To be published as HC 881-v











Evidence heard in Private and Public

Questions 322 – 413



This is a corrected transcript of evidence taken in private and public and reported to the House. The transcript has been placed on the internet on the authority of the Committee, and copies have been made available by the Vote Office for the use of Members and others.


The transcript is an approved formal record of these proceedings. It will be printed in due course.

Oral Evidence

Taken before the Parliamentary Commission on Banking Standards

Sub-Committee H-Panel on Tax, Auditing and Accounting


Members present:

Lord Lawson of Blaby (Chair)

Mark Garnier

Mr Pat McFadden

Lord McFall of Alcluith

Examination of Witnesses

Witnesses: Witness 1, former employee in Structured Capital Markets, Barclays, and Witness 2, former Managing Director, Head of Debt Structuring Group, Bank A, examined.


Q322 Chair: Good afternoon. Thank you for coming to see us. As you will see, there is nobody behind you-you have been told that this is a private session. Its purpose is to enable the Commission to be more informed than it otherwise would be. I want to start with just one straightforward question, unless there is anything that you first wish to add to your written submission.

I will start with one simple question and then I will hand over to Mark Garnier and Pat McFadden. This is all about tax avoidance, a very topical subject at present. May I ask a simple-minded question? How much of this was tax avoidance on behalf of clients or customers-it was a service that the banks that you worked for were providing for them-and how much was the bank itself avoiding tax?

Witness 2: I would say that they were both sizeable. From my perspective, I focused through most of my career on bank to bank trades. In other words, it was intra-UK, or, more commonly, UK-US or UK-European. What you often saw were very large reciprocal deals between banks. There would literally be deals-I am talking about the last decade, in the 2000s, not in the ’90s - where a UK bank would set aside an allocation of tax capacity and do a £5 billion trade on its side, sheltering its own tax in the UK when dealing with, say, a US bank counterparty. It was not contractually linked- but within a short period of time either before or after- a reciprocal trade would be done where the US bank would shelter its own US tax. In effect, they would share the profits between the two trades.

Witness 1 might have a better idea on this; he spent part of his time at BarCap. There were organisations that had a very large corporate reach, but most of the players-most of the banks in the last eight to ten years-were primarily playing in the bank to bank space because it was the easiest place to play.

Q323 Chair: So this was largely a way in which the bank itself substantially reduced its tax bills?

Witness 2: Two things. It would reduce its own tax bill, providing cheap funding to the bank on the other side, and it would get cheap funding by helping another bank in another jurisdiction to reduce its tax bill there.

Q324 Chair: Right. So it was a kind of quid pro quo?

Witness 2: That was a kind of quid pro quo. The real magic was when you managed to get the accounting treatment of both as generating pre-tax profits for the published accounts.

Q325 Chair: So very little of it was an existing client of Barclays-or whoever it happened to be-coming to you and saying, "Can you help me be more tax efficient?"

Witness 1: No. I didn’t see a single example of that.

Chair: Thank you. That clears the air.

Q326 Mark Garnier: I want to talk about the history of how we got to this situation. As I understand it, a structured product was something that came out of the structured capital markets department of the bank some years ago-perhaps one or two decades ago. It would be a pretty straightforward product amalgamating two or three things together in order to provide a customer with an opportunity, and the bank that was writing the product would make a bit of money out of it. That was the purpose. It was clearly just matching the objectives of the customer with a profitable opportunity to the bank in terms of putting something together and then just selling it. Is that a fair description?

Witness 2: When I started at Bank A in the early 1990s, it was tax-efficient lending. That was the euphemism. Why? Because the bank was trying to be important to its corporate client base, or indeed to universities or other such institutional customers. It was finding that its cake was being eaten by investment banks coming in with funky things like interest rate swaps and other such exciting new developments. They wanted to get the business, so they thought about what they’d got that the investment banks didn’t have that they could provide in return for their corporate relationship managers getting invited to nice lunches and getting the opportunity to bid on things they would never really get a lead mandate on, but at least they would get their name on the deal sheet.

They thought they had tax capacity; they had lots of taxable profits from the regular business, so was there anything they could do whereby they could lend the money to the corporate in a way that would be cheap from the corporate’s view, but so that they-the bank-effectively did not pay tax on the profit from that lending? Indeed, it is better than that because they didn’t pay tax on the income from that and they get a deduction for the borrowing cost. If interest rates are 1% above, they could give a loan at LIBOR flat, yet still, in economic terms, make the same returns as if the money had been lent at 100 base points over. It was tax-efficient lending; it was not a financial scientist sitting in a corner, going through tax legislation and coming up with funky derivatives and then packaging it into a product to sell-

Q327 Mark Garnier: Okay. How is that different from normal commercial lending?

Witness 2: Because normal commercial lending is where you borrow money and lend money, and the bit in between-the profit, or the loss if you get it wrong-is taxable or tax-deductible. If, however, you can borrow money at, say, 10%-which was the swap rate when I did it in the early 90’s-and corporation tax is at 30%- that means that the money you have borrowed at 10% has effectively only cost you 7% post-tax. You lend it out at 10%- on the face of it you would never do that in a normal world, because if you borrow at 10% and lend at 10% there is no profit. But, if the 10% you get back in is tax-free, you have made 3% post-tax profit. That’s use of core tax capacity. Every tax structured deal is essentially a variation on this basic theme-

Q328 Mark Garnier: But that is based on the premise that you are making a tax advantage, so it is all about tax advantage.

Witness 2: Yes. When it started off, it was all about the retail banks or wholesale banks- instead of being the victims whom the investment banks came to and stuffed with occasional deals, where the retail/wholesale banks were not really getting paid for it-started to do it themselves. Obviously, because of the massive amounts of taxable profits that they had in scope, they had the capacity to do it.

Q329 Chair: Roughly when did that change happen?

Witness 1: I joined really in the heyday ***, when the market was for giant inter-bank trades. They had realised that you could make an awful lot more money because, to a bank, these things are almost infinitely scalable and the exposure you could take on a counterparty is huge, while a corporate had a very limited size that they could do the deal in. That’s why this evolved into an inter-bank market, with some corporates that were of the size that could compete in that kind of market. Deals were in the several billions.

Q330 Mark Garnier: So you are essentially saying that if you have a number of banks with trillion pound or trillion dollar balance sheets, the market becomes limited so you then have to start dealing inter-bank in order to get-

Witness 1: In order to get the size or to generate the very large sums of cash. In earlier days, when it was more about relationships and having a product to offer, I don’t think the money was quite as important.

Witness 2: No.

Witness 1: But as it evolved into the inter-bank market, where you could make £50 million a year on a single deal that might run for several years, that is where the real money came in.

Q331 Mark Garnier: What sort of sums are we talking about?

Witness 1: A couple of billion.

Q332 Mark Garnier: A couple of billion at a time?

Witness 1: Upwards; there were some very large deals done.

Witness 2: For example, and there would have been similar sized deals at Barclays, there was a £5 billion trade between RBS and Morgan Stanley, executed in 2006, using US tax capacity of Morgan Stanley and providing a spread profit to RBS.

Q333 Mark Garnier: So that was prior to the crisis.

Witness 2: And the clear understanding was for RBS to do an equivalent deal the other way deal the other way round with Morgan Stanley, using UK tax capacity of RBS, though I believe that never happened because of the disruption in the banking and Asset-Backed Securities world that began in mid-2007.

Q334 Mark Garnier: Presumably, I guess that the ultimate limit would be the tax liability of the bank? So, if the bank has a £20 billion tax, that would be the limit?

Witness 1: In theory, yes. But they would have a certain amount of capacity they were allocated to use, in my experience.

Witness 2: I’d put a little more gloss on that. If a financial institution did not have UK tax capacity, how could it take part in this business? The answer is by buying in tax capacity. In other words, it could structure trades. HBOS, for example, did trades in 2003 where they sought to transfer lumps of highly taxable income to an institution that hoped to shelter it. Indeed there was a publicly reported tax case on this in 2009: HMRC v HBOS Treasury Services. You can sell lease tails or in-the-money swaps that become pregnant with the tax liability. The bank that sells it to you can say, "Well look, I’m selling you a £200 million cash flow, on which you expect to pay £60 million in tax, so I should only ask you for £140 million. I’ll tell you what, give me £160 million and you can have it." So you buy the cash flow for £160 million and if you then do something so that the tax liability does not crystallise, you have made £40 million.

While the selling of lease tails happened a lot in the banking world, it was less common to buy in tax capacity from corporates. But I am certainly aware that Barclays did that. Iain Abrahams had a great reach across the corporate taxpayer base and Barclays bought deals from there.

Q335 Mark Garnier: I have a lot of questions coming out of that. One in particular is, when you look at these outstanding interest rate swaps, and I know we are talking about something different-

Witness 2: You mean the LIBOR arguments?

