profitable business, meant that a number of US compa- selves bankrupt — bankruptcy entered the market. In the UK, nies have diversified their portfolios appears to be more socially accept- financial deregulation and the nation-wide, since it is unlikely that able now than it was five years ago. removal of credit controls in the all states are at the same point in the In addition to the attitudes 1980s reduced barriers to entry. economic cycle. However, the towards bankruptcy, changes in US More outlets accept credit cards — recent increase in delinquencies law have made it easier for they can also be used to make may also be a consequence of an individuals to declare themselves purchases over the phone — and industry-generated phenomenon, bankrupt. A survey carried out by this has encouraged their use in a brought on by companies over- VISA earlier this year found that wider range of purchases. reaching and targeting the higher over 65 per cent of people found the Nevertheless the level of card risk categories without adequate bankruptcy process easy and more penetration of the adult population risk-pricing. than a quarter said that they would is behind that in the US. In the UK, Part of the increase in bad debts consider filing for bankruptcy customers have an average of one or in the US has also been attributed to again! More than 10 per cent of two cards, whereas in the US they a rise in overall bankruptcy levels. individuals surveyed had filed for have more. This may reflect the According to Tom Layman, chief bankruptcy at least once before. fact that the level of solicitations

100 CREDIT CARDS

through direct mailing in the UK is More recent influences on credit somewhat lower than in the US card growth in the UK have (where customers can receive a included the substitution of unse- mailshot every week), although it is cured lending for equity withdrawal starting to grow rapidly. and also consumers’consumption in Increase in demand anticipation of windfall gains from Recent research9 suggests that in the demutualisation of a number of addition to growth in supply, influ- building societies. Consumers in ences on the demand-side have also Non-bank card issuers made a substantial contribution to An additional source of competition the UK may the growth in the credit card market in the UK payment card market is in the US. from debit cards and other non-bank be averse to The demand for credit is deter- cards. Debit cards are not very mined by a number of factors, such popular in the US, but they have taking on as the level of wealth in an economy experienced strong growth in the and the proportion of borrowers. UK. Competition from these sources secured credit Research has shown that before the may put further pressure on the after the 1980s these two factors moved in credit standards of card issuers. different directions, offsetting each However, debit cards do not offer r ecession and other and having little overall any competition to credit cards as a impact on demand for credit. source of credit or offer customer negative-equity However, in the mid-1980s, both incentives, so their threat to credit factors began to move in the same cards may be limited. Supermarkets experience of direction, the combination increasing which are entering the financial the overall demand for consumer market also pose an additional the early 1990s credit in the US. threat to the present market share of Wealth levels have increased as UK card issuers. Supermarkets will and instead earnings grew and unemployment be keen to take advantage of the fell, and the proportion of customer loyalty and their estab- opt for the borrowing has continued to rise in lished brand names and extend short-term response to a change in individuals’ this to their own credit cards. attitudes to debt as they are more Supermarkets’ strategy in offering flexibility willing to take on credit. Since these highly competitive deposit rates characteristics also apply to the UK may be an indication of the level of offered by market, it is likely that demand price competition they will under- factors have influenced the growth take in the credit card market. credit cards in the credit card market. Supermarket cards also pose a Consumers in the UK may also threat to debit cards as they are typi- be averse to taking on secured credit cally used for food shopping. after the recession and negative- In the US, there is more compe- equity experience of the early 1990s tition from non-bank issuers such as and instead opt for the short-term monoline or stand-alone issuers, flexibility offered by credit cards. which specialise in credit cards. The

101 CREDIT CARDS

market in co-branded and affinity differentiate between the borrower end up using it anyway (but still pay cards is also far more developed in who intends to use the card to their bills). As they did not expect to the US. increase debt and the borrower who use the card they do not bother to only wants to transfer a high shop around for lower rates. Differences balance. Issuers may be cautious However, less credit-worthy Price competition and reject the applicant or not offer customers, who know they will use US issuers tend to focus on price favourable terms. There is also an the card and potentially default, are competition, whereas emphasis in more likely to shop around for the UK has been on non-price lower rates. Reducing card rates, competition. A number of new US runs the argument, means that entrants have undercut prices of adverse selection applies and the established UK issuers and the “wrong” type of customer is question arises as to how the attracted. market share of UK card issuers If card issuers in the UK have will alter as a result of this US issuers to start competing on prices, APRs increasing competition. Although a and continue with non-price compe- number of issuers currently have tend to focus tition, such as incentive schemes, it brand loyalty in the UK, this may could prove very expensive. not continue as consumers become on price Use of cards more aware of different credit card Although there are a larger number rates. competition, of cards in the US than in the UK However, there may be some whereas (both absolute number and per time-lag before consumers start to capita) they tend to be used less react to the lower APR rates on emphasis in intensively. Furthermore, overall offer, as US research suggests that credit cards are used more as a there are relatively high search and the UK has source of credit in the US. The level switch costs involved in moving to of outstanding balances is typically credit cards with lower APRs. This been more on higher than the industry average of reflects the lack of information that 75 per cent in the UK (about 85 per consumers have about lower non-price cent in the US). interest rates — particularly the Securitisation cost of time and effort involved. competition The securitisation market is far Additionally, card holders that pay more developed in the US than in an annual fee and switch at the the UK. wrong time of the year, could lose money. Switching costs are also THE UK MARKET’S FUTURE involved when transferring to a It cannot be predicted whether the new card. Customers, may face UK has learnt from the US credit higher costs because they could argument that price competition experience, or if it will follow the lose favourable credit limits when introduces an element of adverse trend to the same extent. However, they switch to other cards. selection. Profitable consumers for competition in the UK is undoubt- When switching large balances, card companies are those who do edly increasing and this may lead to card issuers may not be able to not think they will use their card but some developments similar to those

102 CREDIT CARDS

in the US, such as market saturation Chart 5 RISK PROFILE FOR DIFFERENT and a rise in bad debts. SOCIO-ECONOMIC SEGMENTS Market saturation If competition in the UK intensifies there is a risk that entry require- ments will be relaxed and card issuers will target those customers previously ignored. This could Profit increase the potential for default. It is important for card issuers Target Market to be clear about which market they 0 are targeting; one method is to A BCDE differentiate the market according to socio-economic groups. Socio-economic profiles Extended credit takers and full payers have Loss a key influence on the income and costs for card issuers, so card issuers seek a trade-off between individuals who pay their balances each month and extended credit takers. Although the former tive to card issuers: they are people The category targeted by a card are likely to be of minimal risk, that take extended credit, and there- issuers depends on its strategy. they contribute no interest-income fore earn interest income. They do Some issuers may target A, aiming profit to credit card companies pay off their balances eventually. to earn fee income or encourage (indeed, as the income they cont- Segment D includes riskier cus- extended credit-taking through very ribute barely covers administration tomers who take extended credit but low APRs. Others may target D and costs, they may be a cost to the are much more likely to default. E and charge high APRs in compen- credit card company). The latter Segment E may also result in losses sation for the additional risk. It is contribute interest income for the for card issuers: it includes high-risk important that the card issuer is companies, but at the same time the customers who have the greatest clear about which market it is risk of default increases. marginal propensity to default on attracting and can price accordingly. Chart 5 shows the socio- loans. Bad debts economic “risk profile” of different Competition for customers in Developments in the US become consumers10 that a card issuer may segments B and C is increasing even more relevant to the UK when target. as new entrants enter the market one considers that although personal Segment A includes customers and which may encourage more bankruptcies have fallen in the UK, who, more often than not, pay off card issuers to target segments D as shown in Chart 6 they have their credit card bills in full each and even E. Moving to these higher- levelled out at a higher plane than in month. This may result in losses for risk areas, where the potential to the 1980s. the card issuer, which is effectively default increases, requires more The problem of increasing bad having to fund interest free loans. sophisticated credit-scoring models debts may be tackled by differenti- Segments B and C are more attrac- to enable adequate risk pricing. ating between different types of

