Treasury Committee
Oral evidence: Bank of England December 2014 Financial Stability Report, HC 912
Tuesday 6 January 2014
Ordered by the House of Commons to be published on 6 January 2014
Members present: Mr Andrew Tyrie (Chair); Steve Baker, Mike Kane, Jesse Norman, John Thurso
Questions 1-135
Witnesses: Professor Kevin Dowd, Professor of Finance and Economics, Durham University Business School, and Dr Thomas Huertas, Partner and Chair, Ernst and Young Global Regulatory Network, gave evidence.
Q1 Chair: Thank you very much for coming to give evidence to us this morning. We have before us, I think it is fair to say, Professor Dowd, one of the most trenchant and comprehensive critics of the whole apparatus of banking supervision, and I think it is important that Parliament should have an opportunity to hear those views. I would like to begin by asking you some very straightforward questions about the lender of last resort function. In the event of a liquidity crisis, a central bank can act as a lender of last resort to banks that are solvent but cannot access funds to carry out their day-to-day operations. Should banks just be allowed to fail?
Professor Dowd: Yes, of course, I believe they should, Mr Tyrie. First off, good morning. Yes, I believe banks should fail. There has to be some accountability for failure on the part of the management of banks, and banks only fail when the managers fail. A fundamental problem is that we do not hold our bankers to account. If a bank gets into difficulties, then I believe it should be punished—that is to say, the management should be punished.
We have a bizarre situation where we do not want banks to fail. We regard bank failures as somehow unacceptable, but we are quite happy to see—well, we are not happy, but we are ready to accept—failures on the part of other institutions such as manufacturing institutions and so forth. When it comes to banks, however, we reward bankers for getting into difficulties by offering them support that they could not get on a free market. If we reward failure on the part of management, then that is exactly what we get.
Essentially, what we have is a moral hazard where bankers are not held accountable. Look at Northern Rock, for example. Of course, if you look also at the history, when you dissect things afterwards in the cold light of day, you inevitably find that there is a panic overreaction. Decisions are made overnight and so forth and there is no accountability afterwards. We have sent no bankers to jail yet, for example.
Chair: That is a slightly different issue from the lender of last resort issue that I have raised with you.
Professor Dowd: Yes.
Q2 Chair: Perhaps we might come on to that in a moment. Just to clarify the answer to my first question, does this mean that we should explicitly say that central banks should not act as lenders of last resort?
Professor Dowd: I would prefer that the central bank did not act as a lender of last resort.
Q3 Chair: If they are not going to, we need explicitly to say so, don’t we?
Professor Dowd: Yes.
Q4 Chair: We need to say, “There is no lender of last resort in the economy. Banks will fail. We will just let them go”.
Professor Dowd: Yes.
Chair: Even if there is some contagion risk?
Professor Dowd: Well, contagion risk is massively overestimated and I think we have to live with the limited contagion that we experience. For example, in the case of Barings, there was limited contagion. It is not the end of the world.
Q5 Chair: Barings was very small compared to the firms that were put at risk in this crisis after the Great Moderation. Is that not so?
Professor Dowd: Yes, I think that is true. I would still say let them fail, but we must be up front. The bank must have a clear rule about this. We do not want panicked bailouts at the last minute, and that is the problem.
Q6 Chair: Perhaps you could describe what you think the most likely outcome would have been in the United States—and, indeed, in the global financial system—had the decisions not to launch those emergency bailouts been put in place. Have you thought that through and given an estimate about it? Perhaps you could say something about that.
Professor Dowd: Yes. Much of the banking system would have collapsed; in fact, it did collapse. What you need is an emergency strategy to put the banks back on their feet. That might involve temporary nationalisation: go in, clean the banks, get rid of the management, separate out the casino stuff and put the banks back into operation as soon as possible, adequately capitalised and with governance structures that work.
Q7 Chair: How different is that from what in practice has gone on? Isn’t that pretty much what we did with RBS, for example?
Professor Dowd: Well, RBS is still a basket case. What I was proposing was something that would—
Chair: Would have stripped out the bad bank?
Professor Dowd: Yes. Well, also you need to—
Q8 Chair: No, sorry, you are quite right on that, but the first half of what you have just said, which is takeover by the Government in order to maintain the core function—isn’t that what happened?
Professor Dowd: Well, the problem is that RBS is still a basket case. Nobody trusts its accounts and it is not safe to put it back into the private sector fully because it is unsound.
Q9 Chair: There are many who agree with you—probably several round this table—and feel that RBS might have been better split, with the bad assets stripped out so that a much smaller and healthier bank could be recreated to perform core functions. That is not really the question that I am posing to you. I am just saying: what is the difference? You have said you do not want a lender of last resort function, but you now say you want a nationalisation and rescue function. I am asking in the first instance—that is on day 1, day 2, day 3 of this crisis, when contagion risk will always be at its worst—how different your view of how the crisis should be handled is from what happened.
Professor Dowd: My position is essentially a free market position. If the banking system was left alone, not tampered with by the Government, and was effectively unregulated, it would operate just fine. Given that we do not have such a system, we then have to have a way of dealing with the banks that we have. What I am trying to say is that we would have—I am sorry, could you repeat the question again?
Q10 Chair: Let me put the question differently. Do we really need central banks and regulators in this whole apparatus or should we just sweep that away?
Professor Dowd: We do not need a central bank. There is a long history of unregulated banking systems, including in this country, in Scotland in the 17th and early 19th centuries—
Q11 Chair: You would abolish the Bank of England and the Fed?
Professor Dowd: Yes.
Chair: Completely?
Professor Dowd: Yes.
Q12 Chair: Who would perform this nationalisation function and keep an eye on things in between times, in the good times? After all, trying to nationalise something in 24 hours is quite tricky.
Professor Dowd: Yes. Under a free banking system, you would not need to nationalise it in a crisis because you would not have a systematic crisis. You would have occasional banks that got into difficulties and would go out of business.
Q13 Chair: That is because once people are in that position caveat emptor, particularly for investors, would drive out weak banks that might have liquidity crises and reward good banks?
Professor Dowd: Absolutely. Under an unregulated system, the bankers have to worry about keeping the confidence of their depositors. That makes them conservative in their lending. It means that they would be adequately capitalised and the bankers would have their own personal wealth on the line. You have all the incentives to keep and maintain a safe banking system.
Q14 Chair: The development of central banking: may I just explore that with one more question, then I will go over to Dr Huertas? Just to be clear, over the last—I don’t know—almost 300 years, central banking has developed as a consequence of the need for Government to effect some sort of rescue after crises—whether the tulip mania, the Mississippi scheme or the South Sea Bubble, running through various financial crises in the 19th century and, in particular, the big two now of the 20th century. What you are saying is that we would have been and we would now be better off had we not had any of that Government intervention? Is that correct?
Professor Dowd: Yes, that is exactly what I am saying.
Q15 Chair: Therefore, when you offered the view that we should nationalise banks earlier, you are really heavily qualifying that, aren’t you? You are saying there may be a case for nationalisation, but we really should just be allowing them to fail?
Professor Dowd: Yes, banks should be allowed to fail. The only reason why I mentioned nationalisation was that in the middle of a crisis in 2008 something had to be done facing the banking system that we have. Effectively, the banking system had collapsed, so the Government had to do something about it. What it did not do is put the banking system back on its feet, otherwise the crisis would have been over by now and it is not.
Q16 Chair: Dr Huertas, would we be in a better place without central banks and financial regulators? You have been a financial regulator for part of your life. Would we be better or worse off?
Dr Huertas: In my view, we would not be better off. It is, indeed, difficult to imagine the economy functioning as well as it could without a central bank. Central banks perform a central function in terms of providing money, governing the money supply, interest rates and directing economic policy. Without a central bank, it is difficult to imagine how that particular function would be performed by Government.
