Banking Crisis: regulation and supervision - Treasury Contents

Examination of Witnesses (Questions 40-49)


23 JUNE 2009

  Q40  Chairman: In terms of liquidity, how do you define `liquidity', Professor Morrison?

  Professor Morrison: It is a very hard thing to define, which is why we do not have formal regulations over it at the moment, so many people define a bank's liquidity in terms of the asset base that it has, so, if it has lots and lots of gilts on its book, then it is a very liquid institution. The problem with liquidity, as I remarked earlier, is that it is not a constant, but it is something that evolves in line with market expectations and the beliefs of market participants. When market participants stop believing in one another, liquidity dries up, so I think perhaps a better way to think about liquidity would be to think about the exposure a bank has to the drying up of liquidity, and that is particularly to be measured in terms of mismatches of maturities, so banks have always engaged in maturity transformation and that is something that we would like banks to do. When the maturity transformation is not between deposits and long-term investment, but between short-term wholesale funding and long-term investment, it seems that the liquidity risk is much higher, so my inclination would be to measure liquidity in terms of maturity mismatches on a bank's portfolio.

  Q41  Chairman: Why, in Adair Turner's words, did both economics and policy-makers "take their eye off the ball" regarding liquidity?

  Dr Alexander: I think that there was an under-appreciation of the risk that liquidity risk poses to the broader financial system. So much of the policy debate and so many of the academic models looked at market risk and credit risk, and then liquidity risk was under-appreciated. In fact, Alan Greenspan praised the fact that we had a smoothing of risk in the financial system and that securitisation helped facilitate this, so the types of financial instruments that were being used in the economy, credit risk transfer instruments like securitisation, were seen as spreading risk throughout to those who were willing to absorb the risk and, therefore, there was not an appreciation that liquidity risk could arise in such circumstances, so the academic models, the policy, the regulatory frameworks were built upon the fact that credit risk transfer was promoting liquidity, but what we did not count on was the fact that suddenly all the institutional investors could just simply not want to roll over their short-term investments and then the liquidity would dry up. That was something that was not foreseen and it is a major failing, I think, on the part of both academics, policy-makers and of course the risk managers in the banks who should have seen this.

  Q42  Chairman: Dr Lilico, the FSA has suggested a `core funding ratio' tailored to each bank. Now, that might expose the FSA to the accusation of inconsistent regulation across banks. How can it ensure adequate liquidity at each firm whilst ensuring that each firm is treated fairly?

  Dr Lilico: It seems to me that the liquidity problem is a mistake, it is just a symptom of the wider mis-pricing of risk, so, if you got your risk analysis of how much risk there was and if you got your risk assessments right, then the likely fluctuations of the requirements of liquidity would have been less and the liquidity might have proved adequate, so I think that the liquidity point is to be overstated. I think it is important to distinguish between different kinds of crises, so any kind of institution can have a crisis associated with just not having enough cash, sometimes you can have insolvency associated with past losses and sometimes you can have insolvency associated with a lack of future profitability. I do not think that this was just a liquidity crisis and I think that some of the liquidity discussion is just a carryover from the early phases of the crisis when we thought liquidity was a bigger issue, so, although I think liquidity is worth looking at again, I am not convinced that we need to change things all that much.

  Q43  Nick Ainger: Following on from that, the Turner Review sets out the figures in relation to structured investment vehicles which show that in four years from 2003 to 2007 the growth of SIVs and their total assets tripled, and Turner says that this was a major contributor to the highly leveraged situation that certain institutions found themselves in. Dr Alexander, you talked earlier about one of the banks' functions is maturity transformation. How important are SIVs to maturity transformation? Do we still actually need these SIVs?

  Dr Alexander: I think securitisation is an important component of our financial system and that we should not throw it out, but it is how it is regulated and we have to understand the risk that securitisation presents. Again, many experts did not foresee the liquidity risk that an over-reliance on securitisation funding could pose to the broader financial system. I think that, if we properly regulate securitisation, SIVs and the various conduit funding that banks have been using, then they are appropriate ways to raise capital. They are a part of financial innovation and we should not curtail financial innovation, but we have to understand that the funding through SIVs is short-term and that it can dry up quickly, just like in the old days there was a bank run with depositors running for the exit, but now we have got institutional investors that can turn off the funding pretty quickly, and we have to think about how to regulate that and what types of costs to impose on that.

  Q44  Nick Ainger: Coming on to this point about how to regulate, we have been told by academics and practitioners in the financial services sector that part of the problem is the complexity of these various vehicles, these CDOs and CDOs squared, and we even had the Chairman of the Deutsche Bank who did not even know what a CDO squared was. It is all very well saying, "Yes, we should regulate and regulate them better", but with the very fact that they are incredibly complex and that the risk element is in their complexity, how do we expect regulators to understand the risk involved and, therefore, point out to the institutions, "You are at risk because of this" if the institutions themselves do not even understand the risks involved?

  Dr Alexander: Well, if they do not understand it, they should not do it and the regulators should not permit it. Right now under Basel II, regulators have to approve the risk models that banks submit to them for review, for credit risk, for market risk and now liquidity will be included in Basel II, so regulators should not be approving these models as submitted if the bank cannot explain the model and cannot demonstrate an understanding of the model. This is part of the dialogue, the interaction, that the regulator has to engage in with the banks, but it is not the regulator's problem to figure it out. The bank has to explain it and the regulator needs to have maybe advisory experts there to test the models that the bank is submitting for approval. If they have a model that is testable and it makes sense, then they can approve it.

  Q45  Nick Ainger: So what we need is actually far more informed regulators who are able to make an objective judgment of what they are being told by a particular institution?