103 CREDIT CARDS

Chart 6 PERSONAL SECTOR BANKRUPTCIES 1988-96 of marginal customers with credit cards, will reach the same level as 12,000 the US. The key will be the res- ponse of lenders to conditions in the 10,000 market that are less conducive to new entrants. In a fast-growing market, busi- 8,000 nesses can expand their balance sheet without the need to acquire

6,000 market share. Once growth slows, increasing market share is neces- sary to keep a company growing. In 4,000 “winning” customers from other lenders, firms should make sure

2,000 that, in developing attractive pack- ages for customers, they do not lose sight of the need to ensure that they 0 receive adequate reward for the 1988 1989 1990 1991 1992 1993 1994 1995 1996 risks they are running.■

customers, and having multi-tiered NOTES interest rates based according to 1VISA and MasterCard are international payments organisations; they do not issue credit cards directly, but are involved in authorising, transmitting and settling financial transactions. To become cardholders’ creditworthiness. a member of VISA or MasterCard, an organisation must be authorised to accept demand deposits Tiered interest rates are more and be governed by a country’s commercial banking laws (therefore members are usually banks or common in the US than in the UK, building societies). As non-banks are not eligible to join either of these payments organisations it may explain the development of co-branded and affinity cards where non-bank organisations join and allow card issuers to price up with banks to promote a card (and therefore get the benefits of VISA or MasterCard member- according to the risk in different ship). A co-branded card is a card issued by a bank in conjunction with a non-financial institution. “Points” are earned every time the card is used and these accumulate towards a “reward”. The reward segments of the population. This is usually related to the type of non-financial institution involved. Affinity cards are issued by banks trend could develop in the UK — in conjunction with a non-financial institution and are aimed at groups of people “linked” by that organisation and who might wish to support it eg a university or charity. Every time the card is used, potentially to the ultimate of a price the organisation receives some income — on average 0.25 per cent of the value of the purchase. suitable for each individual (“a 2 The credit card “headline” spread is the difference between the “headline” annual percentage rate of charge (APR) and the cost of funding. segment of one”). 3BBA statistics. Some one-off factors that have 4 If the pool of receivables does not deliver the rate of return that is expected, the bond holders cannot hold the card issuers to account. resulted in rapid growth in the credit 5This involves securitised assets being allocated to the investor and to the seller, with any interest (or card market, and consumer credit losses) attributed to those receivables, being allocated to the particular owner. Such a structure ensures that there is no recourse to the card issuer and this transfer of credit risk allows the securitisation to more generally — such as the antic- be treated off-balance sheet. ipation of windfall gains from 6Where interest is divided between the investor and the seller, according to their interest in the port- folio of receivables. demutualisation and substitution 7 Some card issuers charge significantly lower APRs for a limited period (eg six months) in an attempt from secured to unsecured credit — to attract customers. However, after an introductory period, these “teaser” rates are usually replaced are likely to be less frequent. by higher APRs. 8An account becomes delinquent when the customer fails to repay the minimum required amount on It is too early to say, however, time. whether any increase in bad debts, 9 Current Issues in Economics and Finance (1997) Volume 3, No. 4, Federal Reserve Bank of New York, Morgan, D and I Toll. resulting from a rise in the number 10 Each category broadly reflects the standard socio-economic categories.

104 Building societies still have a MUTUALITY AT THE CROSS-ROADS major share of the UK mortgage and liquid retail savings markets. At 1 By Adam Boxall and Niall Gallagher, the Bank of England the end of 1996, they accounted for over 55 per cent of the UK mort- Several of the United Kingdom’s largest building societies and mutual life gage loan stock and 30 per cent of assurance companies have given up their mutual status in recent years or sterling bank and building society are about to do so. This is an important aspect of the structural change deposits (M4). These shares fell which has taken place in the provision of financial services in this country. substantially after the recent wave The alteration in corporate form raises several questions which are rele- of demutualisations, but remaining vant to the Bank because of the potential effects on the behaviour of the societies still account for almost 25 financial sector. Several key questions follow. Have mutuals and public per cent of the mortgage market and limited companies in the same industry sought different objectives — 12 per cent of M4. related to their different corporate forms — in the past? If so, will ex- mutuals behave differently? What are the prospects for the remaining Life assurance companies mutuals? This article explores these issues, using evidence primarily from Mutual life societies were first estab- the mortgage market and occasionally from life assurance. It does not seek lished during the first half of the to provide a value judgement about mutuality or promote one corporate 18th century to provide benefits form over another. Instead, it suggests how different hypotheses about the for widows and other survivors of motives of mutuals might be tested and draws out some implications for members. Initially membership was those that wish to maintain their mutual status. limited both in number and to certain professions and groups. Benefits Origins and market share were originally paid out of annual Building societies subscriptions, but, as life companies Building societies were established developed, they calculated premiums towards the end of the 18th century on the basis of ever more sophisti- as small local organisations whose cated estimates of each member’s members pooled funds to allow life expectancy, allowing them to them to purchase land and build offer the types of policies we observe houses. Membership was originally today. restricted, with similar contributions Mutual life companies also still from each member, and societies undertake a major share of UK life were wound up after all of the business. Unconsolidated figures members had been housed; hence show that, at the end of 1995, mutuals they were known as terminating accounted for over 50 per cent of societies. As societies developed, both the total premiums (£26bn) and they accepted deposits from indi- the total assets (£320bn) of the largest viduals who had no desire to borrow 20 life companies. Although these to buy a home, but simply wished to figures do include Clerical Medical, invest their money. In time such Norwich Union and Scottish Amic- societies became permanent, devel- able, which have now all converted, oping into the building societies that the remaining mutuals still account we know today. for around a quarter of both the