Mark Garnier: No, sorry-

Witness 2: The swaps we brought in?

Mark Garnier: No. You know there is a scandal emerging with the interest rate swaps that have been sold to some of the SMEs. Admittedly this is at the much smaller end of the scale, but what is interesting is that there is a net present value left on the outstanding contracts with you, because the interest rates have now dropped to nothing and they are paid at higher rates of interest. I appreciate this is very small-we are talking about contracts worth £1 million, £2 million or £5 million-none the less, this is exactly what you are talking about. This is where there is a net present value that can be traded, and the margin is essentially the tax liability, which you can turn into something.

Witness 2: Effectively, what I was talking about is, HBOS had borrowed money over the long term, put on a swap, and interest rates were high. They were receiving fixed rate flows-actually, from AIG Financial Products. The swap hedged issued debt. HBOS were receiving say 10% interest over the next seven or eight years and interest rates had fallen to say 3% or 4%. Because of the way our accounting code and our tax code works, that swap was carried on HBOS’s balance sheet and tax books at zero carrying value- because it was on market on the day they did it and you don’t mark to market the thing- you just hedge account the issued debt liability. But if they closed out the swap and crystallised the in-the-money cash value, it was immediately taxable.

In your example-the LIBOR swaps and the SMEs-you couldn’t have done that because those aren’t hedge accounting items.1 But I take your point. The irony of the mis-sold LIBOR swaps- of course people complained, but they wouldn’t have done if interest rates had gone up. It’s one of those unfortunate things when dealing with consumers.

Q336 Mark Garnier: Yes. My next question is, what does this do for financial stability? You are now shifting all this exposure around the marketplace based on a tax consideration, rather than an exposure consideration. How does that work?

Witness 1: For the deals I was involved with, the very big principal trades are not about the underlying exposure. You are talking about things where the credit considerations were very tight. We are not just shifting junk because it has a tax asset attached to it. On the very big deals, the counterparties were very strongly hedged. If they could get away with it, the underlying exposure would be to one of the banks, and then the underlying credit exposure to that asset package would be to the same bank. So the deals were not about taking exposure to an asset. Many of those deals were designed to be pulled apart at a single thread. They would fall apart effortlessly. The swaps would cancel, there would be no payments and it would just split straight down the middle.

Q337 Mark Garnier: Would you deem CDOs and CLOs to be part of the world you are talking about?

Witness 2: No.

Q338 Mark Garnier: So it is something completely different?

Witness 1: Theoretically, they could form part of an underlying package that is nothing to do with the deal.

Witness 2: The other thing to think about in terms of these kinds of deals is the idea that the complexity of the tax-based business maybe brought in assets that people lost lots of money on or did not understand. Actually, in terms of tax-structured finance and the discipline of doing it, you would sit around and come up with a tax arbitrage. Economically, you would want a deal that went from A to B, but you could not go from A to B direct, because the tax arbitrage would not work, so you had to go via C, with a back flip to D and a twist round to E. The skill was, as much as anything, not on the tax side, but on how to make the commercial outcome or the risk outcome-what happens on insolvency, what happens on market change, what happens whenever-have exactly the same risks as, or preferably fewer risks than, if you went from A to B. That was where the invention was and everything that witness 1 talked about. In this field, you become so detail-focused to make the tax work, and the discipline of the people who worked in it tended to mean you did not ignore a potential market risk or a potential credit risk. This was not about taking real market risk; in fact, it was about taking as little such risk as possible, because then you needed less capital and returns on equity looked infinite.

Q339 Mark Garnier: Sure. You are describing a UK tax issue, but how does this work between tax jurisdictions? Was that an opportunity as well?

Witness 1: Almost all my deals were cross-border. There are a number which were UK-only, which witness 2 was talking about. Almost all of them were cross-border, because you are always exploiting an arbitrage between the treatment of the instrument in one country and another.

Witness 2: I would add a caveat. That is definitely the case with the very large intermediated funding trades. To provide perspective, let us imagine you do a £5 billion trade, and interest rates are 5%. With corporate tax rates at 30%, you are really talking about £75 million of tax saving a year-£5 billion times 5% times 30%. To get a £75 million tax saving in a single year, if you could find what euphemistically became known as pop trades, where you could get a one-off deduction-

Q340 Chair: What does pop trade mean?

Witness 2: It means for example turning what is, effectively, a capital payment into a tax-deductible flow. In other words, it is, ideally say, borrowing £250 million, not losing £250 million, yet somehow managing to write it off for tax, so you get a one-off deduction. That is what I mean: it is turning flows of capital, not flows of interest, into deductions.

Witness 1: "Pop" is a term just meaning an instantaneous trade-it just happens on the spot. That is differentiated from an accruals trade, where the benefit arises over time.

Witness 2: You will have seen that there have been cases. There was the First Nationwide case in 2012, which they actually won at all levels including the Court of Appeal, which was based on a structure that was a "pop" trade. Indeed, I do not know if any of you have heard about the TOMS cases-basically FX forward trades that were designed by Ernst and Young back in 2002. The analysis as to whether it worked at the time was, I recall somewhat intricate – and indeed it turned out not to- the ‘Prudential case’. There was a $32bn TOMS FX trade almost done- the Revenue heard about it just in time to issue a press release closing it down, because one corporate that had decided not to do it was aware that another major oil company intended to do it – and warned HMRC. There would, if it had worked, have been a deduction against taxable income of $6.4 billion the very next day if the trade had gone ahead. There were some unbelievable pop trades- but few people had the appetite to do them- and probably, in the overall scheme of things, the long-term drip, drip, drip of the multi-billion pound tax-structured "lending/investment" portfolios that the banks had between themselves took more tax out of the system.

Q341 Mark Garnier: So is it fair to say that as long as you have differential tax rates around the world, there will always be an opportunity for these products?

Witness 1: As long as you have different regimes which treat the same instrument differently in different jurisdictions, then, yes.

Q342 Mark Garnier: I also wanted to ask about the regulatory element, because banks clearly have to look at regulatory capital and that kind of stuff. What element of deemed regulatory capital was part of this?

Witness 1: It was part of the deal. The genesis of a trade-I am talking about very large inter-bank trades, in particular-is that you come up with a tax idea, and the tax idea is the first thing you have. That tax idea is then developed into a working tax idea with legal opinions, and the commercials are actually added around it, and then the client comes last: we ask "What is the fact pattern we could use to make this tax idea work?" and then find somebody who will do it.

Q343 Mark Garnier: Be it a customer or another bank, but latterly another bank?

Witness 1: In my case, it would largely be another bank; in others, they may have found some large customers. At the stage of finding out whether that deal would work or not and adding the commercials, the accounting and the regulatory impact were both examined in great detail, so you would attempt to mitigate regulatory capital.

That was a layer on an original tax idea. You might be structuring to avoid it, and the structure might be as simple as using zone A triple A ‘govvies’-lending to supposedly secure western Governments-under a package that wipes out the regulatory capital charge, simply because the assets were so secure that you could have them in your balance sheet, in effect, regulatory capital-free.

In terms of structuring aggressively to change your regulatory capital, I did not see much of that. The banks were a lot warier of structuring around regulatory issues than they are about tax.

Witness 2: In fact, I can give a few glosses on that. The very nature of why we did bank-to-bank trades originally was, "Wait a minute. If I do a trade with a corporate, I have a 100% regulatory-weighted asset. If I do a trade with a bank, I have 20%, so I need a fifth of the capital for doing the same trade-great!" That is the very basic thing.

Another point on the same subject is that one of the beauties, in a way, of income that comes from tax-rich business was that the Basel regulatory system does not have a section that says, "For tax risk-taking, give a weighting of 32%" or whatever. That is not in there, so that for CFOs-I am not sure how many of them even realised that they had this-this income, for the amount of income that was there, did not really seem to come with any capital cost.

My final gloss point is that you could argue-it is a devil’s advocate argument-that tax-structured finance actually saved Her Majesty’s Treasury putting £2 billion or so into RBS and Lloyds/HBOS in 2008, and saved Barclays having to raise additional billions from Oman or Qatar or wherever it was- because even after compensation and the dividend pay-out effect- it had over a decade created capital-apparent capital-in their balance sheets. If the Revenue can now collect a lot of that back-probably, with a bit of will, they can-that has actually been quite convenient: it represented capital across the crisis that the Government can now collect back.

Chair: I would like to come on to that.

Q344 Mark Garnier: I am conscious that I am taking up quite a lot of time, but I have one last question. You are actually having a trade with another counterparty, so these are proprietary trades on the bank’s balance sheet. It is absolutely a proprietary trade, so-which, given the fact that it is bank-to-bank, had been done for tax repayments-there is no benefit ultimately to any customer of the bank; the benefit is to the shareholders of the bank. It is a very selfish-I am using hyperbole, if you like-and there is no societal benefit.