In terms of whether the activity of the central bank vis-à-vis the banking system could be altered, that is certainly a subject of debate. There are two different functions with respect to providing liquidity to the banking system. There is a normal discount window facility on the basis of eligible collateral to provide for liquidity to the banking system. In the case of the European Central Bank, that is the primary way in which the central bank influences the money supply.
The lender of last resort or emergency liquidity assistance is a hotly debated function. It depends very critically on the dividing line between ordinary facilities and extraordinary facilities. That dividing line in the UK was at a markedly different place in 2007 from where it was at the ECB at that same period. There was a much smaller amount of eligible assets for ordinary discount window facilities in contrast to the ECB. So one moved to extraordinary liquidity assistance or lender of last resort facilities in the UK much earlier than one would have done at the ECB. Indeed, had the Bank of England had the same eligibility requirements for ordinary discount window facilities, it is doubtful that the extension of credit to Northern Rock would have been considered extraordinary liquidity assistance at all.
Q17 Chair: When these arguments were raging—and from what we can tell, the then Governor, Sir Mervyn, now Lord King, was arguing the case for moral hazard, which is that a bailout was risky because it would create moral hazard, the point that Professor Dowd has just made—which side of the argument were you on?
Dr Huertas: I appreciated the statement that Sir Mervyn was making at the time. The theoretical case for moral hazard is quite strong. The practical difference, however, was that the UK had a much stricter standard than any other central bank at the time. It was, as I say, a much earlier introduction of moral hazard than would have been the case in other jurisdictions.
Q18 Chair: So which side were you on?
Dr Huertas: At the point in time when the decision was made that the facility should be extended, I concurred with that at the point in time at which it was extended, but it was made quite clear to the supervisor at the time that the decision about central bank lending was not one for the supervisor to make.
Chair: Yes, but I am really just—
Dr Huertas: Nor was the supervisor’s opinion necessarily requested.
Q19 Chair: This was a big debate going on.
Dr Huertas: Yes, that is correct.
Chair: Indeed, the crucial debate—a debate that then raged and was part of a dramatic alteration of policy over the following 18 months. I am asking you whether, at the time when this broke, you were pressing the moral hazard case put by the then Governor or whether you were arguing against it. Maybe you were not saying anything.
Dr Huertas: To the best of my recollection, I was arguing for a harmonisation of the eligibility requirements toward what other central banks were using.
Q20 Chair: You were disagreeing with the particularly strict line pro greater—
Dr Huertas: Yes, respectfully disagreeing, but disagreeing.
Chair: Disagreeing, and, therefore, you were arguing that the risk of moral hazard should be accorded a lower priority?
Dr Huertas: A lower priority at the time, yes.
Chair: At the time, okay. I think that is clear. There are lots more questions it would be very interesting to ask, but I am going to turn to Jesse Norman.
Q21 Jesse Norman: Dr Huertas, I want to pick up on that last point for a second. Obviously, it is a consistent position to say that one wishes to see no subsidies, implicit or otherwise, to the banking system, high levels of competition and a minimum of regulatory intervention during peacetime, but if there have been those market-distorting factors and in part the result is a crash, all bets are off. That is not an inconsistent position. It is not as though anyone has failed in an intellectual argument if you say that matters are different in wartime from what they are in peace.
Dr Huertas: The circumstances, as you suggest, do make a considerable difference. At the time of Northern Rock, to have simply allowed the institution to have failed would have, in my opinion, exposed the shortcomings in the UK deposit guarantee scheme. There is no way the FSCS could have paid out those deposits with any speed and it would have exposed the extent of the coverage as being less than 100%. The process of resolution of a bank at that time was not a smooth one. What has happened since that time has been a marked revision in resolution statutes in the UK and the EU, together with the introduction of living wills and other preparatory measures that would make it possible to resolve a bank should one fail in a manner markedly different than was the case in September 2008.
Q22 Jesse Norman: Okay, good. That is very helpful. I will come on to that in a second. Just to be clear, you have, as it were, a financial problem and you have a political problem. The political problem is that people are going to the hole in the wall and there is no money coming out or potentially no money coming out, or they are going to take money out of their bank account and there is no money to give them. The financial problem is that that cannot be easily resolved in a rapid way because the mechanisms by which such a resolution would occur simply were not in place.
Dr Huertas: That was the situation in 2007 and again in 2008.
Q23 Jesse Norman: Perfect, good, thank you. Therefore, even a relatively small institution like Northern Rock can have potentially catastrophic systemic consequences?
Dr Huertas: It could have at that time in my judgment and that was the case that I made. In contrast, Bradford & Bingley was resolved in a markedly different manner over the weekend without disruption to the economy or to the financial system.
Q24 Jesse Norman: Yes, in a kind of Barings-type way. Okay, that is helpful, thanks.
Let us come on to the question of too-big-to-fail now status of resolution. Obviously, there is a lot of concern. The IMF is flagging it does not think that too-big-to-fail has been anything like finally resolved. Janet Yellen has said, “I am not positive that we can declare, with confidence, that too-big-to-fail has ended”. That means it has not ended in the Fed’s view. What is your view about too-big-to-fail?
Dr Huertas: Too-big-to-fail is, in my view, too costly to continue. We have made—
Jesse Norman: No, I agree, but how much progress has been made to resolve it?
Dr Huertas: Considerable progress, in my view, has been made with respect to changing the legislation, the framework, so that if a bank were to fail at this point in time it could potentially be resolved without cost to the taxpayer and without significant or catastrophic disruption to financial markets or the economy at large. The key element in that is bail-in of investor capital and provisions are being made to make sure that there is enough reserve capital in place to convert into common equity so the taxpayers do not have to foot the bill.
Q25 Jesse Norman: Okay, let us probe that for a second. There are not many small British banks, but if there was one—perhaps a building society is a less good example because of the way they are capitalised—but, say, a smaller British bank could fail without serious consequences to the system—
Dr Huertas: Bradford & Bingley was a case in point. It failed and was resolved over the weekend.
Jesse Norman: It is a savings and loan. It has a different shape balance sheet, but I know what you mean.
Dr Huertas: Well, let us take the hypothetical. If a large institution were to fail today, one of the first things that would be done would be to convert the preferred stock and subordinated debt into common equity at the point of non-viability. In the case of RBS in 2008, that was not possible. At that point, there were approximately £38 billion to £40 billion of such obligations available at the time, which is on a first order of approximation about what the Government put in in the way of taxpayer money. Investors, not the taxpayer, would have borne that loss.
Q26 Jesse Norman: That is interesting. Let us take the international case. Obviously, these institutions are extremely international—in fact, at almost any size they are. Are you comfortable that there is sufficient clarity as to where assets and liabilities lie and how they are regulated properly, as it were, to be resolved to allow an effective resolution of an international institution to take place?
Dr Huertas: That is the very purpose of the resolution plans being developed in the case of all of the globally systemically important institutions. Certainly, much more progress has been made. There is much more division of responsibility with respect to the individual subsidiaries and branches where those assets and liabilities are held. I am much more confident today than I would have been in 2008.
Q27 Jesse Norman: That is helpful. Thank you very much.
Professor Dowd, do you want to comment on any of the stuff that we have just been talking about? I do not know if you have anything you wanted to say particularly on that.
Professor Dowd: I would say that too-big-to-fail has effectively become institutionalised. Dodd-Frank is a horrendous example of that. In this country, I do not believe any significant progress has been made in that direction.
Q28 Jesse Norman: All the stuff the bank is putting out about stress testing, resolution mechanisms and all that stuff—that is basically hot air?
Professor Dowd: Yes.
Q29 Jesse Norman: Do you have any evidence for that view?
Professor Dowd: About the stress testing?
Jesse Norman: For why you think it is hot air. Why should we believe your view that it is hot air rather than the bank’s exquisitely calibrated documentation?
Professor Dowd: Well, they have always been saying this. If you look at the history of central banking, they have always said, “We will punish you if you get into difficulties. We will not bail you out”, et cetera. Whatever they say, when the crisis comes there is a bailout. That is the fundamental constitutional problem posed by the lender of last resort and we have never resolved it.
