Treasury Committee
Oral evidence: Prudential Regulation Authority Hearings: 2015-16, HC 615
Tuesday 17 November 2015
Ordered by the House of Commons to be published on 17 November 2015
Members present: Andrew Tyrie (Chair); Mr Steve Baker, George Kerevan, John Mann, Mr Jacob Rees-Mogg
Questions 1-99
Witnesses: Andrew Bailey, Deputy Governor for Prudential Regulation and Chief Executive Officer of the Prudential Regulation Authority, David Belsham, External Member, Prudential Regulation Authority Board, and Mark Yallop, External Member, Prudential Regulation Authority Board, gave evidence.
Q1 Chair: Thank you very much for coming to give evidence to us this morning, and welcome to the externals, who are now acting as wingmen for this session. I would like to begin with some questions about ring‑fencing, about which I said something last night anyway. You made it clear in your policy statement in May that you would grant waivers or modifications of ring‑fencing rules in certain circumstances. The worry about ring‑fencing has always been that, although there are opportunities for flexibility, it places a huge burden on regulators to get it right, and that this would create a lot of pressure from banks to vary your decisions. That is why the Parliamentary Commission, on which there was a wide spectrum of views—from those who still hankered after universalism, right through to several who were in favour of full separation—concluded that, at the very least, electrification was required: that you should have powers fully to separate a bank were you to consider that a bank was attempting to game the rules and take advantage of the flexibility given to you in legislation. Have you had any applications for waivers, and how do you intend to handle them?
Andrew Bailey: We have not had any applications yet, because that would naturally follow from the process that we are going to go through, starting in the beginning of the new year, when firms put in their formal proposals for how they are going to structure the ring‑fence and we will then go through those proposals with firms. I would expect that any waiver applications would follow out of that.
Let me just make two points on waivers, which I may have made before. I apologise if I have. We can grant waivers only where those waivers are consistent with the spirit of the legislation. We cannot get out of bed one morning and say, “I do not fancy this any more. Here is a waiver.” To be clear, the waivers have to be consistent with the direction of travel and the spirit of the legislation. That is all we will do, in that case.
The second thing I would say is that any use of waivers has to be transparent as a general matter, but my intention in this case would be that we would make it more transparent than usual. By that I mean that they will be somewhere on our website, but we will be more transparent in this case because of the public interest. We have not had any big discussions yet on the process for transparency for the eventual outcomes. Speaking somewhat personally, my own view on it is that we would be even more transparent in this case than we would be normally, for the reasons that you give, because there is a very clear public interest in this.
Q2 Chair: Turning to the question of bank lobbying, are you aware of any lobbying by banks via the Treasury or via the Chancellor or via other Treasury Ministers?
Andrew Bailey: The banks speak to the Treasury. They speak to Treasury Ministers. I am aware that that goes on. I am not surprised and it is, in a sense, a part of the process.
Q3 Chair: If I may interrupt, are there not two parts to that? There are perfectly legitimate points that banks should be making to the Treasury about the overall conduct of their business. There is what would be the illegitimate attempt to get round an area that, by statute, has now been passed to you as your responsibility.
Andrew Bailey: Indeed. You would have to ask the Treasury what they get. The filter for me is what the Treasury and Ministers do with it. Were we to be approached with what would fall into your “illegitimate” category, then that is a non‑starter.
Q4 Chair: Just to confirm, you have not had any such approach.
Andrew Bailey: No. I was going to come on to say that all the indications I have had from the Treasury and from Ministers is that they remain, as I would expect, firmly committed to the legislation. We have, as you probably are aware, seen some of what I would call lobbying via the press. This goes on as well. It is, frankly, unfortunate.
Q5 Chair: Are you referring to Barclays?
Andrew Bailey: That is one recent example. I can tell you that my advice to banks is: “When you have your dealings with the PRA, do not conduct them through the press. We are approachable people. We have deliberately designed the PRA so that there is senior engagement. This is a change from the FSA. You can come and speak to us. If you seek to do that via the press, it does not go down well.”
Q6 Chair: We made it clear to banks publicly in the last Parliament that if they feel that they are being mistreated by the regulator, they should in the first instance tell the regulator; in the second instance, they should speak to us privately and, if necessary, offer to say so publicly in evidence, and they should not do this via the press. Is that the approach you think they should take?
Andrew Bailey: Absolutely. That is all consistent with our accountability to you, with transparency and with having sensible processes. I am a big supporter of that.
Q7 Chair: I have a couple more questions on ring‑fencing. What do you expect will be the effect of the ring‑fence on borrowing and savings rates, and the fees and terms and conditions faced by ordinary retail customers?
Andrew Bailey: I do not think, frankly, it will be substantial. There is a whole series of changes going on in this period of time. If you think about the introduction of loss‑absorbing capacity in the resolution context, which is going on at exactly the same time, in a sense that is very much the adjunct of the ring‑fence. As we know, the ring‑fence is all about resilience and resolvability. You have to see this as a total package.
As I have said before, there will be some impact, because, frankly, the whole purpose of all these reforms is to transfer what has been a taxpayer subsidy through the public purse—that is implicit most of the time and then became explicit in the crisis—on to creditors. There will be some cost implications of that, and I do not think we should seek to deny that. I do not think the ring‑fence in itself, if one could try to separate that out for a moment, is a major change in itself, because the structure of retail banking is not the thing I would suspect is going to change most. I can tell you the discussions with the firms are all around the future of the non‑ring‑fenced banks, in my experience. The reason for that is not necessarily a narrow point about the non‑ring‑fenced banks; it is the coming together of two things. It is the question about the future viability of some of these firms, the obvious case being investment-banking models, which they are facing up to and have to face up to, in my view—this is not just in this country; you see it in other European countries at the moment—coming together with structural reform. It may be that structural reform, in a sense, is a forcing device for that, but in my estimation, they have to deal with this issue anyway, because the world has changed.
Q8 Chair: You do not see, therefore, the ring‑fence leading to the demise of free in‑credit banking.
Andrew Bailey: You know my view on this. There is no such thing as free in‑credit banking, but it is a question of the public being made aware of how much they are paying for banking. That is another story.
Chair: And you know that it is the view of many of us here that it is a con trick and needs to be described as such.
Andrew Bailey: If you look at the ring‑fencing structure as a whole, will it, in a sense, put pressure on that? It will to a degree, because one of the things that is underpinning this whole concept of free in‑credit banking is cross‑subsidisation within the industry and with institutions, and subsidies that are not well‑understood by the public—which goes back to this point about the public knowing what it is paying—but also we find as supervisors that cross‑subsidies are often not well understood by firms themselves, when we look at things like internal funds transfer pricing.
Q9 Chair: That goes back to the “too big to manage” problem.
Andrew Bailey: Yes. And also, as I say, to the lack of clarity in some of the business models as to what the costs of particular products are. Ring‑fencing may do a bit; it is true. It will do something to, in a sense, force that issue on to the table. Whether it will conclusively force that issue on to the table, I do not know.
Q10 John Mann: Just to explore a little bit further, the independence of a ring‑fenced board could potentially lead to a Co‑op‑style problem, where one board is taking the wrong decisions and the other is unaware of the scale of the problem, could it not?
Andrew Bailey: I made some comments recently in a speech. I sought to set out what I called the limits of the independence of the governance process. I have had people rather over‑interpret what this will mean. It is my view that a ring‑fenced bank, from a governance point of view, has to be independent, in the sense that the board of the ring‑fenced bank has a duty to uphold the ring‑fence and everything that goes around it. What I said was that it does not therefore mean that the ring‑fenced bank just goes off and says, “I am not interested in the strategy of the bank. I am off doing my own thing.” It does not mean that they can refuse to pay a dividend where they have excess capital. It does not mean that the CEO of the ring‑fenced bank can stick two fingers up at the group CEO whenever they feel like it. It is all about how you work that out in internal governance, but it is important to say that the key point here—the governance point—is that you have to abide by the law of the ring‑fence. It does not mean you can make a universal declaration of independence and go off and do your own thing, and then just send your problems back.
Q11 John Mann: Yes, but we looked in detail at the Co‑op problem and this two‑tier, separated board structure appeared to us to be fundamental to the weaknesses that were there.
Andrew Bailey: If we thought that the ring‑fencing structure was creating governance systems like the Co‑op, we should be very worried. Frankly—you have probably looked at this a lot—the problematic aspects of the Co‑op’s governance structure went much broader than what you might call the precise issue on ring‑fencing. I do not think I would make the leap from saying, “There is a parent-subsidiary structure in ring-fencing” or, “There is a sibling structure, with a parent holding company and then a sibling, in ring‑fencing” to saying, “And that will create a governance structure that would immediately lead to the Co‑op problems”. I do not think that follows. Many firms have the holding company and subsidiary structure, because they are operating in other countries where they are bound by the laws of that country. This is not that unfamiliar.
