UNCORRECTED TRANSCRIPT OF ORAL EVIDENCE To be published as HC 881-iv

HOUSE OF COMMONS

HOUSE OF LORDS

ORAL EVIDENCE

TAKEN BEFORE THE

PARLIAMENTARY COMMISSION ON BANKING STANDARDS

SUB-COMMITTEE H

THURSDAY 24 JANUARY 2013

PANEL ON TAX, AUDIT AND ACCOUNTING

RICHARD CARTER, STEPHEN HADDRILL, JIM HARRA, ROGER MARSHALL, PAUL SHARMA and MIKE WILLIAMS

USE OF THE TRANSCRIPT

Oral Evidence

Taken before the Parliamentary Commission on Banking Standards

Sub-Committee H-Panel on tax, audit and accounting

on Thursday 24 January 2013

Members present:

Lord Lawson of Blaby (Chair)

Mark Garnier

Examination of Witnesses

Witnesses: Richard Carter, Director, Business Environment, Department for Business, Innovation and Skills, Stephen Haddrill, Chief Executive Officer, Financial Reporting Council, Jim Harra, Director-General, Business Tax, HM Revenue and Customs, Roger Marshall, Chairman, Financial Reporting Council, Paul Sharma, Director of Policy, Financial Services Authority, and Mike Williams, Director, Business and International Tax, HM Treasury, examined.

Q269 Chair: Thank you for coming to see us this afternoon. I am acutely conscious that, for various diary reasons, this particular hearing involves more witnesses than members of the Commission, but at least both Houses of Parliament are represented here.

Before we start the questioning, would any of you like to make a statement? If not, we will go straight to questioning. I propose to start on the tax issue and then move on to the accountancy issues.

The main issue on tax, so far as the Commission is concerned, was brought to wide attention in the Mirrlees report. This is the different treatment of loan capital and equity capital, whereby loan or debt interest is tax deductable and there is no deduction for dividends paid on equity capital. That, of course, gives an incentive to raise debt rather than equity, whereas the regulators would like to see the banks with more equity rather than debt, so the two are pushing in opposite directions. The Mirrlees report suggested that the playing field should be levelled for the economy as a whole and suggested two possible ways to do that, with which you will be familiar: disallowing the debt interest, or giving an allowance for corporate equity. I am not asking you to say whether you would like to see these, but I will ask you two questions. Do you think that the idea is workable? If it is workable, which of the two options do you think is the more workable-where would you tilt it? Of course, it could be both, which is the proposal from Andy Haldane of the Bank of England.

Mike Williams: Let me start, Lord Lawson. If I can do some ground clearing first of all, I assume that we are talking about debt that is loan capital or regulatory capital-the sort of debt that forms part of the fixed capital of a bank-not, for example, debt arising from a deposit. It is very difficult to argue with the latter sort of debt that you would not relieve the interest.

Chair: You are quite right, Mr Williams. That was what I had in mind; thank you for clearing the ground.

Mike Williams: It is far from clear that the benefits of changing to one of the two systems you have identified would exceed the costs, and that is even before you have considered the costs of transition. In an ideal world, you would level the playing field, and you would not have the asymmetry between the treatments of debt and equity. The difficulty is that actually to achieve that, you have offsetting disbenefits, which I think are quite significant. I would argue that that explains why, in reality, most developed countries have tax systems that relieve the payment of debt interest but, on the other hand, they do not generally give deductions for any sort of remuneration of the equity capital.

Q270 Chair: What is the principal disbenefit in your judgment?

Mike Williams: There are two main categories of them. One just looks at the comparison of the different systems, and the second is the transitional costs of getting there, given that you have a banking sector that has raised capital-either equity capital or debt capital-on the back of the existing system. So I think that there are quite significant transitional costs if you were to change, particularly if you were to start to disallow interest on debt that has been raised on the broad proposition that the interest would be relievable. There are significant difficulties with both ACE and with CBIT. It is probably easier to introduce a system of allowance and corporate equity, and there are one or two countries that have done it, although equally it seems that there are countries that have introduced it and then rather retreated from it, which perhaps hints at difficulties. It is relatively straightforward to do. You either give a deduction for the actual remuneration of equity, or you impute some deduction as a reasonable amount of remuneration for equity. A difficulty with it, inevitably, is that if you maintain the same rate of corporation tax with the allowance for equity, you have significant public finance implications for getting less tax revenue.

In contrast, if you increase the rate of tax so that you remain revenue-neutral, you are increasing distortions on the borderline between what is and what is not taxable. If, instead of a rate that will be 21% in April 2014, you have a significantly higher rate than that, I think that would be difficult. There are also questions that you may or may not want to go on to about whether you would extend this system to the whole of the corporate sector, in which case there would be quite significant winners and losers across industries that have different levels of gearing, or whether you would confine it to the banking sector, but of course you then have boundary issues.

There are also significant issues, given that the UK is a major financial centre, in moving to the system in isolation-unilaterally, if you like-when most of the rest of the world would not have it. You could contrast the position. If we assume that banks are competing to raise debt capital, it would be easier in a world where we had stopped the relief for interest. It would be easier for a bank in some other centre where we had not done that to raise capital-the costs would, in effect, be cheaper for that bank, which would impact on competitiveness. It is also quite difficult to isolate a particular category of interest on which you no longer allow debt relief. You have to go through and make the distinction, which we talked about at the start, between what is debt capital, where the interest would be disallowed, and the whole of the rest of the interest.

Equally, within a group of companies-you must remember that in the UK we tax company by company within a group-if you are going to say that any of this debt is disallowed, there may also be interest receipts of other group companies. If you are saying that the payer does not get a deduction, you have difficulty saying that the recipient would be taxed on that.

