Alasdair Smith on the expected impact of the CMA’s banking investigation

Speech given by Alasdair Smith, CMA Inquiry Chair, at the Beesley Lectures.

Alasdair Smith photo
Alasdair Smith photo

What will be the impact of the CMA retail banking market investigation?

The Competition and Markets Authority (CMA) retail banking market investigation produced its final report in August, following a 21-month investigation. I’m speaking this evening about the very significant impacts that we expect our decisions to have on the markets for personal current accounts (PCAs) and for small business banking. So I’m going to focus on our remedies, and because of the time constraint, on a subset of our remedies. For most of this talk I’m speaking not on behalf of the CMA as an institution, but on behalf of the group of independent members who made the decisions; but at the end I will make some broader, personal observations which represent only my views, not necessarily those of the whole group or of the CMA as an institution.

But before I speak about remedies, I need to outline what it is about the banking market that we think needs fixing.

The first thing to say is that we saw many positive developments in UK retail banking. There has been entry by new banks with some entrants adopting new business models, offering specialist products and exploiting opportunities offered by new technologies, such as digital-only banks. Mobile banking is now widely adopted and growing fast. New types of payment services, lending and financial management services are now available from providers who are not banks.

The main problems that we identified are:

  1. Current accounts for both personal and business customers have complicated charge structures, and the actual cost to the customer depends on how they use the account. Customers generally know very little about the charges and service quality provided by other banks. It is therefore hard for customers to know whether they could get better value and service from another bank or a different product with the same bank. The standard ‘free-if-in-credit’ current account is a reasonable deal for many customers. However, personal customers with high average credit balances can typically make significant savings by switching to an account which charges a monthly fee and pays interest on balances, but it’s hard to shop around intelligently.

  2. It’s easy to switch because there is now a reliable and efficient Current Account Switch Service (CASS), but the service is not widely known, and does not command as much confidence as it deserves. The account opening process for small businesses can also be lengthy and onerous. Only 3% of personal customers and 4% of business customers switch to a different bank in any year.

  3. Charging structures for overdrafts are particularly complicated, making it even harder to compare providers, even though many overdraft users would make large gains from switching providers. Customers worry that if they switch they might not get the same overdraft from their new bank. Moreover we found that many customers underestimate their overdraft use and we know that better management of their overdraft use could make substantial savings for many customers.

  4. Over half of start-up businesses open their current account at the bank where the business owner has their personal account. There is also a strong link between business current accounts and lending: 90% of small businesses get their business loans from their main bank, with no or little shopping around for other lenders. It is hard for small businesses to find out who is the best lender for them. As their existing bank already knows a great deal about them, it is usually easier and quicker to get a loan from their existing bank and harder for other potential lenders to accurately assess and price lending.

High switching rates are not in themselves a sensible policy objective. But low switching rates in the presence of large switching gains is a symptom of weak competition. As a result, the older larger banks, who still account for the large majority of the retail banking market, do not have to work hard enough to win and retain customers and it is hard for new and smaller providers to attract customers. Banks will invest in new products or services or reduce their prices and improve service quality only if they expect to win business as a result, or fear losing business if they do not.

Our remedies package

Our remedies package is grouped into 4 elements. Our ‘Foundation measures’ will underpin enhanced competition in all the markets that we have examined. These will be accompanied by measures specifically directed to certain aspects of the market: current account switching, PCA overdrafts, and the needs of small business.

I don’t have time to talk about all 17 remedies, so I will focus on 5 of them: search prompts, overdraft prompts, the monthly maximum overdraft charge, a prize competition to develop better information services for small business banking, and finally, the element in our package with the most far-reaching implications, Open Banking.

Most financial products have annual renewal dates which provide a natural point for the customer to consider whether they should switch provider. Bank accounts are not like this, so banks are not under the same competitive pressure as for example insurers. We therefore want personal and business customers to be sent occasional reminders (‘prompts’), at suitable times, to encourage them to reconsider their banking arrangements. Some prompts might be triggered by specific events affecting the customer such as the closure of a local branch; others might be periodic, such as a reminder included in an annual statement.

The effectiveness of these prompts will depend on how they are designed, and needs to be based on careful consumer research. We’re aware of the Financial Conduct Authority’s (FCA) work showing that requiring annual account statements has had no discernible effect on customer behaviour. We are therefore not implementing search and switch prompts ourselves but rather asking the FCA to undertake a programme of customer testing to identify which prompts will be most effective in changing customer behaviour and then to implement these prompts.