Q12 John Mann: You describe it as “sibling”, but the Barclays approach appears to be more mother‑daughter than sibling. Surely, what they are trying to do is use the stronger retail bank as a shield to protect the investment bank. You are responsible for prudential oversight. Is this not an inherent weakness and problem?
Andrew Bailey: With respect, that is precisely what they cannot do. The whole point of the ring‑fence is not to create a situation where the non‑ring‑fenced bank is dependent on the ring‑fenced bank as a source of strength. If we end up in that situation, we have got no further than we were in the past.
Let me be clear: Barclays has not formally put to us that proposal. We have seen it in the press; they have come in and discussed it with some of my more junior colleagues. I would have to refer them to things that have been said in Parliament, things that John Vickers said to the Parliamentary Commission and past discussions that, by the way, we have had in this Committee—there was one in October last year and one earlier this year—when we have made very clear our reservations about what you call the mother‑daughter arrangement, and that the sibling structure is the presumption. Be careful with the word “presumption”; it is not formalised as such in the legislation, but there is a pretty strong presumption that that is the model, and indeed it was said in Parliament by a Government Minister.
Q13 John Mann: Can I move on to two other questions on two separate areas? Mr Yallop, you came out of ICAP, which did not come out of the LIBOR situation too well. Do you see the approach now, which most people are categorising as being more lenient and laissez‑faire, as being one that is sensible and prudent for the future?
Mark Yallop: First of all, with respect to ICAP and its position, I cannot really comment in any great detail because the investigation into what went wrong at ICAP commenced after I had left the firm. My knowledge is really limited to what I have been able to read in the FT and the media.
Q14 John Mann: It was a question to you, but it was not personalised in terms of your involvement. It is more that you have a good feel for the culture, the systems, etc. there, and therefore are in a good position to have a view in your current role—hence why you are here today—about what I would certainly characterise, and others have, as a more lenient and laissez‑faire approach. Is that going to be appropriate and effective for the future?
Mark Yallop: I was going to go on to say that I am surprised to hear you use the words “lenient” and “laissez‑faire” to describe the current approach to supervising conduct in markets, because it seems to me that in the last five to seven years since the crisis—not least with the recent discussions around the Fair and Effective Markets Review and the setting up of the FICC Market Standards Board and a host of other measures that we could talk about—oversight of markets by both regulators and firms themselves has moved a long way from the laissez‑faire approach that might have characterised the 1990s, or the period prior to the crisis. A lot of work has been done and significant progress has been made. I am not going to sit here and argue with you that everything is perfect, but the world that I see today as a former practitioner and now an observer from a regulatory chair bears no resemblance to the world that I remember prior to 2007-2008.
Q15 John Mann: The challenger banks and the building societies are kicking up about taxation, as you are very aware. If evidence emerges that competition is weakened by what the Chancellor is doing, how do you intend to monitor that, and how would you react? I am interested both in the non‑execs and yourself, Mr Bailey, in relation to that. What are you looking for in terms of a weakening of competition?
Andrew Bailey: I am looking for two things. The term “challenger bank” is, if you do not mind me saying so, a pretty broad umbrella. A lot of focus has been put on the rate of new entry, which is important—we have authorised 13 new banks in the last two and a bit years, which is up on past performance—but you have to be realistic. They are two things. They are an investment in the future. They can have bigger effects if they are, to use a popular phrase, using disruptive technology—they can sometimes have effects that are bigger than their own size—but they are an investment in the future.
Important though that is, there is a second thing that is more important, in some ways, in the near term. Let me just give you a simple point of comparison. Look at the balance sheet sizes of banks in this country. We have a handful of very big banks, with a total balance sheet often getting in excess of £500 billion. We then have two banks that I would call challengers but many people would not, which are Nationwide and Santander. I say “challengers” because in the current account market they are taking more market share. They have balance sheets in the range of about £160 billion to £200 billion range. The interesting thing is that there is nothing then until you get down to about £30 billion to £40 billion, and you have a set of institutions that are in that range. For me, therefore, one of the tests is: are we going to see some greater bridging of that gap? To run a full‑scale retail bank, £30 billion to £40 billion is quite small, and some of the things that we have seen go on in the past have been due to institutions over‑reaching themselves with balance sheets that do not necessarily support it. That is the second thing.
To answer your point, if we were to observe that the tax issue is going to be an effect here—the tax issue, by the way, kicks in if you make a profit of more than £25 million in a year; I think that is how it works—we would put that to the Treasury and, although we do not have the evidence today because it has not come in, if you were to ask us in the future, we would give you an answer. That would be the answer to that question. I do not know whether my two colleagues want to join in on this.
David Belsham: This is a more general point, but I have been very impressed with how the secondary competition objective is so well embedded in what the PRA does. As an outsider coming in, you know about the primary objective, but the secondary objective is taken into account in all of the policy‑making, which I have been impressed by.
Q16 Chair: I am very pleased to hear that, bearing in mind what a struggle it was to get it on to the statute book in the teeth of opposition from both the regulators and the Treasury.
David Belsham: When I had my interview, I was asked what I thought it meant, because I think the PRA itself was still coming to terms with what it would impact.
Chair: The last Governor was dead against it.
David Belsham: The secondary competition objective and the primary objective, to my mind, do not seem to come into much conflict. If you go down the primary objective route and have risk‑based capital, that in itself creates a level playing field. Removing “too big to fail” allows big companies to exit gracefully from the scene and create space. Having minimum regulatory standards immediately gives public confidence in the new banks that are coming in. The primary objective of what we are doing in terms of prudential regulation is a very sound base for encouraging competition.
Q17 John Mann: My question is more precise. The Government, rightly or wrongly, are making a change in the market. We get plenty of feedback—some of it is public; some of it is private—that this will hit the smaller players. What process are you intending to use to see whether this is impacting on competition, and therefore on your competition remit?
Andrew Bailey: Can I just be clear on our competition remit? Our competition remit is very much focused on the impact of our actions—and also inactions, by the way. Let us be clear on that: it is when we do things and when we do not do things. We are not a competition regulator in the sense that the FCA is or the CMA is. However, we do have a lot of close monitoring and intelligence on the banks, so if we were to observe the point you make in the future—it has not happened yet because it has not come in—we would spot it and I would raise it with the Treasury, and, if you were to ask me the question in the future, I would give you an answer. It is too early because, as I say, it has not come in yet.
Q18 John Mann: I appreciate that. So you would raise it with us.
Andrew Bailey: If you asked me, yes. We aim to be transparent.
Q19 Chair: If you are concerned, you should just come to us.
Andrew Bailey: Yes. I would also go to the Treasury and say, “Look, from our perspective”—and we follow these institutions closely, clearly—“this is what we are observing”.
Mark Yallop: If I may add, from my point of view, I do not think you should be in any doubt that we have the information to be able to see what is going on. We can see what is happening in terms of competitive pressures in the buy‑to‑let mortgage market or SME lending. We can see rates at which deals are being done and underwriting standards. The information is certainly available to make those kinds of judgments that you are talking about. The issues, in my personal opinion, will probably be how the challenger banks could cover the enormous distance that Andrew has just described between where they currently are and the scale that they need to really compete with the incumbent banks, and, equally importantly, whether their customers show the propensity to switch to the challenger banks, as opposed to the overwhelming evidence today being that they prefer to stay with the providers that they are accustomed to.
Q20 Mr Rees-Mogg: I have a couple of little points, Mr Bailey, if I may, before I come on to my main questions. On the relations between the banks and the press, it seems to me that it is perfectly reasonable for the banks to give their views on the whole concept of ring‑fencing, let alone its detailed application, through the press so that the public are generally informed, and that the regulator should not be upset by this. That is a perfectly legitimate way for any public body to behave.
Andrew Bailey: Of course. I am not arguing against a free press, to be clear. You are right to raise the point. My point is this: please do not use talking to the press as a way to talk to us. You are right: the press is important, but we are here to have these conversations directly.
Q21 Mr Rees-Mogg: So in only very limited circumstances you object to the press being used.
Andrew Bailey: No, I am in no sense objecting to the press—
Q22 Mr Rees-Mogg: If Barclays want to say they think ring‑fencing is a potty idea and that it increases their cost of capital, that is absolutely fine.
Andrew Bailey: Yes.