Witness 2: Yes.

Witness 1: Agreed. Certainly, in cases of the large inter-bank trades, which was the bulk of the market by volume and by value, that is true. They will also argue that the money was cheap funding, so it was commercially viable. If you look at where the money went, it often went into the underlying asset package-for example, Government loans-so the bank was using it to buy the assets to sit under the trade, which were then regulatory capital-free, which is what made the thing very cheap to run.

Witness 2: But that argument goes out of the window when you realise that the trade-off was generally reciprocal. In other words, you might say, "On this trade, I’ve managed to get cheap borrowing, but on the other trade, I’ve had to give cheap funding."

Q345 Mark Garnier: We are wrestling with this question: what do you look at in terms of the Volcker rule and of whether or not it is proprietary trading? Of course, people start off by looking at the position on your equity book, for example, as to whether it is proprietary trading or market making. What I am trying to ascertain is whether, if you were looking at any sort of Volcker rule in terms of what proprietary trading is, this would be classified in your opinion as proprietary trading that is different from normal bank funding in terms of going out of the inter-bank market?

Witness 2: It is a difficult one. Unless you ban a bank from dealing with another bank-lending or borrowing with another bank-then I do not really see it happening.

Mark Garnier: But you are not borrowing and lending.

Witness 2: You are, but you are just doing it in a very structured way. Tax is the mechanic by which you do a trade. A CDO you can describe. Clearly, people know what these things are. You know an elephant- you know it when you see it - but try to define it. Arguably, the one thing you might want to think about is where you have the ring fence. If you are going to have a system where you do not have complete segregation of investment banking from retail and other banking, and instead allow a holding company with siblings- then you might want to stop the exchange of group relief between them. The real driver from the ’90s of this business were the retail/corporate banks with a large regular stable tax base. The investment banks do not have that- but- the talent, if that is the right word, on the "dark side" will inevitably sit in the investment bank. If they can get easy access to the tax capacity of the retail bank and can group relieve among themselves across the ring-fence- kill that flexibility and you will kill much of the capacity to do tax structured deals. AIG was a massive player in the tax-base business not because AIG Financial Products itself had enormous amounts of taxable profits- but because Hank Greenberg let them have all the tax capacity of AIG Inc. So they got to play with the toy box of the parent.

Witness 1: What you see with the very large players in the market is that they all had access to a massive pool of capacity. The reason for that was largely the reciprocal trades that were required, so if you sold somebody a trade, you were expected to buy one in return and use your own capacity. If you did not have that capacity, it made life very difficult. You ended up either having to sell a trade at rock bottom rates to somebody who just wanted to do it, or you end up with witness 2’s business, which was, I think, matched book playing where you then have to find the losses to offset your tax-free income.

Q346 Mark Garnier: I have one further last question. By ring-fencing the banks, you are essentially creating different entities, are you therefore potentially creating an opportunity to do the structural trades between the ring-fenced bank and the retail bank?

Witness 2: Not by comparison with not splitting them up. If they are all contained within one corporate tax group, we are really talking about breaking the ability to group relieve. HMRC could come up with clever ways of avoiding that if it wanted to. I can make some suggestions. You could still have your parent and sibling structure, but just create something that causes automatic degrouping of the two. When they are all together-when Barclays as a bank has the BarCap side and the retail side, but they are all in one place-

Mark Garnier: So, there is no opportunity-

Witness 2: If you apply the ring-fence, so they are separately led, in theory they can do tax trades between themselves but, it will get quite difficult because they will not be at arm’s length for tax purposes; it will not be that easy, but perhaps you are right, that they could do it. But in any event, I would still stop them group relieving between themselves because otherwise the IB will do a deal with some new facilitator generating for fees to the facilitator lots of tax-deductible losses, and then surrender them to the retail banking side who will pay them a larger cheque.

Q347 Mr McFadden: I just want to see whether I have understood what you have been describing. Are you saying that effectively what banks were-are-doing here is selling tax liabilities to one another, not on behalf of any client, but on behalf of the bank itself, and for no reason other than to reduce their tax liabilities? There is no other logic or reason to do this trade; it is done purely because it has a tax benefit. Is that the case?

Witness 1: There will be a spectrum of that. In terms of the very large inter-bank market, that is essentially exactly what is happening. They are selling tax avoidance between each other. There was no other benefit to it that I could see while I was there. There will also be elements where they are actually bringing in customer or client trades. Clients are coming to them with a problem and they are trying to solve it. That will also happen, but it muddies the water, because it is much smaller and it is not nearly as lucrative.

Mr McFadden: That’s the small end of this. The big part is what we first-

Witness 2: The big part, I think-yes, if you are looking for the generic, what you have just said is absolutely true.

Q348 Mr McFadden: Well, you use your own words to sum it up; you don’t have to use mine.

Witness 2: Your description of it is essentially that the banks were playing an insular game between themselves that had no real client-driven business and was about swapping tax liabilities. Yes, it was either sheltering your own tax liability and paying someone else to help you to do it, or helping someone else, helping another bank, to shelter its tax liabilities and getting paid for doing it. That was what the business was.

Are there exceptions to that? Yes. Let’s take Goldman Sachs, for example, who often get bad press. Sure, they did some-it’s not that they didn’t do inter-bank trades, but actually they often had the tax structuring people embedded in what you would call the private deal world. Essentially, because Goldman Sachs had such a good business of advising corporates and people on real trades-acquisition or disposal or growth or whatever-their people would look at it and say, "Hey, while we’re doing this, we can do this tax arbitrage," and they would often do it for their own benefit. But that is more difficult. Those are things where, if you were Goldman Sachs, you would say, "Well, I was providing a client service here." It’s relatively uncommon.

Mr McFadden: That’s what I am trying to understand.

Witness 1: Both happened, is the answer.

Q349 Mr McFadden: So when we hear of-I don’t know what these things were called in banks-structured capital markets, is that what they were called?

Witness 1: Yes.

Witness 2: Yes.

Q350 Mr McFadden: On this piece of the matter, at least, we should understand that not to be tax advice to clients in order to make their activities as tax-efficient, if you want to use that term, as possible, or if it is that, it’s a minor part of what that unit does. In the main, these units exist for the sole purpose of minimising the tax liability of the bank itself. Is that the right way to understand it?

Witness 1: Close. They exist for the sole purpose of making money, and the easiest way of doing that is by minimising tax liabilities of the bank or enabling another bank to minimise its tax liabilities and taking a cut.

Witness 2: If you can get paid a lot of money for facilitating a tax shelter for a corporate, people do that. Actually, it was generally quite difficult to get paid a lot of money. You had to spend months working with corporate treasurers-that’s a horrible existence-to get them done, but there was some business done on that side. But the big distinction is that it was very rarely an example of a corporate treasurer coming to a bank and saying, "You know what, Ian," or whoever, "I’ve got this problem and I just don’t know what to do about it. Is there anything you can think of to help me?" It wasn’t that. It was people designing a product and going out with it.

Q351 Mr McFadden: Okay. Let me ask you a couple of other things about this-again, just for my own understanding. You have described what the activity was and what the intention was. I want to ask you about the effect. You touched on this in one of your final answers to Mark Garnier, witness 2. Do you think this had any overall malign effect on how banks were able to calculate risk, exposure and all of that in terms of playing a role in the crisis? Or, as you were saying a moment ago, do you think actually this did not contribute to the crisis and may have had an effect the other way by retaining more capital? We can look at this and say, "That’s not a very nice thing to do, because you’re not really paying all the tax you should be paying," or we can go a step further and say, "Not only is it that, but it actually contributed to undermining the stability of the financial system." Which do you think it is?

Witness 1: I can’t see how it might undermine the stability. The only caveat to that is the fact that you end up with a web of very large lending deals between banks. You end up with a very interconnected web, so when the banks try to deleverage and get everything off their balance sheet, there is more to unwind. That is the only possible structural instability I can think of.

Witness 2: As ever, I am always happy to think of the devil’s advocate argument. The glosses on that include, as witness 1 says, unwinding the thing. Clearly, when people were doing very large trades, you needed raw assets to sit there. Therefore, it tended to be asset-backed commercial paper or short-term ABS. Did the banks end up, because they had these large tax-structured trades, by accident or design, holding more of those kinds of assets than they would otherwise have done? I don’t think so. Actually, there was generally the right of substitution- these deals were just where they parked some of their existing inventory, which was stuff they would have held anyway, but they would just stick it there. That is one argument. The other devil’s advocate argument about how it could have had some connection with the crisis. I am convinced this is true-

Mr McFadden: I hope it is all true.

