Deutsche Bank Markets Research

Industry Date 8 September 2013 Retail Bank Strategy Global

Financial Services

Jason Napier, CFA David Lock Research Analyst Research Analyst (+44) 20 754-74433 (+44) 20 754-11521 [email protected] [email protected]

F.I.T.T. for investors The Future of Banking

Acceleration of change leads a big increase in bank IT spend With the press full of reports on innovation in money matters - think progress by Square, volumes in mobile payments at M-PESA, doubled revenue at Monetise, new bank launches like Metro, big data, digital wallets, mobile banking, crowdfunding, P2P lending - we think most investors see mostly downside risk from change. We look at each topic and conclude that a big increase in bank IT spend is coming which isn't captured in market expectations. But we also think this spend will deliver an ability to process higher volumes at lower unit costs, reduce branch costs and generate new revenue streams. We do not expect material bank disintermediation by tech.

______Deutsche Bank AG/London Deutsche Bank does and seeks to do business with companies covered in its research reports. Thus, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MICA(P) 054/04/2013.

Deutsche Bank Markets Research

Global Industry Date 8 September 2013 Financial Services Retail Bank Strategy FITT Research

Jason Napier, CFA David Lock Research Analyst Research Analyst The Future of Banking (+44) 20 754-74433 (+44) 20 754-11521 [email protected] [email protected]

Acceleration of change leads a big increase in bank IT spend With the press full of reports on innovation in money matters - think progress by Square, volumes in mobile payments at M-PESA, doubled revenue at Monetise, new bank launches like Metro, big data, digital wallets, mobile banking, crowdfunding, P2P lending - we think most investors see mostly downside risk from change. We look at each topic and conclude that a big increase in bank IT spend is coming which isn't captured in market expectations. But we also think this spend will deliver an ability to process higher volumes at lower unit costs, reduce branch costs and generate new revenue streams. We do not expect material bank disintermediation by tech. Industry rate of change IS rising There’s no doubt that retail banking is seeing the start of material change. Mobile banking, for example, has average customers interacting over 20x a month, serving themselves more and creating upside from digital sales, efficiency gains, bank expansion into products like those offered by Groupon generating new revenue, supporting loyalty and (maybe) reducing churn. Re-tooling and IT investment will happen, nudged by regulatory pressure Most big banks need significant IT renewal. Old mainframe systems are the norm, patched for M&A, geographic & product expansion, and regulation, which introduces complexity and inefficiencies while volumes continue to rise. Unless expensed under restructuring charges this initially drives costs as much as 10% higher. Regulatory pressure will increase exponentially here given resolution planning and operational risk assessments, we think. A fundamental branch rethink to happen Digital banking has branch footfall down ~15% YoY. Combined with automation of cash dealings we expect new branches to be 25% smaller with 20% less staff, fewer tellers and more advisors. They’ll be cheaper to fit out, easier to move, faster to break-even, and make for better provision of advice. Outside of overbanked markets (e.g. Italy, Spain), we expect branch design not numbers will be the bigger change. With 50% of retail costs in branches and operations we think expenses here could fall ~5%. Customers and IT firms win… Out of this the customer does best, enjoying better ways to bank and spend. New banks, courtesy of low startup costs will generate profile, some churn, and motivation for incumbents to improve. As a thesis we think IT and data firms – at 15x - 30x EPS – do better out of bank IT change than the banks. …and banks get to keep being banks unless a killer app emerges However, we can’t envisage the emergence of a killer non-bank app that banks are unable to replicate which disintermediates the banking value chain materially. We expect bank IT renewal to increase costs initially but deliver platforms able process higher volumes at falling unit costs, meaning falling cost/income ratios as revenues grow (esp. when rates rise) and reduce customer incentives to churn, driven by better products and better delivery. We value banks using a combination of P/TNAV and P/E, benchmarked against peers. Key upside risks relate to economic growth and interest rates; downside risks relate to growth and regulation in particular. ______Deutsche Bank AG/London Deutsche Bank does and seeks to do business with companies covered in its research reports. Thus, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MICA(P) 054/04/2013.

8 September 2013

Financial Services Retail Bank Strategy

Table Of Contents

Executive Summary ...... 4 This isn’t low-cost-airline-meets-flag-carrier situation ...... 4 The Future of Retail Banking ...... 11 Vanilla banking returns are good, should improve as costs are cut, more significantly when rates eventually rise ...... 11 Every so often it's a good idea to look much further ahead ...... 13 The Rotten Core ...... 15 Key points ...... 15 Banks mostly move and keep track of where the money is ...... 16 Rotten Core – Not all rotten, but generally pretty old ...... 17 Significant implementation risk, little near term gain ...... 18 A twist on the IT transition approach: The parallel bank ...... 21 Conventional competition from the Bank in a Box ...... 22 We expect a significant increase in IT renewal activity ...... 25 Rise of the machines: Mobile & Tablet Banking ...... 33 Technology is changing the way we manage our finances ...... 33 What the Heads of Digital at banks are thinking ...... 37 Mobile Payments: Already in a phone near you ...... 43 Overview ...... 43 M-PESA - Mobile payments (not banking) at its most successful ...... 43 What M-PESA says about the future of retail banking ...... 46 Moving to a less cash & cheque dominated world...... 48 Growth in digital wallets ...... 52 Overview ...... 52 Card model is moving towards a digital wallet, transaction focus...... 52 The key question: Is there a killer wallet out there? ...... 55 Where to for branches? ...... 60 Overview ...... 60 Branches are the “face” of the bank, 50-60% of costs… ...... 61 Falling branch numbers are a structural trend ...... 62 Branches to stay, but in tighter format as advice centres ...... 64 More outsourcing: The ATM case study ...... 66 There will be adjustment pains ...... 67 Big Data: Unleash its commercial value ...... 71 How big is Big Data? ...... 71 Who knows more about your habits than your bank? ...... 71 Investors and customers haven’t been given much reason to suppose banks can harness the data they possess ...... 72 Non-bank service companies show the way ...... 73 Banks must use their data to grow revenues, defend customer wallet share and drive higher customer loyalty ...... 73 Democratisation of your Big Data could increase churn rates and lower margins in the longer term ...... 76 Crowd- & Peer-to-Peer: Cutting out the middlemen ...... 77 Key points ...... 77 There’s a time and place for everything ...... 78 Focussing on bank-substituting lending ...... 79 How peer-to-peer and crowdfunding works ...... 81

Page 2 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Table Of Contents (Cont'd)

Payments: plastic up, cash flat, cheques out ...... 84 Key points ...... 84 Payments: central to all banking and finance ...... 84 What is a payment system? ...... 85 How payments profiles have changed: non-cash is growing fast, transaction volumes rising ...... 86 Rise of the card: A bankable trend ...... 88 The traditional card payment model ...... 91 How fees flow around the card model ...... 92 Large value payment systems – would always be run by banks ...... 97 Regulation – Evolving ...... 100 Cool Innovations ...... 102 No sustained advantage in product, but we liked these ...... 102 Glossary ...... 104

------

Our thanks are due to the management of banks, technology and financial services companies, and consultants and regulators we met while researching this piece. This has proven been one of the most interesting research processes of our time covering banks and, though this report is necessarily only a first step of covering an evolving story, we look forward to tracking the identified trends in future.

We hope that this piece provides bank investors with better insights into trends in technology and technology investors with better knowledge of trends in banking.

Any residual errors are naturally our own.

Finance Director, Aldermore • Chief Information Officer, Aldermore • Chief Operating Officer, Allied Irish Banks • Chief Executive Officer, Australia, The Australia and New Zealand Banking Group • Chief Operating Officer, Special Resolution Unit, Bank of England • Head of Digital Channels, Bank of Ireland • CEO, Barclaycard • CFO, Barclaycard • CEO, Barclaycard Global Business Solutions • Head of Investor Relations, BNP Paribas • Partner, Boston Consulting Group • Digital Banking Lead, Deloitte • Head of Unit, Risk Analysis, European Banking Authority • Director of Investor Relations & Communications, Experian • Managing Director, Marketing Services, Experian • CEO, FirstRand Bank Ltd • Co-Founder, The Funding Circle • Head of Investor Relations, Handelsbanken • Managing Director, Digital, Lloyds Banking Group • Chief Executive, Metro Bank • Chief Operations Officer, Metro Bank • Strategy Director, Monitise • Director of Market Intelligence, Monitise • Senior Manager, Investor Relations, National Australia Bank • Policy Advisor: Innovation and Economic Growth, NESTA • Director, NoteMachine • Managing Director, Digital, RBS • Managing Director, Digital Banking, Santander • UK Group Head of Strategy, SAP • Deputy CEO, Standard Bank of South Africa • Chief Information Officer, Standard Chartered • Director of Strategy and Marketing, Temenos • Head of Strategic Planning, Temenos • Chief Executive, Australian Finance Services, Westpac

Our thanks are also due to Atul-Subhash Hanamante of Evalueserve for his assistance in compiling this report.

Deutsche Bank AG/London Page 3

8 September 2013

Financial Services Retail Bank Strategy

Executive Summary

This isn’t low-cost-airline-meets-flag-carrier situation

Technological change, deleveraging and IT failures see calls for the disintermediation or replacement of existing banks by new more efficient players, including from entrants from other sectors. An array of tech and telco themes around digital wallets, mobile payments, crowd-funding, peer-to-peer lending, big data and core banking system replacements compete for column inches with bank plans to cut costs.

Are retail banks the good compounding long term investments many think they are? Or does change in technology drive incumbent disintermediation by new banks with lower cost structures and non-banks in niche profitable areas like payments? This report takes a bank-analyst look at each issue from a longer-term perspective.

Our key conclusions are:

„ The customer is the clear winner from technological change. Financial dealings with banks, business and friends will get easier. Retail special offers will become tailored to customer habits and circumstances. Financial planning and tracking will become more automated and transparent. Being organised will be easier.

„ We expect market shares in key elements of bank value chain to remain intact. This is because we don’t envisage the emergence of a killer app by a bank or non-bank which incumbents are unable to deploy for the benefit of their already-scale customer base. The speed with which new digital tools become industry standard (and therefore commoditised hygiene factors rather than real drivers of market share shifts) is amazing. Take mobile banking – how many banks don’t offer it now? If anything, banks have advantages over non-banks in scale of customer base.

„ We expect a material increase in IT spend by banks over the next ten years. Core systems are generally old and rely on too many applications patched too many times to cope with rising transaction volumes, regulatory change and digital channel changes in particular. This will drive up to a 10% increase in overall operating costs and will be encouraged by an exponential increase in regulatory attention given to operational IT risk. If this were the only investing thesis, we’d buy the vendor not the bank.

„ Owners of new niche financial service providers will do well where scale and profit is shown (look at the P/E and EV-EBITDA multiples in Figure 5). That said, we don’t expect crowdfunding and P2P operations to take observable market share from banks. We think these models will do well until higher interest rates or an economic downturn intervenes.

„ We expect some new bank promoters in good markets will do well given low start- up costs. Software-as-a-service providers have new banks up and running for US$10m to US$20m, with cash burn minimised by per-customer and lease- arrangements for hardware and software. Though generating customer gains and turning in a decent net interest margin in the current rate environment is a challenge, the hurdle to break-even isn’t so challenging as to keep new players like Aldermore and Metro Bank from generating good returns for their owners, we think.

„ For incumbent banks we expect technology costs to rise and branch costs to decline for a roughly neutral outcome overall. This ignores phasing – IT costs rise faster than branch costs decline. Banks that execute best will cope with much higher customer transaction volumes with little cost increase, will be nimbler in pricing and product launches, more adept at adapting the branch network to changing needs and will drive increased customer loyalty, higher wallet share and

Page 4 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

lower churn. We see Big Data as a key driver of this in time, given the requisite investment by banks. Combined with the very significant upside to returns when interest rates rise we think that retail banking remains a generally attractive investment prospect. Below we summarise the key findings from each chapter of the note.

Rotten Core: A decade of IT renewal is in order; buy the vendor not the bank Most established banks operate on mainframe technology implemented in the 1970’s and 1980’s, added to over the decades to deal with M&A, product proliferation, geographic expansion and (recently) regulatory change. Very few banks have run the gamut of core IT change so far. None we spoke to thought the cost of core renewal could be justified by efficiency gains alone. We look at core projects underway at Standard Bank and NAB. The former took its replacement decision in 2005, isn’t finished and has missed targets. The latter is planned to increase NAB’s depreciation expense by 40% over 2010’s level. We think that increasing transaction volumes (hastened by the rise of plastic and digital banking), increased regulatory pressure (new rules, a reaction to high-profile failures at RBS and elsewhere, and fears around cyber- threats), and further aging of systems will force more banks to act. On IT renewal as a thesis on its own we’d rather own the vendor than the bank. SAP derives 9% of revenues from its financials vertical, Temenos is all-bank and profiled in our case study of “bank-in-a-box” new entrants like the UK’s Metro Bank. We look at new-brand launches at BNP Paribas and NAB as a potentially interesting angle on multi-brand and platform transition.

Rise of the machines: Mobile & Tablet Banking, ATMs The consumer move to the small screen, topical in recent Google and Microsoft results, is evident in banking where mobile and tablet banking is growing much faster than executives expected, with digital customers transacting 20x a month in the UK and Australia. Mobile customer penetration in developed markets is ~ 25% - 50% while online (computer only) is nearing maturity. We expect mobile to continue to grow until virtually all core customers participate, providing good market trends for third party providers like Monitise. Banks will see customers rebalance from phone and on-line banking whilst ATMs and deposit machines increasingly automate branch cash dealings, boosted by outsourcing to firms like G4S and NoteMachine. Digital as a sales channel contributes sub 10% of account sales and will grow, supported by a rethink around branch purpose and design. We think conduct risk will be a growing issue worldwide and don’t see digital as a cast-iron panacea.

Mobile Payments: Telcos won’t displace banks in payments in most markets Pay to mobile and pay to account from smartphone are fast becoming ubiquitous bank services (e.g. Barclays Pingit, Bank of Ireland’s Pay to Mobile, CBA’s kaching, NAB’s nabKiss and RBS’ Pay Your Contacts). Other companies – including mobile operators – are developing competing offerings. We study M-PESA, arguably the most successful mobile payments system in the world, run by Vodafone’s Kenyan affiliate, Safaricom. M-PESA processes payments worth ~ 30% of Kenyan GDP (and growing fast), has two thirds of the country’s adult population as customers, and generated US$255m in revenue last year. Rather than seeing this as proof that banks will be sidelined, we see the real lesson that, aside from certain EM/Kenya-specific drivers of M-PESA’s success, that value added digital services can increase revenue, attract clients, reduce customer churn and help improve customer perceptions of their banks. JPMorganChase’s experience here in digital customer perceptions is a case in point.

Deutsche Bank AG/London Page 5

8 September 2013

Financial Services Retail Bank Strategy

Digital wallets: Lots of wallets, no killer disintermediation app (yet) We see growing convergence of consumer payments around digital wallets. We think mobile wallets have the potential to become the ‘gateway’ through which consumers make the majority of retail and peer-group payments. Here there are new non-bank entrants including tech companies (Google, Apple, Samsung), retailers (McDonalds, Starbucks), payment networks (Visa, Mastercard) and mobile phone operators (Everything Everywhere, Isis). We think the emergence of a single all-dominating wallet is unlikely, given the wariness of all invested parties to the risks here, and bank’s ability to rapidly deploy copies of good ideas to bank’s scale customer bases. The pace at which tools like mobile banking have become commoditised is noteworthy.

Figure 1: Mobile & Digital wallet convergence Figure 2: P2P and crowdfunding industry structure

Digital wallet Account Traditionalservices bank Card issuer management

Mobile wallet Merchant Cash Acquirer payments Credit card stakeholders card Credit Payment Cheques Networks

Mobile Mobile Internet Retailers phone aggregators PoS operators

New entrants Source: Deutsche Bank Source: Deutsche Bank

Future of the bank branch: Smaller, smarter, cheaper Phones put banks in pockets and cards take cash out of wallets so what’s the point of the 689,144 branches in the 66 markets we review? Some countries (Italy, Spain) are over-served by existing networks but overall we expect branch numbers to edge lower but not collapse. In DM we expect new branches will be 25% smaller, show 20% lower headcount, fewer tellers, more advisory staff and more automation. They’ll likely be cheaper to set up, easier to move and faster to breakeven. Branch strategy at JPM, Westpac and NAB are examples. With 50-60% of retail banking costs in the network there is 5-10% of bank cost saves to be had. Upfront costs apply.

Figure 3: How NAB’s channel volumes have changed Figure 4: Core Banking platforms are a complex web

Source: National Australia Bank Source: The Boston Consulting Group

Page 6 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Big data: Who knows more about you than your bank? Advances in computer processing make it feasible to analyse vast quantities of data for better insights into customer needs and habits, yielding better products, marketing and supplier profits. The market appears prepared to pay for data and contact-driven businesses – Groupon trades at 35x 2014 EPS, Experian at 19x, Rightmove at 28x, Moneysupermarket at 15x – but there’s little written on customer data in banks. This is because there’s little evidence it is used for much more than blunt product offers and credit risk. Most banks don’t even provide decent spend-tracking tools. This will likely change. Barclaysbespoke is one of the first offerings we’ve seen to co-ordinate customer data and marketing services. NAB’s PeopleLikeU is interesting. LBG’s agreement with Cardlytics will make tailored retail offers a reality pretty soon. In time we expect almost all banks to provide tailored advertising and offers to customers by mobile: (linked to location: “You’re near Thomas Pink on London Wall (and we know you’re wealthy and have shopped there before), if you go in now and use your ABC Bank credit card you’ll get 25% off”), online and via statements. We think big data will allow banks to price properly for customer wallet share in the same way as motor insurers will move to Pay-How-You-Drive premiums which is good until democratisation of customer data brings much faster churn and margin compression in the long term in some markets. Meanwhile development of marketing service businesses should be most supportive for the larger banks: the more data the better.

Crowdfunding: Cutting out the middlemen Peer-to-peer and crowd-funding businesses are fascinating. Operating using eBay-like auction technology, these aim to disintermediate banks, get credit to borrowers more efficiently, and provide savers with higher risk-adjusted returns. We take a detailed look at The Funding Circle. Other similar businesses include Indiegogo, Symbid, crowdcube, Seedrs and Kickstarter. We think MarketInvoice is clever. We expect returns to owners of these enterprises will outpace those of their customers and the banks with which they compete. With platforms earning around 3% of loans extended with no capital at risk, we see significant value upside for P2P business owners but do not see the model as a significant risk to future bank volumes or returns. We are wary of how customers will fare in a credit downturn and are interested to see the UK government providing business funding via this channel.

Payments – Plastic up, cash flat, cheques out; Buy Mastercard, Visa Global cash outstanding is growing by about nominal GDP lead by EM. Cheques continue to die a slow death and we think medium term governments should assist in eliminating them to improve economic efficiency. The bankable trend in payments is the ascendance of plastic, especially debit cards, which grew by 40% between 07-11 whilst credit cards fell 22%. Transaction volumes are rising and average transaction sizes falling as cards are more widely held, more broadly and easily accepted for smaller transactions in the real world, as on-line grows and countries like Sweden move beyond cash for commerce. These trends underline our Buys on payment networks Mastercard and Visa and support the outlook for bank revenues in this area.

Cool innovations & glossary As we’ve studied some of the advances in technology and product in compiling this note we’ve made a list of innovations we think are particularly interesting. We don’t think there’s enduring competitive advantage in product. Imitation is, after all, the sincerest form of flattery. We were fascinated by a number of advances we came across which we expect will benefit customers generally over time. We list these in a separate chapter in brief, and also include a Glossary for readers to refer to.

Deutsche Bank AG/London Page 7

8 September 2013 Financial Services Retail Bank Strategy

DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB Bloomberg Bloomberg Bloomberg Bloomberg Data source Data source

n/a n/a n/a n/a n/a n/a 1.0 2.7 0.4 0.3 0.4 1.9 0.4 0.4 0.2 0.2 0.2 0.4 1.2 1.1 0.9 0.5 1.1 2.4 0.3 13.5 Debt 131.6 EBITDA/ 135305.8

n/a n/a 0% 0% 0% 0% 0% 0% 3% 37% 15% 87% 97% 10% 65% 94% 93% 97% 63% 37% 37% 40% 56% 24% 25% 44% 428% 109% Debt/Equity

n/a 2.9 8.9 2.7 3.0 2.5 3.3 8.8 5.1 4.5 1.6 3.4 6.8 0.0 2.0 4.6 1.7 3.0 3.0 5.1 7.0 2.9 3.1 3.7 2.9 1.0 P/B NM 10.8

n/a 8% 7% 7% 7% ROE 24% 47% 43% 23% 19% 16% 17% 23% 11% 17% 26% 70% 13% 24% 14% 24% 21% 37% 43% 21% 23% -12% 116%

n/a 3% 23% 20% 57% 32% 50% 22% 63% 29% 28% 21% 30% 31% 38% 38% 16% 16% 21% 15% 54% 34% 29% 16% 24% 20% 39% 30% margin EBITDA

n/a -5% 21% 17% 54% 23% 46% 10% 60% 24% 24% 18% 23% 27% 32% 24% 14% 11% 18% 14% 46% 32% 28% 13% 11% 17% 22% 10% EBIT margin

n/a 8.2 9.3 9.7 9.1 7.3 9.3 7.7 7.0 9.1 6.5 NM 12.3 12.4 28.5 11.9 28.0 12.3 10.0 10.7 14.6 15.9 12.9 15.0 12.1 15.6 27.2 14.5 EV/EBIT

A n/a 7.4 7.8 6.9 7.7 7.8 9.2 4.6 8.1 6.6 5.8 5.1 2.1 11.6 11.4 12.9 11.2 23.9 10.9 12.2 11.3 11.5 15.2 13.9 11.4 14.6 12.1 12.5 EV/EBITD

n/a 1.7 1.7 1.6 6.6 2.2 5.7 2.8 7.1 2.2 6.7 2.3 2.3 2.9 4.7 1.7 1.3 1.8 2.4 0.4 2.1 3.6 3.9 4.3 0.9 2.5 2.0 0.6 2.91 2.91 EV/Sales

n/a 0% 0% 0% 2% 0% 0% 1% 3% 0% 3% 2% 0% 4% 0% 2% 0% 0% 2% 0% 1% 2% 3% 0% 2% 1% 5% 6% Div yield

9% 8% 6% 7% 7% 8% 7% 3% 5% 4% 6% 5% 7% 7% 4% 4% 7% 9% 6% 5% 5% 3% 6% 9% -3% 10% 93% 15% (+1 yr) FCF yield

yr) n/a 9.8 12.1 18.0 20.7 15.4 15.1 11.4 19.9 11.1 28.1 23.0 14.2 14.5 23.1 16.5 16.6 12.7 16.2 13.6 16.1 13.2 14.6 19.6 18.0 14.6 15.0 17.3 P/E (+1 P/E (+1

135 3,741 1,072 5,358 5,538 2,293 3,637 9,852 5,299 4,595 3,305 1,740 1,258 3,590 76,219 10,077 35,268 13,108 14,457 12,189 52,844 23,090 84,197 46,806 17,433 (US$'m) 115,682 154,716 161,905 Mkt Cap

n/a n/a n/a Sell Sell Buy Buy Buy Buy Buy Buy Buy Buy Buy Buy Buy Rec. Hold Hold Hold Hold Hold Hold Hold Hold Hold Hold Hold Hold

nos, Monitise and Safaricom, Priced as on 5 September 2013 8.14 Price 55.00 21.00 25.00 28.00 13.00 22.00 16.00 66.00 44.00 86.00 35.00 35.00 82.00 24.00 92.00 50.00 20.00 32.00 57.00 24.28 72.50 Target 756.00 226.00 520.00 246.00 2150.00 1650.00

7.85 7.85 Price Price 47.78 47.78 24.13 28.26 26.76 20.72 18.03 25.00 72.34 44.61 97.68 38.95 36.01 74.00 33.75 76.00 50.66 19.53 32.26 53.42 22.80 51.50 (local) 629.00 629.00 176.21 468.35 210.55 1996.40 1996.40 1379.25 ock valuationoverview Figure 5: Selected st

Company Payments Inc. Global Payments Corporation Green Dot Mastercard Total System Services Vantiv Verifone VISA Union Western Wirecard Bank Processors Processing Automatic Data Fidelity National Fiserv Paychex Technologies SS&C IT Services & Software Accenture CGI Cognizant Corp Sciences Computer Genpact Oracle Corporation SAP Tata Consultancy Tech Mahindra Temenos Wipro Mobile payments Monitise Safaricom Group Plc Vodafone Bank estimates, for FinanceTeme Bloomberg Deutsche LP Source:

Page 8 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB DB Bloomberg Bloomberg Data source Data source

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 0.5 6.2 2.6 nm EBITD A/Debt

0 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 0% 4% 0% 18% 112% Debt/Equity

0.0 4.3 3.2 3.9 4.2 1.9 1.3 0.7 1.0 1.9 1.4 0.6 1.1 0.8 0.6 1.4 0.7 0.5 0.9 0.8 0.4 0.3 1.6 2.1 1.7 1.3 P/B 195.0

9% 3% 8% 9% 1% 4% 3% 3% 2% -9% -1% -3% -8% ROE 23% 17% 28% 14% 13% 11% 13% 14% 11% 15% 19% 15% 15% 417%

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 7% 33% 39% 42% 30% 74% margin EBITDA

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 4% 24% 31% 38% 15% 73% EBIT margin

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 6.4 14.7 54.5 12.8 26.2 17.2 EV/EBIT

A n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 5.8 10.4 34.8 10.4 13.5 17.0 EV/EBITD

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 3.5 3.5 2.3 4.0 2.4 4.0 12.6 12.6 EV/Sales

2% 0% 0% 0% 4% 2% 6% 6% 0% 2% 2% 2% 3% 5% 0% 0% 4% 4% 2% 7% 4% 3% 7% 7% 8% 7% 11% Div yield

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 5% 4% 6% 7% 4% 10% (+1 yr) FCF yield

yr) 5.2 9.4 6.6 9.4 8.5 7.0 9.3 19.0 34.6 18.4 13.2 15.0 27.5 15.5 17.3 11.4 11.7 10.2 11.5 12.6 11.4 10.3 10.2 12.3 15.0 13.8 12.2 P/E (+1 P/E (+1

429 7,104 1,472 3,695 1,344 15,870 66,574 59,531 83,864 59,521 56,933 80,018 35,005 58,274 81,966 32,893 33,058 74,515 91,445 89,611 70,379 (US$'m) 293,022 251,192 152,336 194,344 219,288 209,483 Mkt Cap

n/a Sell Buy Buy Buy Buy Buy Buy Buy Buy Buy Buy Buy Buy Buy Buy Buy Rec. Hold Hold Hold Hold Hold Hold Hold Hold Hold Hold

9.00 3.30 7.13 5.72 1.80 5.50 Price 17.00 32.00 15.00 64.00 41.00 44.00 77.00 53.00 40.00 33.20 73.20 33.00 35.20 Target Moneysupermarket on 5 Priced as Group, 2013September 970.00 198.55 370.00 730.00 275.00 1300.00 2450.00 1645.00

7.21 7.21 3.57 7.56 5.48 1.52 4.32 luation overview (continued) Price Price 10.66 10.66 31.24 14.37 52.11 36.70 41.82 74.77 48.81 34.08 29.60 72.92 31.48 32.74 (local) 879.56 879.56 174.00 296.00 707.60 335.80 1152.00 1152.00 2382.00 1472.50 Figure 6: Selected stock va

Company Company Customer data Experian Groupon Google Microsoft Group Moneysupermarket Rightmove Online banks Comdirect DAB Bank Major Banks Bank of America Chase JPMorgan US Bankcorp Wells Fargo Barclays HSBC Lloyds Banking Group RBS Standard Chartered Paribas BNP Societe Generale BBVA Santander Intesa Unicredit ANZ CBA Westpac National Australia Bank for estimates, Bank Finance Bloomberg LP Deutsche Source:

Deutsche Bank AG/London Page 9

8 September 2013 Financial Services Retail Bank Strategy

DB DB DB DB DB DB DB DB DB DB DB Data source source

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 0.3 Debt EBITDA/

ty n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 199% Debt/Equi

3.4 2.4 1.7 2.0 1.9 1.2 1.3 1.4 1.7 2.5 3.1 P/B

ROE 13% 21% 14% 18% 19% 21% 22% 23% 15% 23% 10%

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 9% margin EBITDA

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 6% EBIT margin

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 14.1 EV/EBIT

A n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 9.5 EV/EBITD

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 0.8 0.8 EV/Sales

5% 5% 4% 3% 3% 6% 6% 6% 5% 5% 3% Div yield

n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 6% (+1 yr) FCF yield

yr) 9.8 9.2 8.4 8.3 8.2 8.0 8.0 18.7 11.8 15.5 12.8 P/E (+1 P/E (+1

5,587 22,195 16,600 18,050 51,145 62,987 34,558 30,837 (US$'m) 146,154 191,157 238,467 Mkt Cap

Buy Buy Buy Buy Buy Rec. Hold Hold Hold Hold Hold Hold

4.49 7.45 6.80 Price 32.30 31.00 34.00 26.00 Target 148.00 133.00 115.00 212.00

luation overview (continued) 3.49 3.49 5.93 5.29 Price Price 31.08 31.08 12.18 12.68 25.35 (local) 141.10 141.10 115.50 125.10 254.40 Priced as on 5 September 2013

Figure 7: Selected stock va

Company Major Banks ABSA FirstRand Standard Bank Bradesco Itau Unibanco Agri. Bank of China China Construction Bank ICBC BOC HongHoldings Kong Hang Bank Seng Miscellaneous Plc G4S Source: Deutsche Bank estimates,

Page 10 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

The Future of Retail Banking

Vanilla banking returns are good, should improve as costs are cut, more significantly when rates eventually rise

Generally we consider retail banks to be steady compounding investments over the We are fans of retail banks as long term. We think that low customer churn (approx 6% p.a. in the UK), relatively low steady compounding and stable loan losses, and adequate permitted regulatory leverage make for attractive investments over the long & sustainable ROEs. Our Sept 2012 note on UK retail banking, for example, calculated term an average ROTE for the industry – including the impact of the four listed banks which failed or were acquired by larger firms during the crisis1 - of 24%, ranging between a low of 11% and a high of 28% between 2000-2011 (Figure 8).

Figure 8: UK retail banking sector profit drivers over time 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Net interest margin 3.2% 3.1% 2.9% 2.6% 2.5% 2.4% 2.3% 2.2% 2.4% 2.3% 2.6% 2.5% Non interest margin 1.5% 1.4% 1.4% 1.3% 1.2% 1.2% 1.2% 1.1% 1.2% 0.8% 0.8% 0.8% Total income margin 4.7% 4.5% 4.3% 3.9% 3.8% 3.7% 3.7% 3.4% 3.7% 3.2% 3.4% 3.3% Cost/income ratio -53.0% -53.8% -51.8% -51.8% -51.0% -50.6% -47.3% -48.3% -50.0% -49.9% -49.3% -49.7% PPP margin 2.2% 2.1% 2.1% 1.9% 1.8% 1.8% 1.9% 1.7% 1.8% 1.6% 1.7% 1.6% Impairment rate -0.4% -0.4% -0.4% -0.4% -0.4% -0.6% -0.7% -0.7% -1.0% -1.1% -0.9% -0.5% PBT margin 1.7% 1.6% 1.6% 1.4% 1.4% 1.2% 1.1% 1.0% 0.8% 0.4% 1.0% 1.1% ROA 1.2% 1.1% 1.1% 1.0% 1.0% 0.8% 0.8% 0.7% 0.5% 0.3% 0.7% 0.8% Leverage 23x 24x 24x 25x 29x 31x 34x 36x 39x 36x 26x 27x ROE (historic CT1) 28.2% 27.1% 27.4% 25.6% 27.4% 25.5% 27.3% 25.0% 20.7% 11.3% 18.2% 22.5% Source: Deutsche Bank, Company data

We expect these attractive returns will increase dramatically over the next ten years We think these attractive driven by cost management and better deposit spreads when interest rates eventually ROEs will rise over the next rise. The former is linked to more investment in, and better use of, technology decade on cost management demanded by the changing habits of customers, lack of market growth in most and better deposit spreads developed markets and the absence of a shareholder mandate to grow by M&A. when rates (eventually) rise

Interest rate increases in Europe, however, will be a long time coming given huge But higher rates are a distant vulnerability to higher rates everywhere aside from France, Germany and the prospect for floating-rate- Netherlands. Unless the ECB wants to risk another bout of sovereign risk aversion, heavy Europe in particular; Europe needs much stronger growth before hiking rates. Few market participants are costs will be the bigger aware of the degree to which European household debt has tended towards floating- battleground in the near term structures in recent years (Figure 9).

1 Alliance & Leicester, Bradford & Bingley, Northern Rock, HBOS

Deutsche Bank AG/London Page 11

8 September 2013 Financial Services Retail Bank Strategy

Figure 9: The world can’t afford higher official interest rates unless GDP growth is much, much faster Govt ave debt Non-resident Debt/GDP . Household Corporate Pvt sector . Floating rate Mortgage debt Owner maturity share of public debt/GDP debt/GDP debt/GDP household / Gross occupier rate debt debt (% of Disposable total) Income Belgium 6.7 58% 101% . 54% 183% 237% . 15% 76% 78% France 6.9 64% 93% . 57% 105% 162% . 5% 62% 58% Germany 6.5 62% 81% . 59% 69% 127% . 5% 67% 43% Greece 11.1 56% 188% . 61% 63% 125% . 80% 50% 80% Ireland 6.4 61% 123% . 113% 219% 332% . 85% 154% 75% Italy 6.6 35% 128% . 45% 83% 128% . 57% 33% 80% Netherlands 6.8 56% 69% . 127% 95% 222% . 5% 220% 56% Portugal 5.7 54% 124% . 91% 157% 248% . 90% 91% 81% Spain 5.7 28% 93% . 81% 135% 216% . 80% 94% 85% UK 14.4 31% 93% . 95% 112% 207% . 56% 125% 66% US 5.4 30% 112% . 84% 78% 162% . 10% n/a 66% Japan 6.0 8% 245% . 67% 100% 167% . 20% n/a n/a Source: Bank of Italy, Financial Services Authority, Hypostat, National Bank of Greece, Banc Ceannais

Though US mortgage and housing markets are more protected from the costs of rising US housing and mortgage rates by a higher proportion of fixed rate debt – itself the consequence of government markets are more protected policy via Fannie Mae and Freddie Mac, now being retrenched – the sensitivity of against rising rates – but market sentiment, house prices and transaction volumes to changes in the cost of sentiment, housing starts and finance is material there too. Though house prices have posted a decent recovery from HPI are naturally still geared crisis lows (Figure 10), the market will likely remain focused on assessing the importance of the recent deterioration in buyer affordability (Figure 11) and decline in to flow finance costs growth in housing starts as indicators of broader economic activity (Figure 12). Housebuilder and associated stocks provide a decent barometer of the market’s spot view. Some concerns about the direction of travel are emerging (Figure 13).