Q23 Mr Rees-Mogg: Good, okay. The second pedantic point I wanted to quibble about was a constitutional nicety. You mentioned that you were quoting parliamentary proceedings in your dealings with Barclays. I just wonder whether that might not clash with Article 9 of the Bill of Rights: that you are at a place out of Parliament and parliamentary proceedings ought not to be questioned out of Parliament.
Andrew Bailey: You are more of a constitutional expert than I am. I made an observation that a Government Minister—I think it was in the House of Lords—stated that the Government expected the sibling structure to be, if you like, the more normal presumption. I think it was in around October 2013. I am sure we can find this in Hansard, but whether that is a constitutional no‑no, I honestly do not know, because I am not an expert. It is on the record.
Mr Rees-Mogg: It is a constitutional note of caution, that the courts would not use something from Hansard except in very specific circumstances. The Pepper v. Hart judgment is the relevant one in the context of this. Places outside of Parliament—a regulator is quite close to being a court, because you have very specific powers and responsibilities—ought to be very careful in how they refer to Parliament. It is just a note of constitutional caution.
Andrew Bailey: I will take that.
Q24 Mr Rees-Mogg: I now want to come on to the proper questions I have for you, rather than the asides I have made so far. This is on the remit that you are going to get from the Treasury once a Parliament. The Chancellor told us on 22 October that you would not have to abide by the remit, so do you think you will be affected by the remit or that it is just a nice letter from the Chancellor?
Andrew Bailey: We are reasonably familiar with receiving remit letters in the Bank of England. It started with the MPC; there is now an FPC remit letter and, as you know, there is an exchange of letters between the Governor and the Chancellor, and it is a pretty well‑established routine. We take remit letters seriously. It is not something that we would ignore.
Q25 Mr Rees-Mogg: But it is slightly different, because the MPC remit is their remit that they have to follow, and if they do not follow it, they have to write saying why they have not succeeded on the inflation target.
Andrew Bailey: The point I was referring to is that the MPC has a secondary objective, which is subject to the primary objective of price stability, in relation to the Government’s economic policies. The remit letter there states what the Government’s economic policies are taken to be for the period, which I think is usually a year.
Q26 Mr Rees-Mogg: So you will view the remit letter in the context of a secondary remit, because you have your primary remit already by statute.
Andrew Bailey: Yes. In our case, we have a primary remit, which is safety and soundness, and for insurance companies there is policyholder protection. As we were discussing earlier, we already have a secondary objective in respect of competition, but the remit letter, I would expect, would also specify what the Government’s policies were, which are of relevance to the PRA.
Q27 Mr Rees-Mogg: What do you think will be the main thing the remit letter will be setting out for you? Will your risk tolerance be coming from the remit?
Andrew Bailey: We have not seen one yet. I would expect it would set out how we should interpret other areas of Government policy that touch on our activities. I am drawing on my experience as a member of the FPC here. This is more relevant, in some ways, for the FPC and the PRA than it is for the MPC, because there are more what I might call touch points of financial stability policy with other areas of public policy. It is not that monetary policy is an island, but it probably is a little bit more self‑contained in that respect. If the Government have policies on housing, for instance, or on retirement savings, all of those are relevant to things that we are involved in too.
Q28 Mr Rees-Mogg: You are confident that the remit letter will not undermine your independence. It will not be giving you too clear a steer as to what you are supposed to be doing.
Andrew Bailey: I am confident that the concept does not, because we have had experience of it, but we have not yet seen how the Government would propose to structure it. It is too early.
Q29 Mr Rees-Mogg: Finally, it is going to be every Parliament. Do you think that is enough, or would you like one every year, as the MPC has?
Andrew Bailey: There is a provision that the remit letter would be set for a Parliament, which is sensible if you think about the transmission and the timeline and effects of our actions, but I think I am right in saying that there is a caveat in there that it can be re‑interpreted during the life of a Parliament, if that is sensible.
Q30 Mr Rees-Mogg: When are you expecting the first one?
Andrew Bailey: Formally, the first one would not arrive until after the legislation is in place. As you can imagine, there will be conversations before that on the form and shape of it.
Q31 Mr Baker: You have recently consulted on the stress tests. What aspect of the stress tests are banks complaining to you about?
Andrew Bailey: It is worth looking at the stress tests as having two parts to them. We tend to call them the quantitative part and the qualitative part. The quantitative part is all about what numbers come out at the end of the process, and we are into the later stages of this year’s stress test. As you probably know, we are going to announce it in early December.
The qualitative part is where we often have more conversations of the nature you refer to, and it something we have talked extensively to the Federal Reserve about, because they have had a few more years’ experience than us. That is all about things like how they marshal their data, how they run their models, the quality of their risk governance, the quality of their senior management and the quality of the board’s involvement. In the first year—and I do not think this should be surprising—a lot of the conversations of the sort to which you refer were, frankly, about how you marshal the data, because this was a process that had not been done before. We are still having those conversations, and we still have a long way to go in terms of our own management of that side of things, but we have had relatively fewer of those this year than we had before.
By the way, with the quantitative results, we do track the performance of the institutions on a qualitative basis. That is something, again, that the Federal Reserve has done. We are not yet at a point where we are going to publish those results, because, frankly, we are still working out the way to do it, but you may be familiar with the fact that in the Federal Reserve’s so‑called CCAR, which is their stress test, they fail more institutions on qualitative grounds than they do on quantitative grounds.
Q32 Mr Baker: You mentioned marshalling the data. How do they marshal the data?
Andrew Bailey: They have to use their own systems. If you like, they are knitting together the risk management and data systems that they have to do that.
Q33 Mr Baker: Are their concerns just procedural, or are they complaining about the substance of what you are testing?
Andrew Bailey: Some of the concerns are around the scale of the demands that we put on them. My sense is that this was more of a last‑year issue than a this‑year issue. The reason it was a last‑year issue was that the whole thing was new. We run it, as you may know, on a pretty long timeline. We announced the scenario back in the spring; we have been working on it ever since. We are into what you might call the last stages of it now, because there are literally only about two or three weeks to go now; it is approaching its final lap, as it were. We run it over quite a long time, precisely for these reasons; it is not a straightforward thing to do, data‑wise.
Q34 Mr Baker: Could I just bring in Mr Belsham? Do they talk to you about these stress tests? What do they say?
David Belsham: No, I have not had any contact with banks over them. My view, coming from the insurance side, is that stress testing ought to be a core competence of banks. This concurrent stress test does take a lot of resource, and that is because it is in its second year and still bedding down. Over time, one would want to place far more reliance on the bank’s own so‑called ICAAP process, where they do their own stress testing—a wide range of different stresses—to test what is the specific scenario that they are exposed to. We are applying one cyclical scenario each year and an exploratory one every other year, but we really need the banks as well to be casting their net wider, and then they report that into us as part of the normal process.
Q35 Mr Baker: Mr Yallop, what I am trying to get at is where the banks might be vulnerable and therefore are complaining that you are probing them in sensitive places. Have they talked to you about this aspect?
Mark Yallop: Yes, I have had a few conversations—many fewer on this topic than on other things that we do. The few that I have had suggest there are both procedural and more substantive questions, although not necessarily legitimate concerns procedurally. As David says, banks have been used to stress testing for a number of years for the ICAAP process that they are obliged to undertake.
Procedurally, the demands that we are making of them in assembling data, as Andrew says, across multiple systems, multiple locations and probably multiple legal entities for the larger firms, are not trivial, given the generally poor state of many banks’ technology infrastructures. That is a problem, as Andrew says, that will improve over time, because the more they are forced to repeat this, the better their systems will become at managing data uploads to us. It is also fair to say that some of them probably are being pushed to investigate deeper into the risk characteristics of some of their portfolios than they might otherwise do if they were simply performing stress tests according to their own design, rather than the parameters that we are choosing for them and asking them to explore. That is an entirely natural tension between bank and regulator, and you would expect that. It would be surprising if that did not exist.
Andrew Bailey: If I could supplement David’s point, we have observed—we run insurance stress tests as well—that it is easier to run multiple scenarios with insurers than it is with banks. With banks, we are in essentially a one scenario with risks around it thing, for precisely the reasons we have been discussing. To answer David’s point, it has been embedded in insurance—both life and general—for much longer, because of the way the businesses are run; you can run stress tests more easily with multiple scenarios.
David Belsham: There is a far bigger range of risks in insurance, from general insurance through to catastrophe, equity and longevity. A one-size-fits-all PRA stress test would not work.
Q36 Mr Baker: I think I have just heard that it would be easier to conduct multiple scenario stress tests on banks if their internal IT infrastructure was higher quality and they could marshal the data more easily. Is that a fair summary of what I should infer from what you have just said?
