Abstract

JACQUES DE LAROSIÈRE: Charles, thank you very much, indeed, for this extremely thought-provoking speech. I think Professor Goodhart is happy to answer a few questions. I would ask you to be very concise in their formulation because we have very little time, but the floor is open now.

JACQUES DE LAROSIÈRE: Charles, thank you very much, indeed, for this extremely thought-provoking speech. I think Professor Goodhart is happy to answer a few questions. I would ask you to be very concise in their formulation because we have very little time, but the floor is open now.

QUESTION: Charles, I would like to challenge your proposition that we need a model, a single model, of systemic risk in order to account for what lies ahead. I think when you are dealing with risks, you may have an idea of the probability distribution of events lying ahead. One of the characteristics of financial market developments is exactly that we do not have this probability distribution. We have no idea, so I think that basically if you want to have a radar screen of what lies ahead, you do not need a single model. To the contrary, you need a host of models that take account of the various characteristics of financial institutions, their different strategies, the different directions in which they are going and possibly their interplay, so it would actually be dangerous to focus research on such a single model. I think you need a host of things to be running parallel to take account of this underlying uncertainty which will not go away.

CHARLES GOODHART: I entirely agree with that. I probably put my point incorrectly. The problem actually is not that we are trying to focus on a single model, it is that we just really don’t have any models of systemic interlinkages between banks in which default plays a major part. I am certainly not trying to advocate that the particular exercise that I am undertaking should be the focus; far from it. We need a whole series of models, as indeed you correctly said, just as there are a whole series of macromodels. The point is not that we are focusing on a single model and that is dangerous, it is rather that there are just no models out there which incorporate default, financial incompleteness, and heterogeneity of bank and customer behavior. The more models there are that incorporate these conditions, the better, and I would be very keen for a plethora of models, and I am glad to say that such models are being developed. The Austrian National Bank is developing models along these lines, and there are quite a lot of central banks which are thinking of this, and I hope that they will all proceed. The more models the better. As I say, it is not that we are focusing on one, it is just that we don’t have any at the moment.

QUESTION: I just wanted to raise the point you made about the distinction between liquidity risk and capital risk. Surely the point is that when crises come along, when banks get into trouble, it is always a question of whether it is a liquidity problem or a capital problem, and the real issue here is how the regulators come down on the decision making. Lately, of course, they have been much more biased toward capital, but isn’t this really an empirical question that you have to address when a crisis occurs?

CHARLES GOODHART: Yes, but if a bank does not have liquidity, the speed of the crisis will be considerably greater, and the pressures put on the authorities to act will be much more acute. Once a crisis has started, what banks actually need is liquidity rather than capital. The two, I think, are complementary. It is absolutely right to think of capital requirements, but to ignore, or to put entirely on one side, liquidity requirements is not right.

When I started working in this area at the beginning of the 1960s, what regulators thought about all the time was liquidity ratios and cash ratios, and capital really wasn’t considered, and that was wrong. Now it seems to me that all the focus is on capital and there is not sufficient consideration of the liquidity requirements that are needed to support it. So I am asking, in a sense, for more balance.

QUESTION: May I ask you how you will reconcile your defense of fair value accounting with the following example? Suppose that you have a company which would be downgraded by a rating agency, then with full fair value accounting, the value of the debt of this company would be reduced, therefore leading to an increase in the capital base. So it means that what the market would tell you is that the strength, the financial strength of the company increased, whereas the judgment, the supposed intelligent judgment of the rating agency is that the financial strength of the company actually decreased.

CHARLES GOODHART: I am not sure that I follow, because if the fair value accounting has been impaired, then the capital value has also fallen. I do not follow how you will get an increase in the capital ratios under these conditions because the value of the capital is impaired when the assets fall relative to the liabilities.

FOLLOW-UP QUESTION: But in that case the assets have not changed. It’s the value of the liability which has decreased the value of the debt, because the change in the rating means that the actual value of the debt decreases because the implicit interest rate increases.

CHARLES GOODHART: Can I pass on that for the time being?

JACQUES DE LAROSIÈRE: Yes, I think you can pass. It is one of the questions that I had in mind too, and I would add a comment so that you do not need to pass a second time, and that is that I still feel very uncomfortable that variations in the value of derivatives that are basically the hedging instruments of banks have a direct impact on the equity of those banks under some conditions, but this is only a comment and not an invitation to discuss these matters.

Well, thank you very much, indeed. This was a most interesting and thought-provoking talk, and I am sure that we will have in mind your three basic lessons, which I have well understood. The first is that we do need some form of systemic models of default because this is a matter that has to be more rationalized. We have done it in the macrosector. We haven’t really done it so well in the microrisk sector.

The second basic idea is that more consistency—if not coordination, but I don’t like that word—in the fiscal treatment of crisis and lender-of-last-resort lending as a method of dealing with crises has to be established, which is difficult in an integrated world where the forces of the market are worldwide, integrated, global. Those who regulate are basically national, and that’s a very difficult task, and you were right to put the emphasis on that.

Your third point, which I think is most intriguing and interesting for this assembly, is, has the pendulum not gone a little bit too far toward capital adequacy and perhaps we should have more in mind liquidity ratios, which, as you like to say, help at least to get to the next weekend.

Well, thank you so much and applause for Charles.