Q30 Jesse Norman: Okay, that is interesting. Let us look slightly more widely at threats to financial stability. Obviously, there is a serious potential threat now arising in the eurozone. I do not know if either of you has a view particularly on any of the developments there, the extent to which the eurozone might be pushed back into a state of crisis. It is obviously a topic of some current discussion and debate.
Professor Dowd: I think the eurozone is a disaster waiting to happen. We have obvious problems in Greece and so forth. The banks are in a horrendous situation and the recent European Banking Authority stress test was just farcical. Particularly some of the northern European banks, which were signed off as sound, are clearly not, such as Deutsche Bank.
Q31 Jesse Norman: You have a double problem in a way. You have the Grexit-type issues that are happening in Greece and the south and the potential contagion that could occur from that, and then you have a deeper problem that is not being recognised, which is the state of the core European banking system in France and Germany?
Professor Dowd: Yes, and I would go further. I think the stress tests themselves are a joke.
Q32 Jesse Norman: Do you mean the European ones particularly or do you also mean the British ones?
Professor Dowd: Well, especially the European ones because we have seen there is a track record of total inadequacy in Europe. If you look at regulatory stress tests generally, before Iceland collapsed the IMF and the Icelandic central bank and so on did stress tests and their banking system was fine. Then it crashed. The same in Ireland: signed off by stress tests as adequate, then it crashed. Then Cyprus, and Cyprus was on nobody’s radar screen. Essentially, a stress test is just a guess. It is a “what if” scenario.
The number one rule of stress testing is that you should look at multiple scenarios, but there is a tendency to focus essentially on just one scenario. You say, “If this scenario occurs, the banking system is fine” or whatever, but what about some other scenario that you have not considered? The chances of any one scenario actually occurring are vanishingly small. If we are to have stress tests, we need to have multiple scenarios, but that does not really get to the bottom of the problem because essentially the central bank has no forecasting powers that the rest of us do not have.
Q33 Jesse Norman: No, I understand. I do not want to cut you off but I think other colleagues are going to come specifically back on stress testing. Can I ask you, Dr Huertas, about your view of threats to the global financial system? It might be eurozone, it might be the slowdown in China, it might be what is happening with Russia and its territorial claims in Ukraine, et cetera.
Dr Huertas: I think all of the above, along with the impact of oil prices. The Bank of England “Financial Stability Report” has a very extensive list. However, there are additional matters that might also be on that list, such as the possible exit of the UK from the European Union. That would potentially have an impact on the UK economy and potentially on financial stability. That is absent from the report, but is a potential risk.
Q34 Jesse Norman: Isn’t there a deeper problem here, which is that you cannot get the resilience and the heterogeneity that you want in a system if you have essentially one mindset dominating a small number of central banks that are driving change and regulation? Isn’t that a deeper problem? How do you have systematic heterogeneity, if you will pardon the jargon?
Dr Huertas: Groupthink is always the danger, but I can assure you that around the Basel table there was quite a diversity of opinion.
Q35 Jesse Norman: They are all agreed about the importance of thinking about groupthink? Sorry, that is a joke.
Dr Huertas: No, there was quite a diversity of opinion as to what should be done.
Q36 Jesse Norman: Right, so they do not know what to do. What about housing? Can we talk a little bit about the UK housing market? Do you think that the housing market is in a sustainable place at the moment, Professor Dowd?
Professor Dowd: I think the housing market is very risky. Clearly, houses are overpriced and there is going to be a crash. It is unfortunate that the central bank is doing nothing much about it.
Q37 Jesse Norman: You do not think that the new FPC has had rather a useful effect in tamping down some of the concerns?
Professor Dowd: Not really. It is tinkering with the market, but the fundamental thing is that we need higher interest rates.
Q38 Jesse Norman: Right, but on your view, if I understand it, the FPC should not exist at all?
Professor Dowd: That is correct.
Q39 Jesse Norman: Dr Huertas, can you give me your view on where you come out on housing and the effectiveness of the regulator?
Dr Huertas: The question on housing in terms of the impact on financial stability depends very greatly on the loan to value ratio and the loan to income ratio and the degree to which banks have mortgages on their books. That has been certainly—
Jesse Norman: The FPC has started to dial some of that stuff down. You have seen that.
Dr Huertas: They are, and certainly much greater attention has been paid to those fundamental risk factors.
Q40 Jesse Norman: The housing market has cooled off a bit?
Dr Huertas: It has. Whether it has been due to the FPC actions or the trends in the economy is open to question.
Q41 Jesse Norman: It is too early to tell whether the FPC has been effective at all in that?
Dr Huertas: The FPC certainly in my view addressed the right issues. The question that has been raised is whether the focus should be on the borrower as opposed to the lender.
Q42 Jesse Norman: That is really interesting. Sorry, can we just spend half a minute more on that? What do you mean exactly by that?
Dr Huertas: The focus to date of the FPC has been on controlling the actions of the lender rather than on controlling the behaviour of the borrower.
Q43 Jesse Norman: What are the kinds of things you think they could be doing to achieve the latter?
Dr Huertas: Limitations on the degree to which the house can serve as collateral for loans of all sorts.
Q44 Jesse Norman: Okay, that is interesting. Have you written anything about that or is that being discussed—
Dr Huertas: I can send you some material. Peter Sands had an article in the FT that made the case rather well.
Jesse Norman: Yes, I saw that. Okay, good. Thank you for that.
Q45 Steve Baker: Good morning. Before I come on to talk about risk, can I clarify something from the previous questions? When Dr Huertas was talking, you mentioned Barings in relation to resolution, but I remember reading things, Professor Dowd, that you published much closer to the crisis. If I understand it, your temporary nationalisation is a form of receivership, which would include write-downs and extensive bail-in. Is that the key difference between what happened and what you would have done—rather than the taxpayer capitalising banks, it would have been bondholders and depositors who would have capitalised banks?
Professor Dowd: Yes, that is true. Again, I cannot emphasise strongly enough the need to hold bankers to account. Any effort to clean up the banking system must impose personal liability on senior bankers. Without that, you cannot have strong governance in banking and without that you cannot fix the banking system.
Q46 Steve Baker: It has felt to me, although it seemed like you have been a long way from the mainstream in the course of this conversation, what you were saying at the time of the crisis is what is now broadly being proposed for resolution: to bail in those people who have taken risks by lending to banks.
Professor Dowd: Yes, but I would emphasise that people talk about bail-ins, but what we are really talking about is forms of bankruptcy. Why don’t we use that terminology and say let the banks go bankrupt? We have procedures to do that.
Q47 Steve Baker: Turning to attitudes to risk in markets, do you believe that it is possible that those attitudes can be divorced from underlying values and fundamentals? If so, what is the cause of that separation between the attitude to risk and the underlying fundamentals?
Professor Dowd: The attitude to risk depends on the incentives that people face. In current banking, it is essentially gambling with other people’s money, so quite naturally they are happy to take risk. Most derivatives trading is essentially just speculation. Most securitisation is designed to get around the Basel capital requirements and effectively weakens the banking system. If we had the right incentive structures and, again, liability, a lot of that risk taking would be diminished and we would have much more socially responsible risk taking.
Q48 Steve Baker: Dr Huertas, could I ask you to answer that particular point about derivatives trading being used to get round the Basel rules?
Dr Huertas: I believe it was securitisation. Securitisation has a number of features to it, one of which may be to reduce the capital requirement of the bank, but there are also liquidity features to securitisation that are attractive to the bank and to society at large.
Q49 Steve Baker: Do you believe that the rules that you were part of putting in place are, in fact, impenetrable to this problem that Professor Dowd has indicated?
