Andrew Bailey speech on retail banking in the UK - reflections from the FCA

Speaker: Andrew Bailey, Chief Executive
Event: BBA Retail Banking Conference, London
Delivered: 29 June 2017
Note: this is the speech as drafted and may differ from the delivered version

Highlights

• One big lesson from Northern Rock is the importance of understanding business models and where banks earn their returns and how stable those returns are.
• Retail banks are complex because they provide consumers with many products, which can be very long lived and there is scope for cross-subsidies between products and between customers.
• The FCA is undertaking a strategic review of retail banking business models focusing on links between different products and services and their relative profitability.

It is a great pleasure to be here today, at what is something of a fin de siècle event for our hosts – the passing of the baton to the new combined body. I want to start by acknowledging and thanking Anthony and his colleagues for their service and leadership of the BBA. Anthony started at a most unpromising time to say the least. I’m sure there have been times when he felt rather like the mythical figure Sisyphus, condemned to roll an immense boulder up a hill, only to see it come back down again and then start over repeating the action for eternity.

The story has it that Sisyphus was punished for his self-aggrandising craftiness and deceitfulness. To be clear, Anthony has always been a straightforward and helpful person to deal with in my experience, and I want to thank him and his colleagues at the BBA, for what they have done to rebuild reputations. I hope we will be able to look back and say that this has been the period when the foundations of the new British banking system were laid, and they were there shaping the work and the boulder will stay up the hill so to speak.

It is hard to think of many more important subjects in our world than the future of retail banking

It is hard to think of many more important subjects in our world than the future of retail banking – that is, if we heroically put Brexit to one side for a day. There are over 72 million personal current accounts in the UK, with retail deposits of over £1.5 trillion, comprising current accounts, savings products and SME banking. Retail lending is a key driver of economic activity. On the latest evidence, UK households owe £1.3 trillion in mortgages and £198 billion in consumer credit. There are over 5.4 million SMEs in the UK. There are over 61 million credit cards in issue in the UK.

It is not surprising that we can produce very large numbers to represent the scale of retail banking in the UK, or in many other countries, as it is an essential service for all of us. More interesting I would say is evidence of the pace and scale of change in the industry. Sixteen brand new banks have been authorised over the last five years, with 38 in the pipeline, which must be a modern record. The rate of increase of use of digital and contactless channels is fast and unlikely to slow. With that goes a sharp increase in the frequency at which customers wish to extract information from their accounts. And, the scale of operational threats from fraud and cyber-attacks has increased.

Some of these changes are put into sharp perspective when we remember that we are rapidly approaching the tenth anniversary of the outbreak of the financial crisis in 2007. On its own, that is a somewhat sobering thought. But it does provide a perspective on what has changed over that time and what has not. I want to use this speech to describe those challenges that can arise, and their implications for the future of retail banking.

I will start with balance sheets. One of the biggest changes is that ten years ago the pressure was on raising deposits, on funding what had become a very substantial expansion of balance sheets over the previous five years or so. There was a huge pressure on funding, which inevitably pushed up the cost and thus squeezed margins. Alongside that was the growth in the opacity of asset structures which made it harder to assess the likely future creditworthiness of borrowers and thus the solvency of banks. It was this lack of confidence in future solvency, combined with a critical shortfall of capital levels that was the undoing of many of those that failed.

It is almost ten years since I was first introduced to the balance sheet and business model of Northern Rock - a very fast growing and doomed retail bank. Three things struck me about the business model which goes beyond the usual diagnosis of lack of capital, important though that was. First, Northern Rock had a different form of torture to Sisyphus; rather, it was more like the roadrunner on a belt that got faster and faster. Why? Because it had to generate new mortgages at a faster and faster rate in order to feed the demands of its mortgage securitisation master trust, which was nearly half the overall balance sheet. It is hardly surprising that to do this it became over-concentrated at the lower end of the creditworthiness spectrum of mortgages – it couldn’t be fussy about what it took on. Second, by over-concentrating in wholesale and market funding through securitisation, it had a higher cost of funds than many other retail banks with, in particular, current account funding. In order to earn its target returns, it was pushed towards riskier lending. And third, the 50% of the bank that was not in the securitisation master trust was heavily funded by the cash flows of the master trust. But this funding had hard rating triggers, so that if the rating fell too much the funding would have to go elsewhere. It was in that sense a house of cards. And, we know what happened.