That is my view in a theoretical sense. If you look outside the UK, there are countries that have tried reforms, particularly with the ACE sort of system, and they have sometimes reversed the reforms relatively quickly.

Q271 Chair: Who has reversed the reforms?

Mike Williams: I think, from memory, Belgium was an instance where it was introduced-sorry not Belgium. Perhaps Belgium is one of the ones that has introduced it and actually kept it. It has this thing called a notional interest deduction, which in reality is-

Chair: My understanding is that Belgium still does it.

Mike Williams: The ones that seem to have tried are Croatia, Italy and Austria. By "tried", I mean that they introduced that system and then rather retreated from it. If you look at countries that genuinely try to restrict interest more than the UK does-the United States is an example of where there are more restrictions on relief for interest than in the UK-they tend to have quite complex systems for restricting relief that are then quite costly for businesses to operate.

Equally, as I said at the start, I would not underestimate the transitional costs of getting there. If you have raised a load of debt capital on which you assume that you will get relief on the interest, at the margins you may find difficulties paying the interest, and equally if you were to grandfather the existing debt at the time you made a change, meaning that the interest on the old debt was relieved but not that on new debt, it is quite likely that there would be a behavioural impact because banks would tend never to retire the old debt and would try to extend its life as far into the future as they could.

In conclusion, our view is that it is more sensible to stick with a corporation tax system that broadly follows that of most other countries and to rely on other measures, notably non-tax interventions on financial regulation, and also that the introduction of the bank levy will have a greater impact on the financial sector than trying to remove the debt/equity asymmetry.

Q272 Chair: Let me pursue that a little further. As I understand it, you say that you are not keen on either of the Mirrlees proposals but, of the two, you think that the allowance for corporate equity is less problematic than the disallowance of debt interest. Is that what you are saying?

Mike Williams: I think it is probably less problematic to introduce, and that might be why some countries have gone down that road, whereas the numbers that have taken the CBIT approach are rather fewer. There are significant difficulties with it though; if you assume that you would extend it across the whole economy so that you have a corporation tax system that does this throughout sectors, to maintain the same amount of tax revenue you would inevitably have to increase the tax rate. For some-

Q273 Chair: I am sorry to interrupt, but you would not need to do it right across the economy. You have indicated that there is a possibility of doing it just with the banking sector. Although, and I cannot remember your expression, you mentioned that there were border problems-

Mike Williams: The boundary issues.

Chair: Yes, but I recall that there are border problems in pretty much every tax issue, and I would have thought that this is one of the clearest ones, because an institution is either liable to the bank levy or it isn’t. The border is already there, so if you said that you would have this allowance-you see, this is almost as a quid pro quo-for those banks or institutions that are liable to the bank levy, that solves the border problem and the problem that you quite rightly indicated of the cost of extending this to the entire corporate sector. Indeed, if you-sorry for making this a long question, but I will sit patiently while you answer it-or the Chancellor of the Exchequer was concerned about loss of revenue, even if doing it only in the banking sector, he could compensate by increasing the bank levy notionally. So exactly the same institutions would be paying basically the same, but he would do this tilt, because, conceptually, many people have suggested that, in economic terms, the bank levy is very similar to a penalty on corporate debt-bank debt-which is what a disallowance of interest on bank debt would be.

Mike Williams: I think we need to bear in mind, first of all, that when we crafted the bank levy, drawing the boundary was one of the more difficult issues.

Q274 Chair: But you have done that.

Mike Williams: We have done that, but if we then use that boundary for a different purpose, I do not think that we can assume that we have already solved the problem.

To give you an example of the boundary issues, when you have a banking group with a very large insurance operation, do you include the insurance operation or not? It is relatively straightforward to exclude the insurance operation in relation to the bank levy, because an insurance operation does not take wholesale deposits, for example, but if you are looking for an allowance on equity, you would have to look into the group and see whether the equity was supporting the banking business or supporting the insurance business-that is different from what we do now.

Equally, and perhaps this is a more important point, I would not agree that the bank levy or any form of corporation tax has similar incidence. If we look at the bank levy, there are only about 30 payers, because there is a high threshold of deposits before you actually start to pay. There are probably 300 to 400-odd banks within the corporation tax system. Similarly, there are banks that are paying the levy because they have significant amounts of deposits that have triggered the levy, but they are not paying corporation tax; yes, they will do so in time, but at the moment they have very significant losses to relieve against any current profits. If you switch between the two, you get quite significant differences in who pays, so I don’t think you can assume that you can readily switch between the two.

Q275 Chair: I have sufficient confidence in the Treasury to believe that it if wished to do this, it could, with the assistance of HMRC, work out a suitable borderline, although of course that does not dispose of all the other issues that you rightly raised. One last question on this: have you made an estimate of what it would cost to introduce a Mirrlees-type ACE in the banking sector?

Mike Williams: We have looked at ACE across the piece, but we have not, if I recall correctly, looked at it specifically in the banking sector. To go back to your earlier comment, I am sure you are right that we could introduce an ACE, but the question with all these things is whether it would be worth it in terms of the costs. Equally, having struggled before with tax changes that were revenue-neutral, I think it would be much more difficult to get to a position where, within a revenue-neutral change of the sort that you are describing, we didn’t end up with really quite significant winners and losers. That is particularly the case given that the bank levy-or any other offset-would be paid by a rather different population, and within that population each would pay different amounts. That is quite a significant difficulty.

Q276 Mark Garnier: Mr Carter, I have one question continuing on from this. The imbalance, if you like, between debt capital and equity capital is blamed on this tax situation, but what implication does that have for corporate governance? It is equity shareholders who can vote for the board and governance of a company, but the vast majority of the capital of a bank is debt. Do you have a comment on how the tax system can improve-or otherwise-the corporate governance at banks?