105 MUTUALITY

premiums and assets of the UK life to rely primarily on retained surp- sector as a whole. luses. Although joint stock financial institutions also use retained earn- Structure ings to fund growth, the link between The corporate form of mutual finan- growth and surpluses remains more cial firms reflects their origins, when important for building societies due membership was obtained through to their inability to issue equity. acquiring a core financial product, such as a long-term savings product, Mutual society objectives a mortgage or a deposit account, Although joint Mutual societies started as self-help and members were on much the same societies designed to achieve certain footing as each other. stock financial specific objectives for their members Voting rights are not propor- such as the purchase of a house. tional to the size of a member’s institutions Sometimes those objectives extended financial commitment, as members use retained beyond the narrow economic interest typically only get one vote each, of members; for example, building reflecting the character of the orig- earnings to fund societies have been seen as a means inal terminating societies — this of promoting home ownership as a means that the rights to control growth, the link social objective. Now they compete management are widely dispersed. with public limited companies to Members cannot sell or trade their between growth provide financial services. Do they membership rights, so their finan- still have objectives distinct from cial value is difficult to determine, and surpluses those of the plcs? The crucial test is and the closing of an account, repay- not what their managers have said ment of a mortgage or expiry of a r emains more but what the mutuals have actually policy leads to the loss of such rights done. Can they be distinguished by without compensation. important for their behaviour? The capital base of financial building The obvious standard of compar- mutuals is composed largely of the ison for an economist is the behaviour reserves, built up from retained surp- societies because of a profit-maximising company. If luses. Until 1988, building societies the number of people wanting to were unable to raise external capital. of their inability make interest-earning deposits is A building society which intended increasing over time, the profit- to expand had to rely on making a to issue equity maximising bank in the mortgage surplus on its activities, with the market will want to put aside some size of the required surplus being of the profits it earns each period to higher the more rapid the planned allow it to augment its reserves and growth in assets. Income had to be expand, earning more profits in the sufficiently large to cover costs and next period to divide up amongst its provisions and augment reserves in shareholders. line with asset growth. Regulatory However, a mutual which is only constraints were relaxed — in 1988 concerned about its current members and 1991 — but societies continued will be less keen to expand, because

106 MUTUALITY

any future benefits would have to be production technique, economies of divided up among a larger number of scope and differing regulatory constr- members; the new members would aints. But they can help to establish a get a share of the pre-expansion presumption about what the prin- surplus. Established members would cipal objective of building societies prefer lower interest rate spreads to has been in practice. higher reserves. If mutuals are solely In trying to identify the motives interested in the financial interests of any firm, it is important to distin- of their current members, they are guish between the interests of owners likely to grow more slowly than the Principal-agent and the incentives faced by manage- equivalent plc. If the surpluses they ment. In studies of firms’behaviour, could earn are instead distributed theory recognises the interaction between the owners according to the financial commit- and management is the subject of ment of members, then one would that managers principal-agent theory. expect to see both lower mortgage Principal-agent theory recog- interest rates and higher deposit rates may not pursue nises that managers of firms may amongst mutuals, and restrictions on the owners’ not pursue the owners’ objectives new membership (including credit- when owners do not exert full cont- rationing on the mortgage side). objectives when rol over managers’ behaviour and Mutual societies may also have the incentives facing owners and social objectives. If a building society owners do not managers differ. Some have argued wishes to promote home ownership that the normal provision for “one as an end in itself, it will want to exert full person, one vote” in mutuals neces- offer lower mortgage rates than the sarily means that ownership is more bank does. Also, in periods when control over dispersed than a plc. On the other the bank pays out a share dividend, hand, when ownership derives from the building society would use the managers’ use of a firm’s products, members funds available to reduce mortgage may want to have more involve- interest rates and add to reserves so behaviour and ment than ordinary shareholders that it could issue more mortgages incentives facing would, particularly if they share any in the future. The obvious tests of social objectives of the mutual. whether this is an acceptable simpli- owners and How is one to tell if mutuals fication of the objectives of building have fallen prey to the principal- societies are: managers differ agent problem to a greater extent • whether building society mort- than plcs? Again it is useful to gage lending has been at lower investigate what different assump- rates than that of banks, and tions about the objectives pursued • whether it has increased more imply for behaviour. We concen- rapidly. trate on two particular commonly Such tests are not conclusive, suggested managerial objectives: because lots of other factors are growth maximisation and expense important in the real world, such as preference. They have both received access to new technology, choice of particular attention in studies of

107 MUTUALITY

Chart25 1 ASSET GROWTH, RETAIL SPREAD AND THE RESERVE RATIO (PER CENT) Asset Growth 9 Retail Spread and the Reserve Ratio (Percentage Points) 20 8

7

15 6

5

10 4

Reserve Ratio 3 Asset Growth (Annual Percentage Change) When building 5 2

Retail Spread 1

societies face 0 0 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 binding capital Source: Council of Mortgage Lenders constraints, financial mutual behaviour in the in a net present value sense). This is UK and the USA. because, when a reserve require- growth If the management of a firm ment or other capital constraint bites, sought to inflate expenditure on assets can only grow in line with maximisation is staff, buildings and equipment, that reserves. would demonstrate expense prefer- A building society which did likely to lead to ence. If that were a problem for not face binding capital constraints mutuals, we should observe building could expand by attracting deposits the maximisation societies and mutual life assurance and issuing mortgages beyond the companies running higher costs than point where marginal cost equals of surpluses — do joint stock companies with a marginal revenue, at the same time similar range of business activities. running down its reserve ratio. analogous to Growth maximisation would However, such behaviour would the profits occur if management sought to ultimately lead to the mutual being maximise a firm’s size. Such behav- capital-constrained and having to r eceived by iour could lead to a shareholder maximise surpluses so that it could owned firm expanding beyond the maximise growth. plcs point where marginal costs equal marginal revenue, reducing the net Behaviour of mutuals present value of the firm. We now consider some empirical In the case of building societies, evidence to see what light can be growth maximisation can have a shed on the actual behaviour of different effect. When building soci- mutuals and hence on their objec- eties face binding capital constraints, tives. This section is not intended growth maximisation is likely to as a comprehensive comparison of lead to the maximisation of surpluses mutuals with competing shareholder — analogous to the profits received owned firms. But it does establish by plcs — period by period (but not some “stylised facts” which may

108 MUTUALITY

help to focus debate about the special ably. This widening of spreads, and nature of mutuals. the associated rise in surpluses, coincided with a period of financial Growth and spreads — liberalisation, during which banks Building societies entered the housing market, mort- Building society assets have grown gage rationing ended and the rapidly for most of the post-war building societies’ cartel was aban- period, requiring significant and doned. Operating income was in continuing surpluses in order to excess of that required to fund asset maintain prudential capital require- growth and cover provisions, suggest- ments. Chart 1 shows growth in ing that societies were seeking to assets, the reserve ratio (reserves as rebuild reserve ratios after two a percentage of total assets) and the decades of decline. That may have In the early retail spread (the difference between been partly because of regulatory the average mortgage rate and the action requiring societies to increase 1990s building gross share rate) for the building their capital requirements. societies’ society sector as a whole from 1950 From the early 1990s, growth in to 1996. mortgage assets slowed considerably spreads were From 1950 to the late 1970s, because of the housing recession. rapid growth in assets coincided Despite this, building society spreads wider than with a run-down of the reserve ratio widened significantly, leading to and intermittent credit rationing. significant income growth and the those required The decline in the reserve ratio may continued build-up of reserves. Part have reflected the gradual easing of the widening in spreads may have to support of capital requirements during the been in reaction to the severe ’ 1960s. Building societies also faced housing market recession, record societies little competition in their core levels of provisions and uncertainty markets of mortgage finance and over future credit losses at the time. growth and retail deposits during this period. However, the fact that operating provide for The fact that societies sometimes income was always well in excess rationed credit suggests that they of provisions suggests that this is credit losses were not attempting to maximise not a sufficient explanation on its growth. If they had been, they would own. Also, the ratio of capital to have charged more and used the risk-weighted assets was far in excess receipts to build up reserves faster, of the regulatory minimum. The thus allowing more lending in the high surpluses throughout the 1980s, future. Instead they may have sought and the fact that surpluses in the to balance the objective of promoting early 1990s did not fall despite home ownership with that of holding lower nominal credit growth, suggest down mortgage interest rates for that building societies’spreads were existing members. wider than those required to support Between the 1970s and the late societies’ growth and provide for 1980s, retail spreads rose consider- credit losses.