Andrew Bailey: We would expect to make progress over the coming years on having the ability, as I say, for them to marshal the data process more easily. Interestingly, the Federal Reserve has been consulting this year on this whole question—not just the future of their tests but, within that, the question of how many scenarios you can run. They are pretty much in the same place as us.
Q37 Mr Baker: I am just trying to tease out this: is it the case that if they invested more in sorting out their infrastructure, which I have personal experience of—I have seen the mess that they get into over many years—that would help you to run multiple stress tests and have greater confidence in their robustness?
Andrew Bailey: Yes. An improvement in the systems environment would benefit that. It would benefit them as well. One of the key things to say about stress testing—David and Mark have made the point—is that it is not just something we do to them; it is something they do to themselves.
Q38 Mr Baker: Could I ask the non‑executives: what did you learn from 2015, and to what extent have you been involved with setting up the present and future frameworks?
David Belsham: We have had joint discussions with the FPC on this, in deciding how the approach should be adopted, so we have been fully engaged in the process. We are happy with the document that has come out recently, and the stress testing will be used by the FPC to set counter‑cyclical buffers. It will also be used by the PRA to set the so‑called PRA buffer. It is a single approach, but serving two valuable purposes, as well as giving us a complete, transparent demonstration of the resilience of banks, which is fantastic.
Q39 Mr Baker: On that point of resilience, could you characterise how important you think the stress tests are in preventing another crisis?
David Belsham: Stress tests are probably one of the most useful things that firms do to themselves, and that the regulator does, because they are so transparent. All the different insurance firms and banks will have their own array of models, but a stress test allows normal people to judge. They can look at the stress test, take their own view of how strong it is—whether it is one in 50, one in 100 or one in 200—and make their own assessment of whether that means the firms are resilient enough.
Q40 Mr Baker: One of my colleagues, Professor Kevin Dowd—I should say friend, really—has been very critical of stress tests, as I am sure you all know.
Andrew Bailey: I am very aware of it, yes.
Mr Baker: Are you concerned that that degree of comfort that you have just referred to might be based on false premises—that, for example, there is only one stress test scenario being processed? Are you concerned that it could be too simplistic? I would like to ask Mr Belsham and Mr Yallop first; I hope you will forgive me. Is it too simplistic? Is it a false sense of comfort about the resilience of the system, because the data that are necessary and the range of scenarios that are being tested are not realistic?
David Belsham: I have taken great comfort from it. Coming in, I was fairly sceptical to start with. Having been in a firm where you see dozens of different scenarios, to come in and focus on one seemed to me slightly odd. But it is right that for a lot of the UK‑based firms, the stress that is being applied to them is very close to biting stress, and the follow‑up stress that is being done this year, capturing risks arising in the Far East and in Europe, again helps to pick up the biting scenario for other firms. Going forward, at some point, there will probably need to be a stress test of US exposures as well that comes in. I have taken comfort, having started from a degree of scepticism along the lines you are talking about.
Q41 Mr Baker: Would you like to add anything, Mr Yallop?
Mark Yallop: I would echo a lot of what David said. In answer to the question you posed a moment ago, about what I contributed to or learned from the process, one thing that particularly struck me about this year’s exercise was the introduction of the traded risk scenario, of which I have personal knowledge and experience as a result of what I have done in previous lives. I went into a meeting to review the scenario and how it was going to be conducted with a lot of questions in my mind, and was rather pleasantly surprised to find out that pretty much every question that I had about how we were going to treat issue X, Y or Z was being answered in the scenario. You can argue that a lot of the risk in banks lies in their credit exposures and the on‑balance‑sheet assets more than it lies in the traded risk books, particularly now that traded risks have been reduced in scale post the crisis. Nevertheless, I thought that was a very useful innovation, and it is something we should certainly continue. The opportunity to look more into the emerging-markets risks this time than in 2014, when we were more focused on UK economic disruption, was useful learning as well.
Q42 Mr Baker: I am conscious I must keep moving. Monetary policy will tighten at some stage, however limited and gradual the process may be. Andy Haldane has previously said that IFRS was pro‑cyclical. Are you concerned that there will be stresses arising from the mark-to-market process in IFRS when interest rates begin to rise?
Andrew Bailey: That is one of the reasons why we have specifically set up a counter‑cyclical buffer structure: so that we would be able to see that. The point about the stress test is that it is a device to throw up problems like that. Coming back to Mark’s point about the market risk part of the tests, if you saw that happening, it would be one way into closing the problem.
Another example would be this. Let us say that you do the interest rate shock—we have done these tests—and you find that the probability of default is different in the owner‑occupier market than in the buy‑to‑let market. You do not just say, “Well, okay. Fine. Done. That is good. Move on.” It inevitably poses the question, “Well, why is that? What is it telling us?” This is what the stress test is there for, in a sense. It is rather like the MPC forecasting process, but in a different part of the distribution of the curve. It is a device to throw up the problem and say, “Well, what do we make of that? Do we think it is robust? Do we think it is not robust? If we think it is not robust, what does it tell us, and what do we do about it?”
Q43 Mr Baker: In what sense is Kevin Dowd wrong when he says that these stress tests “are highly unreliable and worse than useless because of their tendency to provide false risk comfort”? I find myself, as a member of this Committee, listening to Professor Dowd say things like that, and yet here you are—highly respected, talented people in senior positions—taking just the kind of comfort that he is rubbishing.
Andrew Bailey: The difference is that his argument rather rests on taking a very literal view of the stress test: that you crank it through a machine, you get a number out and that is done. You can say the same thing about the Monetary Policy Committee. If you thought that the forecast was a mechanical process out of which followed a result, out of which followed a policy decision, and there was not much intervention in the middle, you would say that was a very naïve process that was, frankly, open to problems.
Q44 Mr Baker: So it is about confidence in the regulatory process and your ability to adjust to the results.
Andrew Bailey: Yes. It is about what you do with it and how you use it in your more judgmental process. I have not heard Kevin Dowd’s riposte to that, so I do not know what it would be, but that would be my response.
Chair: I am sure you know the riposte.
Q45 Mr Baker: I am conscious I need to ask something about Solvency II. I think you said, Andrew, that it was going to be the biggest single thing you were doing this year, and you have repeated that. What are you hoping to achieve, and how are your efforts going?
Andrew Bailey: It is true. It remains the biggest single thing we are doing this year, and we now only have about six weeks of the year left to finish it, because it has to be implemented by 1 January. We are on course. Frankly—I want to be careful about what is interpreted from this—to a degree, it is “heads down, get to the finishing line” at this point, because we have to do it on time.
Out of that, I have to tell you, has come a lot of learning. I might hand over to David in a moment, because he is an expert in this field. There are, to my mind, revealed flaws in Solvency II that have come out of it, which we need to address, and they mostly need to be addressed in the legislation. It will be done across Europe at the end of this year. I would strike two notes of caution related to what I hear. You will hear firms reporting a number, which is essentially their capital position relative to the minimum requirements; if the minimum requirement is 100, it will be 100 plus whatever. You will hear firms across Europe doing this, and there will be an awful lot of commentary on why one firm is at one number and another firm is at another number. I would be very cautious—and I would strongly advise the market and analysts of this—about over‑interpreting. There are a lot of apples-and-oranges comparisons in those things. How different firms get to their numbers is something I would not over‑interpret. We are interested in whether the firm has met the standards that we require. I might have said this before in this Committee: for us, it is not as big a leap, because we are moving from essentially an interim capital regime that was put in place post Equitable Life, whereas some of the rest of Europe is moving from Solvency I, which is a much older regime. The first point is that I would be really cautious about the apples and oranges point.
The second thing is that I sometimes hear people saying, “You have to have a number that is 160 or 170”. Not in our book. Our book is, “They meet the capital requirements”. I am not interested in individual numbers. Again, because of the things that have been revealed, I would be worried that people were over‑interpreting those results. That is a challenge for us. I have an expert sitting to my left.
David Belsham: One issue that has come up is the risk margin, which is far too high in an environment of very low interest rates. The calculation is specified in the rules; it was calibrated some years ago, but, with interest rates having fallen so far, it is just too high and needs revisiting.
Q46 Mr Baker: Does that not imply a very considerable risk, then, as interest rates start to rise, if one were to adjust the figures to account for a very low interest rate?
David Belsham: You would have to calibrate it so that it worked properly as interest rates moved.