Richard Carter: I agree with your analysis that, assuming that a business is solvent, it will be the shareholders who are doing the corporate governance. That is the basis on which they are the shareholders. The providers of debt will know that if they want to exert influence over the company, they will have to do that in other ways, such as covenants in the debt instruments or whatever. You have correctly set out the arrangements that everybody is familiar with and would be expecting in current circumstances.

Q277 Mark Garnier: A great deal has been said by your Secretary of State-and nobody else-there should be more reliance on shareholders to get involved in this, and yet if you simply have a system where shareholders are the minority interested parties in this type of thing, clearly it is an issue.

Richard Carter: I think that where my Secretary of State is coming from is a desire to make sure that shareholders, as they are the owners of the business, are holding directors to account and that they should be thinking about that with a view to the long-term prospects for the business and that the owners of equity should regard themselves as fulfilling that task, rather than just being people who owned it for perhaps 30 micro-seconds and did not really know what it was that they were owning and did not really care, because they were just looking to make a quick buck, as it were.

Q278 Mark Garnier: Have you any thoughts about bail-in bonds, being a sort of hybrid of one or the other, in terms of how shareholders should vote? Should they vote when the thing goes bust and they are bailed in, or should they have a vote in an AGM for private equity shareholders?

Richard Carter: It seems to me that if you go back to the old rules that you would have seen around preference shares, for example, 20, 30 or 40 years ago, it is perfectly open for the company to have a capital structure that effectively says, "When everything is going fine, it is the ordinary equity shareholders who are fulfilling the corporate governance role. As life gets more complicated, you may have instruments which, with the agreement of the ordinary shareholders, give other people more of a say in the running of the company." That seems to be inherently one of the great flexibilities of the UK model. It enables the providers of capital to agree among themselves whether that is the particular structure that they want to put in place, be it for a manufacturing business, a bank or whatever.

Q279 Mark Garnier: Okay, fair enough.

Paul Sharma, I will address some questions to you, although if the Financial Reporting Council wants to leap in, I will be grateful. One of the big questions with the wholesale banks is that of proprietary trading. There are a number of issues regarding how you identify what type of proprietary trading that is. So if you take something as simple as market making, there is an element of market making, which is where you provide liquidity to the market, where you stand up and you are challenged to make a price, and you make a price. Let us assume that that is a virtuous, useful thing to the wider community.

Embedded within that is an element where market makers may decide they want to take directional bets on the market, which is where it is essentially proprietary trading directly for the banks to try to enhance value for their shareholders. Actually, it serves no useful value to society-it is of marginal interest in terms of liquidity-but they are not doing it to help society; they are doing it to help their staff and their shareholders. This gets slightly more complex by the time we take into account things such as underwriting and various other bits and pieces. In general, do you think that it is possible to identify one type of trading from another in any practical sense, even if that includes redefining what a market maker is in order to leave everything else isolated? Is that possible?

Paul Sharma: I think it is very difficult. As you know, a number of people have tried and are trying. It is not clear that they will be able to succeed.

Q280 Mark Garnier: You are referring to Volcker.

Paul Sharma: Indeed, and it was considered in the Liikanen proposals, and the conclusion was that it would potentially be too difficult. I do not think that we actually know for certain the answer to your question, but we do know for certain that it is extraordinarily difficult-for two or three reasons. First, the boundary is essentially based upon intention. The same acts, as they appear to the outside world, could be inside or outside that boundary, depending on the intention of the parties.

Secondly, the amounts that a market maker in one or other market is formally required or obligated to buy or sell to keep the title "market maker" tend to be much less than what actually needs to be bought or sold in order for it really to be a market. In large measure, the market maker can make a choice when somebody wishes to buy or sell: "Do I wish to buy from them? Do I wish to sell to them?" Distinguishing that choice from the directional choice that you have described is a difficult one.

Q281 Mark Garnier: Can I just ask one quick question? My time on the floor of the stock exchange harks back to the days when you could literally walk up to an individual, and they would be required to make a price, in reasonable size, on demand. If he did not, you could grab him by the lapels. If he continued not to, a well-aimed right knee would get him to do what you wanted.

Paul Sharma: Correct.

Q282 Mark Garnier: That is no longer the case, is it?

Paul Sharma: The key point in what you said is reasonable size. Depending on which market you are in, the reasonable size tends to be relatively small, and somebody who is a serious market maker would often need to be above that size. That is the point at which it is voluntary. When somebody with something slightly larger than the reasonable size comes along and says, "I wish to buy this" or "I wish to sell this", and someone says yes or no, how does one untangle the intention behind that?

The third thing is a certain degree of what some people might call proprietary trading is necessary for price discovery. Other people might call that market making.

Q283 Mark Garnier: Okay, so it is very difficult.

Paul Sharma: Yes.

Q284 Mark Garnier: But, none the less, you will agree that a number of institutions out there are doing proprietary trading for no other reason than to take a view.

Paul Sharma: Oh yes.

Q285 Mark Garnier: Do you find that they will have a separate book, where they have a bunch of prop traders who just sit there doing whatever they like, or is this generally mixed in with market making and the derivatives-writing book? Is that how it works?

Paul Sharma: A little bit of a generalisation, to answer. People will tend to have separate people doing separate things according to the firm’s definitions-you know, "You set of people, you’re meant to pursue the following strategy. You other set of people, you may be trading in the same markets, but you’re meant to be following a quite different strategy for a very different purpose." If you were to take any single regulator or law-created definition and superimpose it on that, you would find that that would cut individuals in two.