109 MUTUALITY

Chart 2 BANK AND BUILDING SOCIETY MORTGAGE INTEREST RATES (PER CENT)

Building societies have Source: Bank of England and Building Societies Commission Chart 3 BANK AND BUILDING SOCIETY earned a return DEPOSIT RATES (PER CENT) on assets at least as high as the banks, despite the lower risk of building Source: Bank of England (internal estimates compiled from Moneyfacts to distinguish between committed and converting building societies)

society Chart 4 RETURN ON ASSETS (PER CENT) assets

Source: CML and BBA

110 MUTUALITY

Comparative spreads and than banks. Variations in expense profitability ratios between banks and building Based on headline rates, the evidence societies reflect differences in busi- about mortgage interest rates sugg- ness mix and diversity of operations ests that there was virtually no as well as differences in cost difference between those offered by efficiency. For example, the Abbey banks and large building societies National, which became a bank in between 1984 and 1997 (Chart 2). 1989, has an expense ratio lower From 1989 until 1995, the larger than many building societies, perhaps building societies appear to have reflecting its continued heavy concen- offered slightly higher deposit rates tration on mortgage lending and the than did the banks . However, as addition of a low-cost treasury oper- shown in Chart 3, from 1995, the ation. Alliance & Leicester, on the deposit rates of converting societies other hand, has an expense ratio converged on those of the banks similar to that of the main banks, while a number of the committed probably reflecting its ownership of Banks in the building societies continued to offer Girobank, a bulk money transmis- slightly higher rates, at least until sion business (see Table 1). UK tend to this summer. Building societies have earned Building society objectives have considerably a return on assets at least as high as It is difficult to conclude that building the banks, despite the lower risk of societies as a whole have consis- higher ‘expense building society assets. Whilst some tently followed any one motive to ’ banks have enjoyed much higher the exclusion of others. Neverthe- ratios than returns on assets in recent years, this less, the above has established some has not been universal, and building stylised facts: building societies have tended to exhibit much • Building society asset growth greater stability in rates of return was rapid for much of the post- societies over time (see Chart 4). war period. This suggests that building societies were not seek- Expense preference behaviour ing to maximise the financial Banks in the UK tend to have interests of current members. considerably higher “expense ratios” • The ratio of costs to assets for than building societies, as shown in building societies was lower than Chart 5, although there has been for banks, but these ratios are not some convergence over time. Chart 6 readily comparable due to the shows that the typical building different business mixes. society branch deals with more • Building societies tended to accounts than the average bank operate narrower retail spreads branch. than banks between 1989-97, However, these data do not the period for which data are show unambiguously that building available, by offering slightly societies operate more efficiently higher deposit rates.

111 MUTUALITY

Chart 5 BANK AND BUILDING SOCIETY EXPENSE RATIOS

4.0

Bank 3.5

3.0

2.5

2.0

1.5

1.0

Building Society 0.5

0 1967 1979 1973 1991 1985 1987 1968 1969 1970 1971 1972 1980 1981 1982 1983 1984 1974 1975 1976 1977 1978 1993 1994 1995 1996 1986 1988 1989 1990 1992 In the cartel Source: Council of Mortgage Lenders and British Bankers Association era, it looks as • Financial accounts data from time banks began to enter the mort- if building 1979-95 suggest building soci- gage market. In particular, rationing eties generated at least as high a was ended, the cartel abolished and societies were return on assets as did banks. spreads and surpluses increased. In the cartel era, it looks as if The evidence is far from conclu- attempting to building societies were attempting sive, but it looks as if, from the early to reduce mortgage costs for both 1980s, building societies increas- r educe mortgage current and future members. But the ingly behaved like profit-maximising impact of competition on pricing, banks, setting mortgage rates at costs for both costs and product diversity was market-clearing levels. However, largely absent. Behaviour appears to the build-up of reserves in the early current and have changed gradually from the 1990s is difficult to reconcile with

future members Chart 6 BANK AND BUILDING SOCIETY ACCOUNTS PER BRANCH

9

8

Building Society 7

6 Bank 5

'000's 4

3

2

1

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Source: BBA Annual Abstract of Statistics

112 MUTUALITY

the theories put forward above. Two amongst mutual insurers, and it is TABLE 1 COSTS AS A possible explanations are: greater PERCENTAGE OF 1996 clear that mutuality does not guar- risk-aversion by management or, MEAN ASSETS antee greater cost efficiency in all simple inertia. cases. Life assurance companies Institution 1996 Recent evidence on life assurance Abbey National 0.9 Why demutualise? companies indicates that there has Barclays 3.0 Why have some mutuals converted been little difference in growth, Lloyds TSB 2.9 to “plc” status in recent years? How when measured by net premium Midland 2.5 might committed mutuals react to income, between mutual and propri- NatWest 2.7 retain their mutual status? In most etary firms. Alliance & Leicester 3.0 of the cases of demutualisation the The slightly lower growth in Birmingham Midshires 1.3 initiative to convert has been taken net premium income of mutuals Bradford & Bingley 1.2 by management not members. The between 1988-95 is likely to have Britannia 1.0 managers of converting mutuals been a result of the demutualisation Halifax 1.2 have put forward a number of of firms such as Scottish Equitable Nationwide 1.3 reasons for their decision: and Scottish Mutual (see Table 2). Woolwich 1.3 • Access to capital markets will

Comparative measures of bene- Source: IBCA allow converting companies to fits to policyholders are not so pursue plans for expansion and straightforward, but one measure diversification more easily. This which is often used is the level of difference in performance may have argument has primarily been put bonuses paid on with-profits endow- narrowed in recent years. by larger societies, who see their ment policies and pensions. The figures on costs for UK core market as mature with only Survey evidence, such as that mutual life companies do not support modest growth prospects. Mutual produced by “Money Management”, the hypothesis that they are less cost life companies appear to have has been used to argue that mutuals efficient than plcs. Chart 8 shows been less eager to demutualise provide better benefits to policy- that two often-cited cost ratios have to expand their operations. holders (see Chart 7). However, been lower for mutuals than for • Mutuality is increasingly inap- performance varies widely across proprietary assurance companies. propriate as a corporate form both mutual and proprietary insurers. However, as with proprietary compa- because the proportion of non- The data also suggests that the nies, cost ratios vary significantly member customers increases.