Overdraft charges are the second largest source (after interest earned by lending out customer balances) of banks’ income from the provision of current accounts. Unarranged overdrafts can be particularly costly, and the charges can be an unwelcome surprise. We will therefore require banks to alert their customers, for example by sending a text message, when they are going into unarranged overdraft. Customers need to be given the opportunity to avoid incurring charges. The alerts that banks will be required to provide will also inform customers of a ‘grace period’ during which they have an opportunity to avoid charges. Many banks do this already, and FCA research shows that it has a big impact – reducing unarranged overdraft charges by 24%; we’re requiring them all to do it.

Banks will also be required to set a ceiling on their unarranged overdraft charges, in the form of a maximum monthly charge (MMC). A bank’s MMC will specify a maximum amount that the bank can charge a customer during any given month, taking together all unarranged overdraft charges including debit interest and unpaid items fees that the bank charges. Banks will be required to disclose their MMC associated with each of their accounts, so that customers are aware of each bank’s MMC and can use it to compare providers. The MMC will be an important aspect of overdraft pricing and will have to be clearly displayed by banks to their customers. Again, some banks already set an MMC, but not necessarily on a consistent basis, and they don’t necessarily advertise their MMC; so again we’re building on good practice here rather than introducing something completely new.

While we are placing requirements on banks to introduce these measures, we are aware that, to have most impact, like the search prompts they need to be carefully designed; and then monitored and improved in the light of experience. We are therefore asking the FCA to follow up by undertaking further research to refine the detail of the overdraft alerts and of the grace periods to maximise their effectiveness; and also to look at how the current account opening process could be used to make customers more aware of the costs of overdrafts. This will complement the FCA’s work to safeguard responsible lending. Also, we have asked the FCA to carry out work on the effectiveness of the MMC after we have implemented it, and to consider how it might be further developed.

It’s worth emphasising that each bank sets the MMC on each account. It’s not a regulated cap applied at the same level to every account. It’s much more of an information remedy than a regulatory remedy. In response to our introduction of this measure, some commentators have asked that the FCA turn the MMC into a regulated cap, particularly to safeguard the interests of vulnerable customers.

We considered a regulated cap, but were concerned about the risk that regulation might reduce the availability of unarranged overdrafts, which are a valued if expensive source of financial flexibility to many customers. One witness to the Treasury Select Committee advocated the complete withdrawal of unarranged overdrafts, implying that a customer already at their arranged overdraft limit might have the direct debits paying their phone or energy bills stopped with all that that would imply. We thought that would be an unwelcome prospect to many bank customers, especially those who are financially stretched.

We have taken the view that Open Banking, together with the full range of our overdraft remedies, has the potential to increase competition for overdraft customers and significantly reduce the problems there are in this part of the market; and we took the view that if problems can be solved by competition, that’s generally a better route than regulation.

Obviously, since we’ve asked the FCA to look at how our overdraft remedies could be developed further, we’re entirely happy with the idea that they should look at turning the MMC into a regulated cap. They’d need to consider whether a monthly regulated cap would be the best protection for vulnerable customers given that the majority of unarranged overdraft users incur charges which may be burdensome to them but might nevertheless fall well below a regulated monthly cap. There is also the risk that a regulated cap will take pressure off the banks to justify the level of their overdraft charges and to compete with lower charges. A customer complaint about the unfairness of charges might be met with the response that the bank is complying with the regulator’s requirements.

Banking services for small businesses

Our foundation measures and current account switching remedies will help business customers as well as personal customers. But we need to do more for small businesses, because the informational asymmetries which characterise small business banking mean that the competition problems in this market are deep-seated.

We are backing the independent innovation charity NESTA’s launch of a ‘challenge prize’ to enable the development and delivery of comparison services that will be both innovative and commercially sustainable. We think this is the best way forward, and, to maximise its potential, we are requiring those banks most active in providing services to small businesses to provide both financial backing and technical support to the NESTA project. We expect the services developed in the NESTA project to make very good use of the development of Open Banking, and we will ensure the 2 projects are properly co-ordinated.

Open Banking

Which takes me to the most fundamental of our remedies, Open Banking.