Dr Huertas: No. The securitisation rules in terms of increasing the capital requirements for securitisation positions held are, in my view, a step in the right direction. The “skin in the game” requirements are perhaps not as effective, but on balance I believe the adjustments made by the Basel Committee were in the right direction.
Q50 Steve Baker: How seriously worried should this Committee be about this problem?
Dr Huertas: Relative to many other issues, I think this is a second order question.
Q51 Steve Baker: Would you agree, Professor Dowd?
Professor Dowd: Not at all.
Q52 Steve Baker: Why?
Professor Dowd: I agree with Dr Huertas that there has been some improvement in Basel III but only in the respect that the capital definition is tighter. It just stands to reason. Basel III is twice the length of Basel II. It was a rushed job. Basel II took a decade to put together. The banking system was signed off as fine and then the banking system crashed. Then we have a panicked Basel III, which, okay, I agree that the capital requirement is better. I agree that the leverage ratio is a better metric, but the rest of it is more of the same and a double-sized rulebook is not a good sign.
Q53 Steve Baker: The Treasury is now consulting on giving the FPC powers of direction over the housing market and leverage ratios. Dr Huertas, are you in favour of such powers?
Dr Huertas: It is certainly a power that the Government has. If the Government elects to delegate that power to the FPC, it is in the Government’s discretion to do so.
Q54 Steve Baker: That is not quite what I am getting at. Do you think it is in the public interest that such a power should exist and be in the hands of the FPC?
Dr Huertas: It is incontrovertible, in my opinion, that the power exists. It is a question of whether it should be exercised.
Chair: That is what you are being asked.
Q55 Steve Baker: Is it in the public interest that it should be exercised?
Dr Huertas: It is in the public interest that the issue be considered. That there should be a leverage ratio as a back-up to the risk-weighted ratio is something that I am on record as favouring.
Q56 Chair: Would you like to have another go at that question because that is still not answering it? The question is: is it in the public interest that we have it or not?
Dr Huertas: It is in the public interest to have a leverage ratio, yes.
Q57 Chair: Is it in the public interest that we have these specific powers to which Mr Baker is alluding?
Dr Huertas: That the FPC exercise those powers is, I think, a matter of discretion as to where the powers should be exercised.
Q58 Steve Baker: Professor Dowd, I feel sure I can get a shorter, straighter answer from you on this one. Do you think that such powers should exist and be exercised in the public interest?
Professor Dowd: Absolutely not. What we should be doing is getting our interest rates to the right levels and this would take care of the housing problem.
Q59 Steve Baker: I think from a political point of view all of us fear higher rates because we know the consequences on our electorate would be quite dire. You are approaching this from an academic point of view rather than a practical or political one. Why do you think it is necessary to put the public through all the pain of higher interest rates to sort out the housing market? What is the benefit?
Professor Dowd: Currently, we have interest rates that are negative in real terms and this is catastrophic for long-term saving. We have to take a long-term view, not a short-term view. Long term, we need capital accumulation. To get capital accumulation, we need incentives to save. To get incentives, we need sensible interest rates.
Q60 Steve Baker: With this in mind, we hear endlessly about the productivity puzzle at this Committee. Do you think that this phenomenon is at the heart of this question about productivity and a lack of rising real wages?
Professor Dowd: I think part of the problem is that with low interest rates we are encouraging the substitution of capital for labour and this is a key part of our productivity malaise here and in the United States.
Q61 Steve Baker: Dr Huertas, would you like to add anything on this particular point of raising interest rates right now?
Dr Huertas: I concur with Professor Dowd that the raising of interest rates would benefit savers. I would submit that many of the electorate are savers; not all are debtors. The political calculation is perhaps more complex than you have suggested.
Q62 Steve Baker: I think perhaps I will move on from price planning then to talk about the risk models. The crisis has exposed problems with the risk models. Before this Committee both Mark Carney and Andrew Bailey have agreed that the risk models are wrong, which I think was quite an interesting observation; the transcript repays close reading. Is it possible to get meaningful information about the riskiness of an individual asset or a bank’s balance sheet from the use of these models, which admittedly are wrong?
Dr Huertas: It is certainly possible to get information about the assets and the way the bank views the risks. My recollection was that there were a number of banks submitting models with respect to mortgages that had the charming view that, although there was a probability of default, there was never going to be a loss given default because the house price would always exceed the mortgage amount. That was a feature of the model that the FSA at the time knocked out.
Professor Dowd: I think risk models are fundamentally flawed for a number of reasons. First, they are unscientific. They have the appearance of science but the risks that really matter are the unquantifiable risks, which are beyond any possible risk model. That is the question we should be asking: how do we address uncertainty, not the quantitative risks that a risk model produces?
Secondly, risk models are based on very, very poor assumptions such as normality in financial returns. Thirdly, risk models are terribly gameable. Essentially, bankers are very, very good at stuffing risk into the tail. They do that because low risk numbers lead to low capital requirements, which means you can do better things with the money—you can pay yourself higher bonuses and stuff like that—and it means the banks are terribly undercapitalised as a result. Lastly, the risk models do not work. There is mounting evidence that the more complicated models perform less well than simpler models. This is why, for example, the leverage ratio is a far better assessment of a bank’s financial health than any fancy value at risk model.
Q63 Steve Baker: When we explored this issue with Mark Carney and Andrew Bailey, they were clear that the models are not accurate and that is why judgment is required. How do you explain this difference of opinion between their acceptance that the models are wrong and judgment is required and, therefore, we will get it right and your view? That is, it seems to me, you are saying that is a futile endeavour. What bridges this gap between your opinion and their opinion that they can make up for the inadequacy of these models?
Professor Dowd: I do not understand their position. Either the risk models work or they do not. If they do not work, then the bank should not be using them. If they do work, we should, but they do not work. I looked at the transcript and I saw Dr Carney’s evidence. I have to say that I felt he contradicted himself.
Q64 Steve Baker: You talked about traders packing risk into that 1%. What do you think are the practical consequences for the financial system of having these risks packed into the 1% that the models do not even look at?
Professor Dowd: It is catastrophic. It means that essentially you have a lot of false comfort that the banks look a lot safer than they actually are. Then everybody runs around in a panic when there is a catastrophe occurring. Let me give you an example. The London Whale—the value at risk on that was something like $60 million in early 2012. It turned out that the actual loss was 100 times that. That is just one example among many where the risk models got it completely wrong, so why are we relying on these things?
Q65 Steve Baker: You are quite clear that if judgment is still required there is just excessive emphasis on the models?
Professor Dowd: Yes. I would go further. There is excessive emphasis on the models, but the models also get in the way. They confuse us. They give us false comfort. If we follow the thread that judgment is what matters, then what we need to do is to enable investors and stakeholders to make sensible judgments. How do we do that? We have to get the accounts right.
Q66 Steve Baker: You are reconnecting accounting, incentives and judgment and putting the risk back on the—
Professor Dowd: Absolutely, and if you rely on proper accounts, pre-IFRS, old-fashioned GAAP under company law and this sort of stuff, you will end up then having to look at the balance sheet, which you can have some confidence in, and then you make a judgment. “Judgment” is the phrase I use—not only about the accounts, but about the quality of the management of the bank. That is how they used to do it in the past and that works. We know it works.
Q67 Steve Baker: Dr Huertas, if we do not know the circumstances of the next crisis, where does all this leave us with risk models and judgment? How will regulators be able to respond in a way that is in the public interest?
Dr Huertas: I think models are helpful as far as they go in terms of dimensioning the risks from known factors. They should, in my view, be supplemented by judgment and that is imperative. Supervisors can partly exercise that judgment, but it is also up to investors to exercise that judgment and the resolution reform will point, in my view, towards investors increasingly taking on that responsibility.
Q68 Steve Baker: Do you believe that the reforms we currently have adequately reorientate incentives so that investors—and, indeed, depositors—will properly assess the risks that they are taking?