Q286 Mark Garnier: Are you anywhere close to coming up with a solution to this problem? Roger Marshall, you have been nodding enthusiastically in the background. I am not suggesting that you do have the answer to that problem, but feel free to leap in if you want to.

Roger Marshall: I am in learning mode, I am afraid to say.

Mark Garnier: Right. Paul Sharma, back to you.

Paul Sharma: We are continuing to look at this question very closely. There clearly is experience to be learned from. I was careful at the beginning not to say the problem is insoluble. I simply say the problem is very difficult and one that we need to continue to look at. The problem is clearly a key one for Government and Parliament in the progress of the Banking Reform Bill.

Q287 Mark Garnier: Would you agree, though, that if it is possible-presumably, this is why we are looking for this-it would be desirable, at the very least, to alert investors and counterparties to those banks that do take this proprietary trading and, arguably, unnecessary risk? Do you agree this information should be reported so that it is available to everybody who wants to deal with that institution?

Paul Sharma: I would agree that it is very important that the risks that an institution is taking are transparent to counterparties, investors and regulators. But bear it in mind that the same act, with a different intentions, can be characterised as proprietary trading or not proprietary trading. The same act, or the same position, potentially gives rise to the same risk, so I am not in the binary world of thinking that activity that we might consider to be of higher social value is necessarily of lower risk.

Q288 Mark Garnier: No, I think there is the same risk-sorry, I want to bounce this back at you-but if you have a higher social value, you can justify that risk for reasons of the greater good. But it may purely be for the purpose of enhancing shareholder value and generating a bonus for the individual who took the risk. That is the other element of it: if you are looking at the incentive to the staff to take a risk and give themselves a bonus to buy the latest 911 come Christmas or January, whenever their bonus is paid, that risk is unjustified in terms of social good. Do you agree?

Paul Sharma: Indeed. That is a very valid point. The only thing I would add to that is, unfortunately, at the moment, I see this as a question of shades of grey, rather than black and white, because of the blurred nature of the distinction. We can think of corner examples that, clearly, anybody would call proprietary and market making-transacting when you are, in a formal, narrow and legal sense, obliged to do so by virtue of whichever formal role you hold in a market. We then have a grey area in the middle, which I think is shades of grey. Should the shade of grey be more visible or transparent to regulators, investors and shareholders? Absolutely spot-on, yes.

Q289 Mark Garnier: So you think it should be.

Paul Sharma: Yes.

Q290 Mark Garnier: You would advocate some type of reporting process where the amount of risk that has been taken by that bank on a day-by-day basis or the amount of capital allocated to that type of risk should be published for the benefit of everybody.

Paul Sharma: Indeed. That is the easy question to answer. The hard question is, "How?"

Q291 Mark Garnier: But you could answer that question by publishing all the risk capital and say, "This is what has been allocated to the market-making book, and this is to the whatever." By the time you have identified it or otherwise, you will have encapsulated the whole lot of the balance sheet that has been committed.

Paul Sharma: Indeed you could. The disclosure that you have just mentioned would have a number of utilities, of which this is but one, meaning exposing more clearly what the capital allocation of the firm is. There is, none the less, a fundamental difficulty. If you ask them, "What capital do you use for market making according to a definition that has been imposed on you, as opposed to proprietary, according to the definition that has been imposed on you?", that might not be wholly clear, even to the firm itself.

Q292 Mark Garnier: But, if there was a pool of cash, some of which was definitely virtuous market making and some of which was definitely bad, evil proprietary trading for no useful purpose, if it was all in the same package, as an investor, counterparty, regulator or indeed the FPC looking at financial stability, you would at least know where you were with potential risk, irrespective of how that breaks down.

Paul Sharma: Yes, you would know. This is a package.

Mark Garnier: That is very helpful. Can I just ask one last set of questions?

Chair: Yes, but I want to come back on this.

Mark Garnier: Why don’t you do that now?

Q293 Chair: I have been very interested in your answers, Mr Sharma; you have been very helpful. The question of what is of social value is not primarily what we are concerned with in this Commission, but we are concerned with-we have been charged to accomplish this-the question of banking culture and standards.

You would agree-I asked you whether you would, but I suspect you would-that the culture of proprietary trading is a very different culture from the culture of serving clients, which is what market making is.

Paul Sharma: I would agree that they are different cultures. The culture of market making and serving clients is, in turn, a very different culture from, for example, serving retail clients.

Q294 Chair: You have said that the difficulty-I am glad that you said that it is not impossible-is because what you are talking about is the intention of a particular transaction, or what is the intention. Paul Volcker, when he was asked that question, replied that if a bank does not know what its intention is, the management should be fired. I agree that that does not solve your problem, but there is clearly a difference of intention, which is known, of those who are responsible for the transactions that are occurring.

You, as the regulator, could make a decision that does not have to be as watertight as something that would stand up in a court of law. You have the power, do you not, to say, "We have taken a view that the intention of these transactions is just pure proprietary trading and speculation"? I have nothing against speculation as such. The question is whether banks should be doing it. You could do that, couldn’t you?

Paul Sharma: On reasonable grounds.

Chair: Oh yes, of course.

Paul Sharma: On the basis of evidence on reasonable grounds.

Q295 Chair: Yes, but you don’t require the same standard of proof, as it were, that a court of law would, do you?

Paul Sharma: Certainly not that a criminal court of law would.

Chair: Yes.

Paul Sharma: If you say that in a civil court of law it is the balance of probabilities, I think we might. Depending on the precise action we took, we may or may not be in the same territory as a civil court of law.