Figure 10: US house prices have bounced… Figure 11: … driven by improved affordability, itself a function of falling finance costs… 200 20% 100 8.00 190 15% 7.00 120 180 10% 6.00 170 5% 140 160 5.00 0% 150 160 4.00 -5% 140 3.00 -10% 130 180 2.00 120 -15% 200 110 -20% S&P / Case-Shiller Index (YoY %) 1.00 S&P / Case-Shiller Index(1Q00 = 100) 30 year year 30 effective non-jumbo mort tate NAR Composite affordability (inverted) affordability Composite NAR 100 -25% 220 0.00 1Q01 3Q01 1Q02 3Q02 1Q03 3Q03 1Q04 3Q04 1Q05 3Q05 1Q06 3Q06 1Q07 3Q07 1Q08 3Q08 1Q09 3Q09 1Q10 3Q10 1Q11 3Q11 1Q12 3Q12 1Q13 1Q01 3Q01 1Q02 3Q02 1Q03 3Q03 1Q04 3Q04 1Q05 3Q05 1Q06 3Q06 1Q07 3Q07 1Q08 3Q08 1Q09 3Q09 1Q10 3Q10 1Q11 3Q11 1Q12 3Q12 1Q13

S&P / Case Shiller (1Q00 = 100) YoY % (RHS) NAR Composite Housing Affordability 30 year non-jumbo effective rate (%, RHS)

Source: Haver Source: Haver

Page 12 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 12: Housing-related activity levels have Figure 13: …with housebuilder stocks confirmation of improved… emerging investor concerns 180 50% 60% 160 40% 50% 30% 40% 140 20% 30% 120 10% 20% 100 0% 10% 80 -10% 0% -20% 60 -10% -30% 40 (%) price performance Share -20% -40% 20

-50% %) (YoY sales and starts Housing 0 -60% Housingstarts andsales (100 at 1Q01) 1Q01 3Q01 1Q02 3Q02 1Q03 3Q03 1Q04 3Q04 1Q05 3Q05 1Q06 3Q06 1Q07 3Q07 1Q08 3Q08 1Q09 3Q09 1Q10 3Q10 1Q11 3Q11 1Q12 3Q12 1Q13

Existing home sales (100 at start) Housing starts (100 at start) Existing home sales (YoY %, RHS) Housing starts (YoY %, RHS) 12 months YTD 3 months

Source: Haver Source: Datastream. Priced on 5 Sept 2013

Every so often it's a good idea to look much further ahead

Shifts in long-bond yields and market sentiment aside, we are positive on the medium Is the promised revolution in term outlook for retail banks. This view, of course, assumes that incumbent banks banking technology a threat continue to dominate household financial dealings. Reading the popular press and or an opportunity for talking to people in the software and hardware, payments processing and mobile incumbent bank banking industries, we hear of a coming revolution in the way that people interact with shareholders? their money and a step change in the manner in which banks can harness IT to service their clients. Is this just hype? If not, what does it mean for customers, and does it pose a threat to banks?

Figure 14: The financial ecosystem around consumer spend

Moving money Helping you What is done Storing money Lending money around spend

Banks, Merchant Banks; Consumer Credit agencies; acquirers; telcos; Banks credit; Peer to Voucher companies; Big Who does it monoline money peer/Crowdfunding data providers agents, ATM providers

Source: Deutsche Bank

To answer these questions we’ve divided this note into chapters grouped into 4 broad areas, the first of which looks at legacy core banking platforms:

„ The Rotten Core: High-profile IT failures at a number of banks and some reduction in regulator workload set the stage for more top-down pressure for banks to refresh their core IT. Why have so few banks done it?

„ Change is on the way: No one we spoke to thought a core platform replacement could be justified on cost grounds alone, why do we think a significant change in IT

Deutsche Bank AG/London Page 13

8 September 2013 Financial Services Retail Bank Strategy

infrastructure will be seen in the next decade? It about time, technological change, relentless volume growth and regulatory pressure, we think.

„ The rise of the bank in a box and the parallel bank: The cost and complexity of full IT renewal at big banks is high, demonstrated in our Standard Bank study. New banks – with Software-as-a-Service architecture – are launched for $15-20m. What do these competitors mean for existing players? What’s behind the launch of new- brand, online-only banks like Happy Bank! and UBank at BNP and NAB? The second area looks at the impact of digital and phone technology on banking.

„ Rise of the machines: Mobile & Tablet Banking: How explosive growth in tablet and smartphone self service is changing bank service patterns and customer demands;

„ Mobile Payments: The case for telco-payment businesses and the lessons these have for incumbent banks. Driven by a detailed look at Safaricom’s M-PESA, arguably the most successful mobile payment network in the world;

„ Convergence towards digital and mobile wallets: Digital and mobile wallets are driving payments convergence across banking, retailer, tech and telecoms. We assess the opportunities and risks for each and implications for banks. The third area looks at the opportunities for banks to cut costs (via branches), expand product offerings (via big data) and explores potential disintermediation by the emerging peer-to-peer & crowd-funding space.

„ The Future of the Branch: If phones put the bank in my pocket, and cards take the cash out of my wallet, what’s the point of the 689,144 branches in the 66 markets in our review? When can banks finally get on with being wide margin tech companies trading on 20x earnings?

„ Big data, banks have more of it than most, right? Who knows more about my spending habits and financial status than my bank? How much upside is there from mining this information better for banks and for other selling organisations?

„ Cutting out the middle-man: Why don’t I cut out the bank altogether and lend direct to people who want personal loans and to SMEs who need finance? My returns would be better than wouldn’t they? The case for crowd funding and peer-to-peer lending. A really fascinating industry in which we think platform owners will make better returns than lenders. We provide historical context around the trends identified above (more card payments, more electronic commerce) by looking at how payments systems work.

„ Payments trends across the world: we give an overview of the payment methods that consumers use, and look at how these trends have changed over with the rise of plastic payments and the decline of the cheque. We also outline how the credit card business model works; Finally, we list some of the more interesting innovations we’ve come across in the process of writing, and provided a glossary.

Page 14 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

The Rotten Core

Key points

We expect many developed market banks to replace core IT in the next decade at significant expense in projects lasting five years or more. Near term this sees costs rise unless expensed under restructuring charges. This should leave firms more nimble, better able to meet rising regulatory requirements, and delivering broadly stable costs despite further increases in transaction volumes. With revenue growth therefore we expect cost/income ratios to fall.

„ Banks are technology companies that move and keep track of money, with linked service and selling arms. Most banks run mainframes installed decades ago and subsequently patched to deal with M&A, product proliferation and innovation, geographic expansion and new regulatory requirements.

„ The Standard Bank case study shows why Core System replacements are fairly rare. The project, begun in 2005, is taking longer and costing more than initially planned: 2012 IT costs were 72% higher than 2008’s despite a significant increase in capitalised costs on balance sheet. Temenos, a bank software company estimates that the average bank devotes 15% of costs to IT: SBK is now at 24%.

„ We expect retail and commercial banks to recoup the 5-10% initial increase in costs from full IT renewal (SBK 70% increase in 15% of group IT = 10%) via cheaper branches and lower unit costs. Phasing is unlikely to match, however, with IT costs up before efficiencies come through. IT is not, in our view, the route to a material decline in big bank costs.

„ We expect regulators to play a pivotal role in some markets in IT renewal, especially given high profile outages at RBS for example. This isn’t to say regulators don’t already impose material operational risk requirements: the 32 Eurobanks we survey have E815bn in op risk RWAs costing 0.7% off the sector ROE to support. Op risk RWAs avg ~14% of credit risk RWAs, with banks such as Credit Suisse (36%) and UBS (56%) carrying higher proportional burdens. We think systems outages, a focus on cyber risk and growing regulator capacity to return to issues other than capital and liquidity will result in more regulatory emphasis on IT.

„ It’s worth examining parallel banks (new digital ventures set up by existing banks like BNP Paribas (Hello Bank!) and NAB (UBank)) as a potential new IT platform (for NAB at least) and as a learning ground for digital and multibrand banking. For NAB at least, the on-line venture will form the basis for the broader group in 3-4 years.

„ We think it is poorly understood that Bank-in-a-Box solutions can get a new bank running for $15-20m. Though market share gains and NIM remain key challenges, the surprisingly modest upfront investment and scaleable cost base means breakeven is mostly about “when” rather than “if” with timing driven by management calibration of growth (paying for customers, opening branches) against profits. We expect owners of these businesses to ultimately do well, but we don’t expect a massive change in market shares for the large banks.

Deutsche Bank AG/London Page 15

8 September 2013 Financial Services Retail Bank Strategy

Banks mostly move and keep track of where the money is

Money performs three functions: Money is a unit of account, medium of exchange and (i) unit of account (so we know how much to pay), store of value

(ii) medium of exchange (allowing commerce beyond bartering), and

(iii) store of value (avoid high inflation if you can).

2 Tim Harford’s newest book provides a good description of the money system of the The island of Yap, near the Micronesian island of Yap, a reminder of how far money has come and also how little Philippines, provides a it's changed. On Yap the currency (the rai) is made of stones quarried on the island of reminder of the basics of Palau, 250 miles away. The bigger the stone, the greater the value. A pig might be money – which haven’t worth a stone a couple of feet across. The largest rai weigh tonnes. Interestingly – and changed over time in common with today's system – those on Yap didn't require delivery or personal possession of this unwieldy currency as proof of ownership. The island’s economy was small enough for the knowledge of ownership to substitute for physical possession even if, according the book the money had sank on its delivery run from Palau.

Day to day, the business of banking is mostly about recording transfers of the “rai” Banks are effectively between company and employee, buyer and seller, saver and borrower. And, with the technology companies with vast majority of transactions taking place electronically banks are effectively tech attached sales and services companies with sales and servicing arms. Over time though, aging bank infrastructure arms: and will increasingly and patchwork software is increasingly struggling to keep up with regulatory demands, struggle to keep up rising volumes (accelerated by the rise of plastic and digital) and more rapid changes in areas like payments.

Figure 15: Rise of the debit cards and online... Figure 16: ... will contribute to further volume growth

7 160 150 6.2 6 140 132 5.5 5.3 Billions 5.2 5.1 4.9 119 5 4.6 4.4 4.5 120 109 4.0 99 4 100

3 80 Billions of transactions

2 60

1 40

0 20 2007 2008 2009 2010 2011 0 Cards with a cash function Cards with a payment function (debit/credit) 2007 2008 2009 2010 2011

Source: Deutsche Bank analysis, BIS data. Adjusted for series breaks Source: Deutsche Bank analysis, BIS data

Fortunately for incumbent banks, churn rates are sufficiently low and products and Fortunately for incumbents, service sufficiently uniform (and replicable), to allow market shares – generally – to customer churn is relatively remain fairly stable over time (Figure 18). Further entrenching this position, for a while low and products sufficiently at least, is the reality that the banks which took most share in the debt bull market to uniform to keep market 2007 tended to do worst during the crisis: who wants to start a market share fight shares fairly stable today?

2 The Undercover economist strikes back, 2013

Page 16 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 17: The volume of bank customer interactions Figure 18: Bank customer churn is relatively low relative increases dramatically with digital channels…. to other consumer industries (UK) 25 70%

58% 60% > 20 20 50%

15 40%

30% month 10 20% 18%

Customer turnover (% p.a.) 13% 14% 6 11% 10% 6% 5 Average visits per active customer per per customer active per visits Average <1 0% UK personal Sky TV UK gas (2011) UK electricity UK contract UK prepaid 0 current a/c (2011) mobile phonesmobile phones Branch Online Mobile

Source: Monitise company data Source: Ofgem, UK Independent Commission on Banking, Vodafone (above metrics are for VOD), Sky Rotten Core – Not all rotten, but generally pretty old

The Core Banking System (CBS) delivers the basic functions of banking – taking Core Banking Systems (CBS) deposits and making loans - and linking customers to front, middle and back office deliver basic banking, with functions. The majority of banks run on mainframe-driven systems installed in the most banks using older 1970's and 1980's which were able to handle large volumes of transactions processed mainframe technology in a fairly rigid format. Subsequent mergers, product innovation and proliferation, patched again and again for geographic expansion and a huge increase in regulatory requirements have driven most institutions to add bridging applications to the core. This has contributed to increases in M&A, expansion, product operational leverage, manual workarounds (with associated error and re-work costs) proliferation and regulatory and the kind of complexity shown in the systems map in Figure 19. change

Figure 19: Which thread do we pull?

Source: The Boston Consulting Group

Deutsche Bank AG/London Page 17

8 September 2013 Financial Services Retail Bank Strategy

Systems complexity contributes to high profile failures, less efficient operations, and Systems complexity sub-optimal service & management conditions: contributes to high profile failures, less efficient „ Many banks, for example, can’t easily tell what dealings they have with a retail operations and sub-optimal customer or group of companies across all products. service and record keeping „ One executive we spoke to told of difficulty in meeting liquidity projection reporting regulatory requirements on its old CBS.

„ Another told of the challenges of writing a recovery and resolution plan that spoke credibly of a weekend re-organisation given the delays of batch-driven processing and silo-ed customer information.

„ Another reported that branch managers until this year could say how many customers they had by product but not how many customers they had overall (another formulation of the single customer view problem). Significant implementation risk, little near term gain

In the minds of many, launching a CBS replacement means underperforming peers on No-one we met thought that costs in the near term whilst running high-profile operational risks in return for CBS renewal could be uncertain product and economic benefits which, even if these materialise, will most justified on cost savings alone likely enjoyed by the next CEO (or the one after that). A 2010 McKinsey IT review reported that less than 30% core business platform replacement projects undertaken in 2000-2010 succeeded, with overall project costs ranging between E50-300m3. One executive described CBS renewal as “like trying to change the wings on a Boeing while in the air”. No-one we spoke to thought CBS replacement could be financially justified on cost savings alone, outside of situations where banks were looking to contract very meaningfully and where re-platforming would cut redundant capacity.

Our Standard Bank (SBK) and National Australia Bank (NAB) case studies give a sense of the costs of CBS programmes.

Figure 20: Standard Bank snapshot Figure 21: NAB snapshot Standard Bank Analyst name Email National Australia Analyst name Email Bank James Freeman [email protected] Stefan Swanepoel [email protected] Spot price: 33 Tgt. Price 35 Rec. Buy Spot price: 116 Tgt. Price 133 Rec. Buy 2010 2011 2012 2013e 2014e 2015e 2010 2011 2012 2013e 2014e 2015e EPS (DB adjusted) 2.1 2.5 2.4 2.5 2.7 2.9 EPS (DB adjusted) 7.1 8.1 8.8 10.7 12.6 14.7 DPS 3.9 4.3 4.6 5.1 6.1 6.7 DPS 1.5 1.7 1.8 1.9 2.1 2.3 NAV 90,755 102,523 113,905 136,651 151,177 167,566 NAV 36,019 39,247 40,835 43,393 46,808 50,543 TNAV 80,372 89,769 99,218 121,825 136,210 154,096 TNAV 37,207 40,493 42,497 45,100 48,514 52,250 Market Cap (US$'mn) 25,795 19,532 20,886 18,050 18,050 18,050 Market Cap (US$'mn) 57,262 53,609 60,886 70,512 70,512 70,512

P/E (DB) 15.0 12.0 15.1 10.8 9.2 7.9 P/E (DB) 11.1 9.4 10.5 13.1 12.2 11.1 Div yield 3.6% 4.3% 4.1% 4.4% 5.2% 5.8% Div yield 6.4% 7.4% 7.2% 5.9% 6.3% 6.9% Payout ratio 55% 50% 49% 47% 48% 45% Payout ratio 71% 69% 75% 76% 76% 75% ROE 12% 14% 14% 14% 14% 15% ROE 14% 16% 15% 15% 15% 15% ROTE 14% 16% 13% 16% 16% 16% ROTE 12% 13% 13% 13% 13% 14%

Source: Deutsche Bank estimates; Priced as on 5 September 2013, Note: All figures in local currency Source: Deutsche Bank estimates; Priced as on 5 September 2013, Note: All figures in local currency unless stated unless stated

3 McKinsey Quarterly: Overhauling banks' IT systems; Marcus Heidmann

Page 18 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

The Standard Bank case study – Longer, more expensive, and still not over SBK decided in 2005 to replace its CBS, working with SAP. Cost and time to complete Standard Bank’s have disappointed relative to management expectations we think. By 2010, SBK had implementation has taken engaged Oliver Wyman to benchmark implementation spend followed by a similar longer and cost more than process by McKinsey in 2011. These concluded that the “change-the-bank” spend was expected with targets around “above benchmarks and would remain so for the next few years…”. The 2011 annual % of costs devoted to IT report undertook that “total IT spend as a proportion of total costs will reduce over the next three years with consequential benefits for the bank’s cost-to-income ratio”. Since missed; IT expense at 24% of then, the proportion of costs devoted to IT has risen to 24% from 21% in 2010 whilst total costs is higher than the the banking cost:income ratio has been roughly flat at ~60% (Figure 22). Temenos, a c.15% average estimate of bank software company, estimates that IT is ~ 15% of costs at the average bank – SBK bank software company is running significantly higher than that at present. Combined with the roughly 50% of Temenos costs in staff remuneration which are also broadly fixed, the operational leverage in vanilla banking is obvious: good when revenues are rising, tough when they’re not.

Figure 22: Standard Bank Group IT spend shows rising cost of CBS replacement decision (ZAR’m) 2008 2009 2010 2011 2012 YoY IT licences, maintenance & related costs 2,686 3,146 3,227 3,183 3,636 14% IT staff costs 1,627 1,709 2,214 2,275 2,691 18% Depreciation and amortisation 799 1,024 1,247 1,683 2,094 24% Other 444 701 562 922 1,184 28% Total IT 5,556 6,580 7,250 8,063 9,605 19%

Proportion of IT spend to total costs 18% 20% 21% 23% 24% Banking cost/income 49% 52% 62% 59% 61% Source: Deutsche Bank estimates, Company data

Figure 22 confirms the sharp increase in depreciation expense which adds to the …adding a future normal high operating leverage of any bank. Depreciation & amortisation last year was depreciation cost burden to ZAR2.1bn, up 160% since 2008. SBK management expect depreciation expense to peak operating costs (and op in 2016 as implementation spend begins to decline. leverage)

Banks weighing platform change probably consider the chief virtues of existing SBK is now holding US$1.7bn infrastructure to be its largely-depreciated nature and proven ability to process billions in capitalised software and of transactions. One executive we met put the life of CBS at 20-25 years whilst most hardware on balance sheet – banks depreciate software over a maximum of 10 years. As Figure 23 shows, SBK is ~13% of equity and ~60% of now holding US$1.7bn in net book value in computer equipment and software, 2012 PBT multiples higher than peers, 2x the net value of its property holdings, ~13% of group equity and ~60% of last year’s PBT.

Software assets are subject to annual impairment tests which require management to assess whether over time the business – facilitated by its IT – will deliver profit sufficient to recover platform costs; this is easier than demonstrating that the additional investment relative to peers has delivered premium product, returns or efficiency.

Deutsche Bank AG/London Page 19

8 September 2013 Financial Services Retail Bank Strategy

Figure 23: It’s no bad thing for EPS that existing infrastructure is already heavily depreciated (FY12, US$’m)

Australia France South Africa UK US$'m ANZ CBA NAB BNP SG ABSA FirstRand Nedbank Standard Barclays LBG RBS Bank Property 1,081 1,151 938 7,309 4,505 471 818 503 923 6,991 11,099 10,080 Other equipment 498 893 491 4,194 1,679 345 248 172 506 2,360 3,320 3,580 Computer equipment 337 370 404 n/d n/d 178 176 82 433 n/d n/d 2,239 Software 1,699 1,640 1,402 1,931 1,072 174 24 291 1,306 2,605 1,064 3,169 Total IT 2,036 2,010 1,806 n/d n/d 352 200 373 1,739 n/d n/d 5,408

Equity 39,753 40,093 43,047 124,637 65,736 8,613 8,260 6,384 13,060 102,311 72,616 114,485 IT / Property 188% 175% 192% n/d n/d 75% 24% 74% 188% n/d n/d 54% IT / Equity 5% 5% 4% n/d n/d 4% 2% 6% 13% n/d n/d 5%

Profit before tax (2012) 7,710 9,609 5,940 13,691 5,776 1,434 2,184 1,265 2,846 11,454 5,805 7,779 Capitalised IT as % PBT 26% 21% 30% n/d n/d 25% 9% 29% 61% n/d n/d 70% Source: Deutsche Bank estimates, Company data

National Australia Bank is renewing its core platform at present NAB is upgrading its IT in the NextGen project as a part of its ongoing investment and NAB’s NextGen will maintenance plan. In common with the Standard Bank experience, NextGen will drive a contribute to a 40% increase significant increase in capitalised software which management expect to peak at A$2bn in depreciation expense over in 2015, seeing annual depreciation rise to A$800m, 40% above the 2010 level. the 2010 base

Figure 24: A$9bn in investment Figure 25: …leading to an increase in Figure 26: …and higher depreciation spend over eight years… capitalised software costs… charges and operational gearing

1,400 900 $2bn peak $800m peak $1bn-$1.3bn 2000 800 1,200 153 700 1,000 199 1500 600 142 288 800 278 447 500 266 352 1000 479 241 600 400 258 371 420 325 300 400

Investment Spend (A$'m) Spend Investment 213 500 1007 200 Capitalised software (A$'m) software Capitalised 773 304 200 271 299 348 578 248 285 100 0 0 0

2010 2011 2012 FY13-17 2010 2011 2012 2015e (A$'m) expense Amort. & Depreciation 2010 2011 2012 2017e NextGen Efficiency initiatives Infrastructure Other (incl. compliance) Australia Other Depreciation Amortisation Source: National Australia Bank Source: National Australia Bank Source: National Australia Bank

Change, as necessary as it may be, doesn’t come cheap. NAB are partnering with NAB have appointed Oracle Oracle under a ten year licensing agreement, implementing the banking platform which under a 10 year license to launched in Sept 2012. As with the “Bank in a Box” case study below, NAB should see implement its new banking scale economies as a foundation customer4: Oracle have 900 staff, including 500 platform developers, working on this platform and is targeting its sale to over 100 tier 1 banks globally. Oracle apparently spends US$5bn in R&D p.a. NAB summarises expected key NextGen benefits below.

4 Pretty much all universal bank software offerings are spin-offs of IT systems of large incumbents, subsequently sold on a more widespread basis. Choice of vendor therefore often comes down to the nearness of fit and existing relationships with other banks in the larger markets in which the bank operates.

Page 20 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 27: National Australia Bank – Benefits of NextGen Focus areas 2009 Future state Understanding customers 20 partial systems/databases Single customer view Time to 'Yes' approval* <120 hours Response within minutes Time to documentation for a simple mortgage* ~20 days <1 hour

Customer Customer Self service <10 capabilities via internet banking >100 new capabilities Processes Complexity impacting productivity Simplified, improved productivity Core consumer and business banking products** ~500 <100 Customer reporting & analysis Multiple manual systems Single integrated system Employee enablement Impacted by manual processes Improved via reduced role complexity and

Employee Employee better processes Operational risk Ageing fragmented systems with high cost and Reduced incidents and technology savings time to make changes Competitiveness Hindered by time and cost to launch products Faster speed to market and greater agility

Group Group Service experience Lower conversion rates, high attrition Better conversion rates, reduced attrition Source: Deutsche Bank, National Australia Bank, *Subject to customer verification and valuation requirements, ** Excludes Wealth and Wholesale Banking products

A twist on the IT transition approach: The parallel bank

A fundamental IT re-organisation probably takes 5-7 years to complete (2 years Banks like NAB and BNP planning, 1 year proof of concept and gapping, 2+ years implementation), costs a lot, Paribas have been increases operational risk before reducing it, and is perhaps something most CEOs experimenting with online would rather not do. A middle-way which may gain traction is the launch of a simpler only banks under new brands digital-only greenfield operation under a new brand. The launch of UBank by NAB is an example. Meanwhile, BNP Paribas started Hello Bank! and which is operating in Belgium, Germany and France, with Italy due to launch in October. The first year budget is E80m in set-up marketing and customer incentive costs. Hello differs from UBank in that it runs on BNP Paribas’ own online banking infrastructure, offering essentially the same customer proposition with different product pricing and without branch access.

Figure 28: BNP Paribas’ digital offering: “Hello Bank!” Figure 29: NAB has UBank

Source: BNP Paribas Source: National Australia Bank

In addition to the usual potential benefits of multi-brand operations (differentiated Similar initiatives may in some pricing, a distinct customer proposition, capturing a proportion of customer churn), cases form the IT basis for a these initiatives may form the IT basis for a broader transition. NAB, for example, is later transition from the old moving its customers to the UBank platform over the coming 3 or 4 years, with current CBS platform volumes and growth proving that the technology is battle-ready.

Deutsche Bank AG/London Page 21

8 September 2013 Financial Services Retail Bank Strategy

Conventional competition from the Bank in a Box

The reason that a parallel bank strategy is viable at all is because it is far cheaper to set It’s far cheaper to start a new up a new bank than re-engineer an old one. One IT consultant we met estimated that bank than refresh the core IT 90% of IT upgrade spend in large banks goes on pre-launch and integration testing to of an established incumbent; attempt to ensure that transitions and implementations don’t fail. As shown in the case up to 90% of upgrade spend study below, greenfield banks can get off the ground for $15-20m and inside 2 years of in big banks goes to planning. This compares with the $1.5bn that Lloyds Banking Group is spending restructuring a part of its retail bank for IPO in 20145. integration testing; a new bank from a box can be up This reminds us of the old joke of the tourist asking a local for directions to a landmark and running for $15m being greeted by the response “well, I wouldn’t start from here, mate”. Is the best way to set up a 21st century bank to start a new one? If so, does this doom incumbents to structural market share decline? We examine Metro Bank, one of the new bank in a box businesses to emerge in the crisis years in the following case study.

Metro Bank, UK Launched in July 2010, Metro Bank was the first new UK High Street bank in 100+ IT providers are becoming years. Co-founded by Vernon Hill, the founder of the US Commerce Bank, Metro Bank more capable of operates on the Temenos T24 platform, the same CBS used by Lloyds Banking Group in implementing CBS non-UK operations, Bank Sinopac in Taiwan and Schroders Private Bank. replacements, but Bank-in-a- Box offerings are improving Figure 30: Metro Bank runs on Temenos T24 too….

Source: Metro Bank; https://www.metrobankonline.co.uk

5 The ‘Project Verde’ work required under state aid remedies imposed by the EU after the post HBOS-rescue recap. See stock-specific research under LLOY.L on the DB equity research website.

Page 22 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

6 Metro runs T24 using a Software As a Service (SOAS) model under which the UK’s Metro Bank is a start-up Temenos’ software is hosted third-party owned hardware (niu Solutions, in this case, which operates on Temenos which also listed Kent Reliance, a JC Flowers associate, as a banking customer) and software hosted by nui accessed via the internet. Solutions on a SoAS model

Figure 31: Software as a Service schematic

Laptop

PC Internet

SaaS

Hardware host

Mobile phone

Source: Deutsche Bank

Metro Bank enjoys scale benefits from Temenos’ implementation of T24 for other This set-up allowed Metro to banks. We see these as the two key advantages of third party software to be weighed get a full service bank up with against the rigidity / inflexibility that comes with less tailored solutions. For a start up superior account opening bank, however, the SOAS / SaaS model has two very important further benefits: performance running for $15m, with a cost base that „ Costs scale with business growth. Metro pays Temenos a monthly subscription based on the number of customers in the business. nui Solutions hosts the scales with customer growth software, with Metro buying hardware at cost price plus a monthly cost run-rate and the R&D and scale based on number of users. This cost structure radically reduces up-front costs and benefits of a big bank cash burn and reduces operational leverage relative to existing banks.

„ The R&D of a big bank. In exchange for the Temenos licence fees Metro Bank pays, it gets the maintenance and product upgrades developed by the company for all clients. Temenos says its spends 20% of annual revenues on R&D.

The bottom line of all this is that Metro Bank was able to get licensed, set up and open Metro Bank was up and in two branches for £10m / US$15m in cash cost. Available products include current running for $15m with a cost accounts, savings, mortgages, unsecured lending, loans, deposits and cash structure which reduces up- management product, with the ability to open an account and get cards and cheques front cash burn and shares on site in 20 minutes (tested by one of this report’s authors). The scaling cost base financial risk with IT service shares the financial risk of the company with service providers if things go badly and an ability to scale quickly if things go well. Metro, for example, grew from 1 store in July providers and which scales 2010 to 19 at present, aiming for 175-200 stores by 2020. with business growth

6 SaaS elsewhere

Deutsche Bank AG/London Page 23

8 September 2013 Financial Services Retail Bank Strategy

Figure 32: Metro Bank’s store locations Figure 33: Metro Bank projected store growth

250

200 200

150 Branches 100

40 50 19 10 12 1 4 0 July 2010 Dec 2010 Dec 2011 Aug 2011 Aug 2013 Dec 2014 Dec 2020 (target) (target)

Source: Metro Bank website. https://www.metrobankonline.co.uk/OurStores/ Source: Metro Bank

A cost effective way of setting up a bank... We expect other banks to We see UK retail banking as one of the most attractive banking markets in Europe, if follow Metro’s lead (see for not the world. We think the IT and operations model adopted by Metro Bank is a cost example, Aldermore) given effective and sensible for managing start up risk. We expect other entrants will set up in attractiveness of the market this way. One such bank, Aldermore7, was subsequently launched using a hybrid of and cost model: market share own- and white-labelled product provision to get to market cost effectively. gains and NIM will ...but not enough on its own: market share gains and adequate NIM needed differentiate winners from Whether Metro and Aldermore are great investments for their owners depends mostly losers on management’s performance in (i) generating market share gains and (ii) delivering a decent net interest margin in a low rate environment.

„ Metro and Aldermore are producing strong growth. Data published by the Bank of UK new entrants are England on its Funding for Lending Scheme shows Metro and Aldermore grew producing good growth domestic loans to £359m and £2.5bn at 1H13, up 360% and 60% YoY respectively. The launch this month of a guaranteed 7-day current account switching service8 which may be boosted by the may generate more market movement. We believe that customer inertia is most introduction of guaranteed 7 reflective of perceptions that all banks provide the same general quality and day switching this month product rather than a view that switching is too difficult and so are sceptical of a sea-change here.

„ NIM: The other key management challenge is the achievement of adequate interest The other key financial margin to cover costs, impairments and shareholder returns, made much more challenge is the delivery of difficult by the low interest rate/QE-affected environment, even before considering whether to pay up for a more price-sensitive customer. enough NIM in a low rate environment

7 Aldermore is owned by AnaCap, runs Temenos T24; savings capabilities provided by Newcastle Building Society. The business has a greater focus on commercial and SME banking and does not have branches. In all management believe it launched with a cost below £20m. 8 Banking reform White Paper, p.51, http://www.hm-treasury.gov.uk/d/whitepaper_banking_reform_140512.pdf

Page 24 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Conclusion: we think incumbents will still retain vast majority of mkt share We expect new banks to We think new banks like these will take share and produce good returns for produce good returns for their shareholders in both private and listed phases of life. But unless we see a radical owners but expect them to change in customer propensity to switch – and without a killer product that can’t be make little difference to replicated by peers we can’t see the catalyst for a change here – we expect incumbents incumbent market shares to retain the vast majority of market share, returning to growth as the market does.

What allows us to be reasonably positive on both incumbents and new entrants is the How can we be relatively low set-up costs of the new players and the low churn amongst the higher cost positive on both the incumbents. If this were a higher churn, shorter duration relationship business – like incumbents and the new low cost airlines meets flag carriers – we would naturally hold a different view. entrants?

Importantly for the incumbents in many markets – but especially the UK and US, with The end to liquidity SA, Australia and parts of EM lagging – the completion of bank liquidity rebalancing, rebalancing in many markets delivered by very successful non-core asset shrinkage, decent deposit growth and will support liability margins; muted customer demand for credit, means that liability margins should be supported. when rates rise NIM and When rates eventually increase we expect retail banking to deliver what will look like ROEs for retail will improve extraordinary ROEs, most likely prompting a shift in equity allocations from IBs to retail. significantly We expect a significant increase in IT renewal activity

Despite the costs and risks highlighted above we expect much more IT simplification and renewal in the coming years driven by the following factors:

„ A lower growth environment has driven a greater focus on costs at DM banks. C- Lower growth increases the suite turnover provides some banks with restructuring budgets. With CBS focus on costs, C-suite unjustifiable on cost grounds alone, a restructuring budget and broader agenda of organisational change is helpful in motivating for significant cost, long duration turnover generates plans like these. C-suite management are also generally taking a more hands-on restructuring budgets role in middle and back office operations, we think9. Two precursors to larger change projects – which we’re seeing at more banks – is a significant reduction in the number of software packages used (see ING, for example, in Figure 34) and products offered to customers (with 80% of the new business volume through 20% of the products this makes sense – NAB and Westpac’s product profile plan is shown in Figure 35).

9 For a detailed run-down on cost management requirements and priorities at European banks please see: Cutting to the Chase, DB equity research, 12 December 2012

Deutsche Bank AG/London Page 25

8 September 2013 Financial Services Retail Bank Strategy

Figure 34: Infrastructure and software simplification: ING Figure 35: Product simplification: NAB, Westpac

100% 0% 600 90% 15% 15% 25% -5% 500 80% 10% 10% 40% 70% 20% -10% 400 60%

50% -15% 300

40% 40% 75% 75% -20% 200 30% 55% %) (YoY Applications 20% Proportion of of IT Proportion infrastructure -25% 100 10% 20%

Number of core bank products (retail) products bank core of Number 0 0% -30% 2008 2011 2015 2018 2009 2012 / Current Target

Isolated Replicated Shared Applications (YoY %, RHS) National Australia Bank Westpac

Source: ING company data Source: National Australia Bank, Westpac company data

„ Consolidation amongst IT providers means banks now deal with fewer vendors Vendors are becoming more which have slowly growing CBS renewal track records. Implementation goals will credible providers of CBS appear more credible, economies of scale more material. CBS implementations typically involve hardware and software suppliers (IBM, Oracle, SAP, Temenos, renewal, though it remains Misys, Sungard, SimCorp, Murex), process and infrastructure consultants (Deloitte, early days BCG, McKinsey), and in some cases, hardware hosts (providing financed and scaleable sale and leaseback hardware).