Witness 2: Fair point. This is subjective; there isn’t hard evidence for it, but I believe it to be true. Banks in the period from 2000 onwards were extremely competitive places internally. It was all about ego and positioning between business heads. Tax-structured finance seemed to create money out of magic. It was magic. There was minimal capital required in it. It made a lot of money. Did that encourage other people in other businesses to say, "How do we make this amount of money?" Did it encourage the structured credit derivatives desk, for example, to say, "Those so-and-sos in SCM seem to have made £500 million this year. If I am gong to remain an important player in this bank and get paid a large bonus, I had better make at least £250m next year. How am I going to do that?"

Mr McFadden: That is quite an important effect.

Witness 2: I think it’s real.

Chair: This is the cultural dimension. That is very important indeed.

Q352 Mr McFadden: I want to ask witness 1 a question, because he has a background with BarCap I think. We are told that in Barclays’ new search for a better reputation, if you want to call it that, under the new management they are going through all the lines of business, asking whether they should keep doing this or not. We are told that this tax structuring unit is to be significantly shrunk or maybe even abolished, I don’t know. Give us your reflection on that. Do you think that this business is reputationally risky for banks because of the driving force behind it that you have outlined?

Witness 1: There is no longer a floor where you can walk through a heavily locked door and find 100 tax structurers beavering away. That has gone. But I don’t know where those people have gone. The tendency you see is that they go and sit at the desks elsewhere. They have fragmented and may have embedded themselves elsewhere in the business.

The tendency of the market also moves on. These giant inter-bank principal trades have disappeared largely, partly because of the availability of capital. That is the main limiting factor-the sheer availability of several billion pounds of funding to take exposure to another bank. The credit crunch seemed to destroy that completely. The asset class has changed. The business morphed into more of an equities-type business; where you buy in equities over dividend dates and it all happens in the flow books of the bank and it is very hard to spot.

You see a move away from the giant structured positions into much more fragmented stuff that is actually happening down on the trading desks. Are the people who were doing that original business still there? I think in large cases, yes they are.

Witness 2: ***

There are a remarkable number of people - I knew in the tax structured business who found themselves called upon, instead, to help deal with the wide range of non-tax complex deals that had gone bad in the crisis. Because they had become so used to complex deals that they were the ideal people to deal with other complex deals that had gone wrong. There has been a morphing of quite a lot of senior talent into that. As for the continuation of tax-structured finance, as I set out in my note in November, there was such an alignment of the stars in the last decade that helped tax-structured finance to take off- and so many of those have gone now. There were something like 15 key attributes I identified and most have gone. I do not see them coming back in a hurry. Even if some people went off and did equity trades, most or all of them tend to be arbitrage on withholding tax. The UK does not have that. Hence it is mainly arbitrage on other people’s tax systems. I am not saying that this is a good thing,2 I am just saying it is not in the UK.

Witness 1: That’s an important point. What you are seeing and what is left is exporting the avoidance. They may be sitting in London, but they are exploiting other countries’ tax regimes. From the UK’s point of view, you might see that things have gone fairly quiet. Whether or not that will be the case, the key is that while these people are very creative, and the good tax structurers are still there, the business always has the capability to come back.

Q353 Lord McFall of Alcluith: I’ll focus on your cultural standards if that is okay with Lord Lawson. Are you saying that with, say, Barclays, people in London may be doing it for elsewhere now instead of doing it for London, but the activity is still going on?

Witness 1: I don’t know what Barclays is doing. I do not have any connection with the bank any more.

Q354 Lord McFall of Alcluith: Witness 2, you made the point in your written submission to us that the structured capital market’s approach will continue to drive a culture that sees rule books as raw materials for arbitrage. Martin Taylor, in a statement he made to a commission that I was on a couple of years ago said, "I don‘t believe that regulators can outwit necessarily determined traders. The traffic wardens don’t break up the drug cartels." So will the drug cartels keep going, and are the traffic wardens innocent cops? We are interested in the relationship between the regulator and the industry. At the moment, it seems that the regulator is chasing but largely getting nowhere.

Witness 2: On the tax side, I would say that it was never really a subject of regulation. The Revenue would look at things, but that is not really regulation; it is ex post facto. It was not really for the regulators, because it didn’t touch on things that they measured. It didn’t touch on capital.

Q355 Lord McFall of Alcluith: Yes, but the regulators have said to us-Martin Wheatley repeated this on the record-that they are going to look at the business model, and P and L. They weren’t looking at P and L, but if they do and see that in the past Barclays generated as much as three quarters of profits at their investment bank operation-

Witness 2: BarCap.

Lord McFall of Alcluith: Yes, BarCap. It will have an effect and the regulator should come into the equation.

Witness 2: I entirely agree. They should ask about the quality of the P and L. Clearly when a bank is reporting a capital base following the accounting standards for tax matters, you recognise the tax benefit unless essentially there is a 50% or greater chance it will not come off. Everyone who did tax deals in the financial institution space had "should level" opinions. As I say somewhere in my note, "should" is a wonderful word because it can sound and be portrayed as 90% certainty to one person but be thought of as 65% certainty to another.

Banks often did not have reserves set aside against their potential tax arbitrages. Sure, regulators should look at that. I have said that many people who did this business have moved off into doing other things. What have they done? Well, some of them ended up sorting out the real deal problem cases in the banks. But many of them sought to ensure the most capital efficiency dealing with the weighting of the assets- to transmogrify what they already had into something that had a lower capital weighting- thus theoretically creating capital.3

Chair: The gaming of the risk weighting is a well-known game.

Witness 2: Efficient capital management, let’s say. For a lot of the time, no one cared about it.4 Another area some have focussed on "efficiency" for the bank levy.

Q356 Lord McFall of Alcluith: We are charged with looking at standards and culture. Is it a bit pie in the sky for us to talk about changing culture and environment where they are perpetually gaming the system? In other words, can we do anything useful in that area?

Witness 2: I don’t know what the end result will be. Are you going to have code changes? I may be naive, but I think that a lot of what you have seen is historic, and I do not think the appetite for gaming will be as great going forward. Not least because people such as non-executive directors or senior executive management who used to say, "This is all far too complicated for me. But I don’t need to know about it", do not want to take that risk any longer. They do not want to take the risk of not understanding but being accountable for what is going on-and the issues arising in tax structured finance are often too complex for them to understand in the time they have available therefore they won’t do it.

Q357 Lord McFall of Alcluith: Yes, but if we come out with a code of practice, or suggest a code of practice, is that pretty lightweight stuff? Would a code of practice achieve a change?

Witness 1: It’s very hard to say, but heads have rolled. Some impregnable businesses within banks virtually toppled as a result of what we have seen, and that is just daylight shining in, the press getting involved and continuous exposure. What has happened in the past are major changes in the SCM businesses, so I do not see why it would not continue to have-

Q358 Chair: We have to close our evidence session quite soon, and I would like to ask a few questions arising out of the very interesting evidence that you have given to us. First, as you said, because of changing circumstances-you have indicated what they are, and we know what they are-this is not happening to any considerable extent at present, but there is no reason why, if circumstances change again, that it might not recur on a large scale, is there?

Witness 1: While the people are still in the banks, you know that they are making money, because otherwise the banks would get rid of them, so you can pretty much surmise that the banks are still making money out of tax arbitrage.

Q359 Chair: Right. Going on to the structural side, you pointed out that on the whole-to take Barclays, which has come up, but it no doubt applied in other cases-it was basically the Barclays retail bank parachute that was being used, but it was the clever people at Barclays Capital, the investment bank, who were doing it all.

Witness 1: Yes.

Q360 Chair: If there is ring-fencing, it may well be-we have not seen the detailed legislation yet-that that kind of relationship between the two parts is forbidden. But then there would be-or may well be-nothing to stop the retail bank having that unit within it. It could do it itself.

Witness 1: In theory, but the activities of a retail bank presumably would be curtailed by the ring fence, so I do not know to what extent SCM-type activities would still be possible.

Chair: That is one of the things that we will have to see, and that we will be able to make a recommendation about.

Witness 1: Yes.

Q361 Chair: Sticking with Barclays, you mentioned, witness 1, that by the mid-2000s, the structured capital markets in Barclays contributed 110% of the profits of Barclays Capital.

Witness 2: Actually I said that.

Chair: Sorry, you said that. So in other words, the rest of Barclays Capital was making a loss.

Witness 2: It was subsidising the building of a fixed-income franchise off the back of the tax profits.

Q362 Chair: Were the auditors aware of that?

Witness 2: They must have been.

Q363 Chair: What did they do about it? Did they inform them?

Witness 1: ***. The SCM teams could often tell the auditors what they could and could not look at. They seemed to have a lot of power in limiting what the auditors were allowed to see. The auditors could pick a trade or two trades and look at that, but they would not be given carte blanche access to everything.

Q364 Chair: Clearly, the auditors did not know what they were doing. I think that is borne out not only by what you have said but by an evidence submission that we have had from KPMG, complaining that it was all too complicated for them to understand.