Q47 Mr Baker: Forgive me, because I really hear echoes here of the scandal we have had over interest rate swap mis‑selling. In evidence, we heard that one of the problems was that those who sold interest rate swaps never accounted for the possibility that interest rates would be this low, and particularly for this long, and that one of the reasons why these products have been mis‑sold is that the interest rate environment has been where it has been. I think what I am hearing from you is the possibility of adjusting the regulatory environment in order to take into account the monetary policy.
Andrew Bailey: No. That is not correct.
David Belsham: The risk margin is supposed to be the allowance for risk in the insurance liabilities, and so it is an addition to get to a cost of transferring the liabilities to a third party. When it was calibrated several years ago, it gave a reasonable cost of transferring annuity liabilities to a third party. Interest rates have gone down very low, and that calibration has not really captured how the change in transfer price would have moved, and we have ended up with a very big risk margin that means that firms can transfer their annuity liabilities to a third party for a significantly lower sum. Because it is mis‑calibrated, it is just not meeting its purpose in the current low interest rate environment. The calculation is specified in the rules, so we need to go back to the rules to revisit that.
Andrew Bailey: It is a little bit a way of saying that the cost of resolution, if you think about it in those terms, is too sensitive to interest rates.
Q48 Mr Baker: Okay. So, if you were to recalibrate the environment, you would just continue to adjust it as interest rates gently came up.
David Belsham: Exactly. It should be interest rate sensitive, but not as interest rate sensitive as we have got. There is another interesting issue with the yield curve between the UK and Europe. In the UK, we use the actual swaps yield curve to discount liabilities, whereas in the Eurozone the actual yield curve is used out to 20 years and thereafter it is extrapolated to a long‑term rate of 4.2%.
Andrew Bailey: This is my apples and oranges point. This is a very important point.
David Belsham: In current market conditions, extrapolating a Eurozone yield curve out to 4.2% is a very sporting assumption, compared to the reality of the yield curve, which goes out to about 1.5% at 50 years. To answer Andrew’s point, when you compare a company with Eurozone liabilities to a UK company, that difference of yield curve is having a very substantial effect on the liabilities. It is items like that that need to be ironed out. Overall, I would say Solvency II is a good framework. It is market consistent, it is risk based, and it is calibrated to a 1 in 200 level over one year. As a regulator, we are very happy with that, but these glitches do just need to be ironed out over the coming years. In the meantime, firms have transitional arrangements, so that protects their balance sheet from any sudden change.
Q49 Mr Baker: Forgive me. You have explained the significance of the process and the regulations and, Andrew, you have set out how much work it is and that heads are down to get through to the deadline, but we had Jim Bichard, a partner at PwC, say, “Whatever happens, Solvency II has been and continues to be a huge distraction”. How would you respond to that?
David Belsham: It has. It has taken a very long time to implement. There were delays in finalising the rules, and that made it very difficult for firms to do. I think firms will be very pleased to implement Solvency II on 1 January and move on to efforts going back to running business as usual. They will all have benefited from a better understanding of their risks. Going into this, I think most firms thought, “We have a Pillar 2 arrangement; Solvency II will not be much different from that”, but most firms, having looked into it, have found that there was a lot more to understand in their various risks, and their business models and their business and risk management will have benefited.
Q50 Mr Baker: So you do not see this comment as a criticism of the regime in itself; you just see it as a comment that the act of changing the regulations has simply required—
Andrew Bailey: It is factually true, I think we would say.
David Belsham: It is true, and I think people want to now move on and let it bed down and get on with it.
Q51 Chair: That is not quite the same point. “Okay, we have taken this pain; we might as well move on.” That is a slightly different point, is it not? That is bowing to the inevitable, rather than saying, “This will benefit us”.
David Belsham: Solvency II, as I say, has a 1 in 200 level of confidence. That is the level that the PRA is happy with, and that is now being applied across Europe, creating a level playing field.
Q52 Mr Baker: On that point, the Dutch are regarded as particularly tough. Do you think there is a level playing field across Europe?
Andrew Bailey: No, because of the point about the euro curve that David has just mentioned. This is a big point.
Q53 Mr Baker: So, our country is categorically different to the Eurozone, then, in this regard.
Andrew Bailey: There is a categorically different treatment of the sterling curve—and this is important from the point of view of discounting—relative to the euro curve. The Europe curve has, as I say, this fixed point in it at the end, which is not in line with current market rates.
Q54 Mr Baker: How seriously should we take threats by insurers to leave the UK because of these regulations?
Andrew Bailey: I do not know what threats you get.
Mr Baker: We have had Prudential say that they would go outside the EU.
Andrew Bailey: Companies that have extensive operations outside the UK—and they do—review these things. I have not had any company come to me and say that they are leaving, but when you are running a very big Asian business and a very big US business, you probably do think about it. We know another institution that is doing the same thing in another industry that we regulate.
Q55 Mr Baker: So you are not concerned that we will see a reduction in the scale and scope of insurance business in the UK as a result of these regulations.
Andrew Bailey: At risk of broadening this out too much, the bigger issue in the UK is what the retirement savings market is going to be in the future. Solvency II is an issue for the industry, but if you think about the longer‑term trends in this industry, one of the telling statistics that we have is that the net cash flow in traditional UK life products has been negative for about the last eight years. Essentially, there has been a switch out of some of the more traditional products, and we know some of the issues they have had with it. There has been a switch out of the traditional with-profits industry, and we have had the end of compulsory annuitisation, which is another issue. The bigger issue the UK industry faces—and different companies will react in different ways to this—is, “What does it look like in the future?” Arguably, it looks more like an asset management industry in some ways in the future, but this is the much bigger issue.
Q56 Mr Baker: I am not sure that is the question I asked you, though. Do you think we will see insurance companies move out of the UK and the EU as a consequence of the Solvency II regulations?
Andrew Bailey: There are two ways to answer this question. There is a purely locational answer, which is: “Where are they going to have their head office?”
Mr Baker: I think we all care very much where they are going to deliver services and where they are going to pay tax.
Andrew Bailey: Absolutely. That is why I slightly switched the discussion, because the bigger issue is what services are going to be available to the UK population for the long term in our retirement savings market. I do not think Solvency II is the big driver of that. There are much bigger issues at stake in the domestic world.
David Belsham: I do not believe that the 1 in 200 level of confidence, which is the benchmark, should be sufficient to drive insurers overseas. It is equivalent to a BBB rating. To go below that, you would have insurance companies with sub‑investment‑grade credit ratings. I feel that one in 200—BBB—is a good standard to apply anywhere in the world. I do not think insurers could argue that it is excessively strong for them and would lead them to go elsewhere.
Mr Baker: We will look forward to seeing if they respond to this evidence.
Q57 Chair: Just to be clear, Mr Belsham, this yield curve differential is not a glitch that is going to be ironed out, is it? We are not talking about half a per cent. difference in those numbers; we are talking about a factor of three, or two and a half, or something.
Andrew Bailey: It is in the legislation.
Q58 Chair: “Yes” or “no” to the point I have just made?
David Belsham: Yes, it is a substantial difference.
Q59 Chair: So it is not a glitch. This makes a nonsense of a great deal of this work, does it not? There is so much difference between our treatment and the treatment of continental Europe. We have not arrived at the point where we are anywhere near a single market.
Andrew Bailey: That is why I made the apples and oranges point. That is important.
Q60 Chair: You came to us and we had an extensive set of exchanges two and a half years ago about the shambles that was Solvency II. You described the process at that time as “shocking”. I described some of this as pretty bad as well. I cannot remember the language I used.
Andrew Bailey: You used language as well.
Chair: Well, that is our stock in trade, I suppose. You also said that the Germans would take at least a decade to adjust to any of this and that the French would be unlikely to agree to anything that you would consider prudentially acceptable.
Andrew Bailey: I do not think I named the countries.
Chair: I have your letter in front of me and I will not read out that passage unless you want me to, but I have more or less paraphrased your letter.
Andrew Bailey: I think you may have paraphrased it.
Chair: I think I had better read it out. You said, “In particular, it is unclear to us that the French authorities will now be able to agree any directive that we consider prudentially acceptable”. Then, swiftly, you moved on to Germany: “Germany in contrast has long been supportive of a relatively prudent agreement on Solvency II but needs a long transitional arrangement (e.g. 10 years or more).” That really is a long transition.
Andrew Bailey: Well, Solvency II transition was 16 years.
Chair: “To allow its domestic industry to adjust.” Then you go on to the description of what is euphemistically described as their “large ‘back book’” of guarantees. This does not look a very happy state of affairs, does it? Was this worth the candle?
Andrew Bailey: Let me just make one point.
Chair: Okay, but then answer the question of whether this whole process has been worth the candle. I would like it noted for the record that both your wingmen are now smiling broadly.