Dr Huertas: Certainly, investors are headed in that direction and the rating agencies support that. They have emphasised the risk given failure or loss given failure and emphasised that for the investor standpoint it depends on the legal entity to which they are exposed and where they stand in the queue. With respect to depositors, I would distinguish sharply between insured depositors, who will not exercise any diligence, and uninsured depositors, who are going to be in a position of risk, and indeed, in the ring-fenced bank, are going to be subordinated to insured depositors—a fact that may come as a surprise to some of them.
Q69 Steve Baker: Can I ask you both, to what extent do you think the current system of regulation has determined the way that banks do their risk modelling?
Dr Huertas: For those banks that rely solely on the regulation, that in my view is a mistake. The better-managed banks will preserve an independent and separate risk function. The danger is that the regulatory pressures are such that they begin to neglect that in the attempt to fulfil all the different regulatory demands on their time.
Q70 Steve Baker: That says to me, by implication, that the act of regulating will tend, through practical imperatives, to herd banks into doing their risk management the same way.
Dr Huertas: It is certainly a danger.
Steve Baker: Professor Dowd?
Professor Dowd: I think regulation is a game as far as the banks are concerned. It creates incentives to have risk models that massively underestimate the true risks in order to get capital requirements down. That is what is driving the whole thing. The banks are allowed to do this also because of deposit insurance, the lender of last resort and so forth, so you have massive moral hazards built into the system. The regulatory system does not control it; the regulatory system endorses it.
Q71 Steve Baker: So the commercial incentives in place on banks, combined with the moral hazard of the regulatory system, mean that the banking system is carrying much more risk that is generally understood—is that what we are arguing?
Professor Dowd: That is absolutely true, yes.
Q72 Steve Baker: If banks are herding towards one type of model or if they have a tendency to herd towards one model, why is it that they report vastly different levels of capital to support similar or identical portfolios of assets? Do you see that phenomenon happening and how would you explain it, Professor Dowd?
Professor Dowd: One answer to that is that the models are no good.
Q73 Steve Baker: Could you elaborate on that point?
Professor Dowd: There is a huge amount of model risk. When you build a VaR model, you can fiddle it in all sorts of ways. There is extensive literature that shows that similar-looking positions can have very different VaRs, but the bottom line is that it is a massive black box. To give you an example, if you look at a bank’s positions, you would need a spreadsheet the size of a football pitch just to list the positions. That of course presupposes that the data fed into the model, into the spreadsheet, is right, is correct, which it is not.
Q74 Steve Baker: So you are arguing that although people are being herded into using broadly similar techniques, because they fiddle those techniques differently, they get different results?
Professor Dowd: Yes, that is correct. The other point I would emphasise is that it is very, very bad practice if the banks are herded into similar models, similar risk management practices, because economic stability requires diversity. For example, in the current system, when there is a crisis, the VaR numbers go up and the banks are pressured to sell. So everybody is selling in the crisis, but you want people to buy. The uniformity, the whole idea of a regulatory risk management standard, is essentially a contradiction in terms.
Q75 Steve Baker: I think the Chairman would like me to ask one more and then hand over. Dr Huertas, do you think that a regulatory regime that allows banks to use their own internal models to determine the riskiness of the assets they hold can provide a robust basis for determining capital requirements—particularly bearing in mind what Professor Dowd has just said about it?
Dr Huertas: Yes, I think it can make a significant contribution. It should be reviewed and controlled by the supervisor, but it does have a significant contribution to make.
Q76 Steve Baker: What about the phenomenon we have discussed of packing risk into the tail 1%?
Dr Huertas: That is certainly a danger and one of the primary things that the supervisor needs to control against.
Q77 Steve Baker: How would they control against it?
Dr Huertas: Through model review, which the supervisors have the capability to undertake and have undertaken.
Q78 Steve Baker: But we have sat here with the senior guys from the regulators telling us categorically that the models are wrong and therefore judgment is required.
Dr Huertas: That is the judgment that the supervisor is exercising in the course of the review.
Q79 Steve Baker: It feels like right now we are in a system where we know that there are potentially enormous risks in the tail 1%, that the models are wrong and we are reliant on the judgment of supervisors to pick up what all that means for capital requirements, knowing that the commercial incentives for banks are to minimise their capital requirements. The whole system then presumably is set up to ensure that they get as much risk into that 1% with the minimum possible visibility. Is that right?
Dr Huertas: That is certainly a danger that one needs to protect against. Whether that is the actual result is debatable.
Steve Baker: I want to explore how one protects against it, but I think I had better hand over.
Q80 Chair: I am surprised, Dr Huertas, that you have leapt to the defence of RWAs, bearing in mind how comprehensively they have been trashed by virtually everybody who has taken a look at them. Indeed, isn’t it a primary argument for using a leveraged ratio as a regulatory tool that we cannot rely on the risk weightings? You have just told us we can rely on them.
Dr Huertas: I believe I said the two should complement one another. The shortcomings of the leverage ratio are well-known. If there is a 100% portfolio of short-term government securities, that is a somewhat different risk than long-term commercial real estate, but the leverage ratio would assign the same capital against the two of them.
Q81 Chair: I was not asking you whether the leverage ratio was any use; I am asking you whether RWAs are any use. We know that leverage ratio drives firms into adding risk to balance sheets to some degree. That is not the question I am asking. I am asking about RWAs.
Dr Huertas: The underlying question, I believe, is about the quality of the models in terms of deriving the risk weights.
Q82 Chair: BIS, the G20, they have all been over this. They have all concluded, based on quite good survey evidence that they have collected from the banks, that the stats are not much use.
Dr Huertas: That it is imperfect is well-recognised, but not much—
Chair: We are not talking about that.
Dr Huertas: But to go away from it entirely, on what basis would one otherwise exercise the judgment with respect to the credit quality of a loan? The fundamentals of probability of default and loss given default, those are capable of being modelled to some extent. Not to do that at all strikes me as somewhat foolhardy. Somewhere, some place that type of modelling, in my view, needs to occur.
Q83 Chair: The question we are debating here is the extent to which we should rely on it, not whether it should be done.
Dr Huertas: I think the judgment is that the exclusive reliance on it is improper; it should be supplemented by a leverage ratio. That has been decided.
Q84 Chair: So we use the RWA as a basic tool and supplement it a bit, or do we use the leverage ratio as a core tool and supplement it a bit with some RWAs? That is the question. What is the relative weight between these two tools?
Dr Huertas: That will depend on the—
Chair: I am asking you.
Dr Huertas: —nature of the portfolio.
Q85 Chair: We have just established that we are not really sure what is in the portfolio. That is why we are a bit worried about the RWAs.
Dr Huertas: I think we are reasonably sure what is in the portfolio. Are we sure about the risks? No.
Q86 Chair: Do we really know what is on the balance sheets of those continental European banks?
Dr Huertas: After the asset quality review, we have a very good idea.
Q87 Chair: I must admit, you are in a minority in thinking that. Even many of the regulators are not sure.
Dr Huertas: I would say the asset quality review has changed things quite dramatically. It is an exercise of unparalleled intensity and detail.
Q88 Chair: Professor Dowd, you have told us at least twice this morning that we need higher interest rates. Is that for financial stability reasons or for reasons of monetary policy?
Professor Dowd: Primarily monetary policy, but I think we have to take a long-term perspective and we need to create an incentive for saving and capital accumulation. We should not put the short term ahead of the long term. That is what we have been doing: we have been living in the short term ever since the beginning of this crisis.
Q89 Chair: If I listen carefully to that answer, that answer is saying both, isn’t it?
Professor Dowd: It is both, yes.
Q90 Chair: I am just trying to get at it. What should drive the setting of interest rates? Should we have a central bank to set interest rates?
Professor Dowd: Ideally, no.
Q91 Chair: Ideally no, but since we have one, should it be doing it on the basis of monetary policy considerations, as was the case until 2007, 2008, 2009, or should we be doing it also on the basis of financial stability considerations?