Table 2 ANNUAL GROWTH RATE FOR NET PREMIUM INCOME (PER CENT)

1988/89 1989/90 1990/91 1991/92 1992/93 1993/4 1994/5 Compound annual growth rates 1988-95 Proprietary 27.8 8.2 19.0 9.9 8.5 -3.2 -3.4 10.7 Mutual 38.7 10.2 19.1 5.3 4.6 -10.7 0.9 10.2 Bank-owned 42.0 27.4 29.1 29.6 35.6 -6.2 26.2 28.2 Total 33.2 9.8 19.5 8.7 8.3 -6.6 0.4 11.5

Source: DTI

113 MUTUALITY

Chart 7 PAYOUTS ON ENDOWMENT POLICIES: MUTUALS VS NON-MUTUALS

50 50

40 40

Mutuals Mutuals

30 30

20 20

Non-Mutuals Non-Mutuals 10 10

0 0 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 WITH-PROFITS ENDOWMENT POLICIES — NUMBER OF BOTTOM QUARTILE WITH-PROFITS ENDOWMENT POLICIES — NUMBER OF TOP QUARTILE RANKINGS BASED ON ACTUAL PAYOUTS RANKINGS BASED ON ACTUAL PAYOUTS Source: Money Management

As mutuals diversify away from that current members benefit at the used to allocate the reserves on conv- lower-risk businesses, towards expense of future home-owners. Or ersion would benefit the majority of newer and potentially more risky it could have been the provision of members. business areas, it may be more savings facilities to those on low Moreover, the conversion of a appropriate for risk capital to be incomes, so that current members mutual society makes a member’s provided by specialist providers. benefit at the expense of future notional ownership of a share of the • Demutualisation is a necessary small-scale savers. reserves a tradable asset, allowing step in the process of merging There is an incentive for existing him or her to adjust the proportions with a stronger partner in the members to realise the net present of different assets that make up his face of increasing competition value of a society’s activities so that or her wealth. in mutuals’ traditional markets. they do not have to share it with If members of the continuing This argument is most often put future new members (who do not mutuals are only interested in their forward by small and medium- have to pay an entry fee). That way financial returns, then those four sized mutual life companies. they do not need to lose all claim on factors will continue to put mutuals Others have drawn attention to any future benefits when they cease under pressure to convert or sell out the financial incentives available to to be depositors or borrowers. For to a plc. What factors could weigh managers of converting societies. The precisely this reason, there is an on the other side? members have been won over in incentive for new members to join a Mutuals could seek to promote nearly all cases because a majority society and then vote for demutual- their social objectives to their memb- of them stood to benefit financially isation. ers. They need to demonstrate that from conversion. Also, to the extent that the bene- they behave differently from other- If the firm’s objective changes fits to current members have accrued wise similar plcs,and in a manner to profit maximisation from some- in the form of reduced interest-rate consistent with their stated objec- thing else, the net present value of spreads, and hence have been skewed tives; that has not always been clear the firm goes up. The “something towards a relatively few large-scale in the past. else” might have been the extension depositors and borrowers, the more They could also try to show that of home-ownership, for example, so egalitarian distributional principle they are better at aligning the inter-

114 MUTUALITY

ests of managers with owners; the cial value of their reserves without committed mutuals was also strong opposite argument has often been changing the mutual status. Instead, (Table 4). advanced, but a brief look at costs any practical form of pay-out would There are several reasons why does not sustain it. only be able to draw on accumu- the managers of life assurance firms Another way in which a mutual lated surpluses over and above committed to mutuality might face firm could benefit existing members minimum reserve requirements. Such less pressure from members for and yet still expand would be to payouts would resemble a policy or demutualisation than their counter- hypothecate surpluses as they accu- member bonus, paid out of the parts in building societies: mulated to members in proportion surpluses of the mutual firm at the • Survey evidence suggests that to their financial commitment to the end of every financial year. mutual life companies have society. Finally, mutuals could attempt tended to provide better benefits When members joined a mutual to distribute benefits more in propor- (in terms of higher returns on society, they would have no claim tion to members’ voting rights. with-profits policies) than non- on existing reserves; instead they Building societies have taken mutuals. would build up a claim on surpluses steps recently to adjust the balance This is important as the average retained during membership, in prop- in favour of the financial interests of size of the financial commitment ortion to their financial commitment. existing members. Such steps have made by members is larger in Ownership rights could even be included highly publicised reduc- the case of life companies. For passed down from generation to tions in lending rates and loyalty example, if a policyholder has a generation. bonus packages which operate in a with-profits pension which is But, as the reserves of a mutual similar manner to a dividend. As a intended to produce £200,000 are its risk capital, and prudential result, net interest margins were cut on maturity, and a mutual life standards require certain minimum significantly in 1996, and reserve company is able to produce a levels of permanent capital, it is ratios have not risen for the first return which is 2 per cent better difficult to envisage circumstances time in several years (Table 3). than a non-mutual, the potential where members of a continuing However, it is important to note net gains from demutualisation mutual could realise the full finan- that asset growth at several of the may not be significant.

Chart 8 LIFE ASSURANCE COSTS

120 10

Proprietary 9

100 Proprietary 8

Mutuals 7 80

6

Mutuals 60 5 Bank Owned

4

40 3 Bank Owned

2 20

1

0 0 1992 1993 1994 1995 1991 1992 1993 1994 1995 1991 ACQUISITION COSTS/NEW PREMIUMS MAINTENANCE COSTS/TOTAL PREMIUMS

Source: DTI

115 MUTUALITY

Table 3 NET INTEREST MARGINS (NET INCOME AS A PERCENTAGE OF MEAN ASSETS)

Society 1989 1990 1991 1992 1993 1994 1995 1996 Bradford & Bingley 2.0 2.0 2.0 1.9 2.0 2.0 2.0 1.4 Britannia 2.0 1.9 1.9 1.9 1.8 1.9 1.9 1.4 Nationwide 2.5 2.4 2.3 2.5 2.5 2.5 2.5 1.8

Source: IBCA

This contrasts with building soci- compared to the benefits of a eties, where for example, an extra windfall gain. Members may be 1 per cent on a £100 deposit is more reluctant to support demu- relatively insignificant. There- tualisation without being certain fore, there is less incentive for of the costs and benefits. new members to join a life However, pressure for consol- company simply to vote for idation within the industry has been demutualisation, since it is new cited as a reason why mutual life members who stand to gain companies may be under pressure to most from any benefits of mutual Survey evidence demutualise. Although, theoretically, status. a mutual life company could merge, • The costs of acquiring member- suggests that having publicly announced its inten- ship of mutual life companies tion to find a partner, it would in all are higher. Generally, those poli- mutual life likelihood become the subject of cies which provide membership takeover offers from plcs. In this rights require a more significant companies have, situation members may be tempted and longer-term commitment by the prospect of “windfall” gains. than opening a share account in the past, In the longer term, committed with a building society. mutual life companies will need to Whilst policies can be termi- tended to demonstrate that they have the nated early, this often results in capacity to outperform proprietary a loss to the policy-holder, since provide better companies on member returns and the premiums paid in the initial benefits than quality of service. Otherwise, memb- stages of the policy would be ers may seek to realise the financial consumed by costs. Again the non-mutuals value of their ownership rights, and effect of this is to deter the there are shareholder-owned compa- taking-out of policies to engi- nies which are happy to assist. neer demutualisation. • Most life products are more Conclusions complicated than the deposit and Building society assets and member- mortgage business of building ship have grown rapidly in the societies, so that it is more diffi- post-war period. This suggests that cult to determine precisely the management decisions were not financial benefits of mutuality motivated solely by the pecuniary

116 MUTUALITY

Table 4 BUILDING SOCIETIES GROWTH IN TOTAL ASSETS

Society 1990 1991 1992 1993 1994 1995 1996 Bradford & Bingley 26.4 31.6 9.6 6.3 5.0 7.5 8.8 Britannia 17.9 14.8 6.9 8.4 4.4 5.9 8.0 Nationwide 14.4 8.8 2.5 1.2 0.9 4.9 7.5