Andrew Bailey: You can see them better than I can. I cannot remember exactly when we had that exchange of letters, but it is a while ago.
Q61 Chair: Yes, so what I am really asking is how much things have improved since February 2013.
Andrew Bailey: To David’s point, the thing that has changed since February 2013 is the Euro curve. I cannot remember exactly what this gap that we pointed to in the long‑term curve was in February 2013, but it would have been a lot less than it is today. That is the thing that has become more acute in the period. Is Solvency II worthwhile? I think the answer is yes, for the reason that David gave earlier—that it is a more robust risk management system than has been there in the past.
The other thing I would say—and I have been saying this for quite a while—is that it is the case that because of the reforms that the FSA introduced post Equitable Life, moving to the so‑called ICAS system, there are two things that follow from that. One is that in some ways, the leap for us to Solvency II is less than the leap for other countries, but the other way of looking at that is that it is less necessary to make that leap, which is really the more direct answer to your question.
Q62 Chair: Do you have anything you want to add, Mr Belsham? Yes, you have.
David Belsham: I would just say that the Pillar 2 regime, as Andrew said, was good. They are not often praised, but the Pillar 2 regime was the best thing the FSA did, in my view—moving to a market-consistent 1 in 200 regime for insurance. If you look back at the credit crisis of 2008-2009, we found that insurance companies had a good capital framework, were well‑capitalised, had strong liquidity and had a good understanding of their risks. Solvency II is an extension of that across Europe, so it is good.
Chair: We will not take that as too bad a criticism of all the other things the FSA did and we will move on swiftly, since the organisation no longer exists.
Q63 George Kerevan: I have a couple of questions while we are on insurance, because I am always aware we do not look at the insurance sector enough. There is still something in Solvency II that I cannot quite get my head around. You are talking to 20‑odd banks, approving or not approving individual risk—
Andrew Bailey: 20 insurers.
George Kerevan: Yes, insurers. If those separate regimes are agreed, are they more or less onerous than the general rules, or Solvency II?
Andrew Bailey: The standard model approach.
George Kerevan: The standard formula.
David Belsham: It will vary across the risks. The standard formula is designed to be correct for an average company with average risks. It does not work well, for example, for a very well diversified company—a company that has lots of different risks that are quite independent of each other. The standard formula would force the company just to add up all those amounts of capital, whereas an internal model allows you to take diversification benefit. It will vary from company to company. Some companies have internal models that give a similar answer to the standard formula, and there will be others that come out higher or lower, according to their specific risks.
Andrew Bailey: There are cases of companies that are using the standard approach, where we are having to say to them—because this is an add‑on to Solvency II—“I am sorry, but there has to be an add‑on, because the standard approach just is not tailored to your type of business”.
Q64 George Kerevan: In what way do we then have a level playing field, either here or across Europe?
Andrew Bailey: It is risk‑adjusted, always. There is a long debate in the banking world about how level the playing field is between the standardised approach and the internal models approach. The one thing I would say about insurance—David is much more an expert on this than I am—is that modelling has always been at the heart of insurance, going back to life tables. The insurance industry has a longer history of modelling in that sense. In some ways, I suspect the choice between standard and internal models for some firms is, “Can we pipeline the models that we do use into the standard approach, rather than have the cost of introducing a Solvency II‑compliant internal model?” That is a reasonable point. For small banks, again, we would say to them, “Please do not assume that internal modelling is a cheap option”. It certainly is not. There are expensive overheads that go with internal modelling.
Q65 George Kerevan: All that said, then, are there circumstances that you can envisage in which you would instruct an insurer to increase their capital as a result of that internal modelling?
Andrew Bailey: The Solvency II regime is a bit different to the bank regime. In the bank regime, we approve the output from the model, which is the capital calculation. In the Solvency II regime, we approve the model. That is quite a fundamental difference.
Q66 George Kerevan: You may need to elaborate on that. I may be misinterpreting, but my immediate response to that would be, “So you go through the exercise, but it does not actually lead to a regulatory result”.
Andrew Bailey: What it does is—to answer your question directly—constrain our ability to come in and say, “Come on, you must be kidding. That is not the right answer. Here is an add‑on.” It is more constrained in the Solvency II world. You have to go back to the model, as it were, and tackle it at that level.
Q67 George Kerevan: Does that mean that in some shape or form the PRA would instruct a company to do something different?
Andrew Bailey: We could, and do, say, “Look, that part of the model is not robust enough”, yes.
David Belsham: The model approval process has to go through a series of standards and tests that are laid down in the Solvency II rules: “Is it used by the company to transact business and set capital, and so on? Has it been properly calibrated? Is it using sensible data? Does it have the right underlying methodology? Has it had an independent validation by someone who has not actually developed the model? Is it properly documented and understandable?” There is a whole series of standards. Once that is met, the output of the model is the capital required for that firm.
Q68 George Kerevan: Is there any point in this process at which the behaviour of the company is modified as a result of the exercise? Forgive me; I am genuinely trying to understand. Does it ever happen that, in the process of the negotiations, a situation emerges in private where they are not conforming to the model, they therefore cannot get approval, and they have to change things before they get the seal of approval?
David Belsham: Yes. If there is something inadequate in the model, that would be fed back.
Q69 George Kerevan: So there is a behavioural modification.
Andrew Bailey: Oh, yes.
Q70 George Kerevan: Would that, for instance, explain why there were market rumours in August about how well several of the larger insurers were going through the approval process, and significant share drops in that period? Is that because the market was interpreting that the firm would have to change its model to get approval?
Andrew Bailey: No, I do not think so. A few things happened in August. There was wider instability in financial markets that would affect insurers as asset holders; there was the Chinese thing. We saw a bit of activity in August, when one or two continental insurers came out and announced what they thought would be their Solvency II capital numbers, going back to my point about the 150s, the 160s and the 200s, and the numbers were coming out lower than the market had expected. That had a bit of an effect. There was a Dutch insurer that did that, and there was a bit of a market reaction to it, and that then got translated as, “If they are in that position, is that carrying over to others as well?”
Q71 George Kerevan: I feel we will come back to this after January, when we see the results. Another point, while we are still on insurance. Could you explain the difference between the stress‑testing regime for insurance companies and that for banks? Is it specific to a company rather than a general market approach?
Andrew Bailey: I will let David reply. By the way, there is one regime for life companies and another regime for general companies; there are different regimes.
David Belsham: There is a stress test running at the moment that has been sent out by the PRA for general insurance. That is co-ordinated across all general insurers who are testing a series of specified scenarios—wind storms and so on. On the life side, every other year a set of scenarios are drawn up by EIOPA to get all the firms to investigate their capital position in that scenario. Generally speaking, as I was saying earlier, I would have expected insurance companies to have been doing stress‑testing for many years, typically by taking one risk at a time—whether it is longevity, interest rate or equity—and stressing that to an extreme point. That shows them what their risk profile is. Then they would build their own scenarios of combinations of those events, and that shows them what their key risks are; what, perhaps, they ought to reinsure; where a new risk line could sensibly be written, taking account of diversification benefit; and so on. It is a key tool for insurance companies.
Q72 George Kerevan: Is it a key tool for the PRA?
Andrew Bailey: Yes.
Q73 George Kerevan: How do you use the information?
Andrew Bailey: Very much the same way as we do for banks. It goes back to the conversation with Mr Baker. It prompts us to ask questions. The nice thing about insurers, as David was saying, is that you have multiple scenarios. In the life sector, you tend to have a situation where more of the multiple scenarios are applicable to one firm. In the GI sector, it tends to be a little less like that. You are trying to come up with multiple scenarios to cover a pretty heterogeneous set of risks. There is a solar-flare scenario, believe it or not. You have to be insuring that business to be affected by that shock, so that is a bit different. Again, it prompts us to ask questions. I will give you an example. You may have read about this. In the GI world, there has been quite an increase in firms writing cyber‑risk policies, so this is essentially business disruption risk resulting from cyber. The question we would put to them is: “Let us understand the risk that you are writing and how it responds to that type of shock and stress event”.
Q74 George Kerevan: Are there very general lessons you draw from the wider experience with stress-testing in insurance that you could apply to stress‑testing in the banking sector?
Andrew Bailey: Taking this back to some of the earlier conversation, I would like to get to the point where we could run multiple scenarios for the banks. We are not there, and we will not be there for quite a few years, because of the data process and the complexity of it. It is more complex, because on the whole, more of the insurance stress test is done on the liability side of the balance sheet, and the risks are rather more homogeneous, whereas in a bank stress test, you are running it on the asset side of the balance sheet, where there are more risks and they are more heterogeneous. That is a more complex business.