Q296 Chair: Before I go back to Mark Garnier, I have one final question relating to the earlier conversations with Mr Williams about the cost of ACE for banks à la Mirrlees. If a written question were tabled, I hope you would answer it and not say that it was too difficult. Can you promise me that you would do that?

Mike Williams: I think it all depends on how clearly specified the question is and how readily we could produce numbers. I would not neglect the boundary issue.

Chair: The boundary would be those banks that are subject to the bank levy. The boundary is the boundary rules you have already drawn.

Q297 Mark Garnier: What I want to concentrate on is the cross-pollination of information between the regulators and everybody who was important. Yesterday, we had Robert Hodgkinson from the ICAEW giving evidence to us. He proudly held up the Bank of England financial stability report from January 2007, if I remember rightly, saying wasn’t it fantastic how stable the banks were and everything was tickety-boo and nothing to worry out and that we were in rude health, robust and all the rest of it and proudly asking how on earth auditors could know what was going on if the Bank of England did not know what was going on. He failed singularly to answer the single question, which was presumably the Bank of England was reading the auditors’ reports. So you have this sort of circularity of information going on. It is no good the auditors saying that the Bank of England thought it would be perfectly alright, when the Bank of England was looking at the auditors. So you get this positive feedback loop.

Clearly there is an issue with the amount of information that is being shared. Many people were looking at different aspects of financial institutions. For example, as a regulator you are looking at-is it quarterly returns for banks or monthly returns?

Paul Sharma: It depends on which bank and which regime.

Q298 Mark Garnier: But at the very least it will be quarterly returns. As an auditor you are looking on an annual basis, which could mean that you are in the bank for nine months, presumably on an ongoing basis. What is going on with the cross-pollination of information? That is a general, loose question. Why does there not seem to be any effective cross-pollination of information going on at the moment?

Paul Sharma: The first thing is to make it clear that achieving a good, stable solution to what you have just described is a key priority for the new PI. It has been acknowledged as such publicly. Secondly, in anticipation of the PRA, the FSA, Bank of England and the major auditors have worked together to produce a code of practice for the sharing of information which is in the public domain. I am sure you have a copy but if not, of course, we can give you one. What that does, together with the law, which permits such sharing of activity, is it creates the right formal structure for the sharing of such information. What we have started doing, and this is a major early priority of the PRA and indeed of the FSA now, in its handover period, is to create the right culture among our supervisors and among the auditors such that there is the degree of trust and the degree of urgency so that we get the kind of good quality sharing of information that is absolutely necessary for a modern, effective, prudential regime.

Q299 Mark Garnier: How does that work on a practical basis? The auditors have a duty to the company-they are who the cheque pays for-yet they talk to the regulator. The layman would see a conflict of interest there. Clearly, some of the information that the auditor hands over to the regulator could arguably be not in the best interests of their customers, albeit that that is a slightly perverse view of what their best interests are.

Paul Sharma: Indeed. I will give two or three aspects in my answer to that. First, it is in our prudential regulatory interest that the auditors do a good job in discharging their role to the company. We do not need the auditors to become shadow regulators and start doing our job or start assuming purposes different from those that the Companies Act has bestowed upon us. The good, successful quality pursuit and achievement of those purposes helps our aims. That motivates us to want to share information with them.

What motivates them to want to share information with us? There is a narrower set of information, which they are obliged by law to share with us. That is not typically information that they are required to seek out, but if that information comes to their attention in the course of undertaking their audit responsibilities, they must bring it to our attention. There is a wider set of information, which they are allowed, but not required, to report to us. The argument here is that it is often in the best interests of the company to do that. So I do not think there is a conflict. There may be confusion-this is one of the cultural issues-between the best interests of the management of the company and the best interests of the company. They are not necessarily-especially in the sort of circumstances we might be talking about here-the same thing.

Q300 Mark Garnier: We have heard that it is beneficial for regulators and HMRC to work together, in particular in their relationship with banks. Mr Harra, do you have any comments on that?

Jim Harra: We work quite effectively together, particularly on policy making. For example, with the introduction of the new tier 2 capital, we were able to advise Ministers, in response to representations from banks, to make sure that they have certainty about tax treatment when that regulation comes in. When it comes to exchanging information about individual banks, there are restrictions on the gateway, which can sometimes be problematic. It is not a very serious matter, but it does sometimes mean that we have to make our own judgments in HMRC when we are told by banks that things are being done for regulatory reasons, because there is no gateway to exchange that.

Q301 Mark Garnier: When you say gateway to exchange that, what does that mean?

Jim Harra: For example, in our legislation for deducting interest, there is an unallowable purpose test. Banks will sometimes say, "The purpose for doing this is regulatory reasons". Mainly in European law, there is a restriction on the FSA being able to confirm that to us or not, so we have to gather the evidence ourselves and reach our own conclusions on whether that is right.

Q302 Mark Garnier: Do you think that is nonsense? You probably can’t use language like that.

Jim Harra: I don’t think it is a serious problem, although it is not ideal from our point of view. It would be ideal if we could have that information. We have ways around it and it has not proved an insurmountable obstacle to us. But it is not ideal.

Q303 Mark Garnier: But you are looking at that from the point of view of your collection of taxes, aren’t you? That is, of course, very important because we have to close the tax gap and we all agree that you should use every mechanism at your disposal to ensure that that happens. The other side of the coin is the regulatory side. It is probably quite difficult for you to think of any examples of when you might have heard of something as a result of your investigations into a financial institution that would have given you stability alarms. Or is it just something you are not looking at at all?

Jim Harra: I am not aware that we have uncovered anything of that nature, but we do have gateways where we can report that if we uncover it.

Q304 Mark Garnier: So you do have a dialogue where you can?