Professor Dowd: I would say monetary policy, because the problem with financial stability policies is that they effectively create instability, so that the harder the regulators try to stabilise the financial system, the more unstable it becomes. I think therefore we need to focus on monetary policy.
Q92 John Thurso: Good morning. I want to ask about stress testing, but before that, having listened to this fascinating discussion, can I quickly ask you each a question? To what extent is this search for formula, mechanisms and models removing human judgment and natural caution from the process, so that instead of us running something and saying, “Not sure about that, so I will not go there” we are saying, “Oh, the model tells me it is okay, so I can”? Dr Huertas, I will perhaps ask you that first.
Dr Huertas: As I have indicated, I think the models have to be used in conjunction with judgment. There are some things that are quite susceptible in the models: the price of a Government bond in response to a change in interest rates can be modelled fairly exactly. The changes in the price of bonds in response to the possible exit of Greece from the euro is a much more difficult matter, where judgment as well as models are valuable.
Q93 John Thurso: Should the model be used as a guide to making judgment or should judgment be the guide to using the model? Which should take precedence?
Dr Huertas: Judgment and the model go hand in hand. One needs to be able to evaluate the model as far as it goes, but bear in mind what the model does not include. For example, many models did not include considerations of market liquidity and the price of the security was established as if there was no difference between the liquidity of an on the run Government security and an infrequently traded corporate bond. All instruments were assumed to have the same type of market liquidity and that was assumed to be constant. The judgment as to the assumptions underlying the model is critical. Which has precedence? I would be hard put to say which should have precedence.
Q94 John Thurso: That was not a great question—that is what you are saying. Professor Dowd, may I ask you the same broad question?
Professor Dowd: I would say judgment is everything and risk modelling is essentially discredited. Of course we need models for valuations, such as Dr Huertas has alluded to, to value bonds and stuff like that, but the risk models have basically been blown out of the water during the financial crisis and we have to come to terms with that. We cannot rely on them for regulatory purposes; we cannot rely on them for risk management.
Again, I cannot emphasise enough that risk models give a lot of false comfort. They also encourage behaviour that itself is highly risky such as, for example, the credit derivatives, CDS and so on. If we did not have complicated models, we would not touch some of these instruments, which means they would not be traded. So it is not just a case of false comfort; it is the fact that we then build ziggurats of risk on top of these models and then we wonder why it all goes very badly wrong.
Q95 John Thurso: Interestingly enough, I remember on the Banking Commission, RBS, I think, giving evidence regarding, “This is a simple model that is a matrix for lending”. Some of us were saying, “Oh, but why can’t we have”—I am talking in shorthand—“Captain Mainwaring back making that judgment, because he knows people?” They pointed out that the Captain Mainwaring model contained a great deal of personal prejudice, and that, properly run, the matrix that they had invented had better outcomes, in that some people who were denied credit previously received it. The point I am making is, is judgment capable of being exercised properly without the assistance of some form of attempt at a rational model of the risk involved?
Professor Dowd: What I am saying is that you cannot measure risk. Essentially what you are dealing with is uncertainty rather than risk. Judgment is essential, but also judgment in the quality of the management. In the old-fashioned banking system, I did not, for example, have to be an expert in something; I would go to my bank, because I trusted my bank to know what was good for me. We have to go back to that model. What is missing in all of this is character and reputation in the individuals involved making the key decisions.
Q96 John Thurso: But the old system did also produce some spectacular bank crashes. If one looks—and one has to go back a bit—an 1878 Glasgow bank bankrupted most of Glasgow’s small businesses and ended up with all the directors in prison, so it was not a uniformly rosy model.
Professor Dowd: No.
John Thurso: That is with unlimited liability, as well.
Professor Dowd: Yes. Look, banks are run by human beings, and like firms anywhere, occasionally they would crash even under ideal regulatory environments.
Q97 John Thurso: The difference being that it was not systemic in any way.
Professor Dowd: Yes, exactly.
Steve Baker: And they went to prison for it.
Q98 John Thurso: Yes. That is another discussion. I am meant to be asking questions about stress tests, but I was sidetracked by the discussion we have been having.
Can I come back to you, Dr Huertas? Do you think that the regulators in general have a reasonable record in identifying the sources of risk for financial stability?
Dr Huertas: The crisis would suggest the answer to that question must be no.
Q99 John Thurso: Do you think they have learnt the lessons of that crisis looking forward or is this a question we will never know until the next crash?
Dr Huertas: They have learned lessons: institutional change, such as the introduction of the Financial Policy Committee and other financial stability boards in other jurisdictions are evidence of that. But I think it is also safe to say that it is entirely possible that there will be another financial crisis and it is also likely that the cause of that crisis will be something that people have not focused on.
Q100 John Thurso: Stress tests usually start by modelling the effects of macroeconomic shocks and then take that through into what would happen to the financial system, but in fact, if we look back, what we nearly always have is the reverse: the financial crash comes first and that then produces the shocks and the macroeconomic downturns. If financial crises are not caused by macroeconomic shocks, what use is there in analysing the effect of such shocks on the financial system?
Dr Huertas: The overall change in the economy is something that will impact the health of the institutions. There is evidence that economic policy, specifically monetary policy, is involved as a precursor to many of the financial problems. Certainly in my view, that was the case in the crisis of 2007/08, the change in US monetary policy. The specific institutional or financial market changes are difficult to posit, but may also be the subject of stress tests that individual institutions could undertake.
Q101 John Thurso: If one took the existing stress tests and went back and applied them in 2006/07, would they have given us any assistance in dealing with what was coming in 2007 and 2008?
Dr Huertas: In my view, had they stress tested at the end of the 2006, beginning of 2007, run forward the actual performance of the economy and the financial markets and suggested that the banks keep capital and liquidity sufficient to withstand such a shock, it would have been of benefit.
Q102 John Thurso: Can I turn to you, Professor Dowd? In your “Math Gone Mad”—lovely title—you said that stress tests should be conservative, plural, with a plausible range, simple models specific to the institutions, non-gameable and transparent and accountable, which is a pretty good list. Do you believe that the European Banking Authority and the PRA in this country achieve any or all of those criteria?
Professor Dowd: Not at all.
John Thurso: None?
Professor Dowd: None of them.
Q103 John Thurso: None of them, right. The fact that they each use only one scenario to test the resilience of financial institutions, given that there are many other scenarios that might arise—how defective does that make the process?
Professor Dowd: It means that they are useless. There is an infinite number of possible scenarios that could occur, so they pick one and they bet everything on that. Take one scenario and let us say you do the stress test adequately against that scenario—the chances of that scenario occurring are vanishingly small. There are a large number of other possible futures, one of which will occur that will be very different, so you can be perfectly well-covered against the scenario that you look at and totally exposed to something else. I would say it is not only useless, but it is worse than useless because it gives us false comfort.
You have to look at the history of stress tests as well. I am not theorising in the abstract, but regulatory stress testing is a catalogue of failures. Just look at the European bank stress tests and the things that they missed, like Cyprus collapsing and so forth. It is a joke.
Q104 John Thurso: This is a daft laddie question, but I will ask it nonetheless: what is the purpose of stress testing?
Professor Dowd: The purpose of stress testing is to comfort the regulators. It has no other purpose. It has the appearance of doing something, the appearance of science and so forth, but in reality we are worse off because of that. If there is one message I would like to get across to this Committee, it is to please encourage the Bank of England not to do this.
Q105 John Thurso: I have to come across to Dr Huertas. Is there any purpose to stress testing, other than the sleep patterns of the regulators?
Dr Huertas: Yes. It is to make the banking system more capable of being resilient under stress in general—in particular the stress being modelled, but stress in general. That is one of the principal changes that has occurred since the crisis. Beginning in 2008, there was a change to force the banks to hold additional capital in anticipation that a stress might occur. That was been intensified over the subsequent years.