Figure 36: Fiserv snapshot Figure 37: Oracle snapshot Fiserv Analyst name Email Oracle Analyst name Email Bryan Keane [email protected] Nandan Amladi [email protected] Spot price: 98 Tgt. Price 86 Rec. Hold Spot price: 32 Tgt. Price 32 Rec. Hold 2010 2011 2012 2013e 2014e 2010 2011 2012 2013e 2014e 2015e EPS (DB adjusted) 4.1 4.6 5.1 6.0 6.7 EPS (DB adjusted) 1.2 1.7 2.0 2.3 2.4 2.6 DPS 0.0 0.0 0.0 0.0 0.0 DPS 0.2 0.2 0.2 0.3 0.0 0.0 BVPS 21.5 22.0 23.1 25.6 30.7 BVPS 6.1 7.8 8.7 9.3 10.1 10.9 ROE 16% 15% 17% 19% 18% ROE 22% 24% 24% 24% 25% 25% EBITDA mgn 33% 32% 31% 32% 32% EBITDA mgn 55% 52% 54% 55% 54% 53% Market Cap (US$'mn) 7,706 8,774 10,400 14,457 14,457 Market Cap (US$'mn) 112,720 144,438 147,595 154,716 152,380 152,380

P/E (DB) 12.6 12.9 13.8 16.2 14.5 P/E (DB) 18.4 16.9 14.8 14.1 13.2 12.2 EV/EBITDA 7.8 8.6 9.2 11.3 10.1 EV/EBITDA 7.4 7.0 6.6 6.9 6.3 5.9 Div yield 0% 0% 0% 0% 0% Div yield 1%1%1%1%0%0% FCF yield 10% 9% 6% 5% 6% FCF yield 7% 7% 9% 9% 9% 10% ROE 16% 15% 17% 19% 18% ROE 22% 24% 24% 24% 25% 25%

Source: Deutsche Bank estimates; Company data, Priced as on 5 September 2013, Note: Figures in Source: Deutsche Bank estimates, company data, Priced as on 5 September 2013, Note: Figures in local currency unless stated local currency unless stated,

Page 26 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 38: SAP snapshot Figure 39: Temenos snapshot SAP Analyst name Email Temenos Spot price: 23 Tgt. Price 24 Kai Korschelt [email protected] Spot price: 53 Tgt. Price 57 Rec. Hold 2010 2011 2012 2013e 2014e 2015e 2010 2011 2012 2013e 2014e 2015e EPS (adjusted) 1.5 0.9 0.9 1.1 1.3 1.6 EPS (DB adjusted) 2.3 2.8 3.0 3.3 3.7 4.1 DPS 0.0 0.0 0.3 0.3 0.4 0.4 DPS 0.6 1.1 0.9 1.0 1.1 1.3 BVPS 6.9 5.1 5.6 6.1 7.0 8.2 BVPS 8.3 10.7 11.9 13.7 15.7 18.1 ROE 16% -7% 7% 18% 19% 20% ROE 20% 31% 21% 21% 21% 21% EBITDA mgn 27% 19% 24% 29% 31% 33% EBITDA mgn 33% 35% 34% 34% 36% 37% Market Cap (US$'mn) 2,950 1,130 1,211 1,740 1,740 1,740 Market Cap (US$'mn) 56,163 68,666 78,552 84,197 84,197 84,197 P/E 42.4 n/a 50.1 21.1 18.0 15.3 P/E (DB) 15.4 14.7 17.0 16.2 14.6 13.0 EV/EBITDA 10.2 9.5 11.4 11.1 9.5 8.2 EV/EBITDA 24.2 14.0 24.2 13.6 11.8 10.4 Div yield 2% 3% 2% 2% 2% 2% Div yield 0%0%2%1%2%2% FCF yield 6% 7% 5% 4% 6% 7% FCF yield 4% 8% 6% 5% 5% 6% ROE 20% 31% 21% 21% 21% 21% ROE 16% -7% 7% 18% 19% 20%

Source: Deutsche Bank estimates; company data, Priced as on 5 September 2013, All figures in local Source: Deutsche Bank estimates; Priced on 5 September 2013; Note: Temenos is not under DB equity currency unless mentioned research coverage. Statistics above, including target price, refer to Bloomberg consensus in local currency unless stated

„ The roll-out of mobile and tablet banking services has significantly increased Online commerce, digital customer interactions with banks and their demands for channel development. This banking and the rise of plastic compounds transaction growth seen from online commerce and card payments noted in the payments chapter below, placing greater burdens on systems. Adding are all compounding mobile capabilities needn’t take long and is already, we think, industry standard: transaction volume growth one bank we spoke to used Fiserv to deploy mobile banking within 8-9 months.

Figure 40: The volume of bank customer interactions increases dramatically with digital channels….

25

> 20 20

15 month 10

6

5 Average visits per active customer per customer active per visits Average <1

0 Branch Online Mobile

Source: Monitise company data

„ We expect bank regulation and oversight will move from capital and liquidity risk to We expect regulators to focus focus again on operational issues given more available capacity. Recovery and on operational risk in a much resolution planning will be a key catalyst for a greater focus on IT capability, we think. High profile IT outages provide public evidence that action is required. Cyber more material way security is also taking on a more prominent role in system risk management.

„ In some markets there is a greater need for efficiency in customer on-boarding and In some markets there is a off-boarding (particularly in EM). Some markets also have more complex pricing greater need to be agile and models in which better systems would confer a greater capacity to price more dynamically. efficient in pricing and customer acquisition We examine the last three of the above arguments starting with regulation.

Deutsche Bank AG/London Page 27

8 September 2013 Financial Services Retail Bank Strategy

Regulatory compliance 1: System outages force regulator reaction RBS suffered a single system IT failure in the summer of 2012 which strengthened the RBS suffered a single system case for regulators to require banks to upgrade or replace core IT. The incident saw IT failure in 2012 which saw over twelve million accounts frozen in Natwest, RBS and Ulster Bank brands for up to a over twelve million accounts week while staff manually updated balances out of sync due to a failure in the bank's frozen for up to a week CA7 batch process scheduling software.

The direct financial costs of this failure so far have been limited. A £175m provision was The direct financial costs of made in 2012, £148m of which had been used by the end of the year. The bulk of costs this failure have been incurred related to staff expenses incurred in dealing with the incident (extended branch reasonably limited with a and call centre hours, opened on Sunday). Customers abroad were transferred funds via £175m provision passed in Western Union. Corporate customers could withdraw limited funds with relationship the year, 1% of group manager agreement. In some cases provision was made for interest and fee waivers. In total however, the £175m expense amounts to just 1% of RBS’ 2012 revenues. revenues in the year

One of the lessons from this failure is that for all the criticisms of banks for running The failure confirmed that multiple product systems on an un-integrated basis, customers and regulators should multiple brands and reasonably plan for all operating subsidiaries to be vulnerable in a material systems geographies can suffer the failure. Within this context we think it right to be sceptical about how much safer same failure from one system; operationally banks will be under ring-fencing proposals to be implemented in the UK we don’t think ringfencing under Vickers/ICB and potentially in Europe under Liikanen. Legal separation is not the same as operational separation. reduces operational risk

The Financial Conduct Authority (the UK consumer watchdog) announced in April 2013 The FCA has begun an that it had begun an enforcement investigation into the RBS incident. The FCA will enforcement investigation reach its conclusions in due course and will decide whether or not enforcement action should follow which may have cost and reputational consequences.

RBS is not the only bank to have suffered outages, with media reports highlighting RBS is not the only bank to 10111213 14 15 service breaks elsewhere . Recent events at Google , Nasdaq and elsewhere have suffered outages confirm – though hardly necessary – that banks are not the only businesses vulnerable to IT failures. We don’t think regulators or the governments to which regulators report are interested in relative arguments (“other countries/industries have problems too”) and expect pressure for action to increase, even though the immediate consequence of systems replacement and renewal is higher not lower operational risks.

Regulatory compliance 2: Cyber threats are an increasing concern Distributed Denial of Service (DDOS) attacks are relatively common in practice and have Cyber attacks are relatively 16 interrupted bank IT operations in the past. ING was affected by a DDoS attack in April common and have 2013, described in its quarterly report17: interrupted operations in the past. Rightly, banks generally ...in April, banks worldwide, including ING, became the target of distributed do not advertise being the denial-of-service attacks (DDoS). During such an attack, a website is bombarded with an excessive amount of traffic. Though a DDoS attack is blocked by a firewall, victim of such an attack

10 RBS and NatWest FAIL downs services across UK, The Register, 6 March 2013 11 24 April 2013, Nationwide Building Society hit by online banking outage, Computerworld UK 12 16 April 2013, Lloyds internet banking service knocked offline, Computerworld UK 13 ING Bank in the Netherlands experienced problems processing payments in April 2013, causing online balance information to appear incorrectly. 14 19 August 2013, Analyzing Friday's Google Outage, Forbes 15 22 August 2013, Nasdaq suffers reputation hit after trading freeze, USA Today 16 18 July 2013, Wall Street tests readiness for hackers and pandemic, Financial Times, 17 ING 1Q13 Quarterly report Press Release, p.12

Page 28 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

the firewall can become so busy as it tries to stop the unwanted traffic that customers can experience difficulties in accessing ING’s system. While the DDoS attacks on ING did cause inconvenience to customers, they never compromised ING’s banking systems and customer databases. ING continues to closely monitor traffic to its website and mobile applications to ensure that the company is well prepared for potential incidents in the future. Actions have also been taken to avoid or minimise disruption for customers. ING is working closely with other banks and the appropriate authorities to take coordinated actions against cyberattacks, if and when necessary.

Not only banks are exposed to these risks: in a report published by the World Federation of Exchanges and IOSCO on 16 July, participants confirmed that half of exchanges repelled cyber attacks in the last year, despite operating a relatively closed loop of information technology systems in which access is mainly limited to brokers and clearing houses18.

Businesses will generally not rush to advertise that they’ve been the target (or The most frequently successful victim) of an attack: we expect cyber crime to be under-reported in public. highlighted operational risk in That said, the most frequently highlighted operational risk in the Bank of England's the BoE’s 1H13 systemic Risk 2013 1H Systemic Risk Survey related to cyber attacks. Sensibly therefore, regulators Survey related to cyber including the UK's Prudential Regulatory Authority (PRA) have increased the resources attacks in this area. This extends to scenario planning and simulations: On 18 July, an industry body conducted a simulated cyber attack called Quantum Dawn 2 involving 50 financial institutions and government bodies.

Regulatory compliance 3: The “how” of the regulatory pressure will be interesting We don’t doubt that the above reasoning will see increased regulatory pressure in some How regulators bring markets for IT renewal. But how this pressure is brought to bear within the context of pressure to bear is going to regulation will be interesting. This is because the current operational risk requirements – be interesting: the 32 banks which not only factor in IT but other issues such as conduct risk, litigation, fraud and we survey hold ~ E87bn of malicious damage – are not insignificant. The thirty-two banks in Figure 41 report a capital against op risk, a total of E815bn in operational risk weighted assets. At our assumed target Basel 3 capital ratio – which we set by bank – this translates into an E87bn shareholder capital 70bps drag off normalized requirement. At a 10% CoE, say, this is an E8bn annual post tax earnings drag, 8% of ROE 2014 earnings and 0.7% off sector ROE, assuming no real yield on this capital. Across the sector, operational risk RWAs average 14% of the credit risk equivalent, with banks such as Credit Suisse (36%), Julius Baer (36%) and UBS (56%) carrying much higher proportional burdens, mostly due to business mix.

18 16 July 2013, Half of exchanges fight off cyber attacks, Financial Times

Deutsche Bank AG/London Page 29

8 September 2013 Financial Services Retail Bank Strategy

Figure 41: Operational risk capital requirements look high relative to recent (elevated losses) Country Bank Basel III CET I ratio Credit Risk RWAs (E’m) RWAs/ Operational risk Operation Tgt Tier 1 Capital Capital Loans RWAs (E’m) al risk against against credit risk ops risk (E’m) (E’m) 2011 2012 2013e 2011 2012 2012 2011 2012 % Credit B3 2012 2012 risk Austria Erste 7.7% 9.1% 10.5% 97,630 90,434 73% 11,210 11,175 12% 10.0% 9,043 1,118 Austria Raiffeisen Bk. 5.4% 6.6% 7.3% 77,150 68,136 88% 9,900 10,569 16% 10.0% 6,814 1,057 Benelux KBC Group 5.3% 5.3% 9.0% 85,786 69,149 50% 10,744 11,045 16% 10.0% 6,915 1,105 Benelux ING n/a 10.4% 10.9% 252,733 209,722 40% 35,450 35,450 17% 10.0% 20,972 3,545 France BNP Paribas 8.0% 10.0% 10.5% 472,780 430,560 68% 54,617 51,154 12% 10.0% 43,056 5,115 France Cred Ag 6.3% 6.6% 7.4% 277,800 257,100 78% 23,100 22,900 9% 10.0% 25,710 2,290 France Soc Gen 7.1% 8.3% 9.9% 273,297 254,134 73% 43,442 41,321 16% 9.5% 24,143 3,925 Germany Commerzbk 6.9% 7.6% 9.0% 187,333 167,301 60% 26,127 22,238 13% 10.0% 16,730 2,224 Iberia Santander n/a 6.5% 7.3% 455,100 447,225 62% 70,013 72,761 16% 10.0% 44,723 7,276 Iberia BBVA n/a 7.5% 8.6% 277,525 271,488 77% 30,063 29,350 11% 10.0% 27,149 2,935 Ireland Bk of Ireland 8.3% 5.9% 5.1% 61,483 51,873 56% 4,530 3,608 7% 10.0% 5,187 361 Italy BPM 8.1% 7.8% 8.1% 37,203 34,915 100% 2,707 2,547 7% 8.0% 2,793 204 Italy Banco 6.2% 7.5% 7.5% 80,154 47,707 52% 6,245 5,919 12% 8.0% 3,817 474 Italy Credem 8.5% 8.0% 8.8% 14,874 14,947 72% 1,467 1,468 10% 8.0% 1,196 117 Italy ISP 8.6% 9.7% 10.3% 277,500 253,313 67% 24,825 25,745 10% 10.0% 25,331 2,575 Italy UBI Banca 8.0% 8.6% 10.4% 84,332 70,145 76% 5,759 5,466 8% 8.0% 5,612 437 Italy Unicredit 7.6% 9.6% 10.1% 376,784 358,553 66% 51,455 51,187 14% 10.0% 35,855 5,119 Nordics Danske Bank n/a 12.5% 12.3% 101,407 91,742 41% 12,043 10,078 11% 13.0% 11,926 1,310 Nordics DnB NOR n/a 12.2% 12.7% 111,403 111,475 63% 8,699 9,871 9% 13.0% 14,492 1,283 Nordics Nordea 8.6% 11.6% 13.4% 161,604 145,340 42% 15,452 16,229 11% 13.0% 18,894 2,110 Nordics SEB n/a 13.1% 14.2% 63,276 55,657 39% 4,742 4,687 8% 13.0% 7,235 609 Nordics Handelsbanken n/a 15.9% 16.7% 50,013 49,490 25% 5,774 6,091 12% 15.0% 7,424 914 Nordics Swedbank 9.2% 15.4% 17.2% 46,667 45,299 31% 6,113 6,302 14% 15.0% 6,795 945 Swiss CSG 4.7% 6.8% 10.3% 113,619 103,586 52% 29,762 37,371 36% 10.0% 10,359 3,737 Swiss Julius Baer 19.0% 18.7% 16.0% 7,034 6,795 41% 2,385 2,422 36% 15.0% 1,019 363 Swiss UBS 6.5% 9.8% 13.4% 95,772 79,155 34% 48,548 44,122 56% 13.0% 10,290 5,736 UK Barclays 7.8% 8.1% 9.9% 293,892 294,503 52% 42,833 66,700 23% 11.0% 32,395 7,337 UK HSBC 9.0% 9.0% 11.1% 738,927 680,813 90% 95,855 92,680 14% 11.0% 74,889 10,195 UK LBG 7.1% 8.1% 10.4% 362,212 317,728 48% 36,660 34,391 11% 11.0% 34,950 3,783 UK RBS 7.2% 7.3% 9.0% 412,536 397,661 72% 45,448 56,442 14% 11.0% 43,743 6,209 UK StanChart 10.8% 10.7% 10.9% 169,959 186,913 85% 22,180 23,311 12% 11.0% 20,560 2,564 Sector total / ave 7.6% 8.6% 10.1% 6,204,661 5,731,953 61% 788,150 814,599 14% 10.7% 606,236 86,971 Source: Deutsche Bank, Company data

Sure, operational risk capital requirements are lower than those for credit risk, but then Even if we add in the very operational risk costs are much lower too. The UK bank association, the BBA, maintains significant conduct risk a database of operational risk losses19. In 2011 – the most recent data – median losses charges taken in the UK for were low (though we couldn’t find the number of incidents to scale this up) with the products like PPI, we see largest costs coming from property losses. If the market is not worried about the cost of annual operational risk losses theft and malicious damage when looking at banks, how will we muster the argument that the regulatory burden associated with op risk is too low? Even adding in the well below the capital conduct risk charges taken in the UK for products like PPI, we see annual operational provided… risk costs as being well below the capital provided.

19 www.bbagold.org

Page 30 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 42: Median operational risk losses by type (2011, E) Risk factor Median loss (E, 2011) Damage to physical assets and injury 215,600 Business disruption and systems failures 50,600 Execution, delivery and process management 48,200 Internal fraud 45,100 External fraud 38,200 Clients, products and business practices 34,400 Employment practices and workplace safety 28,700 Source: BBAGold

In the end, it may well come down to systemic stability arguments and higher board- …it will be interesting to see level accountability for a very complex set of operational issues. Whatever the source of how regulators justify a leverage employed, we are confident that this will prove an increasingly important ratcheting on pressure to driver of change for the banks. upgrade systems that we expect Transaction volumes will likely keep rising – eventually best to re-platform Our work on digital channels and the payments revolution point to customer Banking volumes will keep transaction volumes which will grow meaningfully faster than nominal GDP for the rising (it would be a very bad foreseeable future. We noted an increase in IT depreciation costs of, say, 40% from thing if they didn’t; systems implementation, perhaps offset to some degree in the medium term by cheaper run- upgrades seem to drive a the-bank change costs (new systems being easier to amend than tangled webs like c.40% increase in Figure 19) and the 15% reduction in branch costs we think achievable over this period. depreciation expense

Net net therefore, banks should strive to hold operations costs flat despite rising Though new systems bringing transaction volumes and natural underlying wage inflation, we think. StanChart in its higher fixed costs, we think upgrade to its trade finance platform has managed to deliver this result (Figure 43). these bring headroom to With revenue growth of course, stable operations costs contributes to a declining manage higher volumes for divisional cost-income ratio (Figure 44). flat costs

Figure 43: StanChart trade finance IT upgrades helped Figure 44: ...delivering an improved operations offset inflation and volume growth impacts... cost/income ratio over time

2,000 Trade Finance - income and operations cost (US$'m) 200 10 Operations cost income ratio and productivity 8.32

1,600 8 Inc ome 160

Operations cost 6.22 5.88 1,200 6 5.36

120 3.49 800 4 2.74 2.85 2.89 Total tradeincome Total 80 400 Operations cost 2

- 40 0 2008 2009 2010 2011 2008 2009 2010 2011 Inc ome Operations cost Operations cost income ratio Productivity (volume in 000/FTE)

Source: Standard Chartered company data Source: Standard Chartered company data

Deutsche Bank AG/London Page 31

8 September 2013 Financial Services Retail Bank Strategy

Speed to market and operating agility; depends on market dynamics The strength of the argument around operating agility varies by market. In South Africa The strength of the argument – Standard Bank’s main market – retail banking operates via a relatively complex fee- around speed to market and driven model with higher customer churn amongst lower-value customers. FirstRand, operating agility varies from one of SBK’s major competitors has a 98 page retail banking fee schedule20. SBK market to market advised that under its old IT it would have taken nine months to re-price the product range entirely given a need to recode product platforms. The bank says that the launch of a new SAP core platform in SA and Finacle in its other African markets will “ultimately facilitate our strategic objective of being more customer centric and delivering an excellent, consistent customer experience”. It should also reduce customer on- and off-boarding costs, which is important in delivering decent cost metrics whilst contributing towards better banking penetration for the less affluent.

In most markets we don’t see rate of change or speed to market as key drivers of a decision to upgrade. Where speed to market is an issue, newer “bank-in-a-box” entities may pose a more material competitive threat.

Figure 45: UK banks: a simpler Figure 46: relative to other Figure 47: ...some US peers (JPM) market with lower churn... industries... 70% 20% 12% 58% 18% 60% 10% 16% (36)% 50% 8% 14% 6% 40% 12% 4% 30% 10% 2% 20% 18% Household attrition rate (% (% p.a.) rate attrition Household Customerturnover (% p.a.) 13% 14% 8% 11% account in the previous 12 months 12 previous in the account 0%

% of survey% respondentswho switched 10% 6% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 6%

Main personal current account Credit Cards 0% Jul-11 Jul-12 Apr-11 Oct-11 Apr-12 Oct-12 Jan-11 Jun-11 Jan-12 Jun-12 Feb-11 Mar-11 Feb-12 Mar-12 Aug-11 Sep-11 Nov-11 Dec-11 Aug-12 Sep-12 Nov-12 Dec-12 May-11 Savings Mortgages UK personal Sky TV UK gas (2011) UK electricity UK contract UK prepaid May-12 current a/c (2011) mobile phonesmobile phones SME main bank relationship

Source: Independent Commission on Banking, Figure 7.3 Source: Ofgem, UK Independent Commission on Banking, Source: Company data; Note: Adjusted by company for additional Vodafone (above metrics are for VOD), Sky 25k households in October 2012 due to escheatment rule changes

2020 https://www.fnb.co.za/downloads/pricing-guide/Personal_Pricing.pdf

Page 32 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Rise of the machines: Mobile & Tablet Banking

Technology is changing the way we manage our finances

Though most readers will have been using online banking for a decade or more, few Banks have moved rapidly to will dispute the significant change brought by the transition from online to mobile expand computer-driven self- banking via tablet and smartphone. The move to emphasise mobile and tablet channels service to smartphone and is in keeping with growth in mobile versus fixed computing and internet capacity and is tablet, in line with changes in supported by cost and convenience benefits for the customer and the bank. Last year, customer behaviour mobile subscriptions were up 8% globally versus fixed line which contracted by 2% (Figure 48). Internet usage via mobiles is growing even faster with mobile broadband subs up 3x faster than fixed line (Figure 49).

Figure 48: Fixed line in decline, mobile growing in Figure 49: Mobile broadband outgrowing mobile EM/DM subscriptions – On the move in the ascendance 9,000 30% 3,000 70% 8,000 25% 2,500 60% 7,000 20% 50% 6,000 2,000 5,000 15% 40% 1,500 4,000 10% 30% 3,000 5% 1,000 20% Subscriptions (millions) Subscriptions 2,000 (YoY Subscriptions %) Subscriptions (YoY %) 0% (millions) Subscriptions 1,000 500 10% - -5% 2005 2006 2007 2008 2009 2010 2011 2012* 2013* - 0% 2007 2008 2009 2010 2011 2012* 2013*

Developed - Fixed Developing - Fixed Developed - Mobile Developed - Mobile Developing - Mobile Developed - Wired Developing - Mobile Mobile (YoY %) Fixed line (YoY %) Developing - Wired Mobile (YoY %) Wired (YoY %)

Source: International Telecommunication Union Source: International Telecommunication Union

As discussed in the Future of branches and Rotten Core chapters, mobile banking Mobile banking capabilities capabilities are particularly important in many emerging markets where customer churn are especially important in is higher, revenue per customer is lower, customers are more geographically dispersed EM where churn is higher, and product pricing complexity is often higher than in developed markets like the UK. ave revenue is lower and As EM telco infrastructure improves, the scope for greater mobile penetration at lower geographic dispersion is often customer cost will rise, driving further increases in digital customer penetration. With much lower starting penetration rates in EM relative to DM (1/4 of DM levels in fixed higher; EM penetration rates line, 1/5 in broadband, 2/3 in mobile), strong future growth in digital customers is still low, provides plenty of assured for many years (Figure 50). growth upside

Deutsche Bank AG/London Page 33

8 September 2013

Financial Services Retail Bank Strategy

Figure 50: EM banks have the greater need for mobile competence given growth dynamics and market profile 2007 2008 2009 2010 2011 2012* 2013* Fixed line - DM 45 44 46 45 44 43 42 Fixed line - EM 13 13 12 12 12 11 11 Mobile -DM 102 108 113 115 119 124 128 Mobile - EM 39 49 58 69 78 84 89 Mobile broadband - DM 18 27 37 43 55 63 75 Mobile broadband - EM 1 2 3 4 8 13 20 Wired broadband - DM 18 21 22 24 25 26 27 Wired broadband - EM 2 3 4 4 5 5 6 Source: Source: International Telecommunication Union

Mobile banking statistics published by Denmark’s Danske Bank tell a powerful story: Danske Bank data suggests that mobile is cannabalising „ eBanking customers (online, tablet and mobile) are growing faster than customer desk-top and that eBanking is numbers (+2% YoY in 2Q13 vs. -3% YoY) but not by enough to change significantly the proportion of clients using the bank in this way (46% at 2Q13 from 44% at cannabalising in-branch 2Q12). eBanking overall is maturing. transactions

„ Mobile is growing much more rapidly (40% of eBanking customers at 2Q13 from 23% YoY).

„ Cannibalisation of eBanking transaction volumes is in relative infancy with 27% of digital transactions via mobile and tablet, but growing rapidly;

„ In-branch teller transaction volumes21 are collapsing: branch numbers are down 29% YoY, teller transactions are down 6% more at -35% YoY.

Figure 51: Danske Bank three year digital strategy data 1Q10 2Q10 3Q10 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 Retail customers (000’s) 5,041 5,035 5,037 5,021 4,979 4,962 4,922 4,932 4,928 4,919 4,916 4,852 4,803 4,769 ebanking customers (000's) 1,793 1,833 1,857 1,915 1,980 2,021 2,050 2,103 2,163 2,174 2,196 2,216 2,220 2,213 eBanking as % of customers 36% 36% 37% 38% 40% 41% 42% 43% 44% 44% 45% 46% 46% 46% Mobile / ebanking customers (DK) 3% 3% 5% 9% 11% 14% 17% 20% 23% 26% 29% 32% 36% 40% Retail Branches 720 704 690 670 667 662 661 645 627 578 559 491 469 409 Retail FTE 13,163 13,281 12,931 12,821 12,807 12,885 12,890 12,832 12,640 12,485 12,313 11,941 11,482 11,336 Transaction volumes (000's) In branch teller transactions 12,832 12,448 12,173 11,721 10,895 10,846 10,725 10,427 9,634 9,010 8,790 8,130 5,974 5,830 eBanking transactions 23,872 23,664 23,138 24,697 24,858 24,587 23,900 25,167 25,234 24,454 22,841 24,057 22,474 18,327 Mobile Banking 5 5 14 98 249 474 786 1,150 2,341 2,849 2,960 2,849 3,104 3,808 Tablet Banking 0 0 0 0 0 0 0 61 265 375 520 701 903 1,124 Tablet & mobile as % of eBanking 0% 0% 0% 0% 1% 2% 3% 5% 10% 13% 15% 15% 18% 27% Growth rates (YoY %) Retail customers n.a. n.a. n.a. n.a. -1% -1% -2% -2% -1% -1% 0% -2% -3% -3% eBanking customers n.a. n.a. n.a. n.a. 10% 10% 10% 10% 9% 8% 7% 5% 3% 2% Retail Branches n.a. n.a. n.a. n.a. -7% -6% -4% -4% -6% -13% -15% -24% -25% -29% Retail FTE n.a. n.a. n.a. n.a. -3% -3% 0% 0% -1% -3% -4% -7% -9% -9% In branch teller transactions n.a. n.a. n.a. n.a. -15% -13% -12% -11% -12% -17% -18% -22% -38% -35% eBanking transactions n.a. n.a. n.a. n.a. 4% 4% 3% 2% 2% -1% -4% -4% -11% 34% Mobile banking transactions n.a. n.a. n.a. n.a. n.m n.m n.m n.m 840% 501% 277% 148% 33% 53% Source: Company data

21 In-branch transactions at a teller window, whether deposits or withdrawals.

Page 34 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Similar trends are evident in South Africa’s ABSA digital penetration statistics. Internet Mobile banking penetration in has matured remarkably quickly with virtually all of the digital growth now coming in South Africa is already smartphones and tablets. It is interesting to note reported cellphone customer remarkably good we think penetration in South Africa is in line with that of Danske in Denmark.

Figure 52: ABSA digital customer penetration statistics ABSA 2009 2010 2011 2012 Cellphone banking customers 1.7 2.5 3.2 4.2 YoY % change n.a. 50% 28% 32% Internet banking customers 1.1 1.1 1.1 1.2 YoY % change n.a. 4% 5% 3% Number of customers 11.0 11.1 11.3 10.2 YoY % change n.a. 1% 2% -10% Proportion digital o/w cellphone 15% 23% 29% 39% o/w internet 10% 10% 10% 11% Source: Company data; 2012 annual report, p. 40.

While we’ll see continued growth in internet penetration rates in some emerging Mobile penetration rates look markets the real growth will come from mobile. Indeed, survey evidence is that mobile good, but with further room penetration of customers is already pretty good, with very rapid adoption by customers for growth in the number of post launch reported by almost all banks we spoke to (Figure 53). In one case reported users and the range of uses by Monitise22, a bank client achieved 30% customer penetration within three months of made available to these launch. Post launch of basic digital offerings, however, we think development and delivery of broader mobile offerings which improve efficiency (end to end self-service customers with no-rework being the ultimate goal) will occupy minds most.

Figure 53: Electronic and mobile banking penetration rates by country

70%

60%

50%

40%

30%

20%

10%

0% % of customers active indigital channels

Source: Company data ;ING July 2013 Survey Financial Empowerment in the Digital Age

22 The largest provider of mobile money solutions to banks and other service companies, more below.

Deutsche Bank AG/London Page 35

8 September 2013

Financial Services Retail Bank Strategy

Monitise’s most recent estimates for future industry growth have mobile banking Monitise forecast 26% CAGR customers growing by 26% compound between 2011 and 2016 to a total penetration in mobile banked customers rate of around 30%. We think penetration will rise faster and proceed further. 2011-2016

Figure 54: Monitise projects 26% CAGR in mobile-banked customers 2011-2016 Region Population Adult Pop (15+) Banked population mBanked population Unbanked 2011 2016e 2011 2016e 2011 2016e 2011 2016e 2011 2016e mn mn mn mn mn mn Mn % of mn % of mn mn banked banked WORLD 7,079 7,483 5,228 5,594 2,588 3,057 280 11% 875 29% 2,640 2,536 CAGR 1.10% 1.40% 3.40% 26% -1%

DEVELOPED 1,045 1,074 877 897 803 849 113 14% 420 49% 73 47 CAGR 1% 0% 1% 30% -8% NAmerica 361 376 292 305 259 287 47 18% 158 55% 33 18 W Europe 435 443 369 370 340 348 39 11% 149 43% 30 23 DevAPAC 249 254 215 221 204 215 28 14% 114 53% 11 7

Hybrid/Emerging 6,035 6,410 4,351 4,697 1,785 2,208 166 9% 455 21% 2,566 2,489 CAGR 1% 2% 4% 22% -1% CE Europe 416 419 346 345 165 185 15 9% 37 20% 181 160 Middle East 296 322 208 231 101 127 10 10% 25 20% 108 104 Latam 604 637 440 477 172 219 7 4% 22 10% 269 258 Emerging APAC 3,665 3,851 2,724 2,925 1,202 1,461 123 10% 349 24% 1,521 1,464 Africa 1,054 1,180 633 719 145 216 11 5% 22 10% 488 503 Source: Monitise

Banking is not the only industry seeing a move to smaller screens Banks are not alone in seeing customers increasingly transition to smaller screens in a Impact of customer shift to post-PC cloud era. Microsoft’s 2Q13 revenues were 4% below our forecast with weak smaller screens being backed PC trends our most significant concern (Figure 55)23. Google’s 2Q13 average advertising by banks evident in results rates disappointed relative to market expectations as the mix shifted more appreciably posted by the likes of than expected towards mobile advertising where price per click is lower (Figure 56)24. Microsoft and Google also The DB research view is that mobile clicks are incremental to bigger-screen / higher margin clicks and that the decline in CPC is over-done as a risk. In many ways mobile banking customer use is incremental existing customer demands too.

23 DB Research: Microsoft – PC weakness continues to be a drag, 19 July 2013 24 DB Research: Google – Network Clean Up Clouds A Solid Core, Setting Up Well For 2H, 19 July 2013

Page 36 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 55: We remain concerned about headwinds at Figure 56: Google’s declining cost-per-click partly down Microsoft in the post-PC cloud-driven era to increased share of mobile within the pie $8,000 50% 14,000

$7,000 40% 12,000 $6,000 30% 10,000 $5,000 20% 8,000 $4,000 10% 6,000 $3,000 YoY % change

$2,000 0% 4,000 Revenues (Quarter, US$'m) $1,000 Gross Ad revenue ($'m) -10% 2,000 $0 -20% 0 Q1-12 Q2-12 Q3-12 Q4-12 Q1-13 Q2-13 Q3-13 Q4-13 1Q09 2Q09 3Q09 4Q09 1Q10 2Q10 3Q10 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13

Windows Division Server and Tools Online Services Division Microsoft Business Division Gross Ad Revenue ($'m, RHS) Gross Ad Revenue (YoY %) Entertainment and Devices Division Unallocated and other Average Cost per Click (YoY %) Paid Clicks Growth (YoY %)

Source: Company data Source: Company data

We provide snapshots for the Microsoft and Google in Figure 57 and Figure 58.

Figure 57: Microsoft snapshot Figure 58: Google snapshot Microsoft Analyst name Email Google Analyst name Email Nandan Amladi [email protected] Ross Sandler [email protected] Spot price: 31 Tgt. Price 32 Rec. Hold Spot price: 879.6 Tgt. Price 970 Rec. Buy 2010 2011 2012 2013e 2014e 2010 2011 2012 2013e 2014e EPS (DB adjusted) 2.1 2.7 2.0 2.6 2.4 EPS (DB adjusted) 29.6 36.1 39.9 40.2 47.8 DPS 0.0 0.0 0.0 0.0 0.0 DPS 0.0 0.0 0.0 0.0 0.0 BVPS 5.2 6.7 7.8 9.3 12.2 BVPS 145.1 180.1 219.2 265.6 309.2 ROE 41% 45% 28% 30% 22% ROE 21% 17% 17% 15% 14% EBITDA mgn 43% 43% 42% 39% 33% EBITDA mgn 40% 47% 39% 37% 38% Market Cap (US$'mn) 244,927 221,123 239,215 251,192 257,871 Market Cap (US$'mn) 170,707 183,653 210,318 293,022 293,022

P/E (DB) 13.1 9.6 14.1 11.5 13.2 P/E (DB) 18.1 15.8 16.1 21.9 18.4 EV/EBITDA 7.7 5.7 5.8 5.9 5.7 EV/EBITDA 11.8 10.3 10.4 13.0 10.1 Div yield 0% 0% 0% 0% 0% Div yield 0% 0% 0% 0% 0% FCF yield 9% 11% 12% 10% 10% FCF yield 4% 6% 6% 5% 6% ROE 41% 45% 28% 30% 22% ROE 21% 17% 17% 15% 14%

Source: Deutsche Bank estimates; Priced as on 5 September 2013, Note: All figures in local currency Source: Deutsche Bank estimates; Priced as on 5 September 2013, Note: All figures in local currency unless mentioned unless mentioned

What the Heads of Digital at banks are thinking

We met with the heads of digital banking at a number of banks to better understand plans for this channel and how these impact broader strategy. We think a number of common themes were evident:

„ Mobile solves a customer need that wasn't evident before

„ There is little evidence that mobile banking has impacted plans for other customer service channels much yet, including branches

„ Banks appear very focussed on staying brand-relevant to customers

„ Opinions on mobile as a sales channel are not uniform

„ Digital can reduce legacy product costs - cheques

Deutsche Bank AG/London Page 37

8 September 2013

Financial Services Retail Bank Strategy

1) Solving a problem that wasn't there / wasn't previously fulfilled Customers of all businesses increasingly demand the ability to transact seamlessly with Mobile banking customer service providers via all distribution channels. Banking is no different. When looking at interactions are > 3x the mobile banking however, we found surprising the radical increase in the number of volume of online, > 20x that interactions between bank and customer that mobile allows. LBG and RBS report of branch visits average logins per month of 19 and 31 respectively for active customer; this is corroborated by Monitise’s statistics for the UK where it is service provider to HSBC and RBS, amongst others (Figure 59).