Source: IBCA interests of existing members. The strong those for building societies. advocates of mutuality would argue Several mutuals are constructing that this is appropriate, as mutuals defences against conversion, but if also have social objectives. But the Several mutuals they are to be successful they will accumulation of reserves in the early have to clearly demonstrate the 1990s, beyond regulatory and future are constructing advantages that mutuality has over growth requirements, is difficult to plc form. reconcile with conventional theories defences against They may be able to define of mutual behaviour. altruistic objectives which win the There are powerful financial conversion, but support of their members; in that incentives encouraging members of if they are to case, their behaviour would have to the remaining building societies to differ from that of competing plcs, support conversion, reinforced by be successful for example, by setting lower interest the structure of voting rights. These rate spreads.■ incentives became more apparent they will have following the recent management- led conversions. to demonstrate NOTES In particular, existing members can capture the net present value of the advantages future expected profits growth which 1 The authors would like to acknowledge they would otherwise have to share that mutuality the helpful comments from Alex Bowen, Glenn Hoggarth, Professor David Miles, with future new members (to whom Alistair Milne, George Speight and any bonus or reduction in interest has over Professor Geoffrey Wood rates spread would also accrue), plc form whilst allowing them to diversify their financial wealth. However, it is less obvious that the incentives which are facing members of life companies are as

117 REGULATORY DEVELOPMENTS Compiled by the Bank of England and the Securities and Investments Board

UK SUPERVISORY DEVELOPMENTS main risks, using the most appropriate “tools” of super- vision. One of the main features of the new approaches is to provide more feedback to banks on the proposed New Regulatory Body intensity of supervision, through the provision of a super- visory programme. Details of this programme will be sent On 20 May it was announced that a single regulatory to them after a formal risk assessment has been completed. body would be created in the UK. This will span banks, Although a more systematic assessment of risk will investment firms, building societies, friendly societies be undertaken, the frameworks have been designed and insurance companies. The details of the proposal to allow the application of judgement by individual are set out in this issue. supervisors. The RATE and SCALE frameworks are being tested on 22 banks during the remainder of this year. Lessons Banking Supervision from this exercise and comments on the consultation papers will be assessed. From next year the new frame- The RATE and SCALE consultative documents works will be introduced to all banks. Following the Bank’s Review of Supervision, two consultative documents have been issued on the intro- Building Societies Act duction of risk-based frameworks for the supervision of The Building Societies Bill received Royal Assent before UK incorporated banks (the RATE approach) and UK the UK general election. It increases the range of activi- branches of non-EEA banks (the SCALE approach). The ties building societies may undertake and the financial papers set out how the Bank proposes to revise its current products they may offer. Their freedom of action is still supervisory approach to provide for more consistent and somewhat constrained. better identification of risks in banks and their wider The Act restricts them to undertake only those groups. activities spelt out in their Memorandums: loans for resi- Both frameworks use nine evaluation factors to dential property must make up at least three quarters of assess risk (six quantitative — Capital, Assets, Market groups’ total assets (with certain caveats). Risk, Earnings, Liabilities, Business: and three qualita- The Act does give building societies more scope for tive — Controls, Organisation and Management) and self-determination on the funding side: they may increase seek to ensure that supervisory action is focused on the their use of the wholesale markets. The greater freedom

118 REGULATORY DEVELOPMENTS

the Act allows building societies means that prudentially has been publishing monthly individual progress statis- less reliance will be placed on statutory restrictions tics for the top 24 firms. From October, statistics for the and more on supervision by the Building Societies 17 additional firms are to be published. Commission. Also in May this year, the SIB sought to speed the review further by publishing a list of criteria that benefit guarantees would have to meet, if they were to be used Securities and Investment Regulation as redress for pensions mis-selling. This was an amplifi- cation of the SIB’s 1994 Guidance on the review, which International commodity futures allows for redress by guarantee, subject to prior approval markets from the regulators on a firm by firm basis. Work has continued during this year to follow-up the Copies of the PIA’s announcement on targets are November 1996 meeting in London of international available from the PIA, and copies of the criteria for guar- commodity futures markets regulators. It has focused on antees from the SIB Publications Unit. surveying the way in which commodity futures contracts are designed and reviewed around the world and exam- Tradepoint refinanced ined prevalent techniques of market surveillance and Tradepoint, the order-driven exchange competing with information sharing. Papers have been prepared on best the London Stock Exchange, secured a new £11.4m (net) practice for the design and review of commodity futures financing package in July which, it believes, will provide contracts, and on techniques of market surveillance and it with working capital for the foreseeable future. The information sharing in relation to commodity markets. introduction of new shareholders has been accompanied These papers were considered at the Burgenstock by a number of board and senior management changes. meeting of Regulators in Switzerland on 5 September and Most of the new capital, which has been subscribed are currently being looked at by the relevant IOSCO as zero coupon convertible, has come from three venture Committee. The aim is to complete this work at a meeting capital funds. Apax Partners has put in £6m (which would in Tokyo on 30 and 31 October 1997. give it 29.2 per cent of the issued share capital on conver- sion), Electra has subscribed £1.5m and Smedvig Review of pensions mis-selling £500,000. Other subscribers include three inter-dealer In November 1996, the SIB issued new guidance aimed broker firms (IDBs), which have subscribed a total of at speeding up the provision of redress to people wrongly £2.5m and entered into commercial trading arrangements sold personal pensions. The SIB indicated that, once with Tradepoint. The LSE has since announced signifi- firms had had time to assimilate the guidance, front-line cant changes and reductions to its own charges from regulators would determine realistic new target dates with October, when it introduces its new trading system. firms. In May this year, the PIA announced a new timetable London Stock Exchange for the review. Firms are now required to complete 90 per transparency cent of their most urgent cases by the end of 1997 and the Trade publication arrangements are being put in place by rest by the end of 1998. In addition, the 24 firms with the London Stock Exchange for the introduction of its the greatest number of cases to review were set, and new electronic trading service — SETS — which aims to agreed to meet, demanding individual targets. Targets for ensure that it is one of the most transparent equity a further 17 firms were published in September. Failure markets in the world. The LSE has proposed that all to meet targets could provide grounds for disciplinary trades should be published as soon as they have been action by the regulators. Since the targets were set, the executed. For very large orders, there is to be a special Economic Secretary to the Treasury, Mrs Helen Liddell, worked trade regime which will allow LSE intermedi-