Q75 George Kerevan: Are the two sets of stress tests discussed separately, or have you had discussions with folks within the organisation about how they overlap and how you can draw experience from both sides?
Andrew Bailey: This is one of the things we benefit from in being a regulator of banks and insurers. We benefit from understanding the different methodologies and the different approaches, and we put those to work. They are running on slightly different sequences and timing, so we do not tend to bring the whole thing together into one stress test festival, as it were, in terms of results. We do not do that.
George Kerevan: “Stress test festival”. That is the phrase of the afternoon.
Andrew Bailey: What a thought. There is a lot more to come from this, by the way, in terms of cross‑fertilisation of the lessons from the process and doing it and how you think about scenarios.
Q76 George Kerevan: Beyond insurance, which was my biggest issue, there are a few other policy areas I would like to mop up. I am interested in the relationship between the PRA and FPC now, particularly in relation to the question of capital buffers. When you are setting capital buffers at the PRA, to what extent are you acting independently of the FPC?
Andrew Bailey: This is a really interesting question. This is something of an emerging and evolving landscape, because it is a pretty big focus of attention at the moment as we come to setting a number of buffers. There is the capital conservation buffer, the systemic risk buffer for those affected, and the counter‑cyclical buffer. We will be publishing some more on these things in the course of the coming months, particularly the systemic risk buffer, which we have to consult on in the first half of next year.
The interesting question here is when we think about this, and we think about how we apportion the pieces of the picture and the methodologies, the PRA—and before it, the FSA—has been running stress tests for some years now. We have moved on in recent years, because we are now doing the so‑called concurrent stress tests, but we have been setting what we call Pillar 2B capital planning buffers based on stress tests for some time in micro‑prudential supervision. The really interesting question, which we are thinking our way through at the moment, is: you start with that position and then you say, “I am going to set the counter‑cyclical buffer. I am going to have a methodology for doing that. I am going to set the systemic risk buffer and the capital conservation. How do I think about the future world? Do I just say, ‘Well, we have this PRA buffer, and that is micro, so we will just leave that there and then we have these macro buffers coming in’ or do we say, ‘Actually, in the past, we have been setting this PRA buffer, and that is done from a stress test. How much of that, which we have been doing all under one banner in the past, is really micro and how much of that is macro?’” We will say more about this when we have thought our way through the issue—and that is “thought” in the sense of thinking, not fighting.
Q77 George Kerevan: You sit on the FPC as well.
Andrew Bailey: I do.
George Kerevan: Are we nearing a change in the buffers?
Andrew Bailey: The counter‑cyclical buffer?
George Kerevan: Yes.
Andrew Bailey: The committee is more or less in session at the moment, because we are going into the Q4 policy round, so I cannot presume to judge what the outcome is. We set it every quarter; it is just that we have set it at zero up until now.
Q78 George Kerevan: But you think the relationship between the PRA and the FPC in setting the buffer regime is working well.
Andrew Bailey: It is working better and better. I say that in the knowledge that every time we think about this, the problem gets deeper, but I am an eternal optimist. Genuinely, I think we are getting a much better understanding of the interaction between macro and micro as we do it.
Q79 George Kerevan: I will move on to something sexy: the bonus cap. You are on record as being of the view that bonus caps are bad policy in general, for various reasons—the lack of flexibility in being able to set remuneration that helps you deal with risk and so on—but we have a bonus cap, set by Europe. You expect the banks to be compliant. How does that work? What would you expect the banks to be thinking if you, as a regulator, are saying you are against the bonus cap but there is one here in Europe? Is there not a friction there?
Andrew Bailey: The discussion is quite straightforward in one sense. They know my view; they know my colleagues’ view on this, but they also know they have to abide by European law. I have been quite clear—this is on the question of allowances—and I have said, “Look, you have to do it. We can separate the two discussions as to what the framework should be, but you have to abide by European law. It is very clear.” There are two debates going on at the moment. One is this one, which is the big one. The other one you may have read about is the very important issue of the application of proportionality in this framework and how far the European Commission, particularly, wants to push this down so that it is the same framework applied to small firms and big firms. Whereas there are more people against us on the issue of the bonus cap, there is more consensus outside the European authorities—the European Commission, anyway—that going against the principle of proportionality is a bad thing.
Q80 George Kerevan: Do you regret taking a public stance on this?
Andrew Bailey: Absolutely not, no. I still think that it is a piece of bad policy, because it creates the wrong incentives, and I think—but I cannot point you definitively to this—we are beginning to see the increase of fixed remuneration in European banks. There is some evidence that we are beginning to see that.
Q81 George Kerevan: Can you give me some examples?
Andrew Bailey: I cannot give you examples off the top of my head. There is some evidence emerging and, when we can put it together, I will certainly write to you. I am not sure we quite have it yet, but we may get it when we go into next year after we go through another round of remuneration.
Q82 George Kerevan: You told us in March that you think there should be further legal protection for whistleblowers. Have you thought further on that?
Andrew Bailey: We are most of the way through implementing the recommendations of the Parliamentary Commission. We have discussed this in the past—it was with the previous Committee, before the election. The one that we do not agree with is paying whistleblowers. We do not agree with that one and we have not implemented it, but in other respects we are well advanced in implementing the recommendations.
Q83 Chair: We did not argue for a requirement for payment. We asked you to examine it and consider it.
Andrew Bailey: You raised it as an issue, yes, and we did.
Chair: Just to be clear.
Andrew Bailey: No, indeed. Sorry.
Q84 George Kerevan: My final question, Chair, is moving on to credit unions, which is an interesting one. You have just conducted the annual assessment. I wonder whether you might bring us up to date on any lessons from the annual assessment of category 5 credit unions.
Andrew Bailey: There are over 500 credit unions in this country, with a slightly greater concentration in Northern Ireland than the rest of the country, due to tradition. I always say that we do not operate a “no failure” policy; it is “orderly failure”, and credit unions are the one area where this does happen. Typically, somewhere between five and 10 go out of business each year, and some new ones come in; it is not one‑way.
To be frank with you, it is a good thing that we have credit unions in this country. We have a rather under‑developed credit union industry in this country relative to other countries. I always make the personal observation I am a customer of a credit union in Idaho in the US by virtue of marriage, and I do a lot of business with them and they provide me with a much bigger service. They are a bank, really. If you go to the US, there is a much bigger infrastructure around credit unions than there is in this country. Mr Mann and I have had conversations in the past about mutuals. If we want this in this country, we have to do something more about it, although I am not sure it is really for us to do it. We have 500 credit unions; if we want to make more of mutuals, then we have to do more about the infrastructure and the support for them.
Q85 George Kerevan: How would that involve the PRA?
Andrew Bailey: It would not be led by the PRA. I do not think the regulator is the natural person. This somewhat includes the building societies as well. We have said to them, “Look, there may be legislative changes”. The other thing that you observe in other countries, though, is that there is more of a white‑collar supporting infrastructure. There is a bit more of a shared services thing, so you do not have to develop your own IT yourself, for instance. If you go to Germany, for instance, you see far more of a supporting infrastructure for the mutuals sector. On the whole, in this country, we do not really have that.
Q86 George Kerevan: I was not implying that the PRA should take the lead; I was just probing whether there is anything in the regulatory regime that could be looked at that might encourage credit unions.
Andrew Bailey: Well, we certainly do that under our competition remit. We are currently consulting on credit union capital standards. We get quite mixed responses on this. We have sought, in a sense, to divide the credit unions into those that are more simple and those that are more complex—if you are doing mortgage lending, for instance, then you are taking on greater risks than if you are running a rather simple credit union—and to have different standards for them. I know there has been some criticism of the capital numbers we have come out with. The one caveat I would mention there is that this is an industry that, on the whole, does not have a provisioning framework, so we have to, in a sense, embed that into the capital standard. We are in the midst of consulting on that, and we have a lot of feedback, so that is good.
Q87 Chair: I have a few thoughts that I just want to raise with you, of different types. First of all, do you have any truck with the view—back on prop trading, for a moment—that the decline in prop trading has sucked liquidity out of bond markets? I wrote a letter to the Governor about liquidity on Monday. I would be grateful if you would just comment on this.
Andrew Bailey: I might bring Mark in as well. I tend to divide this question into what I might call the supply and demand for market liquidity. The demand for market liquidity has increased, because if you look at any measure of the scale of asset management globally, it is a pretty continuously increasing pattern. Clearly, there is some shift of activity out of the balance sheets of banks and to asset management; that is, in a sense, an intentional consequence. If you think that is a measure of the demand, then you see it go up. Then if you look at the supply of market liquidity—you touched on this in your speech—what is often looked at there are the so‑called market‑making inventory positions of the banks and their trading books. It is not what it was a year or two before the crisis.