Witness 1: Not if they had the right papers. Internally, there were usually papers that fully explained all the deal. It is whether the auditors would get hold of those papers.

Witness 2: I am slightly lost by the question that there was something for the auditors because of the fact in one year BarCap made 110% of its net profits out of the SCM business. I am assuming that the auditors satisfied themselves with the tax treatments/ provisioning of the deals that were done in some way, however they did it. The fact that a single business area makes 110% of the profits of BarCap-I am not sure that there is anything for the auditors react to in that. It is just a fact. 5

Q365 Chair: You don’t need to comment on it, but why don’t I read out what KPMG have written to us? They say that all the structured capital markets business "created complexity and made it harder clearly to identify and to monitor the risks inherent in the original lending." They were complaining that it was too complicated. What are your comments?

Witness 1: I would have understood it if I had looked at it, so I would expect them to.

Q366 Chair: So it was incompetence on their part, if they were telling the truth-

Witness 2: They could not charge for it, that is the basic problem. We had this classic example all the time: the external auditors would come in and audit it, and what they would want to do was to get their tax specialist to come and have a look at the thing, which would always cost a lot more money, and they did not have the budget within the audit fees, to do that.

Witness 1: They would win the audit by pitching a low fee and then, if they had to go through every last one of these tax trades, it would be vastly expensive for the audit firms.

Q367 Chair: I see. So it was too much work for the charges they had arrived at.

Witness 2: Yes.

Witness 1: Yes. They were writing their charges off against time.

Q368 Lord McFall of Alcluith: Was it the case, then, that the type of people on the audit were junior people, who could not understand it?

Witness 1: ***

Lord McFall of Alcluith: I think that is a feature.

Witness 1: ***

Q369 Chair: Finally, may I ask you a question from a totally different angle? If either of you were the Chancellor of the Exchequer, struggling to deal with an obstinately large budget deficit, and you felt that there might be some help in this area, what would you do?

Witness 2: I would certainly make sure that HMRC tried to collect back the taxes.

Chair: The back taxes.

Witness 2: Yes. In other words, there are a lot of the trades that have not been agreed-they are outstanding-so there are probably several hundreds of millions, perhaps not billions, of potential tax lying around in the banks that could be captured back over time. I would do what I have said on the ring fence, so it cannot easily happen again. I would try to mitigate potential future losses. Apart from that, I guess there is a limit to what the Chancellor can do-unless you nationalise banks, but I am not suggesting that the Chancellor does that.

Q370 Chair: You said that there are matters that have not been agreed. Did you say to the Revenue, "This is what we are doing, is that okay?", or did you hide it from them?

Witness 2: No, and definitely not the latter.

Q371 Chair: How did it operate? Are they up to the task?

Witness 2: They are now, I would say. Relatively. They will always be one step behind because they come and get a raw set of documents, but they do not have the context, so they have to work out what the trade is without having the context of having been there during its creation. The one thing that I think is fairly true-almost uniformly-is that there was no appetite, in any business that I have ever worked for, for hiding the facts. You would do the trade, you would assume that it would get crawled all over by the Revenue-they would look for any factual errors, because if they could find those, that was the easiest way and they did not have to bother with the legislative challenge-then you would submit your computations a year later, you would wait, the Revenue would come through with a few queries and you would spend the next three or four years talking about the trades. Or, rather, you wouldn’t-that was the other great thing about tax-structured business, the people who actually did it at the front end did not have to deal with the Revenue afterwards; that was the poor people in the tax departments who had to do all the hard work. But I do not think that people in the banks hid trades from the Revenue.

Chair: Thank you very much indeed. We could have gone on for some time. What you have done is to explain a very complex business with great clarity, and I would like to thank both of you very much indeed for coming to see us this afternoon.

Witness 2: Thank you.

Witness 1: Thank you.

Examination of Witness

Witness: Sir David Tweedie, President, ICAS, examined.


Q372 Chair: Thank you very much, Sir David, for coming to see us this afternoon. I think you know what we are interested in. Is there anything that you would like to say by way of an opening statement before we come on to the questions?

Sir David Tweedie: Not really, sir. I really thought that it would be quite useful if I were to come, in the sense that for 10 years I chaired the UK Standards Board, after which I chaired the International Standards Board for 10 years, so I have a feeling of how the two sets of standards relate. I very much agree with the IASB submission, and I looked at the evidence presented by my successor, Hans Hoogervorst, and agreed with that as well. I also thought it might be quite useful because, prior to setting accounting standards, I chaired the UK’s Auditing Practices Committee, so I perhaps have one or two views on audit that may be of interest to the Commission.

Q373 Chair: Thank you. I would like to focus, as you will be aware we have in earlier sessions on this subject, on the $64 million question, or whatever it is. Why did the auditors give no warning whatever of the worst banking meltdown we have known in our lifetimes? Why were they apparently completely unaware, or if they were aware, what did they do about it? Was it incompetence? Was it the standards-were they wrong? How did this happen?

Sir David Tweedie: If you want to know why the crisis happened-

Chair: No, I am talking about the auditors’ role. I am not saying that they had the primary role-that obviously lies with the bankers themselves. Nevertheless, why did the auditors completely fail to sound the alarm in any respect?

Sir David Tweedie: The problem the auditor has is the going concern question. The minute you put a going concern qualification on a bank, there will be an immediate run. So a relationship between the auditor and the regulator is required. I think that that was missing at the time. There used to be one. When I chaired the UK Auditing Practice Committee, I remember negotiating with Brian Quinn at the Bank of England. We produced an auditing guideline, called "Audit of Banks", and we did one for building societies and so on. That required the auditor, if he had concerns about a bank, to go straight to the regulator, without the client’s notice, if necessary. He would ask the client generally to tip off the regulator as to the problem. If the client refused, the auditor would do it, and was required to do it. If the auditor was concerned about senior management fraud, or some actions that were taking the institution into serious trouble, he would report to the regulator.

Similarly, we asked the regulator to tell the auditor if he was concerned about various aspects. So we had a two-way communication, which I gather petered out, for some reason. That would have been extremely useful in the crisis. Look at what was happening with the leverage ratios of banks-how can you run a bank with 2% equity? It does not take much to lose that. Of course, they more than lost it in the crisis. So that was one of the key roles-the auditor really did not have the backing that he would perhaps have had before.

Q374 Chair: What do you mean by that?

Sir David Tweedie: The backing of the regulator. I don’t think the relationship was as good as it should have been. I must admit, I was quite staggered to find that that auditing guideline had just been, I presume, suspended, or had fallen into disuse. I think it would have helped enormously.

Q375 Chair: I was surprised too, particularly since I introduced the 1987 Banking Act, which specifically stated that there should be this dialogue. However, it was not made compulsory or mandatory. You said you agreed with Hans Hoogervorst’s evidence, and he agreed with me when I suggested that it should be made mandatory. Do you agree with that?

Sir David Tweedie: I do. In fact, although it was not mandatory in law, the auditing guideline required it, so the auditor had to do it. I have no idea what happened to that guideline.

Q376 Chair: As you say, it didn’t happen. The intention, at the moment, of the Governor of the Bank of England is to introduce a code of practice; my fear, which I do not know whether you share, is that that may-not immediately, but in the fullness of time-go the same way as the guideline. It is therefore safer to make it a statutory requirement.

Sir David Tweedie: I certainly would not like to see what happened then happen again. If that were to happen, it probably would be the case. The other aspect, if I remember rightly, was that there was also a meeting between the regulator and the auditor at which various aspects of the audit became a prime focus for the particular year. One year it might be loans, the next year it might be financial instruments of certain kinds, and so on. So there was very much a relationship with the auditor and the regulator, which, sadly, was not there when we needed it most.

Q377 Chair: So what you are saying is that it is your impression that the auditors were concerned but, because they could not qualify banks’ accounts for the reasons that you gave and because the dialogue with the regulator had ceased, their concerns never led to anything?

Sir David Tweedie: That is the case, and not just in the United Kingdom, because of the 10 biggest bankruptcies in the United States, eight did not have going-concern qualifications, and their capitalisation fell from $75.5 billion to just under $700 million within a year, so a 99% loss of value.

Q378 Chair: On the issue of qualifying banks’ accounts, you are clearly right to say there is a particular problem with banks that does not apply to the qualification of the accounts of, say, a motor car company or whatever else. Do you think there is a case for moving from this black and white, either/or-either they are qualified or they are not-dichotomy to a system of grading a bank’s accounts so that it does not have such a devastating effect? That is rather like, for example, if sovereign debt loses its triple A rating and goes to double A, which doesn’t lead to some great disaster, but it is a signal.

Sir David Tweedie: Well, that’s one way it could be done. I would personally prefer a complete revision of the audit report. When you looked at Northern Rock’s accounts-this was fairly well known in the 2000s-you would see the business model: 75% of their liabilities were due within three months, and they were lending out on 25-year mortgages. The only way Northern Rock was a going concern was if the wholesale markets stayed open, which, of course, they did not. That is what killed Northern Rock. How do you find it in note 43? Why isn’t that right up front?