Figure 59: The volume of bank customer interactions Figure 60: …helping to customers plan better, avoid fees, increases dramatically with digital channels…. and, to some extent, self-serve 25 About Financial: Saving money More the same Less > 20 20 Overdraft fees 7% 44% 49% Late fees 6% 44% 50% 15

month Convenience: I don't About 10 have to: More the same Less 6 Write a cheque 7% 48% 45% 5

Average visits per active customer per per customer active per visits Average Visit a branch 9% 49% 43% <1

0 Contact a call centre 9% 51% 40% Branch Online Mobile Visit an ATM 11% 66% 23% Source: Monitise company data Source: Monitise

We’re confident that virtually no customers were visiting bank branches daily, pre- Digital is helping customers digital. A good deal of this activity, therefore, is using otherwise dead-time on trains self-serve, and be more and between meetings to be more up to date on finances: making transfers, checking organized, reducing late- balances, confirming transfers into and out of accounts and the like. This appears to payment and overdraft costs, have driven down volumes in telephone banking which is positive for customer which should support (quicker, easier, more convenient) and for the bank (end to end processing, zero re- work). customer loyalty

The survey evidence in Figure 60 shows that digital helps customers avoid costs of overdraft and unauthorised limit breaches, which should be helpful for loyalty and promoter scores. Internal data published by JPMorgan Chase at its investor day this year confirms that technology – done well – not only contributes to more sales leads but also supports better brand perception and lower churn (Figure 61, Figure 62).

Page 38 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 61: Technology can generate sales leads and Figure 62: ...and contribute to lower customer churn better brand perception... 100% 140 +33% 89% 90% 83% 85% 120 80% 76% +35%

70% 62% 100 57% 59% 60% 56% 80 50%

40% 60 30% Customer retention 40 20%

10% 20 0% Consider Chase for a Outstanding customer Innovative products Forward thinking 0 mortgage service technology Banking Card

Pre-use Post-use Non-mobile Mobile

Source: JPMorganChase February 2013 Investor Day Source: JPMorganChase February 2013 Investor Day

2) Digital hasn't impacted other distribution channels obviously yet Though people talk about writing fewer cheques, visiting branches less often and so on, Outside of the Nordics there’s conclusive changes to distribution strategy to reflect the new environment are hard to little evidence of a convincing find outside of the Nordics and perhaps Australia to a lesser extent (the Danske teller change to the branch volume declines reported in Figure 51 above are stunning, see more in branch chapter strategy…yet below).

Sensible and less of a leap we think is a rebalancing of call-centre responsibilities. In Less of a leap, we think we be future we expect calls on average will be more important than before, with a higher a rebalancing of call centre proportion of dissatisfied customers, clients with technical enquiries or sales calls. responsibilities to handle Combined with relatively high wage inflation in traditional off-shoring destinations, we tougher, more important expect more banks to repatriate operations to be closer to customers. calls; given inflation differentials this might Figure 63: Higher offshore inflation rates, higher proportion of tougher sensibly be relocated nearer customer calls and a need to rebuilding trust suggests sensible rebalancing to to customers local sites

8.0%

7.0% 6.5% 5.8% 6.0% 5.5%

5.0% 4.2% 4.0%

3.0% 2.2% 2.1% 2.0% 1.5% Consumer Price Inflation (%) 1.0%

0.0% Euroland US UK India Indonesia Philippines Brazil

2012 2013e 2014e

Source: Deutsche Bank Markets Research

Deutsche Bank AG/London Page 39

8 September 2013

Financial Services Retail Bank Strategy

Why the apparent inertia in shifting branch strategy? This is understandable, at least in The inertia in changing part. First, it is clear that the average bank customer was not substituting 20-30 branch branch offerings will give way visits a month for 20-30 online interactions. Second, this technology is fairly new and as customer and operator companies of all stripes are gradually expanding their digital offerings and learning as confidence in the they go what customers want and whether and how to meet these desires. ramifications of digital grows

Figure 64: Ulster Bank transactions by channel Figure 65: Australian bank average customer touch- points per month by channel 100% 100% 7% 90% 21% 90% 31% 27% 80% 80% 2% 32% 70% 3% 70% 3% 60% 31% 60% 7% 50% 39% 50% 45% 40% 40% 40% 30% 30% 30%

% of transactions by channel 20% % of transactions by channel 20% 31% 10% 21% 10% 15% 14% 0% 0% FY09 FY11 1Q13 2002 2004 2006 2008 2010 2011 2012

Mobile Online Phone ATM Branch Mobile Online Phone ATM Branch

Source: RBS company data Source: Westpac company data

Figure 66: National Australia Bank channel volumes Figure 67: JPM consumer cash deposit behaviour

120% 111% 100% 3089% 90% 100% 90% 85%

80% 80% 74%

70% 60% 62% 60% 55% % change % 40% 51% 21% 50% 20% 5% 40% 0% -3% 30% -20% -9% -7% -16% -14% -19% 20% -40%

Branch counters ATM Telephone Internet Banking Mobile Internet tellers through deposits consumer of % 10% Banking Banking 0% 09-12 % change YoY % change 4Q07 4Q08 4Q09 4Q10 4Q11 4Q12

Source: Company data Source: JPMorganChase February 2013 Investor Day

As we show in our separate chapter on branches, we think this inertia will give way to a significant redesign in branch footprints, with significant declines in the number of bank branches in certain markets such as Italy and Spain in particular, given high starting branch numbers, low profitability and low growth. For other markets we expect more change in branch size, layout and staffing levels than in the absolute number of outlets.

3) Banks are focussed on staying brand-relevant to customers. In addition to seeking efficiency gains, we think mobile is a way for banks to defend Mobile offers an opportunity brand relevance and rebuild customer loyalty. As one would expect, surveys have to defend brand relevance banks as amongst the least popular entities with which most households transact (watch out for wallets) and (Figure 68) with developed market banks the hardest hit (Figure 69). rebuild trust

Page 40 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 68: Global trust scores by industry: banks are poor Figure 69: DM banks score far weaker than EM peers, given performance in the crisis

80% Financial Services 46% Banks 49% 69% 69% Media 50% 70% Chemicals 51% 58% Brewing, Spirits 54% 60% Consumer Health 55% 49% Energy 57% 50% Pharma 57% 44% Consumer Packaged Goods 60% 40% Telecomms 60% Food manufacturing 61% 28% Metals 61% 30% Entertainment 62% Aerospace & Defence 62% 20% Food & Beverage 62% Auto 66% 10% Consumer Electronics 70% Technology 73% 0% 40% 45% 50% 55% 60% 65% 70% 75% North America Latam BRIC APAC EU Global

Source: Edelman Trust Barometer; Note: Respondents score 1 for “do not trust at all” and 9 for “trust Source: Edelman Trust Barometer; Note: Respondents score 1 for “do not trust at all” and 9 for “trust them a great deal” to question “Please inducate ow much you trust businesses in each of the following them a great deal” to question “Please indicate how much you trust businesses in each of the following industries to do what is right” industries to do what is right”

But though banks report poor engagement scores as institutions, customer churn Banks almost exclusively still remains low relative to other industries. We think this is because: (i) customers enjoy customer confidence as generally think all banks are equally good / poor; (ii) there is no enduring edge in a safe place for savings; this product design; (iii) Customers are engaged with branch staff if not impressed with and client ambivalence banks as institutions; (iv) banks almost exclusively continue to have customer around product and confidence as a place of safe for their savings. If we see providers of payment services from other sectors (PayPal, Google) begin to achieve similar customer confidence, there institutional differentiation is are grounds for greater concern around future market share. keeping churn low

For now, however, we see mobile as a key strategy in increasing customer Digital gives banks a chance engagement, helped by the addition of other services. Most markets now have banks to offer ancillary services to giving customers the ability to pay friends direct by phone (Barclays call it Pingit, RBS bolster confidence, trust and Pay Your Contacts, Bank of Ireland Pay to Mobile and so on). As described in the Big improve perceptions around Data chapter we expect this to expand to ancillary services like vouchers and special value for money; in all offers which we think will add value for customers and reduce the relative attractiveness of digital wallets and such-like provided by non-banks. markets we studied, it’s free

4) An increasingly important channel for sales Mobile is growing in importance as an account opening and selling medium. Westpac Mobile has more importance reported that 5% of 1H13 account openings were from mobile, 7% of sales were from as an end-to-end self-service digital channels. In five years management expect this to reach 20%. For now, channel at present; sales and therefore, mobile is most important to improve customer service and as a means to account openings are sub promote self-service. 10% of total but growing

As sales via digital channels increase the materiality of associated conduct risk costs Conduct risk issues – a will also increase. Recent experience in jurisdictions like the UK with significant conduct greater focus for costs (cumulative Payment Protection Insurance redress costs to 1H13 rose to £15.4bn management now – can be / US$23bn) has sharpened management focus on product and document design and helped by digital channel selling practices. Some executives we met see online/digital as perhaps the only viable management… means of avoiding conduct risk borne of failure in sales procedure and water-tight documentation of customer acceptance of all terms of engagement.

Deutsche Bank AG/London Page 41

8 September 2013

Financial Services Retail Bank Strategy

Figure 70: PPI impairment charges (£’m) FY11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 Cumulative Barclays 1,000 300 0 700 600 0 1,350 3,950 HSBC 507 295 368 221 180 74 163 1,808 Lloyds Banking Group 3,200 375 700 1,000 1,500 0 500 7,275 RBS 1,065 125 135 400 450 0 185 2,360 Total 5,772 1,095 1,203 2,321 2,730 74 2,198 15,393 Source: Deutsche Bank estimates, Company data

We don’t think online selling provides cast-iron conduct-risk protection. Unless the UK's …but we think there’s some customer protection body, the FPC, and its international peers enter into product kite- distance for banks to travel to marking / auditing akin to that encountered in drug and food testing, we think banks ensure that digital doesn’t will remain open claims of poorly communicated risks associated with products. In raise different but equally online, we think some banks are avoiding implicating themselves in selling by failing to important conduct risks record and control customer behaviour around things like the time taken to read and accept T&C's. Banks aren't alone in this but it is probably too hopeful to assume that caveat emptor will again come to be relevant in bank product sales in the near term

5) Digital can help reduce operating costs An obvious point made above already but worthy of reiteration. Self-service cuts Recent moves by the banking rewaork and costs. In the US, technology is also reducing costs and inconvenience sector have tried to replace or around cheque usage. As a country still very dependent on cheques (see Payments retain the convenience / trust chapter) it is interesting to see banks allowing customers to use their smartphone / culture element of cheque banking apps to bank cheques. Chase’s QuickDeposit has a customer take a picture of usage, whilst reducing the front and back of cheque, the app transmits picture to the bank which clears it; no branch visit, no postage. The technology behind such applications are provided by firms costs e.g. Chase QuickDeposit like Jack Henry & Associates (Nasdaq: JKHY).

Currently this technology cannot be used in the UK due to legal issues (banks apparently are required to file each cheque). We think governments should look strategically to play an active role in reducing the frictional costs of transacting via cash and cheques as rapidly as possible for the benefit of all.

Page 42 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Mobile Payments: Already in a phone near you

Overview

Pay to mobile, pay to Facebook, pay to account, pay to email, all from a smartphone are Pay to mobile, pay to fast becoming ubiquitous alternatives to cheques and are more convenient and easier Facebook, pay to account, to use than internet banking. In many cases this is being driven by third party providers pay to email are fast like Monitise. Examples are Pingit by Barclays, Pay to mobile by Bank of Ireland, becoming ubiquitous kaching by CBA, nabKiss by NAB, Pay Your Contacts by RBS. The UK Payments alternatives to cheques; being Council is also in the process of developing an industry-wide central service which will make it possible to send/receive payments just using a mobile number, no matter which driven in many cases by third bank the customer is with. party providers like Monitise

Does the success of mobile payments in some countries presage significant future bank disintermediation? We don’t think so, after taking a long look at Kenya’s M-PESA, arguably the most successful mobile payments system in the world. We also look at the trend towards increasingly non-cash payments in countries such as Sweden, and the ongoing decline of cheques, though we think extinction remains more than a decade away.

M-PESA - Mobile payments (not banking) at its most successful

M-PESA25 is a small-value payment system launched by Vodafone's Kenyan affiliate Safaricom in 2007. Since then, it has grown beyond all early expectations and operates almost entirely independently of the banking system. M-PESA is an often cited case study in how banks could be disintermediated by non-financial companies in future.

Figure 71: Vodafone Snapshot Figure 72: Safaricom Snapshot Vodafone Analyst name Email Safaricom Spot price: 7.85 Tgt. Price 8.14 Group Plc David Wright [email protected] 2010 2011 2012 2013 2014e 2015e Spot price: 211 Tgt. Price 246 Rec. Buy EPS (adjusted) 0.38 0.33 0.32 0.44 0.52 0.58 2010 2011 2012 2013e 2014e 2015e DPS 0.20 0.20 0.22 0.31 0.40 0.47 EPS (DB adjusted) 16.1 16.7 9.2 11.4 12.2 13.5 BVPS 1.56 1.70 1.81 2.01 1.96 2.13 DPS 8.3 8.9 9.5 10.2 10.2 10.2 BVPS 171.6 169.8 155.0 145.3 149.9 156.5 ROE 27% 20% 18% 23% 25% 25% ROE 10% 9% 7% 8% 8% 9% EBITDA mgn 44% 38% 36% 39% 40% 40% EBITDA mgn 33% 32% 30% 29% 29% 29% Market Cap (US$'mn) 2,871 1,831 1,544 2,807 3,590 3,590 Market Cap (US$'mn) 86,582 107,936 121,128 161,905 161,905 161,905 P/E 14.5 11.4 10.1 13.7 15.0 13.1 P/E (DB) 6.4 7.9 16.2 18.4 17.3 15.6 EV/EBITDA 6.2 4.5 3.6 5.1 5.8 5.3 EV/EBITDA 0.8 1.6 2.1 4.8 4.8 4.5 Div yield 8% 7% 6% 5% 5% 5% Div yield 3.6% 5.3% 6.9% 5.2% 5.1% 6.2% FCF yield 13% 9% 12% 8% 9% 9% FCF yield 3.0% 3.6% 6.2% 5.9% 5.6% 6.7% ROE 10% 9% 7% 8% 8% 9% ROE 27% 20% 18% 23% 25% 25%

Source: Company data, Deutsche Bank estimates, Priced as on 5 September 2013, All figures in local Source: Deutsche Bank, Bloomberg Finance LP, Priced on 5 September 2013; Safaricom is not under currency unless stated Db equity research coverage. Statistics above, including target price, refer to Bloomberg consensus

25 m- for mobile and 'PESA' for money in Swahili

Deutsche Bank AG/London Page 43

8 September 2013

Financial Services Retail Bank Strategy

M-PESA reached critical mass – a key requirement for success in payments – very M-PESA reached critical quickly. Safaricom had an internal target for a million sign-ups in year 1 and achieved mass quickly, with 67% of two million. By year three, M-PESA had 9m customers and by March 2013 this figure pop. over 14 years of age had risen to 17m, 67% of the population over 14 years of age (Figure 73). Ten million of signed up; person to person these were active in the previous month. In the six months to March 2013, M-PESA payments of 30% of GDP now handled an annualised US$11.5bn in person to person payments, about 30% of GDP. This excludes the growing use of M-PESA for payments to and from corporates. 18% or processed this way US$255m in Safaricom revenue in FY13 came from M-PESA, up 30% YoY (Figure 74).

Figure 73: M-PESA customers (m) Figure 74: M-PESA revenues (US$’m)

18.0 17.1 300 20% 17.6% 18% 16.0 14.9 15.8% 13.8 250 14.0 16% 12.4% 14% 12.0 200 9.5 12% 10.0 9.0% 150 10% 8.0 6.2 255.4 8% 6.0 100 203.1 6% 4.0 141.9 4% 2.1 50 97.8 2.0 2% 0.0 0 0% 2008 2009 2010 2011 2012 2013 FY10 FY11 FY12 FY13

M-PESA customers (million) M-PESA revenue (US$m) M-PESA rev. as % of total revenue (RHS)

Source: Deutsche Bank, Safaricom FY13 results presentation Source: Deutsche Bank, Safaricom FY13 results presentation

How M-PESA works M-PESA began as a remittance product marketed as a means for urban dwellers to Customers open and operate send money home. The customer registers an M-PESA account at one of 65,547 M-PESA accounts via a licenced agents (retailers, some bank branches, figure as at March 2013), a distribution network of 65k agents which network which vastly outnumbers that of the banks. Uniform M-PESA branding of all include retailers, ATMs and agents – outsourced by Safaricom – helps create recognisability across the network, as some bank branches well as reinforcing confidence in the safety of customer money. The M-PESA account is linked to the customer's mobile number and accessed via a SIM-card resident application. Funds are deposited and withdrawn via M-PESA agents. Individual agents are organised under Agent Head Offices which manage liquidity and communicate with Safaricom in exchange for a ~30% share of the commission pool.

Figure 75: 30% of GDP in P2P payments in 2H13

Kshs 522bn* payments transacted between customers within M-PESA

Kshs 444bn* Kshs 390bn* deposited into withdrawn from M-PESA via M-PESA via agents agents

* Oct. 2012-Mar.2013

Source: Safaricom FY13 results presentation

Page 44 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Registration with M-PESA and deposits of funds are free. Once deposited, customers Registration and deposits are make real time payments to other individuals (P2P) and companies (e.g. utility free, with transfers and providers), subject to a transaction size limit. P2P and cash withdrawals attract fees on withdrawals attracting a sliding scale at several hundred basis points for small transfers, falling to 20bps for sliding-scale fees which fall to larger ones (Figure 76). Customers do not earn interest on M-PESA balances and 20bps for larger transactions Safaricom’s yield on these funds – which are on-deposited with banks – accrues to a non-profit trust as agreed with the Kenyan Central Bank ahead of M-PESA’s launch.

Figure 76: M-PESA tariff sheet26 (excerpt, converted to US$ at 0.011/Ksh) Transaction Range (US$) Charges in US$ Charges as % of max tn Min Max Transfer Transfer to Withdrawal Transfer to Transfer to Withdrawa to other Unregister from M- other M- Unregister l from M- M-PESA ed Users PESA PESA ed Users PESA Users agent Users agent 10 49 0.03 N/A N/A 6.1% n.a. n.a. 50 100 0.06 N/A 0.11 5.0% n.a. 10% 101 500 0.31 0.75 0.31 5.4% 13% 5% 501 1,000 0.38 0.75 0.31 3.3% 7% 3% 1,001 1,500 0.38 0.75 0.31 2.2% 4% 2% 1,501 2,500 0.38 0.75 0.31 1.3% 3% 1% 2,501 3,500 0.38 1.00 0.56 0.9% 3% 1% 3,501 5,000 0.38 1.20 0.75 0.7% 2% 1% 5,001 7,500 0.63 1.63 0.94 0.7% 2% 1% 7,501 10,000 0.63 1.95 1.25 0.6% 2% 1% 10,001 15,000 0.63 2.51 1.81 0.4% 1% 1% 15,001 20,000 0.63 2.70 2.01 0.3% 1% 1% 30,001 35,000 0.94 3.14 2.13 0.2% 1% 1% 35,001 40,000 0.94 N/A 3.14 0.2% n.a. 1% 40,001 45,000 0.94 N/A 3.14 0.2% n.a. 1% 45,001 50,000 1.25 N/A 3.14 0.2% n.a. 1% 50,001 70,000 1.25 N/A 3.76 0.2% n.a. 0% Source: Deutsche Bank, Safaricom website

M-PESA clearly helped to meet an established customer need in a lower risk, more M-PESA ARPU is a third of convenient and potentially more cost effective manner (many remittances were that of Voice, almost 3x SMS, previously made via taxi companies). Safaricom’s M-PESA average revenue per user contributes to lowering churn, (ARPU) is a third of that of Voice and almost 3x that of SMS. Importantly, and in line supporting overall product with our greatest hopes for digital in banking, it contributes towards the maintenance pricing and customer of market share, reduction of churn and premium pricing of the overall Safaricom mobile offering relative to peers. recruitment

26 http://www.safaricom.co.ke/personal/m-pesa/m-pesa-services-tariffs/tariffs

Deutsche Bank AG/London Page 45

8 September 2013

Financial Services Retail Bank Strategy

Figure 77: ARPU growth across service lines

US$ 7.00

FY12 FY13 5.96 6.00 5.20 5.00

3.96 4.00 3.59

3.00

2.00 1.37 1.08 1.15 1.06 1.00 0.40 0.52

0.00 Voice SMS M-PESA Mobile Broadband Service ARPU

Source: Deutsche Bank, Safaricom FY13 results presentation

M-PESA has grown as a broader channel for digital commerce The significant customer coverage achieved by the system has made it feasible to act Critical mass in customer as a channel for corporates to make and receive payments. Companies make salary and coverage makes M-PESA dividend payments via the system, reducing the need for cash disbursements for field viable as a means for staff, for example. Hundreds of businesses are signed up to receive payments from companies to pay staff and be customers via M-PESA, including utility firms like Kenya Power, insurance companies paid by customers and so on. 32% of Safaricom airtime top ups were bought via M-PESA in 2013.

What M-PESA says about the future of retail banking

What does the success of M-PESA say about the ability of banks to maintain current Part of the success of M- dominance in most payments systems? Part of this is down to the coincidence of PESA is down to a market factors listed below that supported M-PESA which we’ve not seen much coincidence of conditions elsewhere. Other markets such as the Philippines, for example, were earlier to launch relatively specific to Kenya mobile money (Globe Telecom and SMART started SmartMoney and G-Cash in 2000) and some EM peers and have seen more limited penetration.

„ Significant demand for remittance payments (half of households receive payments of this kind), reinforced by urbanisation and a cultural desire to retain strong links with ancestral homes. Most remittances are domestic, limiting the need for cross-border system capabilities.

„ A relatively under-banked population served by a branch network that is sparse enough to make electronic payments very attractive but large enough to allow M-PESA agents to manage their cash positions effectively (Figure 78).

„ Safaricom had a dominant market share, which supports achieving scale more rapidly in a new endeavour (Figure 79).

„ A supportive central bank and regulator.

Page 46 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 78: Customers per branch (1000’s, 2012) Figure 79: Telco market share, Kenya, December 2012

40.0 37.1 Yu Essar 35.0 11%

30.0 Orange 8% 25.0 20.1 20.0

15.0 13.8 Airtel 10.0 17% 5.4 Population per branch (000's, 2012) branchper (000's, Population 5.0 3.2 Safaricom 64% 0.0 Kenya Philippines South Africa United United States Kingdom

Source: Deutsche Bank, IMF, World Bank, FDIC for US Source: Deutsche Bank, Safaricom FY13 results presentation

We don’t dismiss M-PESA’s progress as an EM- or Kenyan-only case, however. We Don’t dismiss M-PESA’s think there are a number of broader lessons which can be drawn from this case study. success as a Kenyan-only case. There are broader „ Customers will pay for a value added service. An obvious point but one worth lessons to be learned revisiting in markets such as the UK and US where the imposition of new fees is anathema to many.

„ In a limited number of situations, new services give service providers an ability to charge more, acquire new customers and reduce back book churn. Safaricom reports higher customer pricing and lower churn as a consequence of M-PESA’s coverage and success. It is noteworthy that the P2P fee structure imposes a fee up to 5x higher on the sender of funds when monies go to a non- Safaricom customer, recognising the leverage the sender has with the recipient to prompt a change in telco.

„ M-PESA’s success isn’t just about the unbanked: product and convenience can drive share gains. Survey evidence27 confirmed that M-PESA early adopters were twice as likely to have a bank account (72% versus 36%), better educated and more affluent. Differentiation in customer convenience and service played a meaningful role in achieving critical mass which now serves as a decent barrier to entry for non-banks at least, in our view. Ease and cost of use has seen the proportion of Kenyan households receiving remittance money increase to 52% from 17% before M-PESA was introduced.

„ Sensible product design, marketing and technology can build customer trust in new financial ventures. In future banks will lose their exclusive claim as the only safe custodian of customer funds. Safaricom supported M-PESA take up via (i) heavy marketing, (ii) transparent pricing left unchanged for at least three years post launch; (iii) No agent-level charges, largely eliminating potential for on-the-ground co-ercion of customers for agent financial gain; (iv) free SMS- driven transaction confirmations to reinforce trust and error correction. Whereas M-PESA customer balances are very low, effectively acting as a pre- funded Paypal, trust levels and regulation have now advanced to a level where Safaricom have launched an interest-bearing savings and loan product suite called M-Shwari. We watch with interest.

27 Mobile Payments go Viral: M-PESA in Kenya, Ignacio Mas, Dan Radcliffe, Bill & Melinda Gates Foundation, March 2010

Deutsche Bank AG/London Page 47

8 September 2013

Financial Services Retail Bank Strategy

„ M-PESA shows the need for banks in general (and new banks in particular) to consider the economic merits of looking at customer's ability to pay based on transaction volumes and profile rather than by size of cash float or loans taken. The bottom line We think digital payment tools like the ones offered by many banks and by Safaricom Banks can offer mobile are another place where the consumer is the largest beneficiary and ubiquity of payments as well as anyone provision means the product becomes a hygiene factor in a short space of time: not a else; the product has become reason to move to a bank, but perhaps a big deal for some if not available. Though we commoditised very quickly remain alert for an un-copyable killer app in banking, we don’t expect to see one and and doesn’t act as a basis for we don’t think non-bank mobile payments is a huge market in most countries. market share shifts in most countries Moving to a less cash & cheque dominated world

Cash: higher in volume, but lower in value Physical cash (banknotes and coins) in circulation within economies has risen over the Most countries are seeing real past 4 years (Figure 80), particularly in Turkey and India even after adjusting for inflation growth in cash in circulation (Figure 81). Others, such as Sweden (which has moved away from using cash).

Figure 80: Change in cash in circulation since 2007 Figure 81: Change in cash, adjusted for inflation, 2007-11

100% Turkey, 99% 80% 73% 62% 55%58% India, 81% 60% 53% 44%45%45%48% 40% HK, 58% 24% 28% 50% 19%20%22% 20% 3% USA, 30% 0% UK, 22% circulation -20% 0% -12% -23% Sweden, -13% -40% South Africa, - % change in banknotes / coins in in / coins banknotes in change % 30% -50% 2007 2008 2009 2010 2011

Source: BIS, Deutsche Bank estimates. Local currency, not adjusted for inflation Source: BIS, Deutsche Bank estimates. Local currency, adjusted for CPI inflation

Cash as % of GDP increased from 8% in 2007 to 8.4% in 2011 for the CPSS countries. Cash volumes are easier to track than cash usage in many markets. We have a decent view of trends for the UK using statistics published by the Payments Council:

„ By volume, the proportion of transactions using cash in shops fell from 75% in 2001 In UK retail cash is 56% of to 56% in 2011. For non-regular or ‘spontaneous’ payments, cash still makes up 3 volume (75% in 01) and 30% out of 5 payments in 2011. of value (45% in 01) „ By value, the proportion of retail spend in cash has fallen from 45% to 30% over 2001-2011. For non-regular / spontaneous payments, cash may be dominant, but 44% of these payments of these are between £1 and £5. So the trend appears clear: cash still makes up a large portion of volumes, but in terms of values it is becoming increasingly less relevant.

Page 48 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 82: Personal non-regular payment volumes in Figure 83: Retail spending by value over time 2011 in the UK 35 100% 90% 30 80%

25 70%

60%

Volumes billions Volumes 20 50%

15 40%

30% 10 20%

5 10%

0% 0 2001 2011

Cash Debit card Credit card Cheque Other Remote Banking Cash Cheque Debit card Credit card / Charge card Other

Source: UK Payments Council, ‘The Way We Pay’ 2013 Source: UK Payments Council, ‘The Way We Pay’ 2013

Governments have an incentive to shift payments away from cash Governments have a real role „ To reduce tax evasion: One of the CIOs we spoke to suggested that one way to and incentive in facilitating a curtail tax evasion severely would be to shift the Eurozone to a cashless economy. shift to a more efficient Though somewhat tongue-in-cheek, this is a big picture reminder that (i) the technology is there and (ii) governments have a real role and incentive to facilitate a digital/electronic economy shift to a more efficient digital/electronic economy.

„ Higher GDP cost: The Swedish KTH Royal Institute of technology estimates a domestic cost of cash of 0.26% of GDP, compared with credit cards at 0.19% and debit cards at 0.09%.28 For many countries, cheques are a good example of an area where government could and should take action.

„ Sweden: an example of an increasingly cashless society: Of all the countries in 65-75% of Swedbank and Europe, Sweden has probably progressed furthest in removing cash as the main Nordea branches don’t do medium of transacting. Here bills and coins represent just 3% of the Swedish economy, and “a small but growing number of businesses only take cards, and manual cash handling; Some some bank offices...have stopped handling cash altogether”.29 retailers are phasing out the acceptance of cash in stores 65-75% of Swedbank and Nordea branches in Sweden are ceasing manual cash handling30. Nordea reports a 20% annual decline in customer demand for cash from branches. Retailers including Kunsaengen, 3 and TeliaSonera have or are phasing out the acceptance of cash in stores31. We think businesses will increasingly move to non-cash models to cut costs. „ Transport for London (TFL): the Oyster Card has been a form of electronic ticketing We expect business will used on public transport in London since July 2003. A contactless RFID (Radio- increasingly move to non- Frequency Identification) chip within the card enables customers to travel on buses / trains / London underground services. In 2007 TfL and Barclaycard launched a cash models to cut costs; the combined Oyster, contactless payment, credit card, called OnePulse (Figure 84). savings here are non-trivial From December 2012 all buses now accept contactless debit/credit payment cards, with 25 million journeys expected to be made via this payment method in 2013. By the end of 2013 customers will be able to use contactless payments cards for all TfL journeys – making London the first major urban transport system in the world to accept contactless.

28 Mastercard Worldwide, Payments Perspectives Blog: Sweden: A cash desert? Or an electronic Oasis? 29 http://www.cbsnews.com/8301-202_162-57399610/sweden-moving-towards-cashless-economy/ 30 Mastercard Worldwide, Payments Perspectives Blog: Sweden: A cash desert? Or an electronic Oasis? 31 Bloomberg, 11 April 2013, Swedish banks make money by saying No to money

Deutsche Bank AG/London Page 49

8 September 2013

Financial Services Retail Bank Strategy

Figure 84: Oyster Cards in action Figure 85: Barclaycard OnePulse

Source: Transport for London Source: Barclaycard

Since the introduction of the Oyster the use of cash to purchase fares on the TFL 25% of London bus fares network has fallen significantly. 25% of bus fares were paid in cash in 2000, today were paid in cash in 2000, this has fallen to less than 1%, and TfL is now consulting on phasing out cash fare payments on London buses at the end of 2014. TfL thinks it will save £24m a year less than 1% today; TfL by going cashless, that it would speed up service, and be cheaper for customers.32 estimates could save £24m a year by going cashless Cash won’t disappear altogether of course Near term, though we think the use of cash will dwindle, the advantages it gives to consumers (through convenience, anonymity, security of finality of transaction) and the important role it plays for some sectors (low income households, charities) should clearly guarantee an ongoing place for cash in most areas of commerce.

Cheques: On the way out... Cheques have been in decline for many years. Data from the UK shows that volumes Cheques have been in decline have dwindled from over 200 million a month back in 1990, down to under 50 million a for many years month in 2013 (Figure 86). Transaction values globally have also slipped at a similar, though slightly slower pace (transactions down c.6-7% per year, Figure 87).

Figure 86: Cheques down 24-28% over 4 years globally... Figure 87: ...and have fallen 80% off peak volumes

45.0 80.0 300,000 140,000 40.0 70.0 250,000 120,000 35.0 60.0 100,000 200,000 30.0 50.0 80,000 25.0 150,000 40.0 60,000 20.0 trillion) 100,000 Values (£m) 30.0 Voumes (000s) 40,000 15.0 Cheque volumes (billions) 50,000 10.0 20.0 20,000

5.0 10.0 values (USD trnasaction Cheque 0 0 0.0 0.0 Jul-99 Jan-90 Jan-09 Jun-07 Oct-94 Apr-04 Feb-01 Sep-02 Dec-97 Aug-91 Aug-10 Nov-05 Mar-93 Mar-12

2007 2008 2009 2010 2011 May-96

Cheque transactions globally Cheque values globally Volumes (000s) Values (£'m)

Source: BIS data, Deutsche Bank analysis Source: Deutsche Bank estimates, UK Payments Council

Much of this decline is due to the inefficiency of clearing cheques: they take several days to clear compared with next day / near-instantaneous electronic transfers. Direct

32 http://www.tfl.gov.uk/corporate/media/newscentre/28413.aspx

Page 50 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

debits are generally more convenient and efficient ways to pay bills for both the sender and the receiver.

Though the elimination of the cheque looks inevitable in the long run driven by The elimination of the cheque technological change, overall extinction is likely more than a decade away. Cheques is more than a decade away – have been around for over 350 years and in some countries remain widely used. Almost with economies such as the 20% of US non-cash payments were made via cheque in 2011, and in China, Mexico, US still remarkably keen on France and India cheques represent over 10% of non-cash transactions (Figure 89). them 3 reasons for the longevity of cheques „ Cheques do have advantages: Sometimes being really efficient isn’t what consumers want: for payers the processing time (typically 4-6 days) can be beneficial as it means retaining the funds for longer; there are no real limits on size (cheques still have the largest average transaction values, Figure 88); you only need to have the recipient’s name (rather than bank details) to make payment. Cheques are still readily used as a way of gifting money to others/charities in a safe fashion, though increasingly utility companies incentivise customers to pay via direct debit.