Application programming interfaces (APIs) are key to the digital services we use on our smartphones and computers. They make life simpler for millions of us every day by enabling us to share information, for example about our location or our preferences. Open APIs allow different applications to share digital information and are the hidden technological drivers behind everyday applications. The maps application on my smartphone knows where I am, can direct me to restaurants of a specified kind within walking distance, and give me real-time information about the trains from the nearest station, accessing data about maps, restaurants and train timetables using open communication standards.

There is already a very active and growing financial technology (‘FinTech’) community which has been developing and introducing new products using existing digital technology. Open APIs can transform the financial services sector.

The development and implementation of an open API standard for banking will permit authorised intermediaries to access information about bank services, prices and service quality. This will enable new services to be delivered that are tailored to customers’ specific needs. The types of new and improved services that will result from this remedy include applications which go beyond standard price and quality comparisons:

  1. Allow banking customers to manage through a single application accounts held with several providers, including an emergency borrowing account which effectively unbundles overdraft provision from the main current account.

  2. Allow customers to authorise the movement of funds between current and deposit accounts to help avoid overdraft charges or to benefit from higher interest payments.

  3. Monitor a current account and forecast a customer’s cash flow, helping to avoid overdraft charges.

  4. Let customers make simple, safe and reliable price and service quality comparisons tailored to their own usage patterns.

  5. Use a small business’s transaction history to allow a potential lender other than their bank to reliably assess the business’s creditworthiness and offer better lending deals than they would without this information.

When it comes to customer data, privacy and security concerns are paramount. Only when customers are satisfied that the right safeguards are in place and give consent will their data be shared with anybody. Open APIs will give customers control over what data is shared and with whom. Open data will be much easier to use and generate a much wider range of services than the current Midata tool, which is very little used; and it will be also much more secure than Midata. And Open Banking will make services which unbundle overdrafts from current accounts much more secure than they are currently.

To ensure that enough time is available to work through the important details of this remedy, particularly those that ensure that customers’ data is secure at all times, we are requiring that the release of information under this remedy takes place in stages. The least sensitive information – for example about banks’ prices, terms and conditions and branch location – will be made available by the end of March 2017. We expect that all aspects of an Open Banking standard will be up and running in early 2018 to coincide with the implementation of the second Payment Systems Directive (PSD2).

The need to implement PSD2 means that the banking sector would be doing some of this work without our intervention; and indeed the Open Banking programme started with a Treasury initiative a couple of years ago. But our intervention will ensure that Open Banking is developed to the challenging timetable we have set, and is genuinely open. The nature of the initiative is such that there will be commercial winners and losers from a radical change in the information sources available to bank customers; current political developments might seem to reduce the imperative to implement PSD2; and the highly technical nature of the programme might give scope for creative foot-dragging. We are therefore putting a structure in place that ensures successful delivery – with the programme overseen by an Implementation Entity and supervised by an Implementation Trustee with strong powers.

Open Banking, behavioural economics and regulation

Open Banking will lead to the development of new kinds of customer information and advice services, far beyond price comparison websites. Such developments pose new challenges for regulators. I’ve already indicated ways in which our decisions have been influenced by developments in behavioural economics – we’re implementing some ‘nudge’ remedies, and we’re asking the FCA to conduct randomised controlled trials (RCTs) to ensure that remedies are designed optimally, because we’ve all learned, for example from the work of the Behavioural Insights Team, that quite subtle changes in the design of an intervention can make a big difference to its impact.

The new services opened up by Open Banking will need to be carefully regulated – indeed the FinTech sector itself is clear that strong regulation is needed before bank customers will share their data with intermediaries. Commercial digital intermediaries will of course use the tools of their trade, and regulators will need to ensure that the well-known biases in human decision-making – framing, loss aversion, overestimation of low probabilities, and so on – are not unfairly exploited. My colleague Mike Walker is giving next week’s Beesley lecture, on behavioural economics, and discussing the implications of behavioural economics for regulatory decision-making: Open Banking gives us still more to think about in this area.

What we didn’t do

Much of the controversy about our decisions has been about what we haven’t done rather than what we have decided to do.

We didn’t break up the banks

We chose not to pursue divestitures – to break up the big banks. We looked hard at this issue.

The larger banks in Britain have a higher proportion of customers who have been with them for a long time, and tend to have higher prices on average for their PCAs, partly because they have a higher proportion of their customers on their higher-priced accounts. This may just be an accident of history. On the other hand, it may be because they have had less incentive to compete as vigorously as smaller banks and new entrants for the more active customers. The most disruptive innovation of recent years – the introduction of a PCA paying generous interest on account balances – came from a bank outside the big 4. Market concentration may therefore be having an effect on competitive behaviour but the evidence is neither strong nor conclusive.