One big lesson I draw from this – for all of us, regulators and practitioners alike – is the importance of understanding business models and thus where banks earn their returns and how stable those returns are.

Let’s roll the clock forward ten years to today. The good news is that the system is more stable, but it faces different challenges. One striking change is that whereas a decade ago the challenge was to fund its balance sheets, today the greater challenge is to find lending opportunities that enable sustainable net interest margins to be earned. Although, it is a simplification, assets often appear to be relatively more scarce than funding. The regulators watch carefully for excessive risk taking in lending because the temptation is there. Fortunately, the experience of the crisis has led to a radical overhaul of the regulatory tools to do this – with stress tests and enhanced asset quality assessment; it feels like a very different world.

But the changes of the last decade have led to outcomes which are an important focus for the FCA as conduct regulator. For example, whereas in the crisis and the aftermath there was a marked trend for more banks to enter the current account market, which was perceived to be lower cost funding, today I see some evidence – not I should say in all cases – for it to go the other way, as current accounts are regarded as more expensive in terms of operational costs. This is, of course, consistent with relatively easier funding conditions.

If we turn to lending markets, we see another feature which has a considerable impact on the returns from retail banking, namely the spread between the returns from so-called front and back book loans, newer and older loans. It is a pretty long established practice in UK retail banking that the returns from back books exceed those from front books. That makes the relative mix of back and front books one determinant of overall earnings. It also means that one set of consumers pay more than another set. I will come back to this issue for the FCA. By the way, Northern Rock appears to have gone some way towards solving the back and front book issue by churning their mortgage book at much higher speed – the two year fixed rate mortgage was a standard product. But this created a dependence on the earnings from fees which came from the churning.

When we look at the whole issue of retail banking business models, it is hard to escape the issue of free-if-in-credit banking. I have learned from experience that this is a subject which attracts particularly forceful commentary. So, as they say, let me be clear! I do not advocate ending free-if-in-credit banking. Why? Because there is no such thing to start with, so it cannot be abolished as such. Nothing in life is free – sorry to disappoint. Some of you may be saying “what is he on about”? Of course, banking isn’t free. So, free-if-in-credit means that costs are recovered and charges levied on some products and services more than others. And, this means some customers pay more or less than others depending on what mix of products they use.

The classical product that appears to be free is the current account itself. Indeed, there is a theory that the incentive to create what was a higher return product like Payment Protection Insurance was fuelled by the imbalance of returns across products. But current accounts are not free because, for one thing, their cost depends on the interest foregone in the rate paid on them. This is not an easy calculation to make – it has to be benchmarked against a cost of funding – and it is therefore opaque. The point is that retail banking is a product for which the cost can be hard to know with precision.

Next in the list of hard to knows in retail banking is the issue of unauthorised overdrafts. This product – if we can call it that – has attracted considerable attention recently, and for us it falls under the broader heading of high-cost credit. It is one of – if not – the most obvious areas where retail banking crosses over into the high-cost credit area. I think it poses a challenge, in relation to a couple of issues.

First, the concept of an “unauthorised” product is odd when described as such. It is perhaps better described as the consequence of an account going into debit. In some countries, this is more often than not covered by the account coming with a pre-agreed line of credit. Moreover, clearly some customers go into their overdraft much more frequently than others, and there is a reason for that, a point I will come back to when I consider the broader issues we face. Moreover, as banks tell me often, advances in technology and payment systems mean that it is now much more possible to control account balances nearer to real time, creating the opportunity to limit the advance of overdraft credit. I will come back to this issue.

The second challenge posed by unauthorised overdrafts concerns their cost. It is quite regularly argued that their cost exceeds the price cap that the FCA imposed on payday lending. We are currently reviewing this as part of our overall assessment of high-cost credit. For retail banks, it is also an issue because it illustrates well the challenge of how costs and returns are spread across products and thus customers – since, as I mentioned, customers with some characteristics are more likely to use overdrafts than others. And, of course, it underlines that one of the meanings of free-if-in-credit is that the reverse holds. It isn’t free in debit. This begs the big question, namely how do we know and assess what is a fair distribution of costs and charges given the pattern of risk to the lender?