Figure 88: Transaction values by non-cash payment Figure 89: Global non-cash payment usages, 2011

1,400 100% 90% 1,200 80% 1,000 70% 60% 800 50% 600 40% 30% 400 20% 200 10% 0% 0 Italy India Brazil China Korea Russia France Mexico Canada Sweden Belgium Jul-93 Jul-00 Jul-07 Australia Jan-90 Jan-97 Jan-04 Jan-11 Germany Sep-94 Sep-01 Sep-08 Nov-95 Nov-02 Nov-09 Mar-91 Mar-98 Mar-05 Mar-12 May-92 May-99 May-06 May-13 Switzerland South Africa Netherlands BACS Faster Payments Saudi Arabia United States United Global average

Cheques & Credit Credit cards United Kingdom Debit cards Cheques Transfers Direct debits Credit cards Other

Source: Deutsche Bank estimates, UK Payments Council Source: Deutsche Bank analysis, BIS data

„ Cheques are still used for some standardised payments: It is also worth noting that Cheques are still used for in the UK, for example, some government welfare payments are still paid by cheque. some standardized payments, 20% of all regular cheque payments by consumers in 2011 were state payments for childcare. This will change we think. and many consumers still place greater faith or „ Convenience / Trust / Culture: Some consumers place greater faith or convenience convenience in them in a payment mechanism which has worked for centuries (requiring just a cheque book, a name, and a pen) over relatively new electronic payments (requiring logins, account numbers, an internet connection etc). For others managing their payments via cheques allows them to track their financial position. We expect there is a generational trend. These consumers have in the past been quite vocal, for example the announcement from the UK Payments Council that in 2009 that cheques would be phased out in 2018 was eventually aborted following criticism from consumer groups and politicians. For example at the time, Mark Hunter, a Liberal Democrat MP, said: “This is a scandalous, self-serving decision that puts the whims of City fat cats ahead of the needs of this country’s most vulnerable people...Many elderly, housebound and disabled people rely on cheques for payment and will be massively inconvenienced...Scrooge-like bankers had the chance to start putting customers first once again, but they have well and truly failed.”33

33 Daily Telegraph, ‘Death of the cheque book announced’, 16 December 2009.

Deutsche Bank AG/London Page 51

8 September 2013

Financial Services Retail Bank Strategy

Growth in digital wallets

Overview

Digital and mobile wallets are two of the more interesting unfolding payment Digital and mobile wallets are innovations. At its simplest a digital wallet is a PayPal-like gateway tool to which two of the more interesting consumers link debit and credit cards in an account which can be used to transact unfolding payment online. A mobile wallet is much the same but is held on a smartphone and increasingly innovations offering benefits including loyalty schemes, retail offers, storage of airline and concert tickets. Wallets can be multi function (Google Wallet, Apple’s Passbook) and single function (McDonalds’ GoMcDo, Starbucks). The key differentiating factor around mobile wallets we think is its increasing functionality in real-world retail in payments and special offers in particular.

The opportunity for wallet providers is straightforward. Increased share of customer The opportunity for wallet spend, greater loyalty and a transaction data set that can inform Big Data propositions. providers is increased share The risks to banks – the owners of the dominant payments value chain in most places – of customer spend, greater is a loss of share in payments (disintermediation – serious) and loss of brand standing loyalty and Big Data; the risk with customers (bank cards “disappear” into e-wallets and are not seen again, with for banks is disintermediation phones deciding which card to use in which store – less serious).

Given the proliferation of wallets coming to market, this is a threat that banks must take We think banks need to take seriously. Overall, however, we expect banks will maintain most of the economics the threat head on by around payments that they enjoy today. The logic here is the same as that expressed in providing better product, relation to mobile payments above. Banks already have scale in storage and transfer of powered by the data customer funds, they have customer trust and expertise in cyber security, in advantage they hold convenience of transacting (contactless or otherwise) they will match other products which come to market and, as we discuss in greater detail in the Big Data chapter, they have a largely unused but vast customer data advantage to deploy here.

In short, we expect customer smartphones to hold numerous wallets that make it easier The customer wins (again) for customers to transact with their favourite coffee shop, airline, cinema chain and and banks should retain much bank. But if no killer version of the wallet emerges – which would have to be terrific as of their market share, with well as impossible to copy – we think banks will retain their role in payments. Though some upside from efficiency we remain vigilant of this fast moving space. gains Card model is moving towards a digital wallet, transaction focus

The early card model focused on plastic as an extension of consumer credit, offering Card as a payment tool convenience for the customer and retailer, and often incentives around loyalty schemes married with smartphone and so on. Since then the model has moved towards a transaction-driven model in line with the strong growth in debit card transactions and declining prevalence of cash and technology promises to cheques within the payments mix (described in the Payments chapter). In the US in significantly change the way particular we read a great deal about innovations in a number of areas, including: that companies market and customers transact „ Digital Wallets: Consumers link debit / credit cards into an online account or identity.

„ Mobile wallets: Mobile versions of digital wallets offering convenience and security of online digital wallets on smartphone. Customers manage debit / credit cards on their phone and increasingly make payments directly using their phones.

„ Mobile-Point of Sale (mPOS): The shift by retailers (particularly in the US) towards using mobile / software Point of Sale systems rather than physical hardware (like a

Page 52 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Verifone/Ingenico terminal). Pioneered by Apple; several retailers, grocery stores, and restaurants are following suit.

„ Big Data: Uses analysis of customer/merchant transaction data to provide targeted offers for customers or marketing campaigns for retailers, amongst other things (see Big Data chapter).

„ Geo-fencing: Virtual perimeter created in real-world means GPS-enabled smartphones can house apps which use the customer’s location for providing targeted offers / vouchers for particular retailers. We expect banks to capitalise on their customer relationships in this area in particular.

„ Near-Field Communication (NFC) technology: Broad term covering contactless technology increasingly seen on cards, phones and PoS terminals. Based on industry standards using RFID chips already present in many smartcards.

„ Contactless cards: Allows payments below £20 / E20 without PINs or signatures. Convenience obvious to any who have used it whilst benefits of adoption to high volume retailers are probably under-appreciated. As with most other things involving money, however, it will take time to increase consumer confidence in the security of transacting on a contactless basis. Survey evidence published by ING shows that 45% of respondents in a June 2013 poll disagreed with the statement “I would feel confident that my money is secure if I used contactless payments”, with a surprising divergence in confidence levels by country. It is interesting that confidence in contactless bears little relationship to mobile banking penetration rates: concerns in this area aren’t just about tech-savvyness (Figure 90). As expected younger customers are more confident – but there is far less sensitivity in the trust metrics to age than we expected (Figure 91).

Figure 90: Poor confidence isn’t just tech savvyness… Figure 91: …and surprisingly low leverage to age

65% 65%

55% 55%

45% 45%

35% 35%

25% 25% % of respondents % of respondents 15% 15%

5% 5%

-5% -5% Europe 25-34 years 35-44 years 45-54 years 55+ years

% doubting money security in Contactless % not using mobile banking % doubting money security in Contactless % not using mobile

Source: ING International Survey, July 2013; Financial Empowerment in the Digital Age Source: ING International Survey, July 2013; Financial Empowerment in the Digital Age

Nevertheless, there has been a steady rise in the proportion of contactless cards in Roll-out of contactless has the UK (Figure 92), with new cards issued typically featuring the technology and been more extensive in the entering circulation as old cards expire. Similarly there has been an extensive roll- out of contactless terminals, with 147,000 in operation the UK now. Adoption in the UK, for example, than in the US has been slower, primarily due to prevalence of mag-swipe/signature payments US which tends to use mag- at PoS terminals which is already a relatively efficient process when compared with swipe/signature PoS chip and pin used on European cards. We expect contactless to continue to grow in terminals Europe in particular.

Deutsche Bank AG/London Page 53

8 September 2013

Financial Services Retail Bank Strategy

Figure 92: 22% of UK cards now contactless Figure 93: Contactless terminals and cards on the rise

25% 1.20% 160,000 35

1.00% 140,000 30 20% 120,000 0.80% 25 15% 100,000 0.60% 20 80,000 10% 0.40% 15 60,000 0.20% 10 40,000 5% rate growth Monthly 0.00% Number of contactlessterminals 20,000 5 Number of contactlesscards (million) % of cards which are contactless are which of cards % 0% -0.20% Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 0 0 Sep-10 Sep-11 Sep-12 Jul-10 Nov-10 Jan-11 May-11 Jul-11 Nov-11 Jan-12 Ma Jul-12 Nov-12 Jan-13 Mar-11 Mar-12 Mar-13 y -12

% of cards contactless Monthly growth rate Number of contactless terminals Number of contactless cards

Source: UK Cards association Source: UK Cards association

Digital/mobile wallets aren’t just about cards Cards have seen a shift towards digital and mobile wallets. But other forms of banking are going mobile including electronic cheque cashing and bank account management. The convergence of card payments, big data, NFC technology and smartphones has seen retailers, mobilePoS, internet aggregators, and mobile phone operators all enter the wallet business. We show this convergence in Figure 94.

Figure 94: Convergence on the digital and mobile wallet

Digital wallet Account Traditionalservices bank Card issuer management

Mobile wallet Merchant Cash Acquirer payments Credit card stakeholders card Credit Payment Cheques Networks

Mobile Mobile Internet Retailers phone aggregators PoS operators

New entrants

Source: Deutsche Bank

Page 54 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

The key question: Is there a killer wallet out there?

It seems clear that increased adoption of non-cash and digital transaction channels will We expect payment networks continue to fuel payments volume growth. This should be good for payments networks to continue to see further and large card issuers who enjoy operating leverage to revenue growth and pre-existing volume growth; as yet it scale for newer businesses like digital marketing services. It is notable that many of remains unclear whether and new entrants (such as Square) operate using incumbent payments networks. which wallet / brand / The important question is whether an iTunes (music), Amazon (books) or eBay gateway app will capture (auctions) of e-wallets will arise to shift market share and competitive advantage. We meaningful market share in can’t yet see a situation where a wallet / similar device emerges which has features that digital payments banks can’t replicate or which stops users acting in a multi-wallet fashion. In fact, all stakeholders in mobile wallets (telcos, banks, payment networks) are wary of ceding control to one tech company given the emergence of platform dominance in other online markets. This should mean that a single killer-app or all-encompassing wallet is unlikely to emerge to the exclusion of others (and those provided by banks).

If this view is correct – our base case – banks which give payments innovation due consideration and investment will maintain their share of the payments industry, consumers will see greater convenience in transacting, and wallet functions (like mobile payments) will soon be something we all take for granted, a commoditised component of the retail and banking experience. In this context, it may well be better to own the provider of wallet technology than the buyer: the conclusion we also came to in respect of core IT.

We look at the opportunities, risks and produce offerings for some of the key players in the wallet market below.

Banks „ Opportunity: Provide better payment experience for customer, capture greater share of purchase wallet, accumulate more customer transaction data, achieve better customer retention and risk pricing, reduce branch operating. We show Barclays mobile payment solution (Pingit, in Figure 95)

„ Risk: If a 3rd party mobile wallet takes share, the banking relationship could become just one of several payment methods in the wallet – depriving the bank of brand promotion, potential sales and potentially customer data. In a world where customer churn is fairly low (e.g. in the UK), mobile wallets which enable customers to easily switch banking services without any change to gateway for interaction (e.g. like a Sim Card) presents a market share risk.

Deutsche Bank AG/London Page 55

8 September 2013

Financial Services Retail Bank Strategy

Figure 95: Barclays Pingit

Source: Company website

The Retailers Opportunity: Increased sales volumes, faster queuing times, lower payment costs, benefit of targeted offers to consumers, customer data, better customer service, extension of brand awareness.

„ Risk: That retailers lose brand identity, margin or control of the payments process to a 3rd party gateway (e.g. Google Wallet) which becomes dominant (like Apple with Music or Amazon with Books) leading to reduction in revenues.

„ Starbucks has launched a mobile wallet which allows customers to pay for coffees and track rewards/vouchers. Seen as the ‘role model’ for retailer mobile payments success has 3m transactions a week; partnered with Square and Paypal.

„ MCX leading US retail companies formed a joint venture in 2012 called ‘Merchant Customer Exchange’ (MCX) to design “a customer focused, versatile and seamlessly integrated mobile-commerce platform”. Collectively the merchants “already serve nearly every smartphone-enabled American on a weekly basis”34 MCX hope to lower funding costs by connecting directly to consumer banking accounts and increasing competition between banks.

„ McDonalds GoMcDo mobile checkout app. Allows users to order and pay on the phone and pick up without standing in the checkout line. Launched as a pilot in 30 restaurants in France, now rolling out to another 1,200. Card issuers / merchant acquirers „ Opportunity: Capture and leverage customer data to sell targeted merchant offers, generating commissions and additional volume, greater transactional data for risk- management and pricing.

„ Risk: Disintermediation, market share losses.

„ Barclaycard has c.25% market share of credit cards in UK and a large proportion of merchant acquirer PoS terminals. This has led it to launch ‘bespokeoffers (https://www.bespokeoffers.co.uk/). See Big Data chapter for product profile.

34 http://www.mcx.com/ Participating merchants include: 7-Eleven, Inc.; 76; Alon Brands; Bed Bath & Beyond Inc.; Best Buy Co., Inc.; Brinker International, Inc.; Circle K; Conoco; CVS/pharmacy; Darden Restaurants; DICK’s Sporting Goods; Dillard’s, Inc.; Dunkin’ Brands; Gap Inc.; HMSHost; Hobby Lobby Stores, Inc.; Hy-Vee, Inc.; Kohl’s Department Stores; Lowe’s; Meijer; Michaels Stores, Inc.; Pacific Convenience & Fuels LLC; Phillips 66; Publix Super Markets, Inc.; QuikTrip Corporation; RaceTrac; Sears Holdings; Sheetz, Inc.; Shell Oil Products U.S.; Southwest Airlines; Sunoco, Inc.; Target Corp.; Wakefern Food Corp.; Wal-Mart Stores, Inc.; and Wawa

Page 56 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

„ Mobile banking apps: many banks have launched mobile banking apps (or used 3rd party solutions like Monitise) which allow customers to monitor spending etc. Payments networks „ Opportunity: Be the gateway for all payments, online and offline, driving volume growth and fees. Greater transaction data; partnering with other stakeholders to take share.

„ Risk: Market share losses from competing offerings.

„ Visa: V.me is a cloud-based digital wallet which lets a customer to use any card (i.e. not just Visa) online under one username and password. Currently online, but plans to use NFC/contactless too. Visa expanded its ‘Visa Offers’ programme which allows issuers and merchants to deliver customized and differentiated real-time offers to Visa cardholders at point-of-sale.

„ MasterCard PayPass & MasterPass. PayPass digit wallet services were announced in May 2012 primarily as a digital wallet. MasterPass, announced in 2013, builds on this and allows the digital wallet to be used online and for retail purchases at checkouts. Offers checkout services, connected wallets (white-label solution for banks/merchants/partners to offer wallets, consumers can use any cards) and value added services (loyalty programmes / offers). Live in the US, Canada, Australia, UK, Spain, Sweden, Singapore, Israel, Italy. Mobile phone networks / manufacturers „ Opportunity: We expect M-PESA-like launches for most mobile operators, generating fees for payments processed from customers and retailers.

„ Risk: That another mobile wallet (e.g. Google) becomes dominant, that NFC is not the preferred technology or (mainly) that customers stick with the vast array of working alternatives already offered.

„ Isis: JV of the top 3 US mobile operators (Verizon, AT&T and T-Mobile) which uses a SIM-card based NFC payments solution. This launched on 22 October 2012 after delays, has suffered performance issues. Wallet resides on the phone rather than in the cloud. Isis has also launched an application which allows consumers to pay/redeem coupons, points and offers (merchants charged).

„ Vodafone UK, Everything Everywhere, and O2 UK: Has won approval to form a mobile payments JV in the UK.

„ Telefonica/O2 UK: O2 UK have launched an O2 wallet, and a virtual O2 Money Visa account card which allows consumers to send/receive money, compare prices, receive offers, top-up phones.

Deutsche Bank AG/London Page 57

8 September 2013

Financial Services Retail Bank Strategy

Figure 96: O2 Wallet app

Source: O2

„ Everything Everywhere: Partnered with Mastercard to offer a prepaid account solution on mobile phones for contactless NFC payments, money transfers and loyalty rewards. The builds on an earlier ‘Quick tap’ version which saw Orange (part of EE), Barclaycard and EE collaborate to create a smartphone-enabled NFC payment app. One issue we see with this is the requirement to ‘top-up’ a separate account for using NFC payments – another account isn’t convenient for consumers.

„ Apple Passbook / Samsung Wallet: allows user to store tickets / membership cards / flight boarding basses / vouchers on the handset. Apple iPhone 5 notably lacks NFC. Currently uses QR-codes rather than NFC, and does not support payments. But our US tech analysts think this is a stepping stone towards mobile payments. Internet aggregation / technology companies „ Opportunity: Online retail spending typically results in a card transaction. Becoming the dominant portal for these card transactions is potentially lucrative (fees for payments and for push advertising).

„ Risk: Someone else gets there first or the offering remains too weak to drive material customer take-up, or phone companies do not allow the digital wallet apps on handsets.

„ Paypal was the most prominent e-commerce digital wallet during the 2000s, popularized via eBay (which now owns PayPal) and other online retail companies. The consumer benefits from protecting their card details and from convenience: just having to enter a username and password (rather than lengthy card numbers, addresses, authentication etc). The ‘aggregator’ Paypal then charges a fee to the person receiving money / the retailer.

„ Google Wallet: Google’s first wallet was launched in May 2011 and based around the phone, facing resistance from mobile network operators, a lack of bank partnerships and could not be used for online purchases. Relaunched in August 2012 as ‘Google Wallet’ (replacing Google Checkout as well) as a cloud-based solution which lets cardholders load cards for use online and in stores through NFC. Google still faces resistance from card issuers who fear brand dilution / disintermediation risk (and it currently blocks cardholder data from being transmitted to issuers which could limit card issuers plans to leverage transaction data); as well as from mobile operators who want control of the so-called ‘Secure Element’ on SIM cards for the use of their Isis platform and have blocked Google.

Page 58 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Mobile Point of Sale / Payments Service Providers Please see the work by our US Technology research colleagues lead by Bryan Keane who have written extensively on the Point of Sale and Mobile Point of Sale market.

„ Opportunity: create a mobile Point of Sale for merchants that is cheaper to operate, offers a better customer experience; reporting, inventory and sales analytics for retailers; and other value-added services like merchant marketing and loyalty systems. Has the potential to disintermediate some of the traditional Point-of-sale providers and encroach on merchant acquirer business.

„ Risk: That traditional PoS providers shift to cloud, margins and value proposition are not sustained, margin pressured by payment networks / card issuers.

„ History: this has been particularly popular in the US, where there has been a move towards using mobile / software Point of Sale systems rather than physical hardware (like a Verifone/Ingenico terminal). Apple pioneered mPOS by eliminating checkout lines at its Apple Stores and using iPads with a card reader / payment app. Several clothing retailers, grocery stores, and restaurants are following suit.

„ Square: iPad PoS solution for retailers in US, simple pricing model vs traditional PoS (based on giving the dongle for free and then charging per transaction). Launched geofencing in the US which means that if a customer walks within 100 metres of a retailer with Square, targeted offers can be sent in real-time, or if the customer pays for something, then their name and details will already be with the retailer, not requiring any presentation of a card / PoS.

„ iZettle: European mobile payments company allowing retailers to accept card payments on smartphones / tablets. Uses Chip & Pin, which is more widespread in Europe (unlike US which where mag swipes still used). In August 2013 the company received mPoS approval by Visa and MasterCard.

Deutsche Bank AG/London Page 59

8 September 2013

Financial Services Retail Bank Strategy

Where to for branches?

Overview

Branches are the face of any established bank and where changing customer service preferences will be felt most. Current branch footprints and layouts are unsustainable in our view, with the most change needed in format rather than number of branches in most markets. We estimate that branches are about 50% of retail banking costs. We think restructuring could reduce this by 10- 20%, cutting overall retail costs by 5%-10% and largely offsetting the increased costs relating to core IT renewal discussed above.

„ Tech change aside, a number of countries in Europe (Spain, Italy, Cyprus, Portugal) have branch networks that vastly over-service customers relative to peers. European rates can’t increase for an extended period given vast quantities of floating rate household debt in the periphery, so we expect material cuts to branches in these countries over time to compensate for tough NIM and LLP conditions.

„ The US, Italy, Portugal and Hungary stand out, effecting little change in network size during the crisis when compared with markets like Denmark (-36%), Bulgaria (-33%), Netherlands (-31%), Belgium (-14%).

„ Trends in branch usage are self-explanatory: ING’s Belgian branch footfall was down 24% in 2012 on digital take-up, Danske’s teller transactions were down 35% last year while digital was up 34%, NAB’s teller transactions were down 16% CAGR 2009-2012 with digital growing strongly.

„ Format wise, we expect future branches to be 25% smaller, with 20% fewer personnel, fewer/no tellers and with more advisory staff. A reduction in cash handling in branch will make fit-out cheaper, reduce operating security costs, make branches easier to relocate and quicker to break even (JPM says six months faster).

„ Automation has a significant role to play. Nearly three quarters of Swedbank and Nordea branches in Sweden no longer deal with cash at tellers. LBG has diverted millions of branch cash deposit transactions to smart deposit machines. ATMs continue to grow in popularity. We expect third party providers like G4S and Notemachine to increasingly own and run ATM estates for banks. These operations marry secure logistics and technology maintenance, neither of which are necessary for delivering bank customer edge. If experiments with de novo self-service formats (drive through; supermarkets etc) prove viable, we expect maintenance of these to outsource also.

„ Changes in branch thinking has been fairly slow in coming given crisis priorities, the somewhat unexpected strength in mobile banking take- up, and upfront costs which not only include “good” capex but also the hidden expenses around lease break-fees and branch property held on balance sheet above market value. Change is inevitable we think and when it comes we expect it will be to the benefit of both customers and shareholders.

Page 60 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Branches are the “face” of the bank, 50-60% of costs…

Branch numbers and associated staff are naturally a sensitive topic. Branches We think current branch are the face of any established bank, readily-recognisable landmarks that most footprints are unsustainable in pass daily, and where most customers remember opening their first bank their current form; in most accounts. But with digital channels changing the frequency and manner in countries it’s branch design which customers deal with their bank, and within the context of broader cost not numbers that needs to rationalisation plans and progress in IT, current branch footprints are unsustainable, we think. As we discuss below, outside of some countries change most... where customers are over-serviced, we think most change needs to come in branch design and rather than in a radical reduction in outlets.

The facts on changes in digital and branch channel usage speak for …driven by changing habits themselves. ING’s incoming CEO says that the take-up of digital banking in its of customers and digital in Belgian business drove a 24% fall in branch footfall in 2012. Danske Bank’s particular branch teller transactions are down 35% YoY whilst mobile and internet interactions are up 34%. NAB’s counter transactions are down 16% CAGR 09- 12 whilst mobile grows strongly off a low base (Figure 98).

Figure 97: Australia: Ave interactions per customer p.m. Figure 98: National Australia Bank channel volumes by channel

18.0 120% 111% 3089% 16.0 100% 1.1 14.0 80%

12.0 5.1 60% 10.0 % change 40% 1.1 21% 8.0 20% 5% 6.0 6.3 0% 4.0 -3% -20% -9% -7% -16% -14% 2.0 -19% 2.2 -40% 0.0 Branch counters ATM Telephone Internet Banking Mobile Internet 2002 2004 2006 2008 2010 2011 2012 Banking Banking Interactions / cust per month by channel by month per / cust Interactions Mobile Online Phone ATM Branch 09-12 % change YoY % change

Source: Westpac Source: Company data

The proportion of retail costs which relate to branches is tougher to estimate We estimate that branches than one might think. Figure 99 shows that Barclays’ 2011 split of UK Retail represent ~ 50% of retail and Business Banking costs has the network at a third of the total before banking cost before refurb & branch property costs and staff performance and retirement pay. We would repurposing costs expect around 75% of these cost pools to relate to branches taking the branch cost share to ~ 50% of the retail total.

Figure 99: Barclays UK RBB cost split (2011) 2011 (£'m) % of total Customer network (ex property) 900 33% Property 300 11% Operations (ex property) 500 19% Performance & Retirement 300 11% Investment and Restructuring 200 7% Head office 300 11% Other 100 4% Total 2,700 100% Source: Company data; Note: Excludes PPI redress

Deutsche Bank AG/London Page 61

8 September 2013

Financial Services Retail Bank Strategy

This 50% share excludes branch renovation and relocation obvious in the UK UK banks have been active in and elsewhere and which is essential to improved staff efficiency, minimising renovating & relocating customer churn and normalising relations between banks and society. It is branches to improve interesting in this regard that Barclays’ management maths had the cost of efficiency, reduce churn and refurb covered if a 2% improvement in back-book retention rates was achieved rebuild brands compared with 25% all-in annual attrition reported at the time. We recognise the sentiment, but not the starting customer churn numbers.

In total we think that ~60% of retail banking costs relate to branches. We think 60% of retail bank costs relate to branches Falling branch numbers are a structural trend

Most developed markets have seen falling branch numbers since the crisis Most developed markets have began with Denmark (down 36%), Bulgaria (33%), Netherlands (31%), Spain seen significant branch cuts (16%), Belgium (14%) particularly noteworthy. The 2% decline in US branch through the crisis: US, Italy, numbers looks modest given 47235 FDIC-insured bank failures since end 07. Portugal, Hungary are outliers Far more out of line, however, given conditions on the ground are the numbers reported in Portugal (up 3%), Italy (down 2%) and Hungary (down 2%).

Figure 100: Branch rationalization in US, Italy, Portugal and Spain incomplete

250%

200%

150%

100%

50%

-

(50%)

(100%) Italy Peru India Chile Israel Brazil Spain Japan Kenya Sudan Ghana Ireland Turkey Poland France Austria Nigeria Cyprus Mexico Iceland Finland Greece Norway Ukraine Uganda Canada Sweden Belgium Ethiopia Bulgaria Portugal Hungary Uruguay Morocco Thailand Australia Malaysia Romania Mauritius Denmark Germany Sri Lanka Colombia Indonesia Cambodia Singapore Cameroon Philippines Switzerland Bangladesh Montenegro Netherlands South AfricaSouth Luxembourg Saudi Arabia New ZealandNew United States Czech Republic United Kingdom Korea, Republic of Russian Federation Venezuela, Rep. Bol. United Arab Emirates China,P.R.:Hong Kong

Source: European Banking Federation, FDIC for the US and APRA for Australia, Branch numbers for Norway and Russia are as on end 2011 and the US and Australia are as on June 2012.

Of course, looking at straight percentage changes in branch numbers doesn’t Straight percentage changes allow for differences in starting branch penetration, country size, geographic in branch numbers doesn’t distribution of populations, bank sector profitability, growth potential and other tell you anything about the obviously important drivers of distribution investment decisions. Switzerland, starting point or other key for example, should rank highly on branches per person and per square drivers of distribution kilometre; Russia and Australia should rank lower on penetration by square decisions... kilometre given more dispersed populations for example. We show time series data for 66 bank markets in Figure 114.

35 http://www.fdic.gov/bank/individual/failed/banklist.html

Page 62 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

To make better sense of the data we take 2012 branch numbers and divide separately by population and land mass size. As you’d expect, there’s a reasonable fit between the two, with the outliers of interest (Figure 101).

Figure 101: Branch density by population and by size

MalawiCameroon 66 Uganda EthiopiaUkraine 60 Cambodia Sudan Vietnam Kenya Ghana Saudi Arabia PhilippinesIndonesia Nigeria 54 Singapore Bangladesh Thailand Malaysia 48 India South Africa UAE Mexico Uruguay 42 Colombia Netherlands Sri Lanka Mauritius Montenegro Chile Turkey Venezuela 36 Israel Korea UK Morocco 30 HK Canada Czech Republic Sweden Ireland Norway 24 Denmark New Zealand Romania Japan Australia 18 US Finland Russia HungaryGreece Population / branch (Rank) / branch Population Switzerland Brazil Belgium Iceland 12 Poland Peru Germany France Austria Bulgaria 6 LuxembourgItaly Cyprus Portugal 0 Spain 0 6 12 18 24 30 36 42 48 54 60 66

Square km/Branch (Rank)

Source: Deutsche Bank, Datastream, European Banking Federation, FDIC, Australian Prudential Regulation Authority, World Bank, IMF

Then, for a rough and ready sense of relative distribution penetration, we add the per person and per km2 rankings in Figure 102: the lower the resulting total, the greater the ubiquity of branches is the basic idea.

Figure 102: Combining geographic and customer density into a single score

70

60

50

40

30

20

10

0 y y y y y g g a ia al ro y an sia p ain g UK US ar g ium ua p eria aria y ines Ital p prus g UAE ublic Peru g g India g anda g Chile y Israel Brazil S apore Kon p Korea Ja g Ken Sudan pp ladesh Ghana Turkey Russia Ireland France Malawi Austria g Poland g Mexico C Ni Iceland Greece Finland Norwa Ukraine U Canada g Sweden Bel Ethio Vietnam Bul Portu Hun Uru Morocco Thailand Australia Mala Romania Denmark Mauritius German Sri Lanka Colombia Indonesia Cambodia Sin Cameroon Venezuela Phili Hon Switzerland Ban Montene Netherlands South Africa Luxembour Saudi Arabia New ZealandNew Czech Re Czech

Source: Deutsche Bank, Datastream, European Banking Federation, FDIC, Australian Prudential Regulation Authority, World Bank, IMF

Deutsche Bank AG/London Page 63

8 September 2013

Financial Services Retail Bank Strategy

Off the bat it’s clear that, even before considering the shift in customer …on which basis Cyprus, demands, Cyprus, Italy, Spain and Portugal have far too many branches per Italy, Spain, Germany and citizen and that a reshaping of industry cost structures is required to deliver Portugal look most in need of better returns before considering technological change: distribution cuts: Italy has more branches/km and more „ Italy has the highest branch penetration on our crude measure of the 66 markets examined, with higher branch/km2 density than branches/person than Luxembourg or Switzerland. Population per branch is lower than all Luxembourg; Spain has more but 3 countries in our sample, partially driven by a co-operative sector branches per person than any representing ~ 60% of banks. Germany has a similar issue on branch of the other 65 markets density given its market structure. examined

„ Spain has more branches per person than any market we track. Though these may be smaller and cheaper than elsewhere (a common refrain), we expect recent industry consolidation and state aid requirements to drive material cuts in networks in the next two years.

„ Swedish and Finnish banks rank well which we believe is a driver of their superior cost/income performance, despite poor relative total income / total average loans income metrics. This restructuring of costs to align with revenue-generation of European markets is the only realistic reaction to the reality that interest rates cannot increase near term without an immediate increase in peripheral European sovereign risk given the vast quantities of floating rate debt in the region’s most vulnerable markets.

Branches to stay, but in tighter format as advice centres

In addition to the slow re-alignment of branch costs and near term revenue Branches are crucial to power, we expect a change in footprints to align with a consumer base which service provision and (i) is more likely to self-service through digital channels, (ii) relies less on cash- acquisition of customers in banked volumes as a proportion of total transactions, (iii) uses cheques less the past and (iv) needs to have trust built in the area of more complex savings / liability- driven products.

We think branch numbers to slowly drift lower and forecast five key changes We see five key changes to to future branch format: future branches

„ Smaller, reducing real-estate costs; NAB and Westpac’s new branches are 25% smaller than previous generation ones. Global peers report similar plans.

„ Less important as deposit venues, reducing cash handling costs (security: (bullet-proof glass, safes, cash bunkers) and tellers) and physical footprints. This will make branches less costly to fit out, easier to relocate and faster to break-even.

„ Less likely to house non-customer-facing staff which will be transitioned to cheaper locations; Overall, JPM expect a 20% reduction in branch headcount.

„ More important as venues for providing advice and sales, a place where face-to-face trust with customers is reinforced and renewed; Almost every bank we spoke to highlighted the intention to increase advisor headcount in-branch.

Page 64 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

„ Using technology far more to improve efficiency; Examples include: o Installation of smart-deposit machines in branch to reduce teller usage. LBG has installed over 300 deposit machines in the Lloyds TSB network, drawing millions of transactions from the counter. Up to three quarters of Swedbank and Nordea branches in Sweden do not deal in cash by teller.

o Move to take in-branch ATMs outside or in a walled-off section of the branch accessible out of hours. o Exploit economies of scale via disposal and outsourcing of ATM estates to independent operators like Notemachine and G4S (the likes of which now own more ATMs than the banks in the UK).

o Employ concierge / mobile technology in branch to reduce unnecessary queuing, improve satisfaction scores. Barclays has iPad-equipped employees in branch to meet entering customers to head off issues like address before queuing takes place; o Use Skype-like video-conferencing in branch to leverage advisors more broadly. In place at Westpac and Barclays, customers in branch can discuss mortgage needs, for example, with specialist staff. At Barclays this links call centre staff to in-branch customers, a sensible means to re-deploy call centre staff which are in some markets seeing lower customer volumes as digital grows. Also helpful in more geographically dispersed countries.

Smaller, advisor-heavy branches occupied less with cash handling is the most We expect 25% smaller meaningful change we expect to see. As noted, Westpac and NAB are branches with 20% fewer deploying branches ~25% smaller than predecessors. In the US JPM is staff despite increased advisor embracing smaller branches with a higher advisory presence. Over 70% of presence new Chase branches are smaller (sub 4,000 square feet), all have self-service banking kiosks, more have in-branch specialists (mortgages, business private client) and all have at least 6 month faster break-even targets than before. JPM target a 20% reduction in “same store” staff headcount by 2015.

NAB is experimenting with no-teller and reduced teller branch formats in Experiments with de novo standalone and de-novo formats in addition to implementing smaller layouts formats are back: drive overall. Figure 103 shows NAB is aiming to hold branch numbers broadly through, malls, in-store constant over the next few years but with a ~ 200 reduction in full service “stores” in favour of a similarly-scaled increase in “smaller advice-driven outlets”. Reduced headcount and floorspace and an increased proportion of automated transactions is targeted to increase personal banking sales hours per branch by 30% and double “time in front of” business customers.

Deutsche Bank AG/London Page 65

8 September 2013

Financial Services Retail Bank Strategy

Figure 103: National Australia Bank’s branch reconfiguration targets

700

600

500

400

Branches 300

200

100

0 Current Target

Full service stores Smaller, advice driven outlets Standalone Business Banking Centres

Source: NAB company data

More outsourcing: The ATM case study

We expect automation to play a bigger role in branch efficiency drives. Even Automation has a role to play automatic teller machines – which have been around for decades – are seeing in improving branch efficiency upgrades as smart deposit machines to reduce teller traffic. This comes on top of increased customer usage of ATMs more generally, with cash withdrawals up 70% by volume and 90% by value over 4 years (Figure 105).