Furthermore, if we had thought that there was a case in principle for ordering divestments to reduce the market shares of the largest banks, we’d have had to confront the problem that the divestments ordered after 2008 by the EU on state aid grounds have proved to be highly problematic and eye-wateringly expensive, basically because of the problems of extracting a divestible business from the entrails of the legacy IT systems of the large banks. The divestment of TSB from Lloyds was expensive and TSB still has some of the Lloyds IT legacy. The divestment of Williams & Glyn from RBS seems to have been even more expensive and problematic and it has still not been accomplished.

In any case, the fundamental problem is that when new and competitive products are introduced, it takes too long to build up customer numbers. It’s best to tackle that problem directly and the remedies we are introducing will enable customers to be more responsive and reduce the advantages of the existing banks. They will also provide stronger incentives on all banks to compete and make the market more attractive to new banks and other providers, as well as facilitating innovation.

We didn’t ban ‘free if in credit’ banking

Some commentators think that the way the true costs of free-if-in-credit (FIIC) banking are hidden is a disincentive to shop around, and even suggest that providing more information about forgone interest would encourage searching and switching. We were not convinced by this. Countries in which PCA models other than FIIC are prevalent do not have higher switching rates. I imagine most FIIC customers know their account is not really ‘free’, because the bank doesn’t pay for their credit balances. But knowing what their existing account ‘really’ costs doesn’t take the customer much further – what they need to know is whether a different account would suit them better. That’s why the better comparison tools which Open Banking will make available are the right way to go.

One bank proposed that we should require the payment of interest on current account balances, and since this would surely then force banks to start charging monthly fees or introduce charges for all transactions, effectively this would be a mandated end to FIIC. I’ve learned that leading an inquiry into banking is not a path to easy popularity, but such criticism in the media as we have faced is, I’m sure, nothing compared to the response we’d have heard to the banning of FIIC accounts. And for good reason – it’s the wrong remedy for the real problem, and one with distinctly unattractive distributional consequences: the losers would be customers with modest current account balances and little or no overdraft use, the Prime Minister’s ‘just managing’ people; the gainers would be those who are managing rather well, thank you, and have high current account balances.

Nor is there any reason to suppose that banning FIIC would reduce overdraft charges: the banks charge high overdraft fees because they are not subject to enough competitive pressure on their fees; banning FIIC would not change that.

We didn’t do anything about ‘too big to fail’

I can be brief about the next 2. The smaller banks feel that the ‘too big to fail’ banks have funding advantages in that they have to pay less for customer deposits. Whether or not this is so, the banking market is in the middle of the ring-fencing set in process by the Vickers report, and it would not have been sensible for us to get involved in any way.

We didn’t do anything about payment systems

Somewhat similarly, it has been recognised since the Cruickshank report of 2000 that access to payments systems gives rise to competition concerns because the major payment systems are collectively owned by the large banks. Cruickshank’s recommendation that a regulator be established has only now been implemented with the creation of the Payment Systems Regulator (PSR), with a primary competition objective. The PSR was just getting underway as we were starting our work, and it was clearly in nobody’s interest that we duplicate or second-guess or supervise their work. Had the PSR been in place for 5 years, we might have thought of reviewing how effectively they had done their job, but they hadn’t, so we didn’t.

We did make strong observations about prudential regulation but didn’t go further

Prudential regulation was a more complex matter. Capital requirements rules are designed to make banks resilient against financial crises and to enable failing banks to be dismembered without knock-on consequences to the wider economy. In ways that and for reasons that I do not have time to go into this evening, different rules are applied to banks with established lending records (the ‘internal ratings based’ (IRB)) approach from those without (the ‘standardised approach’ (SA)). The difference between the 2 approaches for UK banks is particularly wide in the treatment of low loan-to-value residential mortgages, and newer and smaller banks which do not have ‘IRB’ status for their mortgage lending believe that this is a significant barrier to entry and expansion in the UK retail banking market.

Prudential regulation is the responsibility of the Prudential Regulation Authority (PRA), which is part of the Bank of England. The PRA has a secondary responsibility to consider the effects on competition of its regulatory practice, but its primary responsibility is to ensure the stability of the financial system on which we all depend. We had discussions with the PRA, and indeed collaborated with it on a project which sought to estimate the effect in the mortgage market of the IRB-SA difference.