Q88 Chair: But your point would be that a lot of that was prop trading.
Andrew Bailey: I would say a couple of things. Yes, it was prop trading. I do always say to people that if you think we are returning to the years immediately before the crisis as a model of how to do it, then that is pretty alarming. We know what happened. When you look at that picture of banks’ inventories, it absolutely fell off a cliff in 2008. That was before any re‑regulation came in, and that was prop trading.
Q89 Chair: You have answered the question I was angling at, which is: it may have caused the problem, but a return to prop trading on a big scale is probably not the solution.
Andrew Bailey: It is not the answer. In a sense, you pointed to this in your speech: we still recognise that there is an issue in terms of market liquidity—and this is a global issue—but the solution is not to go back.
Chair: I will move on, if I may. We do not want to keep you unnecessarily. If there is something you particularly wanted to say or write to us, by all means do.
Mark Yallop: I could talk for several hours on the topic, but I will keep that for another day.
Q90 Chair: Then I think probably not. You have written me a letter.
Andrew Bailey: I have written you several letters. I always have a guilt complex just before I come here that I have not replied to all your letters.
Chair: Well, we would be grateful if we could have them a bit earlier.
Andrew Bailey: I know. It is human frailty.
Chair: As a gentle marker, it does not give me any time to read it, so I have been reading it in the course of the meeting. I note that one of the points I asked you to look at was the extent to which, if at all, the new corporation tax surcharge has a deleterious effect on retail banking competition by placing new entrants and very small firms at a disadvantage. To be frank, you have not given an answer to this. It is the shortest reply to any of the points I have made, and the least informative. I wonder whether you would be prepared to have another go.
Andrew Bailey: It goes back to the conversation we were having earlier. We cannot answer the question of what the impact of it will be because it has not come in yet.
Q91 Chair: I do not think that is a good answer, Mr Bailey. It is quite reasonable to expect a regulator to try and anticipate what he thinks the reaction might be; after all, that is what you do whenever you introduce a regulation of any type. You now have a duty on competition, and we are asking you, exercising that duty, to give us a view on something that is right in your bailiwick.
Andrew Bailey: With respect, our competition objective is not to give views on the Government’s fiscal policy. Our competition objective is about how it affects us and our activities.
Q92 Chair: I am not asking you whether you like the policy or not; I am asking you what you reckon the effect will be.
Andrew Bailey: As I was saying earlier, for those institutions that are earning more than £25 million a year, it will increase their tax liability and it will therefore, other things equal, reduce the accumulation of capital through retained earnings. How much that will be, we will have to see.
Q93 Chair: So this will have an effect of reducing competition.
Andrew Bailey: Other things being equal, you could conclude that, but there are a lot of other things going on in the landscape at the same time. I am not sure—and this is the open question—how big an impact this is relative to the other things that are going on.
Q94 Chair: So there is an impact; you are not sure how big it is, and we need to take account of other factors.
Andrew Bailey: Yes.
Q95 Chair: What are the main other factors, in summary?
Andrew Bailey: One of the biggest ones is in the letter I wrote to you, which is the thing I that covered at the speech at the Mansion House. We have responded to the European Commission consultation on CRD IV by saying that, in our view, it would be sensible to introduce a capital regime that is tailored towards small institutions and not to just apply the international Basel standards through European legislation to all institutions. The reason is that the Basel standards are aimed at what are called “internationally active firms”. You do not have to apply the Basel standards across the population of small firms. Some countries do; some countries do not. Europe always has. As you know, there is this issue about the gap between internal model risk weights and standardised risk weights. Particularly in an area like prime mortgages, it is quite a big gap. Arguably, the leverage ratio does something there, but if you look at it from a risk weighting point of view, there is still a gap.
We are approaching this on two fronts. As you may have seen from the thing we put in to the European Commission, which is published, we think one way to do this would be to make the standardised approach more risk‑sensitive and to say, “If you are in the prime mortgage lending market, then your standardised risk weight will be lower”. We are also approaching exactly that same question in Basel, in the context of the review of the Basel standardised approach. That is quite a big issue, and it is one that the firms raise with us a lot. It is directly in the scope of our competition objective. Fortunately or unfortunately, we need a change to European law to effect that, but we regard our competition objective as being important in terms of framing our approach towards European policy issues. It is not just what we do at home.
Q96 Chair: That is somewhat more informative, and I am very grateful. I have a couple more questions, and I apologise for chopping and changing. It has been put quite a bit—sometimes behind people’s hands and occasionally publicly—by those in investment banking that the UK’s competitiveness is in sharp decline and that this business is ending up in the US, partly because of scale and partly because the public mood is held to be so discouraging. The UK’s version of Main Street versus Wall Street is something that is relatively new to London‑based bankers, unlike the United States, where it has been going for 80 years or 70 years or whatever. With the bonus cap and the surcharge you have just described, those are the four main factors that tend to be put to me. Do you have any truck with these arguments?
Andrew Bailey: Do you want me to answer? Mark is knowledgeable in this field as well, but I am happy to have a go, if you want.
Q97 Chair: Mr Yallop, are we losing out, and are those the reasons why?
Mark Yallop: It is manifest that the UK investment banking businesses are losing market share. You can measure that in a number of ways, which we can talk about at length another time, perhaps, but I am not sure that the reasons that are adduced for why that market share is being lost are necessarily correct. The problems that UK investment banks have are much more to do with the scale of their own operations, the size of the domestic market that they are operating in, the opportunity that they have to cross‑subsidise earnings in investment banking from activities in other parts of the group that may be available to overseas banks and are not going to be available here in the ring‑fenced world, their poor infrastructures, their high cost base, and the speed with which they deal with legacy asset portfolios. Those are all important factors, alongside the ones that you are hearing about.
Q98 Chair: It is important to be clear that, certainly, this Committee is not arguing—nor am I suggesting—that this is an issue that necessarily should lead to any policy change in these areas. It is just to be clear that we distinguish between the wheat and the chaff in the arguments. Maybe it would be helpful if the PRA could send us a considered view about what constitutes the wheat and the chaff as far as the arguments are concerned.
Andrew Bailey: The only point I would make is that this is more of a European issue than just a UK issue.
Chair: You made that point earlier.
Andrew Bailey: Yes. When you look at the major European universal banks, you see some pretty strong similarities in terms of the issues they are going through. Mark raises a very interesting question about the issue of not having a bigger domestic market, as the US firms do.
Q99 Chair: We do not need to go over the same ground here. The last point I want to put to you is a very general one, about the whole gamut of your work, about which parts of the industry complain a good deal—the intrusiveness, the burdensome nature of it and the cost, among many other concerns. Are we fighting the last war in imposing all this? This point certainly was put to me last night as well. Are we fighting the last war with the work we are doing to bring in all this banking regulation, as has been put to me? I think I should put these points to you.
Andrew Bailey: I would make two or three observations on that.
Chair: And where are the risks going to come from?
Andrew Bailey: One of the things that we are very much into now is not so much fighting the last war as implementing the responses to the last war. If you think about structural reform, ring‑fencing, Senior Managers Regime, resolution and total loss‑absorbing capital, these are some of the big planks of the responses and we are now into the implementation phase. I would distinguish that from new things. That is one thing.
The second thing I would say is that there is, sensibly, a shift from what I might call institutional concerns, which were obviously the most important thing coming out of the height of the crisis, to more of an activities focus. That is directly applicable to your speech on market liquidity. Market liquidity is not an institutions thing; it is an activity‑based thing. That switch of attention that is going on is the right thing. You see this in the work of the FPC, for instance, and you see it in the work of the Financial Stability Board globally.
Thirdly, as I said last week at the Bank’s open forum, in terms of new risks, cyber is probably the biggest thing that has come up the league table in the post‑crisis period. If you go back to the years immediately post the crisis, I do not think cyber was particularly on the agenda, and now it is very big.
Chair: Thank you very much for coming to give evidence to us today, all three of you.
Andrew Bailey: You are welcome.
Chair: We are very grateful. We intend in any case as a Committee in due course to take a closer look at the work of the externals in relation to the importance of maintaining the independence of the PRA, with the new changes and the integration back into the Bank, the PRC and the removal of the subsidiary structure. Those are issues for another day, and we look forward to discussing them with you further. We have asked for one or two things to come to us as a result of this meeting. Thank you very much indeed.
Oral evidence: Prudential Regulation Authority Hearings: 2015-16 HC 615 23