Lord Sharman’s report also suggested something like this, and this is a personal view: instead of the audit report we have at the moment, which is all about why the auditor isn’t responsible for this and why management is responsible for that, with a very small bit on the opinion, I think it would be very helpful if, as is being proposed in America and with the international auditing standard, the auditor starts emphasising the big estimates. What kept them awake at night? Where are the issues of contention? All that would make the audit so much more useful.

One of the things about the crisis in America that shook me is that the 2008 audit report of one of the companies in receipt of the troubled asset relief programme funds cost $119 million, and in 2009 it cost $193 million and yet the audit report was, word for word, the same. So you had no idea what a totally different audit was actually undertaken. There was clearly an emphasis on different aspects of that audit, but an outsider would not know. I think all that should be brought out, which would be very useful for financial analysts and for the general public; it would give us warnings.

Q379 Chair: How would that be brought out?

Sir David Tweedie: By changing the audit report.

Q380 Chair: I realise that. What precise changes are you calling for?

Sir David Tweedie: I would want them to disclose the going-concern assumptions. What makes you say it is a going concern? In what areas did you have disputes with management? What areas were you concerned about? On what areas did you spend a long time in the audit? What are the big estimates that are involved in the audit, because there are always estimates in the accounts? Which are the ones that you had to look at very carefully, and which ones were pretty significant?

Q381 Chair: And this will all be in the published audit report?

Sir David Tweedie: I would prefer it that way. Some would prefer it to be in the audit committee’s report, with the auditor just confirming it, but I think the auditor, in a way, should do that.

Q382 Chair: There has been a great deal of criticism of the IFRS so far as provisioning and the true and fair override are concerned. There are a number of other issues, of which you will be well aware. I realise that as the father of the IFRS it is very difficult for you to criticise your offspring in any way, but we have been through a searing experience. You surely must feel, in light of that experience, that some changes should be made. What are they?

Sir David Tweedie: Not so much offspring as adopted children. We actually inherited the financial instruments standard on bank loans, which you are referring to, but it was none the less in line with what we did in the UK. There should have been provisions under that standard. The incurred loss model could and should have worked. I read in the transcript that someone sitting at this desk said that you have to have a default before you can provide. Utter rubbish! That is not the case. I will give your team the references.6

Basically, what we were trying to do-

Q383 Chair: That was certainly how it was interpreted by a number of people.

Sir David Tweedie: Absolutely. Let me just explain what the position was. In the UK, when I first got involved down here in London, I came down as the technical partner of KPMG. I don’t know whether it was the difference of coming to London, but what was deemed to be immoral in Edinburgh was frowned upon in the Midlands and was damn good business in the City of London. The sort of things that went on in the ’80s were outrageous. We had some multinationals and conglomerates that would create profits very simply. They would buy a company and they would write-down its inventory, even though there was nothing wrong with it, and make provisions for future re-organisations that were never going to happen. That looked very prudent on the balance sheet. When they sold this inventory, they made huge profits. They brought the provision back, which had never gone through the profit and loss account, but it went back through it as a gain, and they would say that that was due to their brilliant managerial skills. It was actually fraudulent accounting.

We brought in rules in the UK that said that you cannot provide unless you have an obligation, and that became the international rule too. We wrote the thing for them. We had a situation whereby an obligation was required before you could provide, because otherwise you can do profit manipulation, whereby you put a provision through in good times and add it back in bad times and disguise it. That happened a lot in Germany. You then had a situation with translating that into bank loans. We didn’t want them simply to put through some huge provision in good times so that you wouldn’t know, when bad times hit, that the bank had had a bad time at all, because you wouldn’t see it. We wanted it to occur when the position was there.

The standard was written to say that you had to have some form of evidence. People have interpreted that as default, and that is wrong. It can be any of the numerous things that we have in that document. It can be adverse financial positions, re-organisations, a change in markets, a fall in prices, an alteration in unemployment rates affecting mortgages, business and financial risk exposure, and local, national and international factors. We listed many factors and said, "Look at all or any of these. Make your judgment. Are you going to get your money back?" That should have worked, but it didn’t.

Part of it may be that, given that the loans were completely mispriced for risk anyway, they didn’t believe what was happening to them. In fact, I suspect that most people at the beginning of the crisis had no idea how big it was going to get. That might have been one reason. Other reasons might include that they didn’t want to do it. They managed to persuade the auditors that everything would turn up. As the markets fell, we heard it again and again: "They’re going to come right back. This is only a temporary blip. Don’t worry about it."

Q384 Chair: Let us focus for the time being on provisioning. There are other aspects, of course, of which you will be well aware. On provisioning, we are really talking about three of the big four auditing firms, because one of the big four does not do much bank auditing; it is all done by the other three. Are you saying that those three firms were concerned about the possible inadequacy of the provisioning that was going on but didn’t do or say anything about it, or are you saying that they didn’t and couldn’t because they perhaps misunderstood, in your terms, what the IFRS enabled them to do and why-or were they just unaware?

Sir David Tweedie: I don’t think they were unaware. The scale of the crisis didn’t hit them till later. Probably, a lot of people hoped that this was the market having a major drop but that it would hopefully come back again. It didn’t, and then you saw in 2009 there was much heavier provisioning than in 2008, because by that time the atmosphere had changed and people realised the gravity of what was happening. A habit had developed, which we were not aware of at the IASB, of interpreting our standard in a certain way, which we would not have approved of. We recognised at that stage that we could probably write another 40 or 50 pages about the incurred loss model and we would still have problems with it if people had already got a mindset about how to operate it. That is where the expected loss came. We brought that in-we didn’t bring it in, because we had to try to get agreement with the Americans, but we started with that in 2008 because we felt that people were not providing heavily enough, and if they weren’t, we would change the standard.

Q385 Chair: Of course, trying to get agreement with the Americans was a big mistake, wasn’t it? It held up progress on a number of fronts.

Sir David Tweedie: It did.7

Q386 Chair: Anyhow, as I understand it, there are some changes being made now on that front with the IFRS, so that, really, is some learning from experience. Do you think that your much more voluminous narrative audit report would also help in this, and that if it had been about then, they would have said that they were concerned about possible underprovisioning?

Sir David Tweedie: They might have done. I was just thinking of that audit I mentioned where the audit fee jumped $74 million: they would have had to explain what their interest was in that area, and I expect very strongly that it would be the mortgage book-that’s what they would have been spending their time on. It would be very clear where the concern would be. Their job at the end of all that, since they gave a clean opinion, would be to make sure that there had been adequate provisioning.

I have heard it said that the IFRS got rid of prudence. It didn’t. It got rid of the title "prudence", but the way the IFRS operates is that standards come first. If you look through those standards, prudence is written all through them. If it comes to a case where there are no standards, you fall back on the conceptual framework. We had a long discussion about where prudence fits in with neutrality. Again, the argument came in that being neutral means trying to get the right answer without bias, and if you bring in prudence you are biasing it. The idea therefore was, "Let’s put in the standards and make them as prudent as we want them to be; then, we want people to try to get the right answer thereafter."

So prudence has not gone; it is right through the standards, and we can give you examples of them. But as far as the international situation was concerned, the view was that we ought to reel it in because we are trying to get the right answer. That is all that has happened. Prudence is still there.

Q387 Chair: Well, not as much as it was.

Sir David Tweedie: If I may contradict you, it is actually in the standard. If you look at IAS 8, prudence is still there. It is out of the conceptual framework, but it is in the standard.

Q388 Chair: It is out of the conceptual framework, as you say.

Sir David Tweedie: But that is a lower level than a standard. It says there that if there is no standard, you fall back on these principles of reliability, neutrality and prudence. The question is then, where do neutrality and prudence come in? And they clash.

Q389 Chair: We have had some conflicting evidence on that, but I would like to move on to another area, which is valuation: fair value accounting, mark to market and mark to model, and all that. This is in many ways very dodgy. I am not saying that the previous system didn’t have its own defects. But let us forget UK GAAP; we are talking about IFRS. The dodginess of that is now fairly widely understood. It is true that in IFRS, as I understand it and correct me if I am wrong, there is a true and fair override that can be used if it looks as if the fair value account of the mark to market, mark to model and so on is producing dubious results. Yet according to a paper, "Audit Policy: Lessons from the Crisis", published in 2011 by the European Commission, "The ‘true and fair override’ is required under accounting standards, but seems never to be applied in practice." Why is that?