The CMA does not have powers to alter the capital funding requirements, which are determined in international forums in Basel and Brussels in which the UK is represented by the Bank of England and the Treasury. The Bank and the Treasury are well aware of the competition concerns associated with prudential regulation and we believe it is important that these concerns are given due weight.

Of course I understand the frustration of the SA banks that a major investigation of competition in UK retail banking did not go further in addressing what they see as one of the most significant competitive distortions limiting their ability to expand. The suggestion was made to the Treasury Select Committee that we might have ‘directed’ the PRA to take some action to address the position. This would have been a mistake. Regulated firms deserve to have a clear regulatory framework in which they know to whom they are answerable. This is particularly true in the regulation of the safety of the financial system. It’s said that one of the factors which contributed to the depth in the UK of the 2008 crisis was a regulatory system with divided and unclear responsibility for financial stability. The current regime was intended to put that right. For the CMA to muddy the regulatory waters by seeking to ‘direct’ the PRA how to do its job would have been a major policy error.

Final reflections

Up to now I have been discussing the agreed, indeed unanimously agreed, decisions of the group of independent CMA members who conducted the market investigation. The CMA will naturally be considering the wider lessons that might be learned from the experience of the energy market investigation as well as the retail banking market investigation. It’s not done that yet, so my closing reflections are strictly my own, not an indication of any thinking by the CMA as an institution.

I’ve already made several references to the ways that our remedies build on initiatives already under way and relate to the work of other agencies: the Treasury’s original Open Banking Working Group, the prompts and grace periods already implemented by some banks, and prize competition being considered by NESTA before we came along; and I’ve also noted how some of our remedies are being handed over to the FCA for further development.

Indeed when you look at our whole programme of remedies, it’s difficult to find one for which we would claim sole authorship. I haven’t spoken at all this evening about the improvements we are seeking in the CASS, whose creation was a recommendation of the Vickers Commission. Here too we are therefore building on what is already in place and are very pleased by the enthusiasm with which BACS, which manages the CASS, has signed up to our recommendations. The NESTA initiative is only one element in our list of remedies for small and medium-sized enterprise (SME) banking – many of the others build on the 2015 Small Business, Enterprise and Employment Act.

I make absolutely no apology for this – on the contrary, in a sector like retail banking that has been for understandable reasons subject to a large number of inquiries and regulatory interventions over the past decade, it’s entirely appropriate that our emphasis should be on the spreading of existing good practice, on the further development of existing interventions, and on the support of existing initiatives, helping to bring them to fruition.

Amelia Fletcher of the University of East Anglia is about to publish the output of an interesting project she has undertaken for Which? on what she calls ‘demand-side remedies’, remedies focused on consumer information, searching and switching; as opposed to ‘supply-side remedies’ of which divestiture is the classic example. In both the energy and retail banking market investigations, the CMA’s focus has been on demand-side remedies (though not entirely so in the case of Energy); I’ve already explained why we did not think that supply-side remedies were what was needed in retail banking.

Now when we look back to the Competition Commission’s market investigations of the past, the most notable cases were ones which resulted in supply-side remedies, with the divestiture of 3 airports from BAA being the classic example. And, of course, such cases often involved hotly contested litigation.

The world of demand-side remedies is different. The CMA’s market investigation powers are engaged, but its powers of advocacy, both with other agencies and with parties, are engaged too; indeed the balance between enforcement and advocacy may be tilted quite far in the latter direction. There will not be unanimous support for the remedies among the parties to an investigation, but the strong opposition may come not from the main parties, but from politicians, the media and consumer groups who think the remedies don’t go far enough; and a good part of the engagement effort may need to be put in those directions. Litigation risk may be low, but political and reputational risk may be high.

It was inevitable once we got into the design of behaviourally informed demand-side remedies that we would collaborate closely with the FCA. The collaboration has been friendly and productive. I think both agencies are very comfortable with the agreed division of labour between us and with the understanding that each has of the constraints upon and powers of the other. But it’s possible to imagine scenarios in future where a market study which points toward a need for demand-side remedies may lead to a differently configured deployment of the respective powers of the CMA and a sectoral regulator. ‘Concurrency’ is discussed at the present largely or exclusively in the context of Competition Act enforcement; in future, concurrency might be as important in the context of Enterprise Act work. And with that speculative thought, I stop.

Published 21 October 2016