I want now to move on from the assessment of these prominent cases such as unauthorised overdrafts, and to step back to draw out the broader issues for retail banks, all of which are highly relevant for conduct regulation. I do not think that these more root cause issues receive enough attention in the analysis of retail banking conduct. They lie at the heart of business models.

The first issue is that banks are more complex as firms because they are multi-product. In other words, they are providing consumers with many products. Banks are managing both sides of their balance sheets, and so they must have a wide range of products. But the number of products creates more scope for less transparency in pricing and the assessment of costs. There are always choices on how to allocate costs across products, something that we have seen, for instance, in work that the regulators have done on funds transfer pricing in banks. Moreover, as I described, banks will price new product sales differently from existing sales, thus creating the front to bank book pricing issue. There is a natural tendency to do this because, like any firm banks need to attract new business.

This illustrates a second feature of banks, which is seen in other financial services firms, namely that the products can be very long-lived, because they are meeting the long-term needs of consumers. A mortgage is a classic example of this. It complicates financial services relative to many other products – and in doing so explains why we have more conduct regulation in finance because the consequences of the terms, for banks and their customers, will often only emerge over a long period of time. This lends itself to a problem of transparency in explaining the terms of products and in judging their appropriateness for customers once events take over.

The third feature of banks picks up a point I made earlier, and is really an outcome of the first two that I have just described. It is that there is naturally scope for cross-subsidies between products and therefore between customers. Moreover, as I also mentioned, it can be inherently hard to allocate costs. But since this has to be done, we should recognise that it involves hard judgements, and as I will come on to describe, inevitably raises difficult issues for a conduct regulator. Some element of cross-subsidy is inevitable in complex businesses with multiple and long-lived products. Transparency over allocation of costs and pricing can help, but it is not a panacea where the underlying judgements are inevitably complex. I have had these challenges myself when running operations – we all know, for instance, that the answer on how to allocate the costs of IT systems doesn’t drop out of the sky.

The fourth issue is a matter of broader public policy. There is a public interest in the supply of financial services. It does not mean that every consumer should be able to get access to every product on the same terms. That would not be realistic. But, there are some very basic challenges which follow from the public interest element of retail banking. There are important issues around access to basic bank accounts, access to bank branches, and access to ATM’s. There is also an issue around access to credit, which for me is at the heart of our interest in high-cost credit. Put simply, it would not be an acceptable outcome to cut consumers off from access to credit when they have a justifiable need for credit, for instance to smooth erratic or lumpy income. Now, before the headlines get written that I am justifying current household debt levels, that is not the point I am making. I am not talking about that subject today. Put simply, there is a point about access to credit which is a broader public policy question, and needs to be considered as such.

I want to end by putting these points into the context of the FCA’s work. But before I do so, I should briefly mention a couple of other important areas of relevance to these aspects of retail banking, namely ringfencing and innovation. These are not the focus of what I have said today because I want to focus on more fundamental root cause issues, but I will offer a few thoughts on each.

Ringfencing, namely the separation in legal entity terms of domestic retail banking, is in my view a useful development, but not on its own the answer to these more fundamental challenges. It is more important as a means to create the wherewithal to separate, and thus preserve, critical retail banking services in the event of a bank failure. It is thus important to solving the too big to fail problem, and as such is welcome. I do not, however, believe that it solves on its own broader cultural and conduct issues because, as we have seen in the last decade, misconduct has been as prevalent in retail as in investment banking. It is possible, that ringfencing might make it easier to identify and tackle the business model issues that I have described by virtue of the clarity that will come from the legal entity separation. I think there will be some assistance here, but I do not think that on its own, ringfencing changes the picture radically.