Figure 104: ATM networks in ascendance Figure 105: Number and value of cash withdrawals

1,400 1,330 25,000 40,000 6,000 1,200 33,872 35,000 1,200 5,000 1,069 20,000 29,980 972 30,000 26,229 1,000 871 22,855 4,000 15,000 25,000 800 19,925 20,000 3,000 600 10,000 15,000 2,000 400 10,000

ATM totals (thousands) totals ATM 5,000 1,000 200 5,000 Point of Sale terminals (thousands) 0 0 0 0 2007 2008 2009 2010 2011 2007 2008 2009 2010 2011

ATM totals POS terminal totals Number of cash withdrawals Value of cash withdrawals USD billions

Source: BIS data, Deutsche Bank analysis Source: BIS data, Deutsche Bank analysis

We think the increase in cash withdrawal volumes indicates changes in the Over 80% of UK cash way consumers get cash rather than a trend towards consumers making more withdrawals are from payments in cash. UK cash machines more than doubled since 2000, with 83% machines, with volumes of cash withdrawals in 2011 from ATMs. The 65% increase in Point-of-Sale growing, despite increasing terminals to 21.6m across the BIS-payments community over 4 years points card acceptance in evidence ubiquity of card acceptance as a cash alternative. We expect both trends to continue, diverting transactions from the branch. globally

Page 66 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 106: Cash acquisition 2001... Figure 107: ...and in 2011 Figure 108: Legend 5% 7% 3% 3% 9%

ATM Cheque/Passbook 28% Cashback Card at the counter 62%

83%

Source: UK Payments Council. £181bn volume Source: UK Payments Council. £230 billion volume Source: UK Payments Council

There has been increase in the number of non-bank owned/operated ATMs, Non-bank and outsourced where non-bank companies (e.g. NoteMachine, G4S) generate fees for ATMs outnumber bank processing transactions. These are either (i) ‘pay-to-use’ and located inside machines in the UK run by retail outlets (e.g. inside a pub), charging the customer ~£2 per transaction; or companies like G4S and (ii) free to customer but with the issuing bank paying ~25p in interchange fee. NoteMachine; we expect NoteMachine, the largest non-bank ATM owner in the UK with c.7000 earned average revenues last year of ~ £10k per machine36. Other operators such as more of this Tesco Bank have deployed large in-store ATM networks to generate interchange fee income from other banks (monetising customer footfall) and reducing interchange fees paid by Tesco to banks for customers paying in store with cards.

Figure 109: UK ATM numbers over time Figure 110: Breakdown of ATM owners

70.0 30,000

60.0 25,000

50.0 12,055 20,000 40.0 15,000 Non-Bank 30.0 Bank 10,000 20,260 20.0 19,274 15,143 10.0 5,000

0.0 - 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Branch ATM Non-Branch free Non-Branch pay-to- Free to use Pay to use ATMs use ATMs

Source: Link Source: Link

There will be adjustment pains

As discussed above, the shift to customer self-service via online and smart- Digital isn’t that new, why phone banking is not so new as to have escaped the notice of bank hasn’t this all happened yet? management. Equally, crisis-declines in interest margins in much of the developed world at the same time as loan losses have risen should have provided ample motivation to drive efficiency gains from the retail distribution network before now, right (Figure 111, Figure 112)?

36 We expect very significant differences between sites.

Deutsche Bank AG/London Page 67

8 September 2013

Financial Services Retail Bank Strategy

Figure 111: European net interest margins by country Figure 112: European average LLPs by country

4.5% 3.50%

4.0% 3.00% 3.5% 2.50% 3.0% 2.06% 2.00% 2.5% 2.3% 2.0% 2.1% 1.50% 1.9% 1.5% 1.00% 0.91% 1.0% 0.89%

0.5% 0.6% 0.50%

0.0% 0.00% 0.05% 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 -0.50% UK Spain Sw iss France 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 UK Spain Sw iss France

Source: Deutsche Bank, Company data Source: Deutsche Bank , Company data

A number of reasons we think explain this apparent lack of material change to We see three reasons for date. None of these, however, preclude serious change ahead. We think there previous delay in action are three key reasons.

„ Online banking is not be new, but mobile banking is growing far faster

„ Banks have had other operational priorities Obviously true and equally obviously temporary, banks are increasingly moving beyond crisis priorities.

„ The upfront P&L pain doesn’t produce instant gains. Like all capex there’s a period of P&L pain before benefits are evident. Also, not all costs of change relate to positive-benefit spend where customers see direct benefits. In markets where property prices have fallen significantly, property book values may well be much higher than market value, requiring a write-down if no longer inhabited for branch- banking. Banks keen on sale-and-leaseback pre-crisis face lease break- costs. Neither of these are capital adequacy concerns but they do highlight frictional costs that delay in the absence of a change in management or some other catalyst for restructuring charges.

Figure 113: Property book value to NAV (%, most recent full year)

18%

16% 15%

14%

12% 10% 10% 9% 8% 8% 7% 7% 7% 6% 6% 6% 5%

4% 3% 3% 2% 2%

Property NBV / Shareholders' equity (%) equity Shareholders' / NBV Property 0%

Source: Deutsche Bank, company data

Page 68 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy 3 8 2 7 9 5 17 17 66 66 33 54 44 40 48 64 53 49 27 23 24 37 58 28 15 30 39 32 65 22 45 16 10 42 Rank 3 4 3 23 23 67 84 46 23 43 92 18 88 67 76 103 103 368 122 136 132 120 235 111 110 416 136 365 207 350 515 205 2012 2012 1,188 1,188 6,866 tralia are June 2012. June tralia are 3 4 3 99 22 72 74 37 22 38 74 17 92 86 59 61 344 110 132 127 114 237 113 374 122 315 198 351 496 177 2004 1,085 6,510 Population / Square km km Population / Square 3 8 5 1 4 6 9 60 17 16 38 23 54 65 64 51 46 27 24 59 44 14 50 26 22 58 25 37 30 18 55 11 Rank 8 1 9 18 16 28 11 36 30 14 10 35 27 33 63 19 10 13 11 126 454 323 225 213 288 857 122 489 2012 1,128 2,626 1,324 1,063 9 6 9 8 1 9 15 19 21 19 43 20 21 38 30 44 73 23 10 n/a n/a n/a n/a 503 190 632 563 228 2004 Square km / branch Square km / branch 1,254 5,628 3,067 1,154 9,240 1,042 2,590 82,450 30,280 77,260 42,430 89,620 68,890 21,640 96,820 130,170 108,760 176,520 472,710 743,530 304,590 547,670 348,770 227,540 128,900 100,250 294,140 364,500 569,140 (Sq. km) 7,682,300 8,459,420 9,093,510 1,109,500 1,000,000 2,973,190 1,811,570 Land area 6 7 2 4 8 5 9 20 54 14 15 60 65 30 40 43 28 25 63 23 57 17 31 16 13 48 51 26 37 21 58 34 Rank 3.3 1.9 2.9 3.0 1.9 5.1 7.6 9.7 1.3 5.0 3.9 3.8 1.7 2.3 3.1 5.4 3.0 2.7 4.2 7.1 1.9 3.4 6.6 2.3 19.4 38.2 97.5 32.1 13.7 16.7 37.1 2012 120.6 7.4 1.6 n/a n/a n/a n/a 3.3 1.9 2.1 1.4 1.0 5.7 2.5 3.3 2.4 1.8 3.2 5.5 3.4 1.6 4.3 7.2 1.9 3.4 22.4 10.9 57.9 16.5 27.9 70.7 2004 210.4 227.2 Pop. / branch in 000s / branch in 000s Pop. 8% 2% 6% 5% 5% 8% 6% 5% 7% 6% 2% 2% 2% 3% 1% 3% 3% 7% 7% 5% 3% 3% 14% 14% 14% 13% 10% 14% 11% (5%) (0%) (1%) (0%) %12/07 %12/07 4% 4% 7% 8% 9% 8% 3% 3% 4% 5% 2% 5% 5% 16% 13% 10% 13% 23% 12% 11% 24% 22% 10% 11% 12% 13% 16% 24% (6%) (1%) (2%) (0%) spectively to eliminate post office banking offices. Branch numbers for Norway and Russia are as on end 2011 and the US and and Aus the US and 2011 end as on are Russia and Norway for numbers Branch offices. banking office eliminate post to spectively %12/04 8 7 1 6 5 7 0 5 8 1 23 11 15 22 35 17 48 11 92 66 82 25 11 10 61 43 50 155 199 247 128 2012 1,237 8 8 1 5 5 7 0 4 7 0 21 11 14 19 33 17 44 10 80 64 82 23 11 10 59 38 49 146 190 231 128 2007 Population (millions) 1,159 8 8 1 5 5 7 0 4 7 0 20 10 13 18 32 16 43 10 74 63 83 21 11 10 58 35 48 141 184 221 128 2004 1,111 5% 4% 5% 3% 6% 0% 18% 18% 55% 50% 24% 17% 13% 33% 25% 53% 12% 61% (6%) (2%) (9%) (6%) (2%) (8%) (2%) 103% (10%) (14%) (33%) (36%) (17%) (25%) %12/07 n/a n/a n/a n/a 3% 7% 9% 5% 11% 26% 56% 18% 45% 12% 34% 86% 17% 18% (1%) 18% 114% 189% 119% 136% (21%) (32%) (11%) (33%) (11%) (20%) (34%) (20%) branch numbers in 2007 and 2012 adjusted by 10.0k and 10.4k re %12/04 233 2012 6,809 4,488 Branches 251 389 251 180 120 877 930 941 463 790 595 117 157 1,112 989 1,163 721 260 2007 2007 6,528 6,528 4,292 7,968 6,727 3,855 4,474 3,837 5,747 6,867 6,651 2,305 1,864 4,921 4,203 2,118 1,876 1,424 2,212 1,424 1,714 3,647 3,872 1,337 1,264 3,340 3,393 14,797 9,680 1,099 1,189 7,623 7,273 56,757 66,946 56,757 38,424 39,239 36,327 39,843 90,315 72,271 32,597 33,301 37,450 37,369 - - - - 84 84 981 981 326 360 178 940 946 493 292 292 2004 2004 6,468 6,468 6,128 6,128 4,378 6,304 4,873 5,610 1,478 1,794 2,134 1,604 3,422 1,228 2,987 7,939 26,425 45,393 67,355 31,000 37,901

Figure 114: Bank branch penetration data

Country Country Australia Austria Bangladesh Belgium Brazil Bulgaria Cambodia Cameroon Canada Chile Colombia Cyprus Czech Republic Denmark Ethiopia Finland France Germany Ghana Greece China,P.R.:HK Hungary Iceland India Indonesia Ireland Israel Italy Japan Kenya of Korea, Republic Luxembourg Source: EBF, FDIC, Datastream, World Bank, Deutsche Bank; France

Deutsche Bank AG/London Page 69

8 September 2013 Financial Services Retail Bank Strategy 4 6 1 21 21 38 46 47 43 59 19 60 55 11 26 29 35 63 62 50 36 12 61 56 18 25 34 20 41 31 14 51 57 52 13 Rank 9 89 89 62 46 73 17 16 23 93 13 42 93 16 23 96 79 34 19 34 169 169 636 497 185 324 127 115 324 200 131 182 110 261 286 2012 2012 7,589 7,589 tralia are June 2012. June tralia are 9 77 56 46 67 16 15 21 94 11 38 86 13 22 87 82 44 32 19 30 133 608 482 149 282 125 115 310 185 127 139 247 263 2004 6,128 Population / Square km km Population / Square 7 2 61 42 43 41 33 13 47 40 49 36 29 19 15 28 53 63 52 12 21 66 45 10 32 31 56 62 35 20 34 57 48 39 Rank 2 11 87 13 62 20 15 40 13 24 12 83 75 98 21 94 160 163 158 227 157 264 107 374 327 216 505 512 256 131 2012 1,257 1,306 4,075 1,282 9 2 14 69 37 17 76 12 51 15 17 n/a n/a n/a 144 249 220 239 261 247 525 817 202 131 126 433 186 100 566 324 Square km / branch Square km / branch 1,652 1,768 5,753 1,427 2004 680 2,030 94,280 13,450 33,760 91,500 62,710 40,000 83,600 328,550 446,300 263,310 910,770 304,280 298,170 304,300 230,000 499,180 410,340 510,890 769,630 199,810 579,340 241,930 175,020 882,050 310,070 (Sq. km) 1,943,950 1,280,000 2,149,690 1,214,470 2,376,000 9,161,920 16,381,230 Land area 3 1 66 50 35 45 39 33 36 24 56 27 11 55 12 22 19 52 53 49 41 61 29 10 47 38 62 64 46 32 18 44 42 59 Rank 6.7 7.3 6.4 6.7 3.8 4.3 2.5 2.5 1.7 3.7 3.3 1.2 7.8 5.0 2.4 7.2 5.4 3.2 9.9 8.7 14.3 10.1 29.1 20.1 17.2 17.2 13.8 63.8 10.8 91.8 10.8 37.6 2012 212.1 100.9 n/a n/a n/a 8.4 4.3 3.7 3.7 4.6 1.9 7.1 4.6 1.1 4.4 2.8 8.1 4.3 3.2 9.7 11.1 14.0 14.7 38.9 35.4 19.5 19.6 14.8 31.4 15.8 74.5 16.7 10.9 35.5 10.8 2004 198.3 Pop. / branch in 000s Pop. 9% 2% 6% 1% 6% 2% 5% 7% 6% 9% 1% 1% 9% 6% 3% 1% 4% 6% 1% 6% 4% 4% 2% 8% 5% 16% 16% 15% 16% 12% 18% 59% (1%) (1%) (2%) %12/07 %12/07 5% 1% 9% 3% 8% 9% 9% 1% 0% 8% 5% 6% 8% 3% 6% 7% 2% 9% 27% 15% 10% 24% 15% 19% 27% 10% 21% 11% 31% 14% (2%) (0%) (4%) 152% %12/04 spectively to eliminate post office banking offices. Branch numbers for Norway and Russia are as on end 2011 and the US and and Aus the US and 2011 end as on are Russia and Norway for numbers Branch offices. banking office eliminate post to spectively 1 1 4 5 5 8 9 3 16 29 33 17 30 97 39 11 21 28 51 46 20 37 10 67 74 36 46 63 30 89 121 169 144 314 2012 1 1 4 5 5 9 8 6 3 14 27 31 16 28 89 38 11 22 26 48 45 20 33 66 69 31 47 61 28 84 114 147 142 301 2007 Population (millions) 1 1 4 5 4 9 7 4 3 13 25 30 16 27 84 38 11 22 24 47 43 19 31 65 67 28 47 60 26 81 109 136 144 293 2004 n/a n/a 4% 1% 5% 3% 9% 6% 4% 4% 21% 26% 86% 37% 13% 31% 22% 87% 34% 42% 21% 34% 45% (5%) (5%) (6%) (2%) 229% 104% (52%) (31%) (10%) (16%) (67%) %12/07 n/a n/a n/a 5% 9% 6% 32% 53% 66% 12% 83% 16% 89% 40% 35% 41% 27% 58% 68% 90% 11% 27% (6%) (6%) (7%) 152% 150% 111% 183% (10%) (35%) (66%) (16%) 1,439% %12/04 branch numbers in 2007 and 2012 adjusted by 10.0k and 10.4k re 75 85 193 308 583 396 452 851 342 2012 2,050 5,113 2,500 1,161 5,810 1,158 4,819 6,280 5,731 1,646 3,712 2,598 1,901 3,330 6,170 3,447 2,360 11,911 11,931 15,190 43,752 38,219 10,234 11,779 97,337 - - Branches 160 160 176 290 411 194 587 329 2007 2007 1,968 1,968 9,458 2,748 3,632 1,155 4,254 1,106 3,623 4,273 6,078 6,350 1,353 1,982 1,935 2,008 3,517 5,081 7,618 1,365 3,299 11,621 39,971 45,571 12,506 99,152 - - - 146 146 775 282 413 140 449 309 2004 2004 2,276 2,276 7,797 2,033 3,820 1,103 3,492 1,234 4,310 8,304 5,397 3,037 1,216 1,486 1,231 2,035 2,627 3,908 6,106 1,338 2,719 31,214 40,656 13,983 92,030

Figure 115: Bank branch penetration data (continued)

Country Country Malawi Malaysia Mauritius Mexico Montenegro Morocco Netherlands New Zealand Nigeria Norway Peru Philippines Poland Portugal Romania Russian Federation Saudi Arabia Singapore South Africa Spain Sri Lanka Sudan Sweden Switzerland Thailand Turkey Uganda Ukraine UAE Kingdom United United States Uruguay Bol. Venezuela, Rep. Vietnam Source: EBF, FDIC, Datastream, World Bank, Deutsche Bank; France

Page 70 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Big Data: Unleash its commercial value

How big is Big Data?

37 According to IBM’s statistics , the world creates 2.5 quintillion bytes of data a day. Daily world data creation is That’s 2.5 billion billion characters. 2.4bn Bibles or 390 billion reports roughly the size of about the same as this report this one. Stacked on top of each other, 390 billion reports like this would (roughly) stacked to the moon 60 times reach the moon sixty times. The volume of data created is growing strongly: today’s daily output is about the same as world data storage capacity in the mid 1980’s. Big data is not just about volume, but also about variety: eighty percent of data produced now is unstructured (email, phone calls, pictures, GPS locations from smart devices and so on), and unstructured is growing much faster than structured data.

The promise of big data is that as computing processing power gets cheaper and faster Computing and storage (today’s standard computer has roughly the same processing power as a power are growing fast, supercomputer of five years ago), and as networks get faster (fibre optic capacity is raising hopes that more can doubling every nine months or so), businesses will be able to harness this data through be done with this data analytics to make and sell greater volumes of better products to more customers.

Who knows more about your habits than your bank? Commentators are overlooking the value of bank For all the apparent promise of big data, we’ve seen little written about the value of customer data, other than in data within banks other than in relation to risk management. We think investors are relation to credit risk ignoring the potential value of this information. Banks are more regulated, carry more management capital and offer less growth than many tech companies, but the valuation differential with data-rich tech firms is noteworthy, we think (see comps sheet at start of report).

Figure 116: Are banks valued like tech / data-rich companies?

30.0 29.0

28.0

26.0

24.0

22.0 20.0 19.5 20.0 19.1 Forward P/E 18.0 16.0 16.2 16.0

14.0 12.6 12.0

10.0 Banks IT Services Mobile Processors Payments Customer Customer Payments networks data ex data Groupon

Source: Deutsche Bank, Bloomberg Finance LP, Datastream; Based on stock universe at the start of this report; Priced at 4 September 2013

37 Understanding Big Data; Analytics for Enterprise Class Hadoop and Streaming Data

Deutsche Bank AG/London Page 71

8 September 2013 Financial Services Retail Bank Strategy

Investors and customers haven’t been given much reason to suppose banks can harness the data they possess

We think the reason the value of bank data is given little attention outside of credit and There’s little evidence banks money-laundering risk management is because there’s little evidence it is being are using transaction data for monitised in other ways. Few banks give customers decent budgeting or wealth anything other than risk tracking tools in their digital channels. Product marketing in most cases appears management rudimentary (Hello, thanks for logging onto our mobile banking app, do you want a savings account or credit card?) and shows few signs of being tailored to customer circumstances.

That said, it is obvious that banks should have better insight into customer financial Banks know more about heath and spending habits than Google can derive based by observing internet search financial health and spending terms or Tesco’s store-card loyalty scheme can know, no matter how high the market habits than Google or loyalty share of each in their respective spheres. schemes can divine

The data accumulated by banks in developed markets covers the majority of economic The large UK banks each activity. As a case in point: the official UK GDP growth estimates are derived from track 25x the number of firms responses to questionnaires sent to 46,000 of the UK's 4m firms. The larger UK clearing used by government to banks each have more than a million business and SME customers: at least 25x the estimate GDP growth sample set used to estimate economic growth.

In retail banking, LBG has 30% of the current account market which allows the bank to SWIFT publishes interesting derive faster (and sometimes better) first estimates of GDP growth than that which the regional GDP growth government process can deliver, we think. Though LBG don’t publish their early GDP forecasts based on estimates, SWIFT 38 does based on payments data from its network (Figure 117). We interrogation of payments expect banks and other financial services companies to make much more of their data from its own network information advantage as IT advances continue.

Figure 117: SWIFT GDP growth estimates (YoY %)

6.0

4.0

2.0

0.0

-2.0

GDP Growth YoY Growth GDP -4.0

-6.0

-8.0 Q1 2003Q1 2003Q3 2004Q1 2004Q3 2005Q1 2005Q3 2006Q1 2006Q3 2007Q1 2007Q3 2008Q1 2008Q3 2009Q1 2009Q3 2010Q1 2010Q3 2011Q1 2011Q3 2012Q1 2012Q3 2013Q1 Q3 2013e

OECD EU27 US UK DE

39 Source: SWIFT

38 Society for Worldwide Interbank Financial Telecommunication, a 10,000 member co-operative formed to process international interbank payments 39 http://www.swift.com/zdoc/swift_index/swift_index.page

Page 72 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Non-bank service companies show the way

Non-bank financial services companies like Experian and Groupon have built significant Experian, Groupon have built revenue and market cap value by leveraging of customer data and linking of buyer and significant market caps from seller. The P/E’s and market caps in the figures below speak for themselves. the customer data

Figure 118: Experian snapshot Figure 119: Groupon snapshot Experian Analyst name Email Groupon Analyst name Email Andy Chu [email protected] Ross Sandler [email protected] Spot price: 1152 Tgt. Price 1300 Rec. Buy Spot price: 11 Tgt. Price 17 Rec. Buy 2010 2011 2012 2013e 2014e 2015e 2010 2011 2012 2013e 2014e 2015e EPS (DB adjusted) 0.6 0.7 0.8 0.9 0.9 1.0 EPS (DB adjusted) -0.3 -0.5 0.1 0.1 0.3 0.4 DPS 0.2 0.3 0.3 0.3 0.4 0.4 DPS 0.0 0.0 0.0 0.0 0.0 0.0 BVPS 2.4 2.7 3.0 3.3 3.3 4.0 BVPS 0.0 1.2 1.1 1.3 1.5 1.9 ROE 27% 19% 23% 8% 23% 25% ROE NA -53% -9% -4% 7% 12% EBITDA mgn 31% 33% 33% 33% 35% 36% EBITDA mgn -54% -13% 7% 7% 9% 10% Market Cap (US$'mn) 8,725 10,810 12,832 15,870 17,588 17,588 Market Cap (US$'mn) NA 13,231 6,676 7,104 7,104 7,104

P/E (DB) 13.5 16.1 16.5 18.7 19.0 17.2 P/E (DB) NA NA 179.8 79.3 34.6 24.6 EV/EBITDA 8.3 9.5 10.4 11.8 11.6 10.6 EV/EBITDA NA NA 34.8 30.2 18.5 12.8 Div yield 3% 3% 2% 2% 2% 2% Div yield NA 0% 0% 0% 0% 0% FCF yield 9% 8% 7% 6% 5% 6% FCF yield NA 2% 3% 3% 4% 5% ROE 27% 19% 23% 8% 23% 25% ROE NA -53% -9% -4% 7% 12%

Source: Company data, Deutsche Bank estimates, Priced as on 5 September 2013, All figures in local Source: Company data, Deutsche Bank estimates, Priced as on 5 September 2013, All figures in local currency unless stated currency unless stated

Experian: Credit scoring business expanded into Marketing Services Experian plc was founded as a credit reference agency. It has around two thirds of the Experian was founded as a UK credit services market, providing credit scores and fraud detection to credit credit reference agency in the providers including banks, telcos, leasing firms and so on. In this capacity Experian has UK and has expanded amassed an immense data set including positive and negative credit data from these geographically and into other customers which is used to divide customers into 250 demographic categories. 30% of businesses since… Experian’s total revenues are from the financial services sector.

Figure 120: Experian FY2013: 26% EBITDA margin, 8% organic revenue growth, 94% cash conversion 2006 2007 2008 2009 2010 2011 2012 2013 Revenue (US$'m) 2,930 3,407 3,712 3,790 3,803 3,859 4,456 4,713 Organic revenue growth (YoY %) 12% 8% 4% 3% 2% 8% 10% 8% EBIT margin (%) 21.0% 21.9% 22.8% 23.6% 24.5% 25.7% 26.2% 26.6% Operating cash conversion 102% 100% 99% 101% 100% 98% 96% 94% Source: Company data

Experian has expanded by geography (into Brazil, Turkey, Australia, Russia, South …delivering a 27% EBIT Africa, amongst others) and into other businesses like marketing services. This business margin on £4.7bn in revenue uses anonymised customer transaction data to profile customers for marketing which grew 8% YoY with a purposes at a post-code level. 94% operating cash conversion in FY13 Banks must use their data to grow revenues, defend customer wallet share and drive higher customer loyalty

It makes sense that - even managing for very significant reputation risks and customer In future banks would better sensitivity around the use of their most personal information - banks will follow the exploit their information set in customer data mining pursued by the likes of Experian in time. We expect banks will three ways better exploit their information set in three ways:

„ Providing customers with better financial planning tools and advice;

Deutsche Bank AG/London Page 73

8 September 2013 Financial Services Retail Bank Strategy

„ Marketing services for retailers, benefits for customers;

„ Moving to wallet share pricing of customers; Providing customers with better spending and financial planning tools We expect all banks to improve financial information provided to customers to include Bank customers would get decent expenditure tracking versus budget. In the developed world we see a need for better digital tools to control this to be integrated with wealth tracking and retirement-need profiling, integrated with current financial behavior and liability-product sales. What proportion of DM customers properly understands the plan for the future; DM required NPV of savings needed to live comfortably for 15-20 years post retirement? For customers need to some banks this product provision will come with IT renewal. For others it will begin with new online bank ventures set up on fresh IT platforms. For all it requires understand retirement investment and careful design for the management of conduct risk. savings requirements better

Marketing services for retailers, benefits for customers We expect banks to get into the business of linking retailers and customers, made We expect banks to get into possible by data analysis (a US$1bn revenue business for Experian). We like NAB’s the business of linking People Like U site. Billed as the world’s first “econography” tool, retailers and customers, www.peoplelikeu.com.au uses analysis of a billion NAB customer transactions to show made possible by data the financial habits of similar people to the user down to post-code level. The level of analysis detail provided is interesting: food spend, for example, is split into dining out, groceries, liquor, groceries, pubs and clubs and so on; the site shows which retailers are popular, preferred and emerging, with average spend per visit in each also listed.

Figure 121: National Australia Bank’s People Like U is based on analysis of over a billion anonymised datapoints

Source: National Australia Bank

Page 74 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

The site allows customers to click through to the retailers listed for online shopping. The crucial next step would The next step from here, which we expect to see pursued by many banks, will be to roll be to launch retail offers to out voucher and discount offers to customers. At the slightly more cutting edge these customers to generate can be location-driven with offers pushed to smartphones based on customer revenue, grow loyalty and transaction analysis: “Hi; you’re near Thomas Pink on London Wall (and we know defend against non-bank you’ve shopped there / at its largest competitor before). Pop in now and pay with your ABC Bank credit card for a 25% discount on fitted shirts”. digital wallets

Operated sensibly on an opt-in basis and carefully managed to limit “big brother” fears40, we think these businesses will:

„ Generate incremental bank revenue;

„ Add value to customers, increasing loyalty and net promoter scores;

„ Reduce the marginal benefit of retailer-specific e-wallets and so slightly cut the (manageable) threat of bank digital disintermediation;

„ Reinforce bank merchant acquiring businesses for some banks;

As with mobile banking and mobile payments, this technology can be acquired from The Cardlytics – LBG deal is external providers. Cardlytics, a US company, recently launched in the UK offering interesting, as is the Barclays customers voucher and special offers based on behavioural data, signing Lloyds bespokeoffers launch Banking Group with its 25% market share. Other banks will develop such products in house: here we note Barclays’ launch of Bespokeoffers earlier this year, which we expect to expand in this direction from a Groupon-like start.

Figure 122: Barclays bespokeoffers

Source: www.bespokeoffers.co.uk

Moving to wallet share pricing of customers, with risks and rewards Perhaps the most fundamental change which we expect in the medium term will be a Perhaps the most move to price properly for wallet capture. The upside from success is obvious: Figure fundamental change which 123 shows Westpac reporting that core customers have 8x the retention rate and 4x we expect in the medium the revenue take of non-core equivalents. However defined, there is upside from term will be a move to broadening the core customer base of any bank. What better means of doing this than properly price for wallet sharing the gains of higher wallet share with the customer? There are low-tech ways of doing this – see for example, the Sainsbury’s Bank Double Nectar Points offer which capture grants loyalty points on in-store purchases multiplied by up to 6 based on the number of products the customer holds from the bank41.

40 We know of one European bank that has purposely built delays into its service to attempt to mitigate this risk. 41 http://www.sainsburysbank.co.uk/doublenectar/index.shtml?source=NETGLOBNAVISOURC0006

Deutsche Bank AG/London Page 75

8 September 2013 Financial Services Retail Bank Strategy

Figure 123: Wallet-share pricing is around the corner – Figure 124: The move in motor insurance to Pay-How- why aren’t we sharing the upside? You-Drive is in its infancy 9.0x 200 8.0x 80% 180 8.0x 180 increase in 6 months 7.0x 160 140 6.0x 5.4x 120 5.0x 4.3x 100 100 4.0x

customers 80 3.0x 2.4x 60 2.0x 35 1.5x 40

1.0x ofNo. UK telematicpolicyholders (000s) Westpac 'MyBank' vversus non-'MyBank' 20 12

.0x 0 Retention Footings ($) Revenue Products Advocacy 2009 2010 2011 1H12

Source: Westpac company data Source: BIBA, Deutsche Bank

Democratisation of your Big Data could increase churn rates and Medium term (10 years+) lower margins in the longer term bank margins could come under pressure if customer Later on (say 10 years+) we think bank margins could come under pressure if customer transaction data becomes transaction data becomes portable in a standardised format. Look at the general portable in a standardised insurance case study: margins came under pressure when comparison websites like format Moneysupermarket made shopping around easier (Figure 125). Longer term, telemetric driver monitoring will see more motor premiums move to a Pay-How-You-Drive model42. Estate agents are seeing pressure from on-line property portals like Rightmove (Figure 126).

Figure 125: Moneysupermarket snapshot Figure 126: Rightmove snapshot Moneysupermarket Spot price: 174 Tgt. Price 199 Rightmove Analyst name Email Patrick Kirby [email protected] 2010 2011 2012 2013e 2014e 2015e Spot price: 2382 Tgt. Price 2450 Rec. Hold EPS (adjusted) 5.0 7.0 9.0 11.0 12.0 13.0 2010 2011 2012 2013e 2014e 2015e DPS 3.8 4.5 5.7 16.5 7.2 8.0 EPS (DB adjusted) 34.3 42.3 59.2 72.3 86.7 99.7 BVPS 36.8 32.7 37.2 28.0 26.0 27.0 DPS 14.0 18.0 23.0 27.5 32.5 37.5 ROE 4% 9% 14% 38% 46% 50% BVPS 25.8 23.5 7.4 -2.7 -8.4 -14.7 EBITDA mgn 27% 29% 30% 35% 36% 37% ROE 241% 200% 417% NA NA NA Market Cap (US$'mn) 619 829 1,370 1,472 1,472 1,472 EBITDA mgn 70% 72% 74% 75% 76% 76% Market Cap (US$'mn) 1,086 1,823 2,401 3,695 3,695 3,695 P/E 51.6 31.9 32.8 16.6 15.0 13.5 EV/EBITDA 9.1 9.6 13.5 11.7 10.6 9.6 P/E (DB) 18.9 25.7 25.1 33.0 27.5 23.9 EV/EBITDA 11.8 15.9 17.0 22.8 20.0 17.8 Div yield 5% 4% 4% 9% 4% 5% Div yield 2% 2% 2% 1% 1% 2% FCF yield 8% 8% 6% 6% 7% 7% FCF yield 7% 5% 5% 3% 4% 5% ROE 4% 9% 14% 41% 53% 50% ROE 241% 200% 417% NA NA NA

Source: Deutsche Bank; Bloomberg Finance LP, Priced on 5 Septembert 2013, Moneysupermarket is Source: Deutsche Bank estimates, company data; Priced as on 5 September 2013, Note: Figures in not under DB equity research coverage. Statistics above, including target price, refer to Bloomberg local currency unless stated consensus. All figures in local currency unless stated

We don’t think it a stretch to imagine regulators requiring banks to make customer data available (similar to a Freedom of Information request in the UK) in a format which would allow a consumer portal to facilitate all banks bidding on all of your financial product needs. In a crude manner, credit report agencies already do some of this by pointing customers to financial products based on their financial characteristics. We think these offerings will become more sophisticated too.

42 For more on telematics please see our insurance team’s detailed 4 Sept 2012 report: Technology race

Page 76 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Crowd- & Peer-to-Peer: Cutting out the middlemen

"I see opportunity knocking for finance. Hopefully, the growth of peer-to-peer lenders and those involved in crowd-funding will help solve the problems we have with lending for small and medium enterprises ... The banking middlemen may in time become surplus links in the chain."

Andrew Haldane, Executive Director, Financial Stability, Bank of England43

Key points

„ Peer-to-peer finance and crowdfunding businesses are emerging worldwide as low yields on cash, growing investor trust of online commerce, maturing auction technology and cyclical bank deleveraging converge to bridge investor demand and borrower needs.

„ P2P generally relates to unsecured lending to retail customers. Crowdfunding primarily backs SMEs, protected by personal guarantees and/or collateral. Crowdfunding is also used to support creative projects such as music and film in exchange for social-media recognition and/or purchases of the end product.

„ Both models use eBay-like online auction technology to allow lenders (investors with cash) to bid for small parts of loans sought by individuals or customers. Lenders pay an annual management fee of around 1%, borrowers pay the ruling interest rate on the funding they accept, plus around 2% in fees to the facilitating website. Lenders are also generally able to trade their loans at a cost of about 25bps providing secondary liquidity.

„ There new models of financing are interesting: readers would be well advised to spend some time on the major sites seeing how much funding is being raised on a cost effective basis by creative endeavours in particular.

„ Though historically low interest rates and some supply-side constraints amongst banks make the nexus between borrowers and lenders more fertile than usual, we suspect this is true of any recession which involves significant bank trauma. What makes this cycle different is the role of online auction technology to marry these parties, reducing costs and placing scale without expensive distribution requirements within reach.

„ Overall, we note returns to owners of scale P2P and Crowdfunding enterprises will far outpace those of their customers and banks with which they compete. In May 2013, The Funding Circle lenders were earning 5.8% net on their loans. Borrowers are paying between 7.1% and 9.4% depending on risk bucket in interest. Funding Circle earns a 2% fee from the borrower and a 1% fee from the lender plus 25bps on each loan bought or sold. This amounts to 300bps+ of volumes arranged with no capital at risk. We see similar economics at the peer group. We see significant value upside for P2P business owners but do not see the model as a significant risk factor to future bank returns.

43 The Telegraph, Crowdfunding could revolutionise lending, says Andrew Haldane; 17 December 2012

Deutsche Bank AG/London Page 77

8 September 2013

Financial Services Retail Bank Strategy

„ The 5.8% reported net return depends on age and diversification of the portfolio, but compares with risk-free 1 year savings rates of about 1.9%44. The 300bps revenue margin for the arranger compares with average bank risk- adjusted margins of 140bps. P2P and Crowdfunding sites provide users with impressive transparency in the counterparties which they are supporting. How conduct risk evolves for the industry will be interesting to see, for example in the UK as the new Financial Conduct Authority beds down. It is interesting to see the UK government backing lending through Funding Circle and MarketInvoice with public sector funds.