119 REGULATORY DEVELOPMENTS

aries to offer clients a guaranteed base price for very large Guidance should assist investment firms to run their busi- orders and then to work those orders for improvement ness in accordance with regulatory standards and may be before agreeing the final price. Key aspects of the used where appropriate for the purpose of intervention Worked Principal Agreement (WPA) which the LSE has and discipline. agreed with the SIB are: - orders qualifying for use of the WPA regime must be Cost-benefit analysis at least 8 x Normal Market Size (which equates to 20 In June the SIB completed a new text on regulatory effec- per cent of a full day’s customer trading). tiveness, Protecting Investors by Enhancing Markets: A - orders executed as part of the working must be Guide to the Role and Nature of Regulatory Cost-Benefit published immediately, regardless of size. Analysis. The guide draws lessons, and uses examples, - the originating order must be executed — and from the SIB’s recent experience of cost-benefit analysis published — once 80 per cent has been worked or at and related disciplines. It characterises cost-benefit the end of the day, whichever is sooner. analysis as an aid to policy makers in the form of an The Exchange estimates that no more than 15 per cent impact study. It shows how the market analysis and quan- of trading by value should fall within the WPA regime. titative techniques involved can help the development of new regulatory measures and assessment of existing Responsibilities of senior management ones. It describes the range of benefits that regulation can The SIB published in July a consultative paper on bring. the responsibilities of senior management in financial The guide notes that cost-benefit analysis and its institutions for establishing, understanding and applying related disciplines are not exact sciences. Their value lies adequate internal controls. It proposes to focus the boards in making explicit all the likely impacts (costs and bene- and management of investment firms on the need for fits) of a regulatory measure and in indicating the broad adequate internal controls in two ways: extent to which the benefits exceed the costs (or vice - by issuing guidance on management standards in versa). The guide also sets cost-benefit analysis in its regulated firms; wider context, which includes the PIA’s work on effec- - by requiring firms to draw up a statement on their tiveness and the government’s extensive use of regulatory management structure. impact studies. These initiatives, and those of the SIB, A firm’s statement on its management structure have shifted the focus from the limitations of these disci- should be available at all times for the regulator to see plines to how much they can deliver. and will have to provide details of the individual respon- Copies of Protecting Investors by Enhancing sibilities of the firm’s directors and senior managers. Markets: A Guide to the Role and Nature of Regulatory They in turn will be expected to acknowledge that they Cost-Benefit Analysis are available from the SIB’s cost- are aware of and understand their responsibilities. The benefit analysis department (tel: 0171 638 1240, ext 2145). SIB believes that this requirement will impose no more than a minimal burden on well-run firms. The level of Open-ended Investment Companies detail which the statement will need to go into will In May the SIB published proposals for regulation of the depend on the complexity of the organisation; some second stage of open-ended investment companies small firms may be exempted from the requirement (OEICs). Currently, only OEICs which invest in trans- altogether. ferable securities are permitted. The second stage would The SIB Guidance is intended to help firms ensure bring the range of investments available to OEICs in line that their management standards comply with the SIB with those available to authorised unit trusts, (ie money Principles, especially Principles 9 (internal organisation), market funds, funds of funds, derivatives and property 1 (integrity), and 2 (skill, care and diligence). The funds). That consultation is now complete. Further devel-

120 REGULATORY DEVELOPMENTS

opments are dependent on the government’s decision on equate instructions on how to identify pension cases whether it wishes this second stage of OEICs to be requiring a review and for failing to provide adequate progressed. procedures and resources to monitor its appointed repre- In August the SIB made further proposals on collec- sentatives’ conduct of the review. tive investment schemes. First, proposals to develop a set Other PIA firms which were disciplined and the of draft standards which a scheme would have to meet amount of their fines are: Glinthurst Insurance and if it wanted to call itself “guaranteed”. Key among these Financial Consultants (£5,000); John Wood Services Ltd would be the need for a legally enforceable guarantee, (£3,500); BMA Services Ltd (£30,000); Grosvenor with the guarantor being subject to prudential supervi- Butterworth Financial Services Ltd (£2,500); Taylor sion. Second, the SIB proposed permitting single pricing Graham Financial Planning (£2,500); The Independent for unit trusts, thus aligning the unit trust regime with that Consultancy Group (£2,500). applicable to OEICs. Further work is being done on John Jackson Insurance Services was expelled from whether single pricing should, ultimately, be mandatory. PIA membership. The SIB also proposed minor amendments to the invest- ment rules in the light of events at Morgan Grenfell. Fidelity Brokerage Services (FBS) On 9 May the Securities and Futures Authority (SFA) Disciplinary action by the Personal Investment fined Fidelity Brokerage Services (FBS) £200,000 with Authority (PIA) a contribution of £162,500 to the SFA’s costs. This was Between January and September 1997 the PIA fined 10 the culmination of a series of enforcement actions over firms a total of £716,000 and expelled one firm from the previous year which arose from accounting and membership for delays in carrying out the Pensions Review. reconciliation problems and large numbers of customer DBS Financial Management Plc was fined £425,000 complaints associated with a new computer system and plus £19,450 costs for rule breaches in connection with increased volume of business. On 31 October 1996 the the review of past pensions business. The firm admitted SFA announced that FBS had entered into undertakings that it had failed to take all reasonable steps to carry out not to take on new direct customers or introduce new its review of the past pensions business transacted by its business services until the SFA was satisfied with its appointed representatives and to monitor the review of customer service performance. These undertakings were pensions business transacted by its appointed represen- renewed at the end of January 1997, at which stage the tatives prior to joining DBS. SFA announced that it intended to initiate disciplinary The M & E Network Limited was fined £100,000 action once the outstanding problems had been resolved. plus £25,000 costs for failing to take all reasonable steps The SFA closely monitored FBS’s progress in to carry out a review of past pensions business in accor- resolving the problems, as well as its handling of customer dance with the standards and specifications prescribed by complaints and payment of compensation. The SFA lifted the PIA. In particular, the firm admitted that between the restrictions on 9 May 1997, at which time FBS 8 March and 21 August 1996 it delayed the mailing of acknowledged breach of SIB Principles 2 and 9, and pension review questionnaires. disciplinary proceedings were brought and settled. The Lincoln Independent Limited was fined £75,000 plus SFA acknowledged that investors money and securities £10,000 costs for failing to monitor adequately its were not in any jeopardy. employees’ compliance with the firm’s procedures in relation to the pensions review and, therefore, the firm’s Swiss Bank Corporation own compliance with the rules of the PIA. Berkeley On 28 August 1997 the SFA announced the settlement Independent Advisers Ltd was fined £70,000 with £15,000 of two disciplinary cases against Swiss Bank Corporation costs for issuing its appointed representatives with inad- (SBC). The SFA severely reprimanded SBC and fined it

121 REGULATORY DEVELOPMENTS

£300,000 with a contribution of £121,095 to the SFA’s IMRO rules. IMRO considers there to have been inade- costs in relation to transactions involving shares in quate management control and that the affair plainly regional electricity companies and cash performance illustrates the dangers of ignoring clear and repeated notes referenced to those shares during late 1994. SBC warnings. IMRO also makes clear that it expects that acknowledged that it failed to observe high standards of other investment managers will ensure that they are not market conduct in its acquisition of a long position in exposed to the same risks. Yorkshire Electricity Plc, that it failed to implement, monitor and control its Chinese Wall procedures, and that it failed to operate well defined compliance and EU SUPERVISORY DEVELOPMENTS supervisory procedures. The SFA for its part accepted that the breaches were not wilful or intentional, and that SBC EC Commission proposals for new measures acted in good faith. to protect consumers of financial services The SFA also severely reprimanded SBC and fined In June this year the European Commission published a it £180,000, with costs of £55,000, in relation to the liqui- “Communication” on Financial Services: enhancing dation of the Kleinwort European Privatisation Trust Plc consumer confidence. It sets out its proposals in response (KEPIT) in October 1996. SBC admitted that it failed to to the public consultation exercise it organised over the act with due skill, care and diligence, and that it failed past 18 months on what new action at EU level might be to ensure fair treatment to KEPIT’s fund manager in rela- required to strengthen the rights and protections of prin- tion to its disposal of KEPIT’s portfolio holdings in cipally private consumers in the Single Market in France and Spain. The SFA for its part accepted that SBC financial services. The wide ranging action programme did not set out to disadvantage its client and took prompt for the coming months includes a new directive on the steps to investigate the matter internally and to correct the “distance selling” of financial services, an updating of position. a 1976 Directive on insurance intermediaries, consider- ation of a new directive on “unregulated financial Morgan Grenfell Group intermediaries”, a review of a 1987 Directive on The Investment Management Regulatory Organisation consumer credit, an update of the 1988 Recommendation (IMRO) announced on 16 April that it had fined two on new means of payment, and Commission monitoring companies in the Morgan Grenfell Group £2m in relation of a number of other areas with a view to deciding to breaches of IMRO’s Rules and the SIB Statement of whether further formal measures are necessary at EU Principles. The companies were also required to pay £1m level. The various UK authorities will in turn be moni- in costs. The level of fine reflects in part the number toring developments and contributing to negotiations. of investors affected and amounts of compensation involved. It also takes into account Morgan Grenfell’s ISDN Directive prompt response in agreeing the compensation package The UK regulatory authorities have worked with HM announced by IMRO on 20 December 1996. The fine Treasury and the DTI to ensure that current practice in relates to the management of Morgan Grenfell European the UK of recording of telephone calls by firms, Growth Trust, Morgan Grenfell European Fund and a exchanges or regulators for legitimate commercial and Dublin-based fund, Morgan Grenfell European Capital regulatory purposes, can be maintained under an EC Growth Fund. Directive on the processing of personal data and protec- IMRO’s charges related to breaches of the SIB’s tion of privacy in the telecommunications sector; and to Principles 2 (skill, care and diligence), 9 (adequate minimise any adverse implications of the Directive for compliance arrangements) and 10 (keeping a regulator “cold-calling” rules to which firms may be subject in the informed), as well as breaches of a number of specific UK and in other Member States.