Jim Harra: Yes.

Stephen Haddrill: As do we, with the FSA. Quite often through the work of our audit inspection team and so on we find things that give us cause for concern, and we raise that with the FSA. The FSA no doubt does the same if they want us to focus our inspections on a particular area.

Q305 Mark Garnier: This is my last question. How markedly different is this from pre-2007 crisis? What was the situation like then and how different is that from now? Can you give us an assurance that we will not have a situation that allows financial collapse to happen without anybody seemingly being able to spot it?

Paul Sharma: To answer your question in respect of all the different relationships that you have talked about, the most interesting one of course is that with auditors. Pre-financial crisis, there was a well-documented lessening of regulatory attention on prudential issues in general. The falling off in communication between the prudential regulator and the auditors was a part of that overall lessening of attention on prudential issues. I am quite clear, and the whole senior management designate of the PRA is quite clear, that we cannot be an effective prudential regulator without significant, good, strong, ongoing, timely two-way communication between the auditors and the supervisors. That is why we have acted to put guidance in place. Reference the comments I made about putting the right culture in place.

Will there be another lessening of the attention on prudential supervision? Well, part of the answer to that is that Parliament has now established two regulators, with the PRA having the sole focus on prudential precisely to address that problem. On the Financial Reporting Council and HMRC, the answer is roughly similar to the one that I have just given, albeit the falling off was not as great. The main difference between post-2007 and pre-2007 is that there was actually a lot more to talk about.

Mark Garnier: There certainly was.

Q306 Chair: I have no doubt that the lessening of attention and the cavalier attitude to supervision, and auditing too, in the manner that you have described was deplorable, but I am slightly puzzled as to how it came about. I recall that, when I was Chancellor, I was very concerned about the inadequacy of supervision and that is why I introduced the 1987 Banking Act, which made a number of changes, including the creation of the Board of Banking Supervision-some of you may recall it. It was quite effective when it started, but then, after my time, it became less effective and was abolished. In the 1987 Act, I also made it very deliberately possible for there to be this dialogue, which I felt was crucial, between the auditors and the supervisor. I made a big mistake. I encouraged it, but I wanted to make it mandatory. At the time, the Bank of England advised me that that was quite unnecessary, so after a few years it stopped happening, which was appalling.

You now say that you have this code of practice, but, based on the experience that we have had, do you not see that that might not be-I would be interested in the view of Mr Marshall and Mr Haddrill-enough and that it may be necessary and wise to make it a statutory duty, as indeed the House of Lords Economic Affairs Committee recommended and as Hans Hoogervorst of the IASB agreed would be a sensible move when he gave evidence? Would you agree with that?

Stephen Haddrill: Yes. The arrangements that Paul has described are satisfactory, but you are raising a different matter about how we can rely upon them being maintained for all time. It is very hard to identify any disadvantage to those arrangements and any reason why we would want to remove them in future. You might say that Parliament therefore does not need to get involved, but there is no downside to having that as a mandatory arrangement.

Paul Sharma: My view is yes, wholeheartedly.

Q307 Chair: Thank you very much.

I have one question for Mr Sharma before I move on to the pure accounting issues, which we need to discuss. I was slightly alarmed when you said that, in this dialogue between the auditors and the supervisors, there was a problem of trust. What is that problem?

Paul Sharma: No, it is more a question of establishing regular meetings where they get to know each other very well. You refer to a past period of banking supervision. I was not a banking supervisor then; I was an insurance supervisor at DTI. Believe it or not, even in humble DTI we had that close relationship of trust by virtue of knowing each other well and meeting regularly. I was referring to recreating a structure in which people meet regularly, rather than people not knowing each other and distrusting each other.

Q308 Chair: I see. Thank you for clarifying.

I now move on to the accountancy issues. Again, this is a question that I would like to put to Mr Haddrill and Mr Marshall. A number of serious problems emerged with IFRS. I am not saying that UK GAAP was all hunky dory, but a number of serious problems have emerged with IFRS: the move away from prudence; the move, as it has been interpreted, by many accountants to box-ticking, rather than the exercise of judgment; and the weaknesses, and in many cases the complete artificiality, of fair value, mark to market and mark to model. When did you first become aware of those problems?

Roger Marshall: I am not sure I would wholeheartedly agree with the premise. Perhaps I could respond to the premise. I do think there are issues with IFRS, and I will deal with prudence first.

The basis of IFRS was, up to 2010, prudence. I think most of the detailed standards in IFRS come from a prudence standpoint, but some do not. The period when the IASB were co-operating with the US standards setter to try to harmonise accounting standards was not a period in which prudence was emphasised. Their most recent statement of principles, on which they are still working and which were launched in 2010, did not include prudence as a concept. We have gone back to them quite recently saying that we do not agree with that; we think it is important that prudence is a basic concept of accounting standards.

Although the IASB do not want to go back and reopen those particular sections, they have come up with a new word, which is "caution." They say that caution is the bedrock of IFRS, and they want to emphasise caution in their standards going forward. So I think they are possibly changing direction now they have stopped co-operating with the US standards setter in trying to produce harmonised standards. The standards, certainly up to the credit crunch, were based on a statement of fundamental principles that included prudence; it is only more recently that they have not.

Q309 Chair: And on the other matters I asked about.

Roger Marshall: On the box-ticking approach, I think that IFRS, as a whole, is principles-based. There are a lot more anti-avoidance rules in IFRS than there were in UK GAAP, and I think that is unfortunate. We have consistently pushed back, saying, "What we need are principles." Anti-avoidance could be achieved in other ways, rather than lots and lots of rules. IFRS is, none the less, much more principles-based than, say, US GAAP, which is, I suppose, the main competitor.