„ Exit valuations on the sale of crowdfunding businesses will depend on scale and the credit performance of the loan book. When risk-adjusted returns for retail-driven participants disappoints (as it does when any credit cycle turns down) we see risk of a rapid and significant decline in volumes, valuations and perceived standing of the business amongst a tech-savvy, social-media driven client base. There’s a time and place for everything

Crowd-driven activities are enjoying strong growth driven by advances in technology Crowd-driven funding is and growing acceptance by individuals and businesses of transactions conducted enjoying unprecedented entirely on the web. Social media and online auction technology put P2P business scale growth as advances in within reach for business founders whilst the debt supply deficit is more acute than technology, user acceptance usual given bank deleveraging and de-risking in large parts of the world. We divide / of web-commerce and bank split the crowd-driven industry by differentiating between the nature of “returns” offered to the providers of finance or skills, as shown in Figure 127. deleveraging converge

Figure 127: Making broad sense of the P2P and crowdfunding industry structure

Creative funding Social benefits platforms e.g. Kickstarter, Indiegogo

* P2P (Consumer Return requirement Interest unsecured) * Crowdfunding (SME)

Equity * Crowdfunding

Source: Deutsche Bank

44 www.moneysupermarket.com

Page 78 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Recent examples of available propositions in each category include:

„ Social benefits: Help fund a Nick Scott film “Bad day at the Office”; donate £2 and get a Facebook mention; contribution options rise up to £1,000 or more which is rewarded by Executive Producer credits, digital download of film and soundtrack, poster etc45. See www.kickstarter.com, www.indiegogo.com

„ Interest: Peer-to-peer consumer lending: Borrow £5,000 for two years for 5.8% all-in rate; Lenders are offered a 4.8% return on their invested funds46. See www.zopa.com, www.lendingclub.com, www.prosper.com.

„ Interest: Crowdfunding of SMEs: Businesses with > £100k in turnover and two years of lodged financials in September 2013 could borrow for 7.1% - 9.4% plus 3% fees on 2-3 year loans; Lenders via Funding Circle earn 5.8% net of losses, according to the site47. Total advanced to 5 September 2013: £146.6m.

„ Interest: Crowdfunding of receivables: Businesses factor receivables via MarketInvoice providing borrowers with transparency on identity of the debtor, indicative cost 2.45% p.m. including MarketInvoice’s 99bps fee. Lenders subject to minimum participation of £50,000 and must be HNW or deemed sophisticated investors. Total advanced to 5 September 2013: £65.7m.

„ Equity: Investment in business ventures: Own 9% of “Linguistadores”, an online customised language learning product for a total of E26,00048. See www.symbid.com, www.crowdcube.com, www.seedrs.com. All of the above businesses / web-portals are fascinating in practice. We encourage readers to visit at least some of them to see them in operation.

Focussing on bank-substituting lending

As a “Future of banking” piece we focus here on the disintermediation of bank lending Dreadful real returns on cash highlighted in prospect by the Bank of England’s Andy Haldane above. It’s obvious why savings drive greater investor increasing numbers of savers are interested in this potential means of increasing the interest in crowdfunding at yield on their savings given the compression in returns from low interest rates through this stage of the cycle… the crisis. Returns which are poor in nominal terms (Figure 128) have been consistently negative on a pre-tax real basis since 2008 (Figure 129).

45 http://www.kickstarter.com/projects/1235861071/bad-day-at-the-office?ref=category 46 Uk.zopa.com; 5 September 2013 47 https://www.fundingcircle.com/businesses/cost-of-finance 48 http://www.symbid.com/ideas/2370-linguistadores#prettyPhoto

Deutsche Bank AG/London Page 79

8 September 2013

Financial Services Retail Bank Strategy

Figure 128: Nominal returns to savers have been poor… Figure 129: …and real returns substantially worse

4.00 3.00 3.50 2.00

3.00 1.00

2.50 0.00

2.00 -1.00

1.50 -2.00

1.00 -3.00

0.50 -4.00

0.00 -5.00 Real ,pre-tax money ,pre-tax mkt /ratemoney o/n deposit Real -05 -06 -08 -09 -10 -13 -05 -06 -07 -08 -13 Money rate /bank depositmarket overnight Money y y y y y y y y y y y Jan-04 Jan-04 Jan-05 Jan-05 Jan-06 Jan-07 Jan-09 Jan-10 Jan-11 Jan-06 Jan-07 Jan-08 Jan-09 Jan-08 Jan-12 Jan-10 Jan-13 Jan-11 Jan-12 Jan-13 Sep-04 Sep-04 Sep-05 Sep-06 Sep-08 Sep-09 Sep-10 Sep-05 Sep-06 Sep-07 Sep-08 Sep-07 Sep-11 Sep-09 Sep-12 Sep-10 Sep-11 Sep-12 May-04 May-04 Ma Ma Ma May-07 Ma Ma May-11 Ma Ma Ma May-09 Ma May-12 May-10 Ma May-11 May-12 Ma

US UK EU US UK EU

Source: Haver Source: Haver

Banks’ needs to build capital, re-balance from higher risk lending and reduce wholesale …as bank rebalancing and funding dependence has combined with cyclically low borrower confidence, seeing lower borrower confidence significant declines in corporate debt across much of the world. Some of this quite has seen non-financial normal in an economic downturn (we count six periods of contracting US commercial & corporate credit contracting industrial loans since 1945, with another 11 occasions where growth fell below 10% in the US, EU and UK through from significantly higher levels in Figure 130). the crisis The fact remains that in the US, UK and Europe non-financial corporate loans are / have been contracting during the crisis (Figure 131). Within this there is good evidence that small and medium sizes enterprises (SME) fared more poorly than large corporates, lacking bond market access and with negative consequences for growth49.

Figure 130: US C&I loans over the long run (% YoY) Figure 131: US, UK, EU non-financial corp loans (YoY %)

40% 25% 20% 30% 15%

20% 10% 5% 10% 0% -5% 0% -10%

-10% -15% outstanding (YoY %)

Pvt Non-Financial (YoY loans %) -20% -20%

US CommercialUS & Industrial loans -25% -04 -05 -07 -09 -10 -11 -30% y y y y y y Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Sep-04 Sep-05 Sep-06 Sep-07 Sep-08 Sep-09 Sep-10 Sep-11 Sep-12 Ma Ma May-06 Ma May-08 Ma Ma Ma May-12 May-13 Jul-55 Jul-68 Jul-81 Jul-94 Jul-07 Jan-49 Jan-62 Jan-75 Jan-88 Jan-01 Mar-51 Mar-64 Mar-77 Mar-90 Mar-03 Sep-57 Nov-59 Sep-70 Nov-72 Sep-83 Nov-85 Sep-96 Nov-98 Sep-09 Nov-11 May-53 May-66 May-79 May-92 May-05 US UK EU Source: Haver Source: Haver

49 Analysed in detail 15 Mar 13 UK SME lending - Reviewing the demand/supply debate ahead of next week's Budget

Page 80 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

How peer-to-peer and crowdfunding works Case study: The Funding Circle Funding Circle was founded in the UK in August 2010. Investors sign up to view credit Lenders on Funding Circle profiles and financial information of businesses wishing to borrow funds. Credits are can see the financial profile categorised by Funding Circle’s eight assessors as between A+ (Very low risk) and C and credit need of borrowers, (Average risk). Investors may conduct due diligence on the borrows – including open and bid to lend money at a bulletin-board discussions with peers – before bidding to lend money, by specifying the rate they specify rate at which they will lend.

Figure 132: 7-9% gross interest last 100 loans (13 May Figure 133: The view from the borrower 13) 16.0% Mean Min Max 13.9% 14.0% Loan amount 35,078 5,000 75,000 12.0% 10.2% Number of lenders 418 72 872 10.0% 9.3% 9.3% 8.1% 8.3% Interest rate 8.0% 6.5% 10.0% 8.0% 7.2% 7.4% Max interest accepted 8.5% 6.7% 10.6% 6.0% 4.0% 4.0% 4.0% 4.0% Min interest accepted 5.5% 4.0% 8.8% 4.0% Loan term (months) 33 12 36 2.0% Rates on last on last loansRates on 13 100 2013 May 0.0% A+ (Very low risk) A (Low risk) B (Below average C (Average risk) risk)

Minimum Average Maximum

Source: Company data (https://www.fundingcircle.com/statistics) 50 Source: NESTA

Borrowers may access between £5000 and £1m for between six months and five years Borrowers may apply for and must be LLPs or incorporated businesses, have turnover above £100,000 a year, between £5,000 and £1m in and have lodged at least two years of financial information with Companies House. This debt up to 5 year duration and other pertinent credit quality data are provided to Funding Circle’s credit analysts provided accepted by from Experian. Once accepted by Funding Circle an auction for the funds goes live. The Funding Circle, and subject to borrower may accept the cost of the loan as soon as the required funding is committed or may leave the auction open for two weeks to see if the cost of the loan declines as £100k turnover requirement new bidders emerge willing to provide funds at a lower cost. and 2 years of financial information at Companies Expected return, diversification and other surprising information House Borrowers are achieving extraordinary diversity of funding with loans backed by over Individual exposures are 400 lenders each in the NESTA survey data (Figure 133). In that data set, the median modest at ~£60/borrower lender loaned £2,000 to 35 companies, a rough average of £60 per loan. The return with £2, 000 lent; diversified track records of investors with more diversified portfolios has been better (Figure 134). portfolios are doing better

Overall, investors are averaging a return of 6.2% (May 2013) net of 1.2% in bad (which Investors are earning 6.2% is itself substantially below the 3.9% loan loss the firm cautions investors to expect) and after bad debts and after the 1% p.a. fee charged. Part of this bad debt outperformance we believe is down management fees, partly due to seasoning (cumulatively only 24% of loans lent have been repaid). Inclusive of annual to loan loss outperformance management cost, one trade per loan and current estimated loan loss rates we relative to FL’s guidance – calculate net returns of between 5.3% and 7.4% depending on risk of borrower. On our numbers, risk-adjusted, A+ (Very low risk) was outperforming A (low risk) (Figure 135). some of this will erode we believe as the portfolio ages

50 Banking on each other: Peer-to-peer lending to business: Evidence from Funding Circle; Based on information on 57 loans participating in a NESTA survey

Deutsche Bank AG/London Page 81

8 September 2013

Financial Services Retail Bank Strategy

Figure 134: Portfolio diversification is important Figure 135: DB estimates of net lender returns

60% Average Running Trading Net

49% 50% rate cost cost Bad debts annual A+ (Very low risk) 7.2% -1% -0.25% -0.4% 5.6% 40% A (Low risk) 7.4% -1% -0.25% -0.9% 5.3% 30% 30% 28% B (Below average risk) 8.3% -1% -0.25% -1.1% 6.0% 23% 22% 22% 20% 20% C (Average risk) 9.3% -1% -0.25% -0.6% 7.4% % of borrowers% 20% 16% 14%13%

7% 8% 10% 6% 6% 6% 4% 5% 0%1% 1% 0% < 0% 0 - 2% 2-4% 4 - 6% 6 - 8% 8 - 10% > 10%

Max exposure 1% Max exposure 2% Max exposure 10%

Source: Company data to 9 March 2013, downloaded 13 May 2013 Source: Deutsche Bank estimates

The volume of available information on for lenders is impressive with members able to Transparency in see all credit performance and yield information, and credit profiles and statistics of crowdfunding is impressive borrowers. Whether borrowers are using this information to make informed decisions, and commendable… and whether the investing community would react reasonably to a deterioration in credit performance is harder to say.

The NESTA survey data showed that 86% of lenders have post-school qualifications …with post-school and that 9% of their financial wealth (savings plus investments) was deployed in this qualifications and portfolio fashion which suggests some sophistication and diversification. diversification in evidence…

But the survey also showed an average 15 minutes of due diligence per loan (probably …but survey evidence symptomatic of the small amounts lent in each case) and that 45% of lenders use the suggests modest due ‘autobid’ tool to lend. This function allows investors to specify the interest rates at diligence and use of autobid which she will lend in each risk bucket and the amounts available for investment in implies reliance on Funding each bucket, with the system automatically bidding at these rates in auctions which are Circle’s credit gating and held. In this regard investors are perhaps relying on the P2P / Crowdfunding hosts’ measurement and classification of relative credit risk. scoring…

With a very web- and social media-savvy clientele (Funding Circle hosts open …with web-savvy clients discussion boards on which investors can discuss their experience with the service and likely in our view to voice with the loans they have made) we naturally see business volumes as vulnerable to a discontent if credit downturn in performance. performance disappoints

Will this ever meaningfully usurp bank lending? Technology, interest rates and bank deleveraging make this the perfect time to launch This is the perfect time to crowdfunding platforms. If current low rate conditions remain in place for an extended launch crowdfunding period (and we think they will have to stay low – see Figure 9) then we think scale and platforms… profitable P2P and crowdfunding platforms will emerge. Government plans to circumvent/bolster the monetary transmission mechanism will help in some countries (£100m of the UK government’s Business Bank £1bn in government funding for business has been earmarked for crowdfunding platforms including MarketInvoice and The Funding Circle).

We expect some crowdfunding platforms to achieve profitable exits for their owners. …which have far superior As shown above, platform revenue runs at about 3% of credit advanced in the case of economics to banks with no The Funding Circle. This compares with an average net interest margin of 1.9% in 2012 credit risk, wider revenue for the banks we cover globally – with credit losses reducing this yield by 0.5% before margins and lower capital fee income and before considering banks’ more costly capital and regulatory and regulatory requirements requirements. Platform owners do not run credit risk and clearly web-lead

Page 82 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

crowdfunding has no need of physical branch infrastructure and so should be able to achieve suitable scale at reasonable cost.

Figure 136: The revenue model for crowdfunding Figure 137: Crowdfunding scale modest in market terms compares very well with conventional banking

6.0 1,600 5.4 1,350 5.0 1,400 1,200 4.0 3.7 3.4 1,000 2.8 2.8 3.0 2.6 800 685 1.9 2.0 1.5 1.6 600 1.0 0.8 0.9 1.0 0.6

(2012, DB universe)(2012, DB 0.5 0.5 400 0.2 146 - 200 65 Net interest margin, loan losses/loans

Loans extended / Outstanding (£'m) extended / Outstanding Loans 0

Net Interest Margin (2012, %) Loan losses / Loans (2012)

Source: Deutsche Bank estimates Source: Company data, as per websites in Sept 2013, Haver; Note: Crowdfunding platforms disclose cumulative lending not credit outstanding, overstating market share in the above.

Though the solid economics of being a successful crowdfunding platform seem and though we don’t expect straightforward – until and unless credit quality or customer malfeasance intervene – their scale to threaten bank we do not expect them to change the market rate for bank credit services or capture a pricing or loan growth we do material part of the overall market (Figure 137). expect some to achieve very profitable exits

Deutsche Bank AG/London Page 83

8 September 2013

Financial Services Retail Bank Strategy

Payments: plastic up, cash flat, cheques out

Key points

In this section we look at how payment systems within the financial sector work, and the key global trends seen over the past half-decade. We explore the rise of plastic card payments across the world, and show how a typical card transaction works.

„ Non-cash transactions are growing fast. Transaction volumes rose from 209 billion to 271 billion over 2007 to 2011 (30%) in the main global economies. Based on the 8% growth rate recorded in 2011, we estimate transactions would have reached 316 billion by the end of 2013. The main driver of this growth is higher card transaction volumes, which now are 55% of non-cash volumes. This growth has come almost entirely at the expense of cheques, which now represent 1 in 8 payment transactions globally vs. 1 in 5 in 2007. As a consequence, values transferred per transaction have been falling.

„ For a typical credit card transaction we estimate the following revenue breakdown: customer pays 19% net (credit card interest and fees + retailer fees), retailer pays 1% net (merchant acquirer fee less customer transaction fee), merchant acquirer earns 0.4% net (retailer fee less interchange + payment network fees), payment network earns 0.2% net (via service/data fees), and card issuer earns 12.3% (before funding costs but after service fees, customer income, fraud, impairments). This is stylised only, complexity of fee pricing makes almost every transaction different.

„ The trends we are seeing in card transactions underline our Buy recommendations on Mastercard & Visa. We also think transaction growth should support banks which participate in card payments as issuers or acquirers. Though regulatory changes in Europe could represent a risk for issuers, we expect revenues to be recouped elsewhere in the payment model. Payments: central to all banking and finance

The act of transferring, exchanging and transacting goods and money are fundamental Exchanges of value are as old activities within an economy. Whereas in older or less sophisticated economies this as commerce, but the rise of might have taken place predominantly via bartering or cash, today’s developed global technology now sees 98% of economies (and banking sectors) are now built on foundations of electronic payments payments in the UK take and settlement systems. According to the Bank of England, today in the UK 98% of place electronically payments by value are electronic.51 This inexorable shift towards electronic payments has had many consequences: central and commercial banks now manage their liquidity balances in real time, businesses can transact with suppliers and customers more efficiently, and consumers can pay for products easily and quickly from multiple vendors, channels and in different countries and currencies.

The central role of payments systems today makes them both a vital economic and financial infrastructure from a security and operational risk standpoint. But payments are also the key medium through which consumers and businesses interact with the financial sector.

51 http://www.bankofengland.co.uk/markets/Documents/paymentsystems/boesettlementaccounts.pdf

Page 84 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

What is a payment system?

There are a multitude of payments systems in the financial sector. It is important to There are a multitude of distinguish between different payment systems types and objectives; the balancing of payments systems, which we central bank liquidity (e.g. via Target2), for example, clearly has a different operational group into five broad types and commercial aim than the processing of retail payments (e.g. Visa / Mastercard).

We see 5 broad types of payment systems „ Cheques / paper credit systems

„ Electronic transfer (direct debit, credit transfer, e-money payments)

„ Card payment systems (credit, debit, contactless, online)

„ Cash (banknotes, ATMs)

„ Large Value Payment Systems (e.g. CHAPS, TARGET2) We summarize these payments methods in Figure 138. Dotted lines indicate where these institutions generate revenues / charge margins.

Figure 138: Payments methods in a box

Cash • Cheques • Credit transfer • Electronic transfer • Large value payments • Direct debits Payment user Bank Payment system Bank Payment user

Card network

Merchant Credit / debit cards acquirer

Source: Deutsche Bank

The last 2 on this list are somewhat distortive to the overall payments picture:

„ By its nature (and one of its advantages), cash transaction volumes and values are more difficult to track than electronic.

„ Large Value Payment Systems also distort the overall picture, given that many of the payments through this system are for wholesale financial use, typically of significant value but very limited volume (in the UK CHAPS represents 0.20% of transaction volumes, but 92% of values). We analyse these areas separately from other non-cash transactions where we have a more consistent and longer-term dataset. First we put these non-cash transactions in context by volume, value and value per transaction.

Deutsche Bank AG/London Page 85

8 September 2013

Financial Services Retail Bank Strategy

How payments profiles have changed: non-cash is growing fast, transaction volumes rising

Transaction volumes: increasingly dominated by cards Non-cash transactions have risen 30% in the past 4 years across the world. For the 23 Non-cash is growing fast – up countries included in the BIS Committee on Payment and Settlement Systems 30% in the last 4 years (hereafter CPSS) study of payments systems52, transaction volumes rose from 209 billion to 271 billion over 2007 to 2011 (Figure 139). Based on the 8% growth rate recorded in 2011, we estimate transactions would have reached 316 billion by the end of 2013.

But the growth in transactions has not been uniform across payment types (Figure 140).

Figure 139: Global non-cash transactions are up 30% in 4 Figure 140: ...and are increasingly dominated by cards years... 400 9.0 300 8.0 7.3 8.0 350 250 7.0 Other card 5.8 5.9 6.0 300 6.0 200 Credit card 271 250 5.0 Debit card

250 236 Billions oftransactions 150 221 E-money payments 209 4.0 Cheques 200 3.0 100 Direct debits Billions of transactions

2.0 %increase in transactions 150 50 Credit transfers 1.0 100 0.0 0 2007 2008 2009 2010 2011 2007 2008 2009 2010 2011

Source: BIS, Deutsche Bank estimates Source: BIS

As we show below, some 55% of non-cash transaction volumes are now processed by Cards are now 55% of card, up from 49% in 2007. This growth has come almost entirely at the expense of volumes from 49% in 07, cheques, which now represent 1 in 8 payment transactions globally vs. 1 in 5 in 2007. replacing cheques…

Figure 141: 2007 Figure 142: 2011 Figure 143: Legend 6% 5% Credit transfers 17% 17%

18% 17% Direct debits Cheques

14% 14% E-money payments Debit card Credit card 25% 12% Other card 19% 33% 1% 2% Source: BIS data, Deutsche Bank analysis. Adjusted for breaks in Source: BIS data, Deutsche Bank analysis. Adjusted for breaks in Source: BIS data, Deutsche Bank analysis time series time series

52 Countries on the BIS Committee on Payment and Settlement Systems: Australia, Belgium, Brazil, Canada, China, France, Germany, Hong Kong SAR3, India, Italy, Japan, Korea, Mexico, Netherlands, Russia, Saudi Arabia, Singapore, South Africa, Sweden, Switzerland, Turkey, United Kingdom, United States

Page 86 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

So the main driver of growth has been a move to plastic, and within this – debit card …with debit cards the real transactions, up from 53 billion to 88 billion over the period. area of strong growth

Figure 144: Growth rates in types of transaction Figure 145: Closer look at plastic trends 25.0% 160 20.0% 140

15.0% 120 Credit transfers 100 10.0% Direct debits Other card 80

Cheques Billions of transactions 5.0% Credit card Debit card 60 Debit card 0.0% Credit card 40 -5.0% 20

-10.0% 0 2008 2009 2010 2011 2007 2008 2009 2010 2011 Source: BIS data, Deutsche Bank analysis. Adjusted for breaks in time series Source: BIS data, Deutsche Bank analysis. Adjusted for breaks in time series

Transaction values: dominated by credit transfers In terms of non-cash transaction values, credit transfers dominate, with c.77% of the Despite their important role in values. This is down slightly from the 80% recorded in 2007, due to the growth in direct retail, card payments are less debits, which now have overtaken cheques in terms of value transferred. Card than 2% of non-cash payments represent less than 2% of total payment values. Overall transaction values are payments; credit transfers are lower (over the dataset we have), but this is driven entirely by lower credit transfers and the largest system by far somewhat distorted by FX changes.

Figure 146: Transaction values across non-cash Figure 147: Transaction values over time payments 600 10% 500 Credit transfers Other card 11% 400 Credit card Direct debits Debit card Cheques 300 E-money payments E-money payments Trillion USD transactions Cheques Debit card 200 Direct debits Credit card 100 Credit transfers Other card 77% 0 2007 2008 2009 2010 2011

Source: BIS data, Deutsche Bank analysis. Adjusted for breaks in time series Source: BIS data, Deutsche Bank analysis. Adjusted for breaks in time series

Average value per transaction: have fallen over the past 4 years Unsurprisingly, credit transfers have the highest average value per transaction, 5-6x the Average transaction values size of average direct debit and cheque payments. The average debit card transaction is are falling as volumes E50, and credit card is E89. With the exception of direct debits, the transaction values continues to rise and mix of all payment types have fallen over the past 4 years. shifts to card commerce

Deutsche Bank AG/London Page 87

8 September 2013

Financial Services Retail Bank Strategy

Figure 148: Average payment value per transaction Figure 149: Change in payment values over time

9,000 5% 7,855 8,000 0% 0%

7,000 -5% -5% -7% 6,000 -10% Credit transfers -12% 5,000 -15% Direct debits 4,000 -20% Cheques 3,000 -25% Debit card 2,000 1,553 1,314 Credit card

Average USD transaction value transaction USD Average -30% 1,000 9 50 89 36 -35% 0 -37% Credit Direct Cheques E-money Debit card Credit card Other card -40% transfers debits payments 2007 2008 2009 2010 2011

Source: Deutsche Bank estimates, BIS data Source: Deutsche Bank estimates, BIS data

Rise of the card: A bankable trend

Fantastic Plastic has been around since the 1970s Plastic cards have been used for many years in markets like the US and UK. In the UK Plastic cards first issued in the the first plastic cards were introduced in 1972. By 2010 200 million were in circulation, 70’s; there are now more about 4 per person in the country. Transaction volumes and values have undergone a than 4 for each UK resident… similar increase over the past 30-40 years (Figure 150).

Figure 150: Number of cards in the UK 1972-2010 Figure 151: Increase in volumes and values in UK

250,000 90 3,000 80 2,500 200,000 70 60 2,000 150,000 50 1,500 40 100,000 30 1,000 Thousands of cards 20 50,000 500 10

0 0 0 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Credit cards Debit cards Other cards Transactions / card (LHS) Value / transaction (LHS) Value / card (RHS)

Source: Deutsche Bank analysis, BBA data Source: Deutsche Bank analysis, BBA data

4 key recent trends in plastic cards: 1. Growth has continued, particularly in EM: Data from the UK above might suggest a …with cards in issue still stagnation in cards in recent years, but globally card volumes and transactions growing strongly in EM… continue to gain momentum (Figure 152, Figure 153). As at 2011 there was around the same number of payment cards in circulation as the global population.

Page 88 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 152: Volume of cards in circulation Figure 153: Volume of transactions

7 160 150 6.2 6 5.5 140 132

Billions 5.2 5.3 5.1 4.9 119 5 4.6 4.4 4.5 120 109 4.0 99 4 100

3 80 Billions of transactions

2 60

1 40

0 20 2007 2008 2009 2010 2011 0 Cards with a cash function Cards with a payment function (debit/credit) 2007 2008 2009 2010 2011

Source: Deutsche Bank analysis, BIS data. Adjusted for series breaks Source: Deutsche Bank analysis, BIS data

By country, the number of cards in the traditional card-using economies (US, UK, …with penetration catching NL) has fallen since 2011, most likely due to saturation of card providers (in 2007 up quickly in the BRICS in each US inhabitant had an average of 3 cards with a cash function, or 5 with a particular payment function). These countries remain amongst the highest in terms of card penetration. However, emerging markets have seen significant and rapid growth over the period – particularly in the BRICS (Figure 154).

Figure 154: Change in cards in circulation 2007-2011 (cards with cash function) Millions of cards 2007 2008 2009 2010 2011 Change Cards per 2007-2011 inhabitant United States 982 952 855 795 805 -18% 2.58 United Kingdom 165 168 162 165 165 0% 2.61 France 92 94 95 96 93 1% 1.42 Germany 124 126 130 130 133 7% 1.62 Australia 51 55 58 58 58 14% 2.61 Singapore 9 10 9 10 10 14% 1.97 Belgium 17 19 19 19 20 14% 1.82 Sweden 10 11 11 11 11 16% 1.21 Switzerland 11 12 13 13 14 25% 1.78 Saudi Arabia 11 12 14 12 14 28% 0.50 Italy 40 47 45 50 53 32% 0.88 Turkey 34 38 39 41 45 33% 0.60 Mexico 77 82 83 98 110 44% 1.01 Brazil 195 223 237 295 325 66% 1.67 Russia 103 119 126 144 200 93% 1.40 China 1,499 1,800 2,066 2,415 2,949 97% 2.19 India 130 162 201 246 296 128% 0.25 Total cards 3,551 3,930 4,162 4,600 5,303 49% 1.44 Source: Deutsche Bank analysis, BIS data, cards with a cash function

Plastic as a payment mechanism is global, which is important when considering the global networks which underpin the plastic card business model.

2. Spending on cards has increased: Consumers are spending more on plastic than Card spend is up even where they used to. US card numbers in circulation may have fallen slightly, but consumer card numbers are down: US spend on plastic has increased by c.20% in 2007-2011. We show the USD spent on values up 20% 07-11

Deutsche Bank AG/London Page 89

8 September 2013

Financial Services Retail Bank Strategy

cards per inhabitant in 2011 in Figure 155. EM remains low compared with traditional card using economies (Sweden, UK, US, Canada, Australia).

Figure 155: USD spent on cards per inhabitant, 2011 Figure 156: Number of card transactions per inhabitant

20,000 250 18,000 16,000 200 14,000 12,000 150 10,000 8,000 100 6,000 4,000 50

USD spent per inhabident per spent USD 2,000 0 0 Card transactions per inhabitent transactions Card Italy Italy India India Brazil Brazil China China Korea Korea CPSS1 Russia Russia France France Turkey Turkey Mexico Mexico Canada Canada Sweden Sweden Belgium Belgium Average Australia Australia Germany Germany Singapore Singapore Switzerland Switzerland South Africa South Africa Netherlands Netherlands Saudi Arabia Saudi Saudi Arabia Saudi United States United States United Kingdom United United Kingdom United

Source: Deutsche Bank analysis, BIS data Source: Deutsche Bank analysis, BIS data

3. Debit cards have grown significantly, credit cards have peaked: When we break Debit cards are where the down the increase in cards over the past few years, it is clear that the growth has growth is, up 40% 07-11, been driven by debit rather than credit cards. Debit cards are up c.40% during whilst credit cards are down; 2007-2011, whilst credit cards in circulation have fallen from 2.3 billion to 1.8 it’s the mechanism not the billion. On a per inhabitant basis, the trends are similar – cards with a cash function finance that’s most attractive or a debit function have been growing. This is an important trend as it is evidence of the change in business model for cards from a credit-customer to a payment- in cards retailer model, which we discuss in more detail below.

Figure 157: Debit and credit cards in circulation Figure 158: Number of cards per inhabitant

5 1.8 4.3 4.5 4.1 1.6 4 3.7 1.4 3.4 1.2 3.5 3.1 1 2007 3 0.8 2008 2.5 2.3 2.3 2.2 Debit cards 0.6 2009 2 1.7 1.8 Credit cards 0.4 2010 Billions of cards in circulation 1.5 0.2 2011 1 0 0.5 Cards with a Cards with an Cards with a of which: of which: cash function e-money payment cards with a cards with a 0 function function debit credit 2007 2008 2009 2010 2011 function function

Source: Deutsche Bank analysis, BIS data Source: Deutsche Bank analysis, BIS data

4. Values per transaction have fallen in real terms: Though transaction volumes are up Transaction values are falling (50% globally 2007-2011, Figure 159; and in the UK Figure 160), the value per in real terms – UK credit card transaction has remained relatively stable on a nominal basis, falling in real terms. average £64, debit card £44; In the UK the average credit card transaction is c.£64, the average debit card As sizes fall, ease of use transaction in c.£44; and though this moves seasonally (esp around Christmas), it becomes a greater issue for hasn’t really changed for the past 10 years. Globally, debit card average transactions were at $48 vs $52 in 2007, and for credit card cards $90 in 2011 vs customers, efficiency a $89 in 2007. The transactions that consumers are using cards for are getting greater focus for providers smaller. This again has implications for the payments business model – volumes, efficiency and ease of payment have become more of a focus.

Page 90 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 159: Transaction volumes per card, UK Figure 160: Average £ spent per transaction

90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

Credit card transactions per card Debit card transactions per card Credit cards Debit cards

Source: Deutsche Bank analysis, BBA data Source: Deutsche Bank analysis, Payment Council state

The traditional card payment model

In Figure 161 below we show the steps which take place between the various stakeholders when a customer makes a purchase from a retailer.

Figure 161: How the payment process works for cards

1 Payment card / method presented

Customer Retailer 9 Goods / services

Funds Monthly Transaction deducted from credit card put into 10 11 debit account / bill / reward physical or Price + Interest 2 points mobile POS + Fees paid system

Payments Card issuer Service Provider

POS sends transaction details Price minus Transaction to Merchant Authenticates Merchant details sent for 8 3 Acquirer via a 5 6 transaction Discount fee authorisation Payments + interchange fee Processor

7 Authentication sent, funds cleared Payment Merchant Network Acquirer Transaction details sent 4 to the payment network + Interchange fee

Source: Deutsche Bank

The payment process typically works in the following way:

1. Upon paying for goods, the card is presented to the retailer.

2. The transaction details are entered or loaded onto a Point-of-Sale (PoS) either physically (via a terminal), mobile (mPoP, using software on an iPad for example), or online (online PoS).

Deutsche Bank AG/London Page 91

8 September 2013

Financial Services Retail Bank Strategy

3. A Payments Service Provider (PSP) securely transmits this data to the Merchant Acquirer

4. The merchant acquirer (the retailer’s card agent) receives the transaction, then sent on over a payment network/association (with interchange fee, more details later).

5. The Payment Network / Association (Visa / Mastercard / American Express) then routes the transaction to the card issuer.

6. The card issuer authenticates (or not) the transaction (e.g. PIN correct, credit limit not breached, funds available), sending authorisation via the payment network.

7. This is sent to the Merchant Acquirer. Funds are also settled/cleared via the payment network (though technically this doesn’t happen immediately).

8. The Merchant Acquirer then sends on the transaction to the retailer via the PSP, less the merchant discount fee for using the service.

9. The customer receives the goods / service.

10. The customer is billed for payment (immediately for debit cards, typically a month afterwards for credit cards).

11. The cardholder pays the full amount for the goods / services with interest (if not paid off immediately), and any fees incurred (setup, issue, foreign exchange, payment protection insurance, lost card insurance, late payments, administration).

How fees flow around the card model

Another way of looking at this is to show how provider revenues flow around the Another way to look at how model. The fee structure for cards is complex and non-standardised, reflective of the the card transaction model many different types of cards (debit, credit, store, loyalty), customers (e.g. poor credit works is to track the revenue history, abroad) and retailers (online, large-scale, airlines). There is no ‘one size’ model generated by the card issuers, for plastic payments or what revenue % is picked up at each point. payment network and We show a schematic of how revenues and fees typically flow around the model in merchant acquirers – the key Figure 162 below and describe the roles of each of the stakeholders below. In the links in the chain traditional card payment processing model, the key stakeholders are the card issuers, the payment network, and the merchant acquirers.

Page 92 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Figure 162: How fees, revenues and costs work within the payment system

Card transaction fee (often)

Customer Retailer

Non-$ Interest + benefits PSP fee / POS Fees paid (airmiles rental etc)

Fraud costs Payments Credit losses Card issuer Funding costs Service Provider

Service fees + data Interchange Merchant discount fee processing fee fee

Payment Merchant Network Acquirer Interchange fees + service fees + data processing fee

Source: Deutsche Bank

Customer: pays anything from 0% to 20% „ Benefits: The customer is incentivised to use a card for a number of reasons, including convenience (ease of use, ability to track spending), security (typically credit cards provide consumer insurance), credit, or other benefits (e.g. airmiles, cashback, discounts).

„ Costs: The customer pays for these benefits through fees / interest paid to the card issuer (particularly credit cards), or through a transaction fees paid to the retailer. For many retailers now transaction fees are relatively rare (retailers normally want customers to use cards, more below), though for airlines (2% fee on Ryanair flight bookings53), and smaller independent retailers charges remain. Over time, we note that the proportion of balances which pay interest in the UK has fallen from 80%+ to 60% (Figure 163). This has been accompanied by higher credit card APRs (Figure 164), meaning the overall rate on balances has remained in the 10-12% range.

„ Example: Customer pays 1% transaction fee; credit card APRs are at 17.9% in the UK, so a net charge of c.19% if the credit card user doesn’t pay off every month, or 1% charge overall if a debit card or credit card which pays off balances each month.

53 http://www.ryanair.com/en/terms-and-conditions

Deutsche Bank AG/London Page 93

8 September 2013

Financial Services Retail Bank Strategy

Figure 163: Credit card APRs UK Figure 164: Proportion of interest-bearing card balances (UK) 24.5 85%

22.5 80%

75% 20.5 70% 18.5 17.9 % 65% 16.5 60% 60%

Annual Percentage Rate (%) Rate Percentage Annual 14.5 55% Proption of balances interest bearing interest balances of Proption 12.5 50% Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Q1-03 Q3-03 Q1-04 Q3-04 Q1-05 Q3-05 Q1-06 Q3-06 Q1-07 Q3-07 Q1-08 Q3-08 Q1-09 Q3-09 Q1-10 Q3-10 Q1-11 Q3-11 Q1-12 Q3-12 Q1-13

Source: Haver Source: Haver

Retailer: „ Benefits: The retailer benefits from (normally) faster, more efficient payments (less cash in a till required), and potentially wider customer base. For online retailers, card payments are integral to the business model.