122 REGULATORY DEVELOPMENTS

European standards of fitness G30 Report on “Global Institutions, National and properness Supervision and Systemic Risk” Through the Informal Group of Chairmen of EU A Group of Thirty report on the management and super- Securities Commissions, the SIB, together with the Bank vision of global financial institutions has been published. of England and the SROs, has launched an initiative to It concludes that an industry initiative is needed to explore with other EEA competent authorities the scope promote consistently high standards of risk management for agreement on certain common standards and proce- and control for global institutions with substantial inter- dures in assessing the fitness and properness of national operations. A detailed description of its contents investment firms (and relevant individuals in them) to can be found in the article, Global Institutions, National provide investment services. This is intended to flesh out Supervision and Systemic Risk, in this issue the relevant requirements laid down in the Investment Services Directive and Second Banking Directive, and to G10 Core Principles strengthen arrangements for co-operation and sharing The Basle Committee on Banking Supervision has now information on firms and individuals, as far as national finalised its “Core Principles for Effective Banking laws permit. Recent meetings have provided an oppor- Supervision”. These represent the first time that the inter- tunity for participants to gain a better understanding of national supervisory community has attempted to each other’s approach, criteria and formal powers in this formulate comprehensive principles for the conduct of area, and to identify points of similarity and difference. banking supervision in general, as opposed to particular, Work continues. aspects of supervision. The Core Principles were drawn up by representatives of the Basle Committee and of a number of emerging market economies. DEVELOPMENTS IN OTHER The 25 Core Principles cover all aspects of banking INTERNATIONAL FORA supervision, from licensing through to the powers of supervisors and co-operation between supervisors from Meeting of the G7 in Denver different countries. They are designed to operate as high At Denver, the G7 Heads of State and Government level principles. They do not require that the practice of welcomed the further work that has taken place to supervision follow any particular model, but can be enhance co-operation among supervisors, particularly of adopted to fit a wide variety of national circumstances. globally active financial institutions, on an ongoing basis The Core Principles were formally launched at a and in emergencies. meeting held in Hong Kong during the Annual Meetings The Joint Forum of banking, securities and insurance of the World Bank and the International Monetary Fund supervisors has agreed that, in appropriate circumstances, a co-ordinator should be identified to facilitate the exchange of information. Work is continuing to develop REGULATORY DEVELOPMENTS IN OTHER the roles of such a co-ordinator. COUNTRIES G7 finance ministers agreed to support changes in laws or regulations that would improve information exchange for supervisory purposes, while preserving the Australia confidentiality of information exchanged. Finance ministries are now assessing impediments to the sharing The Wallis Inquiry of information, on the basis of work by banking, securi- On 2 September 1997 the Australian Government ties and insurance supervisors. This will be on the agenda announced it’s response to the Wallis Inquiry report on for the Birmingham G7/G8 Summit next spring. Reform of the Australian Financial System which had

123 REGULATORY DEVELOPMENTS

been published in April. Responsibility for financial regu- In particular, it abolishes the restrictions on commer- lation will be divided between three agencies: cial banks’securities activities and permits them to derive - The Reserve Bank of Australia will be charged with a limited amount of revenue from non-financial activi- looking after the stability of the financial system. As ties. The bill does however maintain many restrictions on well as financial stability, it will be responsible for the commercial banks’ insurance business. It would also regulation of payments systems (as well as conduct allow non-financial firms (other than, in effect, the top of monetary policy). 1,000 largest US companies) to derive a limited amount - The Australian Prudential Regulation Authority will of their revenues from banking activities. undertake the prudential regulation of deposit taking The bill would create a body called the National institutions, life and general insurance companies and Council on Financial Services which would have the superannuation funds. authority to define financial products and to determine - The Australian Corporations and Financial Services permissible financial activities. Commission will have responsibility for regulating Substantial amendments were made to the bill when the integrity of market conduct, consumer protection it was further debated in the House Commerce Committee. and corporations. These changes may alienate some of the bill’s existing A further body called the Council of Financial supporters and could adversely affect its chance of Regulators will be established on which each of the success. Even if approved by the Commerce Committee, individual regulatory bodies will sit. Its purpose will be the bill must overcome a number of obstacles before to ensure co-operation between the three new bodies and, facing a vote in the House later this year or early in in addition, it will aim to minimise the compliance costs 1998. to firms of the new regulatory arrangements. It is currently envisaged that the changes will be introduced by the middle of 1999. A number of changes Japan to the payments system are also planned. Access to the payment system is to be widened beyond licensed banks Big Bang and deposit taking institutions. A body called the In November, the prime minister, Ryutaro Hashimoto, Payments System Board will be created within the proposed a wide-ranging set of reforms to the regulations Reserve Bank of Australia. This body is to be charged that govern financial activity in Japan. with responsibility for improving the efficiency of the These proposed reforms have come to be known as payments system. Legislation is also to be introduced to “the Japanese Big Bang”. They aim to improve Tokyo’s minimise systemic risk arising from the failure to settle standing as a financial centre, to make it “free, fair and by individual participants in the payments system. global” and comparable to London and New York by the year 2001. The initiatives are wider than was the case in the UK United States Big Bang in the mid-1980s (which was limited solely to the Stock Exchange). They include the liberalisation of Glass-Steagall reform foreign exchange, as well as revisions to the tax system The Banking and Financial Services Committee of the and accounting standards. US House of Representatives narrowly approved a finan- A number of measures will become effective from cial services reform bill in June. This bill would April 1998, one of the key elements of which is imple- dismantle the provisions of the 1933 Glass-Steagall Act mentation of the Foreign Exchange Law. This will which separate commercial and investment banking in completely liberalise cross-border transactions, and so the US. accelerate the trend towards deregulation.■

124