  Dr Alexander: They need to be able to make an objective judgment, but the burden is on the bank to make that model work, to prove that under certain conditions of stress-testing and various other ways, and the regulator has to be there to be vigilant and it is not the regulator's job to figure it out, but it is the bank's job to explain it and then the regulator, I think, will rely on expertise to make sure before deciding whether to approve it.

  Q46  Nick Ainger: Do the others agree with that?

  Professor Morrison: Well, no one could argue against a well-informed regulator, it is like arguing against commonsense, but I think we need to be aware of our own systemic limitations. Some of the models which were generated were very elegant and very clever and had very little economic content; they were essentially physics rather than economics. We made the mistake of thinking that, as a physical model can tell us what happens when water boils, these models would tell us what happens when a default starts to occur in financial markets, and sometimes the problem here was excessive reliance in the regulatory arena upon models which were really intended as devices to help managers to understand what was going on, so, to the extent that there were regulatory failings, I think perhaps the regulatory failings were in taking some of these models too seriously. When you do that, the incentives to the people who create the models are bifurcated. To some extent, they are concerned with creating accurate models and, to some extent, they are just concerned with reducing their capital charges, so one problem here was not appreciating the limitations of some of the tools that were being used in regulation and that, by using the regulation, it may have changed their nature, and that is a big problem. Another problem was failing to recognise, and this is a simple problem that can easily be fixed, but failing to recognise that banks that provide backed-up lines of credit to SIVs, even if those lines of credit are not legally binding, actually find it very hard to walk away from them and, hence, had a liquidity exposure that was not recognised.

  Q47  Nick Ainger: You used the phrase "physics, not economics", but what do you actually mean by that?

  Professor Morrison: What I mean is that, if you take a simple model of a securitisation, say, I have a securitisation where I take two loans that I have made to a corporation, bundle them and then sell another piece of paper that defaults when both of these loans default together, then, in order to put a price on the securitised asset in the absence of a liquid market, I have to make assumptions about how often each of these loans will default individually and how often they will default at the same time. People do this using elegant methods that come from the mathematics of fluid mechanics originally, so I have a model which shows the price of one of these things moving up and down and another one and, when they both move down far enough, we see a default, so those are models that have this feeling of precision. You can calibrate them by going out and gathering data and you get a figure for simultaneous defaults and you get a figure for individual defaults and you treat those things like physical constants in the same way that you have a number that tells you when water freezes and when it boils, but those things are not physical constants, they are economic numbers that depend upon the behaviour and the expectations of market participants. When market participants suddenly realise that everyone is doing the same thing and they suddenly realise that, because of the immense opaqueness of these markets, they do not know what one another is doing and exactly what the exposure of different people is, those parameters can change rapidly, so assumptions built and hard-wired into models about correlation parameters just turned out to be completely incorrect in 2007. It is not that they had been estimated in the wrong way, the estimation procedures were good, but there was a structural change in the way that people saw the world. I think failure to understand these feedback mechanisms as being integral to the way that financial markets operate has got a lot to do with the misuse of these mathematical models, so, when I say "physics, not economics", I mean that in physics we have some things which are constants and we can make predictions that are going to be correct over time, and in economics we can make statements about how people react to incentives, but straightforward correlation parameters are not physical constants, they are summaries of how people respond to incentives and what those incentives are.

  Q48  Nick Ainger: Dr Lilico, alongside the SIVs and so on, we also had the mutual funds and so on taking part in maturity transformation as well, so this shadow banking system developed. Turner says that this again added further risk into the whole system and much of it appeared to be almost beyond the regulator. Is that a fair analysis?

  Dr Lilico: I do not think that that is really right, no. It seems to me that a key discipline for innovation in any sector is that you put a lot into some new, fancy thing where you do not know what is going to happen about it, you get it wrong and you go bust, but, if you do not have that discipline that you go bust if you get it wrong, then you should not expect innovation to be efficient, so that is one key factor here. Also, I think that people have been a little bit harsh on some of the SIVs in that, if the companies had actually gone so far as going into liquidation, I do not believe that as much would have come back on to the balance sheet as it actually did because, when you have some legal separation, the creditors of various bits would have objected to being lumped together with creditors on other parts, so I think that that thought that it is an entirely artificial separation is a mistake. I also think that there is little evidence that actually hedge funds and the wider sector contributed anything negative to the financial crisis. I think that, in some cases, they were the canary in the mine and, in other cases, they were the messenger, the bringer of bad news and I think that, if you did not want to hear the bad message, then you objected to the hedge fund and its practices, and I think that is just a mistake.

  Q49  Nick Ainger: Anyone else on shadow banking?

  Dr Alexander: I am less suspicious of the structure of finance as it evolves. It evolves in response to government concerns and the financial system has evolved because of the regulation as well. What we have to do is try to ensure that the regulators try to impose a proper cost on risk-taking, and what we had not understood was the type of social cost that this so-called `shadow' banking sector posed to the financial system. It is not that that shadow banking sector is bad to have, that we should prohibit it, but it is that we want to put a price on it so that the risk-taking is internalising the costs that it is creating, and that, the regulator had failed to do.

  Professor Morrison: The shadow banking sector, like most innovations, there are good things and bad things about it. One needs to be careful in financial markets to distinguish, although it is very hard to do so precisely, between innovation which is there to encourage the efficient use of capital and the efficient deployment of resources, which is what we would like financial markets to accomplish, and innovation that is there to get round regulation, and there is no doubt that a good proportion of the shadow banking was about getting round regulation. This may reflect the fact that regulation became excessively complex and there were massive whole regulations and we are now aware that this is a problem, and I suspect that it is being addressed, but a good part of financial innovation is the response to poorly designed regulation.

  Chairman: Can I thank you for your evidence this morning; it was very helpful to us in advance of Lord Turner's appearance before us this morning, so thank you.

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2009
Prepared 31 July 2009