Q106 John Thurso: I leave aside the argument that there should be none against there should be appropriate or whatever, but isn’t the purpose of regulation to try to ensure that banks hold sufficient capital to cover the risks they have written, and that is broadly what we are seeking to achieve? Therefore, in looking at the stress test, there is an attempt to work out the risk side of the equation. To what extent would it not be easier to simply go back to the judgment side of the equation and say, “They need more capital and that is what we will regulate. If you want to be a big bank operating in a big multinational stage, you simply must have a minimum level of capital”? In other words, you are going back to the leveraged ratio argument.
Dr Huertas: The result, I think, has been more capital and significantly more capital as a combination of the increase in the minimum and the introduction of the stress test and the requirement that banks have capital in advance sufficient to absorb the losses that could result if the stress were to materialise.
Q107 John Thurso: The PRA’s stress test assumes a rise in inflation to 6%, while the EBA assumes inflation remains moderately below target. That, to a layman like me, would indicate a pretty massive divergence of view in the basis of the stress test. Is that a reasonable assessment?
Dr Huertas: Yes. However, there is also a different currency in many cases.
Q108 John Thurso: What should I be looking for in this to try to formulate a judgment as to the quality of the stress tests in the two different institutions?
Dr Huertas: Two factors: first, the severity of the stress and how much more severe the stress is than the likely economic environment or the baseline forecast. The judgment is, as you have indicated, that the Bank of England had a tougher test than the EBA.
The second is getting the starting point right. One of the problems with the EBA’s stress test, certainly in 2011, was that the starting point was not sufficiently nailed down, so you had a “garbage in, garbage out” problem. That was, in my view, corrected—in the most recent exercise, at least—with the ECB attempt to introduce the asset quality review.
Q109 John Thurso: One of the things that has come out of this is that at one end you have highly liquid, highly tradeable assets—nothing is 100% secure, but it is pretty close to it—and at the other end, you have some of the most exotic instruments created by computer geniuses in the basement, which even the chairmen of their own banks did not understand, which have to be at the absolute other end and have proved to be extraordinarily risky, at least in the short term, and very illiquid. As I think you both pointed out, the modelling works better at one end than that other end. To what extent should regulators stress test and just say, “There is a point beyond which we are not prepared to allow risk weighting or anything else. Those things are just so complex no judgment can be made. They must be assumed to be toxic and you must carry the capital accordingly”?
Dr Huertas: There has been that judgment in the case of a number of instruments where the 100% deduction is required, related, if I recall correctly, to some securitisation positions. The complexity or the riskiness of the assets is by no means restricted, in my opinion, to trading assets, but long-term commercial real estate lending on properties without tenants qualifies in most people’s book as every bit as risky as a derivative.
John Thurso: Any comment on that, Professor?
Professor Dowd: Again, I would say that we need to be talking about uncertainty rather than risk. We would be far better without the models and just acknowledge that the world is very uncertain and we had better be careful what we do. In terms of assessing, how can you know whether a stress test is any good? The answer is you do not. The best you can do is look at the historical record of stress testing, and the historical record of regulatory stress testing is essentially zero.
Q110 John Thurso: If I may, just one quick question on that. All of what you say makes tremendous intellectual sense if you choose to go down that path, but ultimately we have very, very big institutions that cover the globe and we have a great many people who do not think about moral hazard or making wise judgments, who just have £5,000, £10,000 or £20,000 that they wish to invest or deposit safely. It would not be possible or credible for a politician to endorse a system where somebody in the high street could walk away, as they did in the Glasgow days, bearing that loss, which is the natural consequence of the end of your logic, of your argument.
Professor Dowd: It may be politically difficult to do this, but I am saying that that system would be a lot safer than the system we have now, but you do not even have to go back to unlimited liability. The key thing is to have some form of liability of the bankers concerned so that we can start to trust them again, and that is what the person on the street needs to be able to safely invest. Currently, banking essentially is just a rip-off. Nobody trusts their bank manager any more; the reputation of bankers is in total tatters. In the old days, we did trust our bank manager—“My word is my bond” and stuff. That really mattered.
Q111 Chair: That was because we had unlimited liability?
Professor Dowd: That is because we had liability—not necessarily unlimited. If you look at the history of banking, there has been joint stock banking with limited liability for a long time as well, but we now have a situation where there is no liability for the main bankers. They can run their banks into the ground and at worst they might lose a knighthood.
Chair: Just to clarify, a central theme of the Parliamentary Commission on Banking was to reintroduce by a number of proxy measures a measure of full responsibility on individuals that in some ways could bring us back towards unlimited liability for high-risk investment banking decisions.
John Thurso: I was going to say, we could discuss this for a very long time, and I am tempted, but I think I had better stop. We had better stick to what we have before us.
Q112 Chair: To take up Professor Dowd’s point with you, Dr Huertas—that stress testing is so weak and unreliable that it could be positively counterproductive, creating an illusion of something being done, when in reality we are offered no protection. Isn’t it the case that when you were involved in all this, that is what the world turned out to look like? We had a tripartite that was supposed to be looking at all this, which subsequently turned out to be doing virtually nothing of use when it was tested.
Dr Huertas: With due respect, Chairman, I think those were quite turbulent times. It is difficult to say that without the efforts of the tripartite matters would have been better or worse.
Chair: That does not sound very good.
Dr Huertas: In my opinion—
Chair: You are saying it would not make any difference—that it would be better or worse in the end?
Dr Huertas: No, in my opinion, the efforts of the tripartite from 2007 on, through 2010/11, made the financial system in the UK substantially better.
Q113 Chair: Yes, but that is too late. The horse has bolted, hasn’t it? The tripartite was there to do the preparatory work for 2007, not to set to work on it from 2007.
Dr Huertas: Yes, and the tripartite did identify certain issues in advance of 2007. As Professor Dowd pointed out in one of his articles, a decision was made not to attempt to change the resolution procedures prior to the failure of Northern Rock. That would be an example.
Q114 Chair: But you would agree, would you not, that the tripartite was an illustration of Professor Dowd’s point—that we had a body that created the illusion of something being done to manage systemic risk, when in fact if anything was being done, it was manifestly inadequate?
Dr Huertas: That it turned out to be inadequate is without doubt.
Q115 Chair: What about the rest of what I have just said in that?
Dr Huertas: My recollection of the discussion was that the authorities bought into the assumptions underlying the Great Moderation and underestimated the risks that a crisis could occur.
Chair: That is a very measured remark. The minutes of the Court are being published tomorrow, and without lifting the veil, I have no doubt there will be interesting things on the tripartite to take a look at. Mike Kane.
Q116 Mike Kane: Thank you, Chairman. I have a series of questions about stress scenarios and how we could regulate them better, but I think I might be in the wrong company with the Professor and the Doctor, because effectively what you are saying is having a stress test is like placing all your bets on Papa’s Moustache in the fifth.
Professor Dowd: Sorry, I missed the last part.
Mike Kane: Having a stress test is as good as a gamble—it is like placing your bets on the horse in the fifth race of the day.
Professor Dowd: I do not think it is as good as a gamble: when you gamble, you know you are gambling, but a stress test gives you false comfort. That is a mortal sin in the risk management business.
Mike Kane: Doctor?
Dr Huertas: I think the stress test, in terms of the result of the stress test—namely more capital and better liquidity—does provide some protection, not only against the scenario being modelled but against other possible sources of stress.
Q117 Mike Kane: Any organisation I have been involved with has to have some form of risk log. What does a risk log look like for a bank then, if we are not having stress tests?
Dr Huertas: The risk log is a quite extensive list of the credit market, operational and other risks that would confront the organisation, as well as the factors that will cause those risks to change.
Q118 Mike Kane: But in terms of known knowns, known unknowns and unknown unknowns, what was 2008 then, in your opinion?
Dr Huertas: I think that 2008 was a factor of essentially gyrations in the US resolution policy. With all due respect, I do not think anyone in the UK expected the US to do what it did in the case of Lehmans.