I do feel, and I think we and the FSA said this in our written evidence, that the period after IFRS was introduced in Europe was a period when everyone was trying to get comparability. What they did not want was IFRS launched in a hundred different ways, reflecting the national practices beforehand. There was a lot of emphasis on consistency and on trying to make everyone do the same thing. That was a slightly unfortunate time. Since then, it has reverted more to looking at principles.

However, I do not believe that IFRS is just a box-ticking exercise. Indeed, it is hard to say that and at the same time accuse companies or auditors of not being prudent enough. If it was just a question of ticking the boxes, there would be no choice and as it is there is a fair amount of choice in how you apply these standards. You can apply them in a more or less prudent way. I know that a number of people feel that they are not being applied prudently enough in the UK in some cases. I am not sure that I wholeheartedly subscribe to the premise that it is just box-ticking, but there is some of that.

Q310 Chair: And the third problem that I referred to: mark to market and all that, and fair-value accounting.

Roger Marshall: I think mark to market and even mark to model are unavoidable in some cases. Accounting standards have gone through a rather wandering path of going fairly wholeheartedly into mark to market and then retreating from that. I think that where the IASB is going with its current financial instrument standard gets it about right. Assets and liabilities in the trading book are mark to market, because those things churn over very quickly. Also, there are things like derivatives, which have no cost, be it positive or negative. The UK GAAP perspective was that you did not value derivatives in the balance sheet, but that is wrong. The only way to value derivatives is to do it at market.

Chair: If there is a real market, yes.

Roger Marshall: But sometimes there may not be a market, because the institution is the only market maker. Even a model value is closer to reality than zero. There may be a large loss sitting there, so not putting anything in the accounts is not necessarily very helpful. I am afraid that it is partly the complexity of the modern world that means that one has to mark to market with certain assets and liabilities. Even if that means marking to model, that is unavoidable. I do not subscribe to the journey that the standard assessors were on, which was to fair value everything. As I say, they have now gone away from that. A loan book will continue to be valued at cost, less impairment provisions. It will not be valued at market. I know that the FSA also have their views on this.

Paul Sharma: Perhaps I can look at the same three themes: prudence, box-ticking and mark to market. For me, on the prudence question the most important thing is that the accounts need to be what they purport to be. The propaganda of prudence in the standards does not do anything for me if the actual numbers are not prudent. In fact, I would much prefer numbers that were not prudent, but were what they said they were, that a reader, including the regulator, could add prudence to, to something that was described as prudent, but had a variable and opaque level of prudence in it. If one gets to, let’s say, provisioning, provisions are, if you read the words of the standards-okay, you can fault them in terms of whether it should be expected loss, rather than incurred loss, but there are a lot of good words in the standard. What does it actually mean and, in particular, is it comparable between institutions and is it comparable between jurisdictions?

On mark to market, frankly there are quite a lot of things that banks and other financial institutions do where only the current value matters. There are quite a lot of things they do where historical cost with adequate provisioning is perfectly satisfactory, and indeed preferable. With the things where only the current value matters-derivatives is, of course, a very good example-we want the current value, and if there is a market, the best current value is to mark to market. If there is not a market, the least worst solution that anyone has come up with is to mark to a model that seeks to approximate the market. The last thing that I want as a regulator is a huge amount of risk in a bank balance sheet that is present but not visible because, when a swap was put on six months ago or six years ago, its cost was zero.

Q311 Chair: I understand the difficulties-they are very real. I am rather more sceptical of the notion that you can believe a bank when it says that this model shows that the value is whatever it is. That seems to me to be just as fanciful as saying the value is zero.

Paul Sharma: If you treat the question of belief as a binary question, and you think that the only reaction of the reader-to some extent, one would have a point on this, at least for some readers-is either wholly to believe the material in front of them or wholly to disregard it, you would have a different answer. But if you say that we want useful information from both the regulatory accounts and the financial statements about a bank’s financial position, with respect to derivatives and a number of other things, I want their best effort at the current value. Then-you refer to us as supervisors being able to come in and take action on a lesser standard of proof than, say, a criminal court or even potentially a civil court-our supervisory judgment can come in when we look at the information, but at least we have something to start with and something that it is relevant for us to consider, among other things. That is not the sole piece of evidence we take, but something that is indispensible for any sensible, modern look at a modern bank’s balance sheet.

Q312 Mark Garnier: I follow all that, and I completely accept your premise that you start off with something that is raw data and then you can address that, but why does Basel III talk about risk-weighted assets, because presumably this is taking exactly this valuational approach to the assets of the lending cycle of a bank? You are not doing that with the trading books.

Paul Sharma: There are rules for the trading book that do limit risk-weighted assets, but Basel III starts off from assuming that assets and liabilities have been valued. For the loan book, it is an easier-it was supposed to be an easier-proposition to know what the value was. We discovered, with the difficulties of knowing what actually is being represented by the provisions, that perhaps that was not as robust a point as first supposed. What would be hideously complex would be to have a set of regulatory rules that one could use-

Q313 Mark Garnier: Do banks not do internal risk-weighting of their trading books? The internal risk assessors will look at trading books dependent on whether the position was realisable-or some banks, certainly.

Paul Sharma: We are perhaps talking at cross-purposes. There is the question of how you value and there is a question of how you quantify the capital you need to hold for the possibility that things might turn out differently from the valuation. Everything that you have said is entirely correct with respect to the second question.