Sir David Tweedie: It is supposed to be in exceptional cases, not in the generality of cases. The reason you don’t have an override in countries like Australia is that it was used very often. People used it just if they did not like the standard. They would use an override and go past it. Eventually the Aussies abolished it for that reason. So you can actually devalue the whole accounting system by it. I have used it personally. When I was a partner we refused to allow a company to obey a standard which would have given them an extra £50 million profit, simply because the standard was not dealing with their particular situation. You could read the standard in the way the company did originally, but clearly it was not right and they did not make the profit. The profit was not there and we refused to allow them to show it. So they had to use an override, otherwise we would have qualified the accounts. So it can be used but only in rare situations. It is not meant to be used just willy-nilly. The idea with the standards is that we consult pretty widely. We take three years to produce the standard and at the end of it, that is the international consensus of what should happen. So we don’t want companies willy-nilly just blowing them away. So we put a fairly tight constraint on it.

Q390 Chair: But banking is different, isn’t it, particularly banking as it had developed in the last decade? Instruments became more and more complex and harder and harder to value. This did not apply much outside the banking system, but within the banking system these financial instruments were so complex that the valuation became more and more artificial in many cases-not in all, but in many-and therefore the need for the true and fair override must have increased commensurately. Yet nobody ever used it.

Sir David Tweedie: In 2008, we took the opportunity to try to get rid of the financial instrument standard. We knew that there were problems and that we would like a new one by the end of 2008-for 2008 year ends-and we did it. IFRS 9 was ready in November 2008, and the European Commission at the last minute, by intervention by the French Trésor, didn’t ratify it. So it is sitting out there, and what it says is very simple. The principle is that if you know the cash flows and you are keeping the instrument to get those cash flows, you can then show what it costs. So we are talking about debt; we are talking about loans. If you don’t know the cash flows, use the market values.

On the alternative, equities you can see fairly clearly, generally from the markets. Derivatives are the big issue, and how do you value derivatives? If you don’t value derivatives, there is no cost. It is a contract, so you and I decide to exchange £10 million at whatever the current dollar exchange rate is for the number of dollars that would give you today. The minute that exchange rate changes, one of us wins and one of us loses. If you don’t show that, which is the price of the derivative, you will not know what is going on in that company until eventually that contract expires. You could be running up colossal losses or colossal gains, and it would be all off balance sheet. That is why these things are at value.

The problem, as you correctly say, is some of these things are so complicated that it is very difficult to value them. The alternative is either to show no value, and just put a little note saying, "By the way, we haven’t been able to value these things," or there is some other assessment, and what is that going to be? Generally speaking, that is a model. Only 3% of banks’ assets in the UK are modelled. None the less, in that-this is where prudence comes into the standard-it quite clearly says in the relevant standard, IFRS 13, that you have to have a discount for model risk. Most banks would also put one in for liquidity risk, so they write it down.

Apart from that, what are we going to do? I remember the French banks arguing, even on the equity markets, "This market has gone down. We know that the market is showing 40, but we think that the real intrinsic value is 80." I think if we had let companies do whatever they liked in that situation, using an override, the whole thing would be out of control. Nobody would know any benchmark. At least they could decide that the market was too low and to write up the assets, but if you give someone complete freedom to pick any number they like, I think it will be a complete mess, and that was why we did not do so.

Q391 Chair: But the auditors, you say, can apply a discount.

Sir David Tweedie: Not the auditors; the company. The company would do it.

Q392 Chair: But if the company does not, what do the auditors do? It must cause the auditors considerable concern.

Sir David Tweedie: It says explicitly in that standard that if they do not do it, it is not a fair value, and that breaches the standard.

Q393 Chair: But there was no case where that happened, was there?

Sir David Tweedie: I presume, in that case, that they did the discounts.

Q394 Chair: Or else they did not, and the auditors did nothing about it.

Sir David Tweedie: That is the alternative view, but I would not have thought that an auditor would wave that one through. These things are too dangerous.

Q395 Chair: You do not know, anyhow; you simply surmise.

Sir David Tweedie: I don’t know. I did not do the audits.

Q396 Chair: One last point on this issue. The Financial Policy Committee gave evidence to the Treasury Committee the other day. The FPC stated that IFRS materially overstated the values of banks loans. What do you have to say to that, because the Financial Policy Committee could hardly be a more respected authority?

Sir David Tweedie: But that does not mean that it is right. Basically, I think the situation was that it is not IFRS. When you look at the paragraphs within IFRS that deal with the incurred loss model, they fire up all these different factors to look at, and those were all present. I have heard the governor of the Irish Central Bank say, "We were not allowed to provide." Oh yes, you were, and I am surprised that you did not. The same thing happened in 2011 when I was on the "Today" programme. We were talking about Greece, and I said publicly, "I do not know why they are not providing. Greece is in trouble."

You have to look at the financial status of a loan. For some reason, I remember the French banks were writing 20% off Greek debt, and most of our banks were writing 50% off, but it was quite clear. You had to write it down to what you estimated the future cash flows would be. I would have thought that it was pretty obvious in 2011 that you were not going to get 100% back from Greece. It was there; it was all in the standard. Just do it.

Q397 Chair: So you are saying that the standard is impeccable.

Sir David Tweedie: No. It could have worked and should have worked, and we fully expected it to. Once you get a situation like that, however, you have to say, "Okay, behaviourally it is not working, so let’s go to the expected loss model." We produced one very quickly, and it was very simple. What we really said was, "For this tranche of loans, whereas you normally lend at 4%, because of the risk involved in this, you are going to lend at 6%-expecting, therefore, 2% losses-and that you will get later on." What we proposed was, "Since you are not providing quickly enough, let’s put this in and therefore every year you will provide the 2%. So you do not take 6%; you will just produce the 4% that you ultimately expect to get." That would have been fairly simple, but we ran into the American block, and the G20 was telling us, "You must agree with the Americans." That makes it difficult.

Q398 Chair: But as I said in another session of this panel, you should have told the G20 to take a running jump, because here was clearly a case of the best being the enemy of the good. Some ideal of the global agreement was actually preventing a much-needed accounting standard that we could have used in this country.

Sir David Tweedie: Well, eventually, of course, Hans did tell them to take a running jump, and it was on the cards probably for a year before I left the IASB-when do we say, "This is enough"? We kept hoping that the Americans would come round and take IFRS. More than 120 countries use it. I think that it was important to try to ensure that, as happened in the crisis, we did not get arbitrage, with people looking at the US standard and saying, "Oh, that’s weaker than IFRS; we’ll have this bit," and vice versa.

There is a big price to pay for having different standards and I think generally the world wants a global standard. I am sure that eventually the world will get it, but Dodd-Frank and the American political system got in the way.

Q399 Chair: But there are other things that may be even more important than getting global agreement, even though global agreement in itself would be handy, if it was on the right basis.

Finally with IFRS, as I understand it you have been saying that there is really nothing wrong, except possibly this adjustment you have made from incurred loss to expected loss-a slight adjustment. You don’t think there are any more adjustments that are desirable-nothing else; it’s otherwise perfect to you?

Sir David Tweedie: No, it is not perfect; there are lot of things. It is one of the frustrations-I expect you felt this in your parliamentary career-that you leave without things done that you would have liked to have done. I think that it was the same in UK GAAP when I left, and it was the same when-

Q400 Chair: What was your biggest single regret?

Sir David Tweedie: I think that the biggest single regret is that we didn’t get IFRS 9 out in Europe; I think that was a big mistake that the Commission made. We also regret that we couldn’t pull the Americans along with us at a much higher speed. If we had known what was happening, I think that we would have pulled the plug quicker, and that is a major regret. But there are lots of things in accounting that are pretty archaic and have to be changed, and IFRS is not immune from them either; they are in British accounting as well.

Q401 Chair: In addition to the imperfections in IFRS such as they are, which you have conceded, they were made worse, if I understand what you are saying-tell me if I am wrong-by the way they were interpreted by the auditors. They were interpreted in a way that was too mechanistic.

Sir David Tweedie: I think that is too strong. Basically, I think the original problem was the mispricing of risk. I don’t think people realised in 2008 how serious this crisis was, and that led to-

Q402 Chair: Is that people including auditors?

Sir David Tweedie: Including auditors-companies and auditors, and certainly regulators, too. And I think that afterwards, in 2009, you saw a big hike in the provisions that were being made. I really feel, looking at the standards, that the words were there; it was the way people interpreted them. And I think genuinely they probably didn’t realise just how bad this was, and they certainly didn’t in 2006 and 2007.

When we look at expected loss, it is not going to be the panacea, in the sense that in 2006 people did not anticipate the crisis and they would not have made provisions for that sort of thing way up front. And the danger, of course, with expected loss is profit manipulation-people over-provide too much up front.

There are other ways of dealing with some of the issues, and I suggested them both to the Financial Stability Board and to Lord Turner. One of the problems that we had in the boom, according to the Basel Committee, was that the banks paid out everything. So, all the profits that they got went out on dividends, on buy-backs of shares to keep the price up, or on compensation. We were under a lot of pressure at that time. The regulators asked "Why don’t you put in something like a through-the-cycle provision?" That was what they called it. So, in good times, you get banks to provide a great chunk, and in bad times you feed it back in.