„ Costs: the retailer pays fees to the Merchant Acquirer and the Payment Processor for being able to accept card payments. Charges also accrue for renting the terminal equipment (this can be rented separately from the manufacturer), connectivity to the card network, customer support and payment guarantee. Example: Barclaycard Business advertises that businesses can pay ‘from 1.5%’ on card transactions, with terminal rental/online management fees ‘from £10 a month’. We assume c.2% Merchant Acquirer Fees, less 1% transaction charge passed on to the customer. Point of Sale terminal provider „ Generates revenues by selling or leasing the terminal hardware/software which card transactions are processed on. Example companies here would include Verifone (Figure 165) and Ingenico.

Figure 165: Verifone Snapshot Figure 166: Global Payments Inc Snapshot Verifone Analyst name Email Global Analyst name Email Bryan Keane [email protected] Bryan Keane [email protected] Payments Inc. Spot price: 20.7 Tgt. Price 13 Rec. Sell Spot price: 47.8 Tgt. Price 55 Rec. Buy 2010 2011 2012 2013e 2014e 2010 2011 2012 2013e 2014e 2015e EPS (DB adjusted) 1.3 1.9 2.7 1.5 1.8 EPS (DB adjusted) 2.8 3.1 3.5 3.6 4.0 4.3 DPS 0.0 0.0 0.0 0.0 0.0 DPS 0.1 0.1 0.1 0.1 0.1 0.1 BVPS 2.3 12.4 11.9 13.1 18.4 BVPS 10.5 14.7 14.7 15.6 19.9 23.9 ROE 83% 24% 8% 2% 0% ROE 27% 24% 24% 24% 22% 20% EBITDA mgn 16% 13% 22% 20% 19% EBITDA mgn 24% 23% 23% 22% 21% 21% Market Cap (US$'mn) 1,809 4,096 4,398 2,293 2,293 Market Cap (US$'mn) 3,699 3,543 3,716 3,741 3,741 3,741

P/E (DB) 15.6 22.1 14.6 14.3 11.4 P/E (DB) 16.1 14.3 13.3 13.1 12.1 11.0 EV/EBITDA 11.2 24.1 12.9 9.1 7.2 EV/EBITDA 9.0 7.2 7.4 8.2 7.2 6.1 Div yield 0% 0% 0% 0% 0% Div yield 0% 0% 0% 0% 0% 0% FCF yield 8% 4% 4% 6% 8% FCF yield 7% 9% 7% 6% 9% 10% ROE 83% 24% 8% 2% 0% ROE 27% 24% 24% 24% 22% 20%

Source: Deutsche Bank estimates, company data, Priced as on 5 September 2013, All figures in local Source: Deutsche Bank estimates, company data, Priced as on 5 September 2013, All figures in local currency unless stated currency unless stated

Payments Service Provider „ Generates revenue by acting as the processor of card transactions, through ‘front- end’ (authentication, directing traffic to card networks / merchant acquirers / issuers) or ‘back-end’ (settlement, clearing). Example companies would be Global Payments Inc (Figure 166) or WorldPay.

Page 94 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Merchant Acquirer „ Revenues: Fees from retail merchants, providing incentive to build a large base of retailer customers for economies of scale.

„ Costs: aside from the cost of upkeep of equipment / network / customer service, the main charges a merchant acquirer faces are service and data processing fees paid to the payment network. These are typically calculated as a % of transaction value (service fee) and per transaction (data processing). Also, an interchange fee is paid to the card issuer and covers the credit risk of the product/service paid for and varies according to card, transaction and location (e.g. ‘customer not present’ transactions are more expensive). We show the charge structure of VISA UK interchange fees in Figure 167 (it’s complex).

Figure 167: UK Visa Interchange fee breakdown Immediate Credit & Charge Card Debit Deferred Debit Contactless (£0.00 - £2.00) £0.01 0.65% 0.65% Contactless (£2.01 - £10.00) £0.04 0.65% 0.65% Contactless (£10.01 - £15.00) £0.08 0.65% 0.65% Contactless (£15.01 - £20.00) £0.08 0.77% 1.02% Parking (£0.00 - £35.00) £0.08 0.77% 1.02% Vending (£0.00 - £15.00) £0.08 0.77% 1.02% EMV Chip £0.08 0.77% 1.02% Card Not Present - CVV2 £0.11 1.10% 1.35% Recurring Transaction £0.11 0.65% 0.90% Secure Electronic Commerce £0.10 0.87% 1.12% UK V.me by Visa £0.10 0.87% 1.12% ECommerce - Low Value Payment (£0.00 - £2.00) £0.12 n/a n/a Airline - Chip terminal n/a 1.00% 1.25% Airline n/a 1.10% 1.35% Card Not Present (CNP) £0.18 1.30% 1.55% Standard / Non-Electronic £0.18 1.30% 1.55%

Standard Refund Interchange Fees Refund - Card Present £0.09 0.80% 1.05% Refund - MOTO and CNP £0.12 1.10% 1.35% Refund - E-Commerce £0.15 1.10% 1.35% Source: VISA Europe, http://www.visaeurope.com/en/about_us/our_business/fees_and_interchange.aspx

Payment Network „ Revenues: Service + data processing fees from merchant acquirers, card issuers. Visibility on a huge amount of transactional data (more on ‘Big Data’ later). We show a breakdown of Visa’s fees in Figure 168: around a third of revenues come from data processing, a quarter from international fees, and the bulk of the remainder from service fees charged to the merchant acquirers and card issuers.

„ Costs: Upkeep and expansion of the payments system (people + technology), brand marketing costs (e.g. Visa at the Olympics), and incentives which are effectively ‘negative’ revenue as they are rebates / incentives given to the payment networks’ customers. See Figure 169.

Deutsche Bank AG/London Page 95

8 September 2013

Financial Services Retail Bank Strategy

Figure 168: Visa breakdown of revenues, 2012 Figure 169: Visa breakdown of costs/incentives, 2012 Other revenues, Professional / Administrative / 704, 5% consulting fees, other, 453, 8% 385, 6% International Volume and transaction support fees, 3,025, incentives, Service Fees, Advertising / 24% 2,155, 36% 4,872, 39% marketing, 873, 14%

Network & communication s, 414, 7% Data processing fees, 3,975, Personnel, 32% 1,726, 29%

Source: Company data Source: Company data

We show snapshots of the two major payment networks in Figure 170 and Figure 171.

Figure 170: MasterCard Snapshot Figure 171: Visa Snapshot MasterCard Analyst name Email Visa Inc. Analyst name Email Bryan Keane [email protected] Bryan Keane [email protected] Spot price: 629 Tgt. Price 756 Rec. Buy Spot price: 176 Tgt. Price 226 Rec. Buy 2010 2011 2012 2013e 2014e 2010 2011 2012 2013e 2014e EPS (DB adjusted) 14.0 18.7 22.0 26.1 30.3 EPS (DB adjusted) 3.9 5.0 6.2 7.6 8.8 DPS 0.6 0.6 1.2 1.2 1.2 DPS 0.5 0.7 0.9 1.3 1.3 BVPS 39.7 45.8 55.0 59.6 82.7 BVPS 33.8 37.4 40.7 42.7 52.8 ROE 42% 43% 43% 45% 42% ROE 12% 14% 16% 18% 18% EBITDA mgn 52% 55% 57% 59% 59% EBITDA mgn 60% 63% 63% 64% 64% Market Cap (US$'mn) 30,034 38,283 54,206 76,219 76,219 Market Cap (US$'mn) 59,208 55,385 76,728 115,682 115,682

P/E (DB) 16.3 16.0 19.6 24.1 20.7 P/E (DB) 20.5 15.7 18.3 23.2 19.9 EV/EBITDA 8.9 9.1 11.6 14.5 12.4 EV/EBITDA 11.4 9.1 11.2 14.7 12.7 Div yield 0.3% 0.2% 0.3% 0.2% 0.2% Div yield 1% 1% 1% 1% 1% FCF yield 5% 7% 5% 5% 6% FCF yield 1% 6% 6% 1% 7% ROE 42% 43% 43% 45% 42% ROE 12% 14% 16% 18% 18%

Source: Deutsche Bank estimates, company data, Priced as on 5 September 2013, All figures in local Source: Deutsche Bank estimates, company data, Priced as on 5 September 2013, All figures in local currency unless stated currency unless stated

Card issuer „ Revenues: Interchange fee from Merchant Acquirer. Any monthly fees / charges / interest due from the customer (Figure 163). For more detail on the credit loan model of credit cards, please see our detailed report on retail banking54. If the Card issuer = the merchant acquirer (e.g. Barclaycard in the UK), then the transaction is ‘on us’ and the interchange fee is effectively not paid. Large card issuers therefore have an incentive to be large merchant acquirers to be on both sides of the transaction.

„ Costs: Card issuer typically assumes costs of:

„ Fraudulent activity: in the UK fraud amounted to £388m in 2011, on £500bn of credit card transactions, so c.8bps, Figure 173;

„ Impairments in the event of a customer default: in the UK this is currently running at c.4%, peaked at 10% during the crisis, Figure 172. This is typically the most volatile item in the credit card cycle.

„ Funding costs for credit and operating costs for card issuance / systems / customer service.

54 UK Retail Banking 2012, Past, present and future, 7 Sept 2012

Page 96 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

„ The card issuer also pays the service and data processing charges to the payment network.

Figure 172: Credit card write-offs in UK over time Figure 173: Fraud losses are low

16% 0.18% 0.16% 900 0.15% 14% 0.16% 800 0.14% 700 12% 0.11% 0.12% 600 10% 0.10% 0.09% 500 0.08% 8% 0.07% 0.08% 400 6% 0.06% 300 4% 4.08% £m losses Fraud 3.68% % transaction values 0.04% 200 2% 0.02% 100 0% 0.00% 0 2007 2008 2009 2010 2011 2012 1Q94 1Q95 1Q96 1Q97 1Q98 1Q99 1Q00 1Q01 1Q02 1Q03 1Q04 1Q05 1Q06 1Q07 1Q08 1Q09 1Q10 1Q11 1Q12 4 quarter rolling average Quarterly Fraud losses Fraud as %

Source: Deutsche Bank estimates, Bank of England Source: UK Cards Association data

So in stylised example of fees paid on a £100, interest-bearing, credit card transaction:

„ Customer pays £100 + 1% fee to retailer, 18% to credit card issuer in APR.

„ Retailer pays 2% to merchant acquirer / payment processor

„ Merchant acquirer pays 1.5% in interchange fees to card issuer, 0.1% to payment network

„ Card issuer pays 0.1% to payment network, 0.1% lost due to fraud, 7% lost due to impairments. So, overall: the customer pays 19% net, retailer pays 1% net, the merchant acquirer earns 0.4% net, the payment network earns 0.2% net, and the card issuer earns 12.3% before funding costs. Though, of course, the revenues can vary significantly depending on the type of card, customer, product or transaction that takes place. Large value payment systems – would always be run by banks

The majority of the flow of money around the financial system goes through Large Most money flows by Large Value Payment Systems (LVPS). Although a very small proportion of transactions Value Payments Systems (0.19% in the UK for CHAPS, Figure 174), payments through this mechanism dwarf which will always be bank- other retail and commercial payment systems (Figure 175). LVPS are typically dominated… operated/managed by central banks and used by individual banks for making large value transfers, in real time, on a gross basis (i.e. there is no netting of positions).

Deutsche Bank AG/London Page 97

8 September 2013

Financial Services Retail Bank Strategy

Figure 174: Chaps: just 0.19% of transactions... Figure 175: ...but 92% of transaction values in the financial system CHAPS 0.19% BACS Faster 5% BACS Payments 31% 1%

Debit cards Cheques & 47% Credit 1% CHAPS Faster 92% Debit cards Payments 1% 5% Credit cards Cheques & 0% Credit cards Credit 14% 3%

Source: Deutsche Bank estimates, Payments Council Source: Deutsche Bank estimates, Payments Council

Examples of use: LVPSs do have commercial uses, but their most important role is for enabling the flow of capital around the financial system in an efficient and speedily manner.

„ Wholesale / liquidity / central bank uses: To balance liquidity or treasury positions when intraday repos are opened or closed. Central Banks also use such systems (e.g. in the European Union via the TARGET2 system) to make payments to each other. …generating a fairly significant revenue stream in „ Commercial/Retail use: typically for high value, secure transactions which need to markets – such as the UK – be made immediately, e.g. house purchases and same-day, guaranteed payments. As a proportion of LBPS this represents about 78% of transactions, and 23% of where these payments attract values. We estimate that on the basis of 27 million transactions a year, £25 fee per customers fees of up to £25 transaction, the total fee pool for the UK from CHAPS is ~£675m per transaction

Figure 176: Retail vs Wholesale CHAPS use Figure 177: CHAPS represents around £675m of fees for the UK Banking sector 90% 78% 80% 77%

70% 60% UK 50% Retail Retail transactions per annum 27 million 40% Wholesale Fee per transaction £25 30% 22% 23% 20% Total fees per annum £675m

10% 0% Volume Value

Source: Deutsche Bank estimates, UK Payments Council Source: Deutsche Bank estimates, UK Payments Council

UK example: CHAPS In the UK the large value payment system is called Clearing House Automated The UK LVPS, CHAPS, gives a Payments System or CHAPS, operated by a private company (CHAPS Co) in which the decent case study of how settlement banks are shareholders. The Bank of England acts as settlement agent and these systems work owner of the system’s infrastructure.

Page 98 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

How does it work? CHAPS processes sterling payments in real time, with no restrictions on value of transactions. It is open each working day from 06:00 to 16:00 and for a further 20 minutes afterwards for interbank funding payments.

„ Payment instructions are sent via SWIFT (Society for Worldwide Interbank Financial Telecommunication, a secure messaging system used by financial institutions globally) to the RTGS / CHAPS system run by the BoE.

„ Once payment has been settled (debit / credit in settlement accounts) a confirmation is returned via SWIFT. Who uses it? There are 19 settlement banks which are direct members and shareholders of CHAPS, who co-operatively govern all areas (including rules on entry, pricing, operational procedure). A further 4,500+ other financial institutions make CHAPS payments and settle through agency arrangements with these direct members (e.g. a small bank like Metro Bank will settle CHAPS through one of the direct members, who in turn will settle through the Bank of England).

Figure 178: Large Value Payments System in action

Bank of England

Settlement RTGS infrastructure / Settlement Bank A Settlement accounts Bank C

Participating Participating Settlement Bank A Bank B Bank B

Retail Market (Business & Personal)

Source: Deutsche Bank / OFT

What are the risks? There are no credit risks (during normal operations) as payments are real time, irrevocable and final, and the paying bank has to have sufficient funds in the BoE RTGS settlement account to make a payment.

Liquidity risk is more of a risk (e.g. the order in which these payments are processed/scheduled), and therefore members have access to real-time information on their settlement balances at the BoE.

How often is it used? In July 2013, CHAPS processed 3,156,258 payments worth £6.3 trillion over 23 settlement processing days.

Deutsche Bank AG/London Page 99

8 September 2013

Financial Services Retail Bank Strategy

Regulation – Evolving

European Payments For a thorough overview of European Payment regulations, please see work by our US Technology colleagues, ‘EC Proposal Likely Drawn Out’, 24 July 2013.

„ The Single Euro Payments Area (SEPA) is the payments integration initiative of the EU, and covers 3 payments instruments: credit transfers, direct debits, and card payments. SEPA aims to improve cross-border efficiency and promote a single market for Euro payments.

„ The Payments Services Direct (PSD) is the legal foundation of SEPA. Recent proposals (not final) from the EC covering revisions to the PSD and Interchange fee regulations cover a number of areas, including:

„ Third party payment providers: could be granted access to customer accounts to check balances / initiate transactions. This is aimed at lowering the barriers of entry and increasing competition. However, there are security concerns over this.

„ Legal separation of card schemes & processing entities: this may mean that card networks have to split their processing functions into a separate organisation. There are a number of domestic card networks with a monopoly over domestic transactions e.g. in Italy (Bancomat) and France (Carte Bancaire), so this could be beneficial for Mastercard / Visa if these monopolies are broken.

„ Limiting interchange fees for all consumer debit and credit card transactions to 20bps and 30bps respectively. These will be effective 2 months after the regulation is enforced for cross-border and 2 years for domestic transactions. Interchange fees don’t go to the network - they are paid from the acquirer to the issuer so issuers would be most directly impacted (and issuer-acquirers less so). A report by Europe Economics (commissioned by Mastercard)55 estimates a loss to card issuers to be up to £2.4bn for the UK alone with £2.2bn of savings for large retailers but with “no evidence that these savings are passed on to the consumer in lower prices”. This report says that evidence from other countries where interchange fees have been artificially lowered has been that issuers “seek to recover costs by increasingly prices and reducing rewards and services for consumers”. At the same time we think there could be pressure to lower network fees paid to payment networks from card issuers in order to compensate. The report suggests that cardholder fees could rise by up to £11 for debit cards, and £25 for credit cards – which would clearly have implications for the ‘free banking’ model in the UK. Overall, these proposals have the potential to change the current model

„ There are likely to be positives overall for large payment processors (from levelling of playing field, breaking domestic network monopolies).

„ We think the proposed changes to interchange fees are likely to see issuers attempt recoup any lost revenue from charging consumers more, or negotiating lower payment network fees.

„ Moreover, if interchange fees mean bigger savings for retailers and so attract more retailers to shift to card payments, this should benefit all stakeholders in the card- payments cycle through increased transaction volumes.

55 The Economic Impact of Interchange Fee Regulation in the UK, Europe Economics. http://www.europe- economics.com/userfiles/library/FINAL_Europe_Economics_Media_Exec_summary_280613.pdf

Page 100 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

„ MasterCard management have said that the EC proposal poses no revenue or increased competitive risk; that the proposal provides an opportunity for MA to gain share from the domestic networks as issuers have freedom to co-badge and the card schemes and processing is unbundled; and that legal separation may not require a split-up at the company holding level.56

„ Our US analysts think that the timing of upcoming European Parliament elections and EC commissioners means that the new regulations will likely be delayed; and they think the recommendations could changed materially given the lengthy legislative process. We expect push-back from the UK, particularly given the implications that interchange fee caps could have on ‘free banking’ at present. US Payments For an overview of current US payment regulatory themes, please see work by our US Technology colleagues in ‘Fed Appeal Sets Stage for Year of Uncertainty Around U.S. Debit’, 21 August 2013.

56 MasterCard: ‘Management Comfortable with the EC Proposal Risk’, 25 July 2013

Deutsche Bank AG/London Page 101

8 September 2013

Financial Services Retail Bank Strategy

Cool Innovations

No sustained advantage in product, but we liked these

We don’t believe banks can deliver sustainable competitive advantage in retail banking We don’t believe banks can via product design or technological prowess in most developed markets. Imitation is, deliver sustainable after all, the sincerest form of flattery. We did think the following offerings were competitive advantage via interesting, in alphabetical order of the provider we found (in all cases there are no product or technological doubt others): prowess but we did think these products were cool „ Pay mates by smartphone, fast becoming ubiquitous: Pay to mobile, pay to account, pay to email initiated and managed by smartphone are fast becoming ubiquitous assisted by service providers like fiserv and Monitise. A clear win for customers. Branded Pingit by Barclays, Pay to Mobile by Bank of Ireland, kaching by CBA, nabKiss by NAB, Pay Your Contacts by RBS and included as standard in mobile banking apps at Bank of America and Wells Fargo, amongst others. Press reports57 say Google is trialling a money-by-email product in the US.

„ Talking ATMs for the visually-impaired – Co-Operative Bank, UK: ATMs in the Co-Op network are being upgraded with high contrast screens and spoken feedback delivered via a headset plugged into the unit.

„ Sale and financing of phones, tablets, accessories – FirstRand, South Africa: FNB sells Apple, Samsung, Blackberry, acer, HP and Google phones, tablets and accessories to bank customers. Product pricing and obligations to qualify (volumes of monthly cash deposits, for example) dependent on customer segmentation58. To 1H13, around 12,000 smart-phones and 70,000 tablets had been sold.

„ Remote cheque deposits – JPMorgan Chase, United States: Customer takes a picture of front and back of cheque, mobile banking app transmits picture to the bank which clears it with no branch visit, no postage, no delay. Software solution provided by firms like Jack Henry & Associates and marketed to bank customers under names such as QuickDeposit (JPM). We understand that deploying this in the UK requires changes to bank sector rules: we think this would be a sensible measure to take to reduce transaction costs.

„ GPS- and camera-enabled smartphone househunting and mortgage-lead app – Lloyds Banking Group, UK: Last year LBG launched the Home Finder app under its Halifax brand, powered by properties-for-sale data from Zoopla. The app uses GPS to show properties for sale in the vicinity and, in a neat twist, superimposes those on the market on a live camera heads-up view. Barclays now also has a similar offering.

„ M-PESA – Safaricom, Kenya: Mobile person-to-person and person-to-company payments platform. Has 67% of Kenyan adult population as customers, processing payments of ~ 30% of GDP last year, earning Safaricom US$255m in revenue, up 30% YoY. Arguably the most successful mobile payments platform in the world.

57 http://news.yahoo.com/blogs/upgrade-your-life/paying-phone-conveniences-cautions-141559523.html 58 https://www.fnb.co.za/promotions/smartphoneTabletOffer/Smartphone-Tablet-Offer.html

Page 102 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

„ People Like U, NAB, Australia: Billed as the world’s first “econography” tool, www.peoplelikeu.com.au uses analysis of a billion customer transactions to show the financial habits of similar people. Includes comparisons of how much Australian and local-postal-code peers spend and where they spend it. Interesting level of detail proves the point around big data in banks: food spend, for example, is split into dining out, groceries, liquor, groceries, pubs and clubs and shows which providers are popular, preferred and emerging, with average spend per customer visit. Well worth a visit.

„ Cardless ATM withdrawals – RBS, United Kingdom: RBS uses a Monitise platform to offer customers the ability to withdraw cash from an ATM without a debit card using a code provided by the bank by phone or text/SMS. Branded “Get Cash” is helpful for customers with lost or stolen cards.

„ Hands-free checkout – Square, US: Merchants operating Square and designated as “Favorite” by the customer allow customers to pay by saying their name at check-out59. The customer’s smartphone connects with merchant Square account, letting the vendor know you’re in the store, customer photo appears on the cash register, if the same person store authorises payment and you’re done. Smart, convenient and better than contactless.

59 https://squareup.com/help/en-us/article/3906-square-wallet-payments-and-receipts

Deutsche Bank AG/London Page 103

8 September 2013

Financial Services Retail Bank Strategy

Glossary

„ Average Revenue Per User (ARPU): Per capita revenue for a defined period and product, usually a month, mainly used by mobile telco firms. See for example, Voice ARPU, SMS ARPU, M-PESA ARPU.

„ Big data: Uses increasingly rapid and cost effective analysis of large volumes of customer data to better inform product design, customer marketing and product pricing in particular.

„ Core Banking System (CBS): Hardware and software which deliver the basic functions of banking to customers.

„ Cost-per-click (CPC): Google’s largest revenue pool is advertising / referral income, driven by volume of internet user click-throughs multiplied by average advertiser cost per click. Google is growing clicks, average revenue per click is declining, overall top line is growing. Part of CPC decline is from an increased proportion of mobile in the mix.

„ Crowdfunding: Web-hosted, online-auction-driven lending and funding of small business and creative projects. See for example www.crowdcube.com and www.fundingcircle.com.

„ Digital wallet: Consumers link debit, credit and loyalty cards to an online account / identity. Shields account details, some improvement in transacting convenience, drives loyalty schemes and directed offers. “Digital wallets” encompasses both e-wallets/online-wallets like Paypal and mobile wallets on smartphones such as Paypass and Google Wallet.

„ Distributed Denial Of Service attack (DDOS): Method of cyber attack which aims to make web sites unavailable or so slow as to be unusable by directing a barrage of requests from numerous sources at a site, rendering it unable to cope with legitimate access requests.

„ Geo-fencing: Virtual perimeter in real-world allows GPS-enabled phones can house apps which use customer’s location for providing targeted offers / vouchers for particular retailers. We expect banks to capitalise on their customer relationships in this area.

„ Mobile Point of Sale (mPOS): The shift by retailers to mobile / software-driven Point of Sale systems rather than physical hardware exclusively devoted to payments, favoured by the likes of Verifone and Ingenico. Apple pioneered mPOS to reduce checkout lines, others are following.

„ Mobile wallet: Mobile subset of digital wallets. Offers convenience and security of online wallets on smartphones. Customers manage debit, credit and loyalty cards; increasingly able to make payments by phone.

„ Near-Field Communication technology (NFC): Broad term covering contactless technology increasingly seen on cards, phones and PoS terminals, based on industry standards using RFID chips already present in many different smartcards.

„ Point-of-Sale (POS) terminal: Machine/mechanism registering customer payment, typically in a retail environment, mainly accepting credit and debit cards.

„ Software-As-A-Service (SAAS or SOAS): Information technology delivery model under which software is hosted on third party hardware accessed by the client over the internet. Customer typically benefits from economies of scale in product

Page 104 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

innovation and a cost and infrastructure base that scales with business growth. See Metro Bank case study.

„ Peer-to-peer lending: Web-hosted, online-auction-driven unsecured consumer lending. See for example www.kickstarter.com, www.zopa.com.

Deutsche Bank AG/London Page 105

8 September 2013 Financial Services Retail Bank Strategy

Appendix 1

Important Disclosures

Additional information available upon request

For disclosures pertaining to recommendations or estimates made on securities other than the primary subject of this research, please see the most recently published company report or visit our global disclosure look-up page on our website at http://gm.db.com/ger/disclosure/DisclosureDirectory.eqsr

Analyst Certification The views expressed in this report accurately reflect the personal views of the undersigned lead analyst about the subject issuers and the securities of those issuers. In addition, the undersigned lead analyst has not and will not receive any compensation for providing a specific recommendation or view in this report. Jason Napier

Equity rating key Equity rating dispersion and banking relationships

Buy: Based on a current 12- month view of total 1600 share-holder return (TSR = percentage change in 1400 46 % 48 % share price from current price to projected target price 1200 plus pro-jected dividend yield ) , we recommend that 1000 800 investors buy the stock. 600 39 % 34 % Sell: Based on a current 12-month view of total share- 400 6 % holder return, we recommend that investors sell the 200 24 % stock 0 Hold: We take a neutral view on the stock 12-months Buy Hol d Sel l out and, based on this time horizon, do not recommend either a Buy or Sell. Compani es Covered Cos. w/ Banking Relationship Notes: Global Universe

1. Newly issued research recommendations and target prices always supersede previously published research. 2. Ratings definitions prior to 27 January, 2007 were: Buy: Expected total return (including dividends) of 10% or more over a 12-month period Hold: Expected total return (including dividends) between -10% and 10% over a 12- month period Sell: Expected total return (including dividends)

of -10% or worse over a 12-month period

Page 106 Deutsche Bank AG/London

8 September 2013 Financial Services Retail Bank Strategy

Regulatory Disclosures 1. Important Additional Conflict Disclosures Aside from within this report, important conflict disclosures can also be found at https://gm.db.com/equities under the "Disclosures Lookup" and "Legal" tabs. Investors are strongly encouraged to review this information before investing. 2. Short-Term Trade Ideas Deutsche Bank equity research analysts sometimes have shorter-term trade ideas (known as SOLAR ideas) that are consistent or inconsistent with Deutsche Bank's existing longer term ratings. These trade ideas can be found at the SOLAR link at http://gm.db.com. 3. Country-Specific Disclosures Australia and New Zealand: This research, and any access to it, is intended only for "wholesale clients" within the meaning of the Australian Corporations Act and New Zealand Financial Advisors Act respectively. Brazil: The views expressed above accurately reflect personal views of the authors about the subject company(ies) and its(their) securities, including in relation to Deutsche Bank. The compensation of the equity research analyst(s) is indirectly affected by revenues deriving from the business and financial transactions of Deutsche Bank. In cases where at least one Brazil based analyst (identified by a phone number starting with +55 country code) has taken part in the preparation of this research report, the Brazil based analyst whose name appears first assumes primary responsibility for its content from a Brazilian regulatory perspective and for its compliance with CVM Instruction # 483. EU countries: Disclosures relating to our obligations under MiFiD can be found at http://www.globalmarkets.db.com/riskdisclosures. Japan: Disclosures under the Financial Instruments and Exchange Law: Company name - Deutsche Securities Inc. Registration number - Registered as a financial instruments dealer by the Head of the Kanto Local Finance Bureau (Kinsho) No. 117. Member of associations: JSDA, Type II Financial Instruments Firms Association, The Financial Futures Association of Japan, Japan Investment Advisers Association. Commissions and risks involved in stock transactions - for stock transactions, we charge stock commissions and consumption tax by multiplying the transaction amount by the commission rate agreed with each customer. Stock transactions can lead to losses as a result of share price fluctuations and other factors. Transactions in foreign stocks can lead to additional losses stemming from foreign exchange fluctuations. "Moody's", "Standard & Poor's", and "Fitch" mentioned in this report are not registered credit rating agencies in Japan unless “Japan” or "Nippon" is specifically designated in the name of the entity. Reports on Japanese listed companies not written by analysts of Deutsche Securities Inc. (DSI) are written by Deutsche Bank Group's analysts with the coverage companies specified by DSI. Russia: This information, interpretation and opinions submitted herein are not in the context of, and do not constitute, any appraisal or evaluation activity requiring a license in the Russian Federation.

Deutsche Bank AG/London Page 107

David Folkerts-Landau Global Head of Research

Marcel Cassard Ralf Hoffmann & Bernhard Speyer Guy Ashton Richard Smith Global Head Co-Heads Chief Operating Officer Associate Director CB&S Research DB Research Research Equity Research

Asia-Pacific Germany North America Fergus Lynch Andreas Neubauer Steve Pollard Regional Head Regional Head Regional Head

International locations

Deutsche Bank AG Deutsche Bank AG Deutsche Bank AG Deutsche Securities Inc. Deutsche Bank Place Große Gallusstraße 10-14 Filiale Hongkong 2-11-1 Nagatacho Level 16 60272 Frankfurt am Main International Commerce Centre, Sanno Park Tower Corner of Hunter & Phillip Streets Germany 1 Austin Road West,Kowloon, Chiyoda-ku, Tokyo 100-6171 Sydney, NSW 2000 Tel: (49) 69 910 00 Hong Kong Japan Australia Tel: (852) 2203 8888 Tel: (81) 3 5156 6770 Tel: (61) 2 8258 1234 Deutsche Bank AG London Deutsche Bank Securities Inc. 1 Great Winchester Street 60 Wall Street London EC2N 2EQ New York, NY 10005 United Kingdom United States of America Tel: (44) 20 7545 8000 Tel: (1) 212 250 2500

Global Disclaimer

The information and opinions in this report were prepared by Deutsche Bank AG or one of its affiliates (collectively "Deutsche Bank"). The information herein is believed to be reliable and has been obtained from public sources believed to be reliable. Deutsche Bank makes no representation as to the accuracy or completeness of such information. Deutsche Bank may engage in securities transactions, on a proprietary basis or otherwise, in a manner inconsistent with the view taken in this research report. In addition, others within Deutsche Bank, including strategists and sales staff, may take a view that is inconsistent with that taken in this research report. Opinions, estimates and projections in this report constitute the current judgement of the author as of the date of this report. They do not necessarily reflect the opinions of Deutsche Bank and are subject to change without notice. Deutsche Bank has no obligation to update, modify or amend this report or to otherwise notify a recipient thereof in the event that any opinion, forecast or estimate set forth herein, changes or subsequently becomes inaccurate. Prices and availability of financial instruments are subject to change without notice. This report is provided for informational purposes only. It is not an offer or a solicitation of an offer to buy or sell any financial instruments or to participate in any particular trading strategy. Target prices are inherently imprecise and a product of the analyst judgement. As a result of Deutsche Bank’s March 2010 acquisition of BHF-Bank AG, a security may be covered by more than one analyst within the Deutsche Bank group. Each of these analysts may use differing methodologies to value the security; as a result, the recommendations may differ and the price targets and estimates of each may vary widely. In August 2009, Deutsche Bank instituted a new policy whereby analysts may choose not to set or maintain a target price of certain issuers under coverage with a Hold rating. In particular, this will typically occur for "Hold" rated stocks having a market cap smaller than most other companies in its sector or region. We believe that such policy will allow us to make best use of our resources. Please visit our website at http://gm.db.com to determine the target price of any stock. The financial instruments discussed in this report may not be suitable for all investors and investors must make their own informed investment decisions. Stock transactions can lead to losses as a result of price fluctuations and other factors. If a financial instrument is denominated in a currency other than an investor's currency, a change in exchange rates may adversely affect the investment. Past performance is not necessarily indicative of future results. Deutsche Bank may with respect to securities covered by this report, sell to or buy from customers on a principal basis, and consider this report in deciding to trade on a proprietary basis. Unless governing law provides otherwise, all transactions should be executed through the Deutsche Bank entity in the investor's home jurisdiction. In the U.S. this report is approved and/or distributed by Deutsche Bank Securities Inc., a member of the NYSE, the NASD, NFA and SIPC. In Germany this report is approved and/or communicated by Deutsche Bank AG Frankfurt authorized by the BaFin. In the United Kingdom this report is approved and/or communicated by Deutsche Bank AG London, a member of the London Stock Exchange and regulated by the Financial Services Authority for the conduct of investment business in the UK and authorized by the BaFin. This report is distributed in Hong Kong by Deutsche Bank AG, Hong Kong Branch, in Korea by Deutsche Securities Korea Co. This report is distributed in Singapore by Deutsche Bank AG, Singapore Branch or Deutsche Securities Asia Limited, Singapore Branch, and recipients in Singapore of this report are to contact Deutsche Bank AG, Singapore Branch or Deutsche Securities Asia Limited, Singapore Branch in respect of any matters arising from, or in connection with, this report. Where this report is issued or promulgated in Singapore to a person who is not an accredited investor, expert investor or institutional investor (as defined in the applicable Singapore laws and regulations), Deutsche Bank AG, Singapore Branch or Deutsche Securities Asia Limited, Singapore Branch accepts legal responsibility to such person for the contents of this report. In Japan this report is approved and/or distributed by Deutsche Securities Inc. The information contained in this report does not constitute the provision of investment advice. In Australia, retail clients should obtain a copy of a Product Disclosure Statement (PDS) relating to any financial product referred to in this report and consider the PDS before making any decision about whether to acquire the product. Deutsche Bank AG Johannesburg is incorporated in the Federal Republic of Germany (Branch Register Number in South Africa: 1998/003298/10). Additional information relative to securities, other financial products or issuers discussed in this report is available upon request. This report may not be reproduced, distributed or published by any person for any purpose without Deutsche Bank's prior written consent. Please cite source when quoting. Copyright © 2013 Deutsche Bank AG

GRCM2013PROD030184