Mike Kane: Professor?
Professor Dowd: You can look at 2008 in two ways. One is that it is a massive unknown unknown, which is exactly why the banks, had they relied more on judgment and less on faulty models, would not have been so exposed. However, one has to say—and it is easier to say in retrospect—a lot of it had to do with policy: loose monetary policy in the United States, Fannie and Freddie, the housing policies and so forth.
The problems in Europe were partly due to the risk models. One of the problems with the banking system in Europe was excessive exposure to sovereigns. That was encouraged by zero risk weights. In retrospect, you can say that there was a kind of disaster waiting to happen, but it was off the scale of any model that we had. Again, I keep coming back to the track record of these models. It is zero, except for Northern Rock, I think, and then the regulators did nothing about it anyway.
Q119 Mike Kane: So banks could manipulate their capital requirements?
Professor Dowd: Look, any system is gameable, but the models make it a lot more gameable. One of the problems with risk weights is precisely the gameability. This is why simpler metrics are better than complicated metrics. They are not perfect—you can still game them a bit—but you cannot game them to anything like the same extent.
Q120 Mike Kane: So stress tests are also manipulable?
Professor Dowd: Yes, of course. It is just a black box done by a regulator, and then that is supposed to produce confidence. One of the most absurd things about, for example, the European stress test is that at one point they were proposing secret stress tests to promote public confidence. It is a joke; you could not make it up.
Q121 Mike Kane: John Thurso mentioned something that you said previously about transparency. When we go to a bank, I have no idea what I am buying. I base it on the basis of whether I am an ethical banker and I bank with the Co-op or Unity Trust or I like the choir that sings in the advert or the offer of the day. I suppose at least in the US there is some transparency about what goes into a bank, what comes out to a bank locally, regionally. Would anything like that be better in the UK or Europe?
Professor Dowd: I do not think you can get away from judgment. Any black box is manipulable and goes against transparency, so there is no transparency but the appearance of transparency only. At the end of the day, what you want is a system where you can trust financiers. In order to get that, you need liability. We used to have that system. We would go to Captain Mainwaring and you could trust him to give you good advice. Now the only thing I can be confident about with my bank is that they are trying to rip me off.
Q122 Mike Kane: Yes. I was a Farepak customer, or my wife was, and I am just thinking about how the High Court and the judges slammed Lloyds and HBOS about their treatment of those customers—and then the state came along and gave them a huge bailout. I suppose that is effectively your argument: they should have let it fail.
Professor Dowd: Yes, another moral hazard. We need to make people responsible and we need the incentives to do it.
Mike Kane: Anything to add, Doctor?
Dr Huertas: The resolution changes that I have mentioned will produce investor responsibility. With respect to the situation of the depositor, I come back to Mr Thurso’s comment that the small depositor is unlikely to exercise that type of diligence, nor is it clear to me, or to many, that society would wish that type of responsibility to be placed on the small depositor. It is by giving insured depositors preference that the risk of those deposits and the risk to the FSCS in guaranteeing those deposits has been substantially reduced.
Q123 Mike Kane: Do you fundamentally believe that the Bank of England cannot estimate the impact on balance sheets through stress tests as a regulator?
Professor Dowd: No, it cannot.
Dr Huertas: I would disagree. I think it can certainly estimate the impact with respect to the scenario selected and will have, with judgment, a view as to the adequacy of the capital with respect to the possible stresses that the bank might experience.
Q124 Chair: I would like to end with just a few questions that relate to specific points you made in your book, Dr Huertas. If I may say so, this book is a first-rate read; I read it at the time it came out. I think it is fair to say that today we have had before us a radical free market challenge, but we have also had in front of us a pillar of thoughtful orthodoxy pretty much—
Dr Huertas: Thank you.
Chair: —on a large part of this. But I would just like to go through a few of the issues that you raise. You say in your book that the risk management function in a bank should be in the business, but not of the business. Are you not also worried that now we have given so much more power to the Bank of England, we are accumulating financial stability risk at the Bank of England and that therefore we need much more scrutiny of the way they are managing risk?
Dr Huertas: That was certainly discussed at the time the change was made and I would concur that that is necessary. I believe this—
Q125 Chair: Do we have enough?
Dr Huertas: That is for you to judge, sir. I believe this Committee is—
Chair: No, I am asking you to give us guidance. You are very well-placed to do so.
Dr Huertas: Certainly this Committee is in the forefront of exercising that scrutiny.
Chair: We are doing our best. I am asking you whether we have enough and, if not, what extra tools we might need.
Dr Huertas: Let me come back to you with a more considered response, Chairman.
Q126 Chair: In another part of the book, you say that we relied far too much on the risk rating agencies and I think this is generally accepted. Have we really put that right? Aren’t we still relying on their wayward judgments about credit risk? They did better on the commercial side than they have done on the sovereign side, but even so.
Dr Huertas: We have lessened the reliance on the rating agencies.
Q127 Chair: Have we really?
Dr Huertas: With respect to formal reliance or delegation to the rating agencies for the characterisation of the risk weight in regulation, certainly that has been the case in the US as a matter of law.
Q128 Chair: So the weakness that you identify in your book you think is now being addressed or not addressed?
Dr Huertas: It is on the way to being fully addressed.
Q129 Chair: You also elsewhere in the book—although I do not have the page reference, I remember reading it—have a pretty good go at audit and the accounting approaches that were taken in banks. Do you think we have that sorted out? We have just had Tesco; it does not seem that the audit does a very good job there either. That is a much easier operation than a bank. What is the point of all this auditing? I am sure we need to do it, but what are we getting for it?
Dr Huertas: An assessment of the state of the balance sheet at a period of time.
Q130 Chair: Are we getting just false confidence?
Dr Huertas: The audit is a point-in-time snapshot; it is not a guarantee that the institution will survive over a forthcoming period.
Q131 Chair: Can I ask you whether you think we have addressed the severe weaknesses in that function that you set out in your book?
Dr Huertas: I think there has been some progress in that direction, but there is further to go.
Q132 Chair: In a percentage between nought and 100?
Dr Huertas: I would hesitate to put a figure on it at this stage.
Chair: I am going to ask you. After all, “some progress” does not tell us a lot. Are we talking about 20% progress or 80%?
Dr Huertas: Some 60% to 65% would be off the top of my head.
Q133 Chair: All right. We are two-thirds of the way there.
Can I end by asking you a question where I began, Professor Dowd? Let us take a bank that is illiquid but not insolvent because of some exogenous shock that could not reasonably have been foreseen by that bank, for example. There are three options, aren’t there? We can let it fail, we can nationalise it, as you were earlier implying—or more than implying—or we can allow it access to a lender of last resort at high or penal rates. Just to clarify, are you absolutely clear that we should either let it fail or nationalise it and eliminate the option of providing lender of last resort facilities to such a bank?
Professor Dowd: I would prefer that we eliminate it.
Q134 Chair: No, it is not a preference. We have to take a decision here; this is policy-making.
Professor Dowd: Yes, let it fail. Let it fail.
Dr Huertas: May I disagree?
Q135 Chair: You have disagreed. I will give the last word—since Professor Dowd had the first word in this hearing—to you, Dr Huertas. Why? This seems to me to be the most fundamental question in this field.
Dr Huertas: The premise on which you made the question, namely illiquid but not insolvent or meeting capital requirements?
Chair: If you could detect the difference.
Dr Huertas: There are instances where you can detect the difference—for example, where the bank’s system has failed and it is unable to receive payments, but it can make payments in such an instance. A lender of last resort facility can keep the bank in operation.
Chair: We have had quite a detailed canter around a very, very important field that has exercised the minds of this Committee and a number of others, including the Parliamentary Commission, for some years. I am very grateful to you both for coming in. We have had an interesting contrast between a radical challenge and the orthodox view. Thank you very much.
Oral evidence: Bank of England December 2014 Financial Stability Report, HC 912 21