Well, we say that really isn’t our job. Our job is to try to show what happened, but we agree with the problem. What we suggested to them was that in the accounts you show the accounting profit, and then you say, "Now, regulators want a certain amount held back as an undistributable reserve." That gets deducted from profit in what we called regulatory income. You could also deduct from that all income from modelled fair values, and then you end up with a level that is suitable for distributions, compensation calculations and so on. That is different from the accounting profit, because once you start bringing in phoney provisions, you have a problem, whereas what we are saying is, "You can get exactly what you want," which is retaining profit by having undistributable reserves, and that does not distort-it is clear-and that can be settled by the regulators.

Q403 Chair: I remember you wrote a letter to the Financial Times saying that you had suggested that there should be two different concepts of profit: accounting profit and regulatory profit. I think that led to a certain confusion. It is understandable that people were confused but, again, I suppose that if you had this lively dialogue all the time between the accountants and the regulators, that confusion might not have been so bad.

There is one other aspect-one other dimension-of this: do you not agree that although auditors have a duty to the shareholders, they also have a wider public duty? There is a wider public interest in having reliable audit. Would you agree with that?

Sir David Tweedie: Yes; I think our conceptual framework actually talks about the fact that the objective of accounting is to give financial information to informed investors, but we also include debt holders and all the other people who are involved in the company. It is a much wider brief than just the equity shareholders. One of the problems you had in the crisis was: where were the debt holders? If you are going to have 98% of a bank funded by debt, where were they? Were they assuming the regulators or Governments would step in and deal with the issue? When you looked at the credit default swaps, the premiums on them were falling as the crisis was coming nearer and nearer, so they were completely mispricing the risk-or perhaps they were just expecting Governments to step in.

Q404 Chair: And the auditors were aware of this, were they?

Sir David Tweedie: I do not think that is the auditors’ job. That is what the debtors-

Q405 Chair: But were they or were they not aware of it?

Sir David Tweedie: I would not know; I am not an auditor. It is a long time since I have been there. It is not their job, really.

What struck me about the crisis was watching how this leverage just rose and rose. You had equity of about 5% in 2000 and it was down to about 2% to 3% in 2007. That is very, very dangerous.

Q406 Chair: We are all aware now of what went wrong. The question is: why was this not noticed in advance? Why were the alarm bells not ringing? The auditors have a role to play in this because we are concerned that this should not happen in future.

Sir David Tweedie: Well, whose job is it to sound the alarm when leverage gets low? I would have thought that that was a regulatory issue. What the auditors did was to show the accounts,8 and the accounts showed very clearly that we were heading for a problem, and that is one of the things it did show-the explosion of the balance sheets.

Q407 Lord McFall of Alcluith: In the previous commission I was involved in, Jon Danielsson of the London School of Economics stated, "The problem with banking is [it is] so complicated that any financial institution can make any number look any way it wants, meaning that if the government starts to target anything, the banks find a way to bypass the target". Is there not an element of truth in that, and therefore perhaps a regulatory case for smaller banks?

Sir David Tweedie: The interesting issue you raise is: when it occurred in the crisis, the markets froze, so what was the fair value at that stage? The answer was that it most certainly was not distress fire sales, but what was it? We looked to see who was going to come up with answers to this, and there was no one. We had to set up an ad hoc group, which dealt with illiquid markets, to come up with prices.

Going on from there, I nearly got involved with the International Valuation Standards Council. One thing that became very clear very quickly is that although there are certain rules for real estate valuations and business valuations, there are no rules for financial instruments. The Accounting Standards Board can say, "You do fair value and you use models," but what models? You get something like Lehman’s marking at a totally different level from Goldman’s. Why do you get these massive differences; there is a big crack that everything is falling down? I fully agree with the point that we have to do more on the valuation side, and that is what we hope to do in the next two or three years. We have to close that gap. There are no rules for that at the moment.

Q408 Chair: You mentioned that one of your regrets was that IFRS 9 had not come in, and IFRS 9 is intended to address a number of problems, including the volatility issue. There is no sign now of IFRS 9 being agreed. Do you think that we should go ahead on our own and try to get a waiver to enable us to do that, or do you think that we should just carry on with an imperfect-a seriously defective-system?

Sir David Tweedie: We were pretty angry when IFRS 9 was rejected by the Commission, but at the moment, we in the UK have signed up to using IFRS as adopted by Europe and we are stuck, so you really have to get round that agreement.

Q409 Chair: That is a problem, isn’t it?

Sir David Tweedie: Yes, but I think it would be a mistake if the UK went off on its own. I would like to see the UK persuading other countries that we have to go after the Commission and say, "Get on with it."

Q410 Chair: In an ideal world, I am sure that that is so, but we have to deal with the real world in which we have a severely imperfect system. You know what, in your opinion, would deal with the problems, or at least some of them-namely, IFRS 9-and you cannot get agreement.

Sir David Tweedie: No, we have a difficulty there. It has been adopted in Canada, Australia, Brazil and Japan. Other countries use it; Europe doesn’t.

Q411 Chair: But Europe doesn’t. Well, this is a serious business, isn’t it?

Sir David Tweedie: But it is a political issue, and not one that IFRS could deal with.

Q412 Chair: It is also a practical issue.

I have one other thing. In the evidence that you have given, it has been clear that there are very specific problems with the accounts of banks that do not apply to the rest of the economy or the rest of the corporate sector. That is partly connected with the qualification of banks’ accounts that we have discussed and partly with the complexity issues, which are likely to get greater and greater-they are going to be discussed-and there are also a number of other things that will be very familiar to you. In the light of that, Andrew Haldane of the Bank of England has suggested that it might make sense to have a special variant of IFRS to deal with banks’ accounts that is tailored to dealing with these specific problems. Do you think that that might be of value?

Sir David Tweedie: I have just had lunch with Andy. No, I don’t; I think it would cause more problems than it would solve. Most of the difficulties over the financial instruments standard were with banks. We spent a lot of time with banks discussing it and trying to get an agreement and so on. The danger you have is that first you have to know how to define a bank. When you look at companies such as GE, they have almost a bank within the whole company, so how do you split that off?

Q413 Chair: Well, a bank is defined, because a bank has to be licensed as a bank.

Sir David Tweedie: But then you have parts of other companies licensed and you get into a mixture. Why would we have different rules for the same sorts of financial instruments? I think that banks can be coped with within the system. Interestingly enough, in America, the Pozen committee, reporting to the SEC, said exactly that: we should not have industry standards; we should have one standard and everyone should adhere to it.

Chair: Do any colleagues wish to ask any questions?

Thank you very much indeed, Sir David, for coming before us this afternoon. It is a complex but important subject. We will be reflecting on what you have said, and that will be fed into the conclusions that we reach. Thank you.

Sir David Tweedie: Thank you; it was nice to meet you.

[1] Note by witness: You could not do this now anyway with hedge accounted swaps because tax legislation has changed.

[2] Note by witness: Indeed I do not think it is a “good thing”. It can be seen by other countries as “London” helping in raids on their tax base. When deals go wrong it can severely damage relationships of banks with each other. For example, Alessandro Profumo (former CEO of Unicredit) is facing criminal charges in Italy – for what was essentially a "vanilla" tax-structured Pref deal sold to them by BarCap. There are 6 US banks now facing enormous tax costs for just one structure (“Stars”) sold to them by BarCap. I somehow doubt that BarCap are winning many corporate finance or DCM mandates from any of these institutions.

[3] Note by witness: Protium is the classic public example, though it seems it did not work.

[4] Note by witness: The reason they did not care was because banks were not actually, or rather not transparently, short of capital until post-crisis. Deal teams did it before then because it improved their internal Return On Equity and thus tended to boost the apparent business performance and probably their bonus base.

[5] Note by witness: SCM profits were unsurprisingly never split out but simply reported as part of the Fixed Income business. To the extent investors and analysts perceived BarCap as breaking into the big league on Fixed Income, assumed it was repeatable and diversified and applied a PE ratio greater than 1.0, then perhaps there was something rotten in the State of Denmark . Perhaps auditors might have thought this material to understanding of the accounts and should have insisted on additional disclosure. Of course, the same point goes for every other bank with SCM type activities - but the relative weighting was far, far smaller.

[6] Note by witness: The relevant references are IAS39 paras 58-64, IAS39 Implementation Guidance and IG Para E.4.1.

[7] Note by Witness: I would like to add that the G20 wanted the same rules worldwide so we had to try to align the US rules and IFRS.

[8] Witness Correction: I should have said ‘give an opinion on the accounts’ and not ‘show the accounts’.

Prepared 13th March 2013