Technology and innovation can and should be harnessed to enable change

Innovation, particularly technological, is having a major effect on retail banking, and will continue to do so. This is a good thing, and at the FCA we are keen to enable change to occur consistent with our public policy objectives. This involves change within banks, and competition from non-banks in some products. My own view is that technology and innovation can and should be harnessed to enable change to assist the issues that I have described, but on its own it is necessary rather than sufficient. In other words, we won’t solve the issues by just leaving them to innovation; rather it needs to be harnessed to help.

That brings me to the role of the FCA in all of this. When we published our Mission in April, we framed it in terms of how the FCA serves the public interest through the objectives given to us by Parliament. Through this, we have sought to provide firms and consumers with greater clarity about how and why we prioritise and intervene in financial markets of all sorts. Parliament has given us a single strategic objective – to ensure that relevant markets function well – and three operational objectives to advance: to secure appropriate protection for consumers; to protect and enhance the integrity of the UK financial system; and to promote effective competition in consumers' interests.

Our aim is to serve the public interest by improving the way financial markets work and how firms conduct their business

As a public body our aim is to serve the public interest by improving the way financial markets work and how firms conduct their business. By doing this, we provide benefit to individuals, businesses, the economy, and so the public as a whole. To deliver our objectives, Parliament has given us a range of tools. It has also given us independent powers to make decisions about how best we should use these tools. We can use them to serve the public interest in different ways, but we must be targeted when we decide when, where and how to act.

Regulation is not cost free. The FCA is funded by fees paid by market participants, and both direct and indirect regulatory costs are likely to be passed on to individuals and business through higher prices. So our aim is to use tools efficiently and cost-effectively, in a way that delivers the greatest value to the public.

The FCA’s task can be difficult but is always fascinating

The issues that I have described in relation to retail banking illustrate why the FCA’s task can be difficult but is always fascinating. They are inherently hard questions. They go to the heart of questions about fairness in society, because access to banking services is a necessary service to the public. There is no easy answer to questions about what is fair access to credit, and a fair cost for any consumer. Nor is there a long history of regulators enabling these questions to be answered.

I want to end by briefly describing three pieces of work that we are doing to improve our ability to enable answers to be produced to the issues I have set out. The first piece of work is our strategic review of retail banking business models. This is very much a piece of discovery work to start with, focusing on links between different products and services and their relative profitability. It includes the impact on different groups of customers. It should enable us to assess better the impact of changes – for instance in technology – on retail banking business models. It is very much an empirically driven piece of work which picks up where the Competition and Markets Authority (CMA) left off, but is not a full evaluation of the functioning of competition in retail banking markets. It is broader in scope than the CMA’s work, which focused on personal current accounts and SME banking. We are covering both sides of the balance sheet of banks – deposit taking and lending. We are already in touch with stakeholders on this work. It is, I should say, an ambitious project. My hope is that we can lay out a body of evidence from which conclusions can start to be drawn. But, let me manage expectations! We expect to be working hard for the rest of this year and into the first two quarters of next year to get the evidence laid out.

We are also taking action on the recommendations made to us by the CMA following their investigation into retail banking. We are making good progress on these recommendations which include measures to improve service and to increase customer engagement with their current account. We will bring forward the first set of proposals on service metrics shortly.

The second piece of work is our review of high-cost credit including overdrafts. Last November we issued a Call for Input covering high-cost products, overdrafts, the high-cost credit short-term price cap and report and multiple such borrowing. We are looking at all high-cost products to build a full picture of how these are used, whether they cause harm and, if so, to which consumers. We will then be able to decide if we need to intervene further. We are also reviewing the overdraft market in detail following the CMA’s review which identified problems in the market. We are also reviewing the price cap on short-term high-cost loans which came in force in January 2015. We intend to publish our findings on this over the summer.

The third piece of work will appear in the near future, and is a thematic review by our supervisors which responds to a recommendation from the Parliamentary Commission on Banking Standards report in 2013 that banks should assess customers understanding of the products and services they deliver.

As you can probably tell, retail banking is a big agenda of work for the FCA. But, as I hope I have described reasonably accurately, the underlying issues are not small ones. And, finally, if you feel disappointed that I have not set out a view on the likely conclusions and responses, I will say very deliberately – let’s get the evidence together to enable a better assessment of these important issues.