Stephen Haddrill: There is another aspect to this, which Paul very briefly alluded to, which is the inconsistency with which different banks pursued valuations in this period. Auditors were sometimes challenging people because they were at the upper end of the acceptable range-which in a sense shows that there is a principles-based approach-but the need to put some real pressure on the management to justify their valuations and answers is quite different from what an audit firm is seeing in relation to similar assets elsewhere. I think we felt at the time, and certainly subsequently, that not enough advantage was being taken by the big audit firms of identifying inconsistencies and demanding evidence-and treating sceptically that evidence for why inconsistencies arose.

Q314 Chair: At the present time, it is clear that risk weightings are not worth the paper they are written on, because there is evidence to show that different banks attribute wildly different risk weightings to virtually identical assets.

Paul Sharma: That is really a question for me. Risk weighting is not about the accounting balance sheet. There are problems of comparability on risk weightings between banks, to which you have just alluded. There are some severe problems in not using risk weighting-in effect treating all exposures as though they were the same risk. Neither of them alone gives us the necessary answer.

The position of the Bank of England’s Paul Tucker and Andrew Gracie, and myself at the prudential business unit of the FSA, is that we need something that does not rely at all on risk weights, leverage ratios and floors to the amount of capital on particular books of business, but that we need risk weights so that we can capture two different sorts of arbitrage that banks potentially engage in. On top of that, we need what Lord Lawson referred to earlier: the exercise of supervisory judgment on the basis of evidence recognising that we don’t necessarily need the same degree of proof that might occur in the criminal courts.

Q315 Chair: I think that we must draw to a close. I apologise that you were asked to come here at such an uncivilised hour; I hope that you have an excellent lunch tomorrow.

This question is particularly for Mr Marshall and Mr Haddrill, but if Mr Sharma wants to chip in, that is great. Are there any changes in IFRS which, in the light of a pretty awful experience, you would like to see?

Stephen Haddrill: I suppose the first and most important of those is that the changes that the IASB has been developing in IFRS 9 around valuations and the re-writing of that standard are adopted as soon as possible by the European Union. The EU has delayed that, pending the completion of all the work that the IASB thinks it needs to do. We think that what has been done so far should be adopted. That is the first step. Given the amount of time it takes to bring about change through the international system, implementing what has already been agreed would be a good step.

Roger Marshall: May I add to that? First, the IASB for the past 10 years has tried to harmonise with the US. It has now stopped that overtly, but it is still being urged by the G20 to come up with harmonised solutions. That is actually delaying the whole process. The reason why we are still here without a finished new loan impairment standard is because for years the IASB was trying to end up with an agreed solution with the US. Even though it is now trying to do one on its own, it is still being urged by the G20. They are trying to drive timeliness, quality and convergence all at the same time. I think that getting something out quickly that is not necessarily perfect but good is almost the most important thing.

Q316 Chair: I think you have identified a major problem where the best is the enemy of the good. That has got to be cracked somehow. I hope that you will focus on that. After all, the global agreement is not going to happen; the Americans are not going to change. They have made that absolutely clear, so the G20 can take a running jump, and I hope you will tell them to do that. We are then left with Europe, where it may be possible to reach agreement, but it may take a very long time. Because the United Kingdom is a more important banking and financial centre than the whole of the rest of Europe put together, it makes no sense for us to be held up if, after a reasonable time, we can’t get the agreement we would like to see. Do you agree with that?

Roger Marshall: I do, and we are frustrated that eminently sensible things are being delayed, first of all by this convergence process and then by the Commission not wanting to implement them. So there are things that could be done now which aren’t, because of these two problems.

Q317 Chair: Can you tell us what are the eminently sensible things that you would like to see done that you believe are in the public interest and in the interest of what your task is, but are being held up?

Roger Marshall: As Stephen says, I think the rapid conclusion of the various finance instrument work streams is important. We are talking about loan impairment, we are talking about a better method of hedge accounting and we are talking about fair valuing. We are still required in Europe to fair value our own debt, even though the IASB has already changed that standard. Although it has changed the standard, it has yet to be ratified by the EU, so in Europe we are not actually allowed to use the updated standard, and we are still showing profits on the fact that our own debt is getting less in value, perhaps because the business is deteriorating. We are still showing profits through the profit-loss account as a result of a standard which has since been overhauled.

I think we cannot on our own get the IASB to change its stance both on prudence and on stewardship, but certainly the FRC has made strenuous representations to the IASB to pay more attention to both those issues in future standards. Now, it is listening to us, but I think the more people who actually urge it to do that, the better.

Q318 Chair: Do you want to add anything to that, Mr Haddrill?

Stephen Haddrill: No, I think that summarises it.

Q319 Chair: Pursuing that a bit further, am I right in thinking that even if we wished to go it alone on these issues, we would not be permitted to do this legally?

Stephen Haddrill: We cannot.

Roger Marshall: That is European law.

Q320 Chair: That was what I thought. It has been suggested by Andy Haldane at the Bank of England that there may be a way round it-that it may be possible to achieve the same objective through regulatory means. Have you thought of that?

Stephen Haddrill: We have more discretion in terms of what is disclosed in the report of the directors, as opposed to the way in which the accounting is done. So if the regulator requires a different approach to be adopted, I think we would probably be able to arrange some way of that being disclosed in the report, but I don’t see any way round the basic accounting standards.

Roger Marshall: I agree.

Q321 Chair: I think it would be very helpful to the Commission if you could let us have a note on what changes and developments you would like to see that would, in your judgment, improve the system. This would be a sort of three-part list: first, which changes would you like to see; secondly, which are being held up because of the difficulty of reaching agreement; and, thirdly, which you think could be achieved after a fashion by this regulatory route? Could you do that, please?

Roger Marshall: We can.

Chair: Thank you very much.

Thank you very much indeed to all of you for coming to see us this afternoon.