My Lords, as a recently retired member of the Select Committee on Economic Affairs, I give a warm welcome to its report on banking supervision and regulation. I recognise in its clarity of language and cogency of argument and in the balanced nature of its recommendations the hand of the skilled drafter and expert adviser, as well as the assembled knowledge and wisdom of the committee.
This report, like others that the committee has produced, serves an important educational function: it should be required reading for all those who want to understand the origins of the banking crisis and the steps that might be taken to prevent such events in the future. However, its remit allows it to cover only part of what went wrong and what needs to be put right. Reform of the financial system has to be fitted into a larger framework of reforms covering the conduct of macroeconomic policy, the liquidation of the global imbalances between China and the United States, and, I would argue, the distribution of wealth and income in our society.
The report's discussion of the causes of financial fragility is excellent but, in my view, does not go quite far enough. The reason is that the report remains captive to what I would call the risk modelling paradigm. That is the belief, to quote from the FSA's Turner report that:
"The risk characteristics of financial markets can be inferred from mathematical analysis, delivering robust quantitative measures of trading risk".
If that were so, financial markets would never crash. The fact that they do must be due, therefore, to a failure in the transmission of information somewhere along the line. That leads inevitably to policy recommendations based on devising better methods of risk management.
It is true that the committee has doubts whether mathematical risk models can be applied in all situations. There is an important passage in paragraph 62, which rightly says that statistical models are no substitute for judgment based on analysis. In paragraph 36, the report recognises that mathematical forecasting models rely on parameters which are assumed to be fixed but are actually shifting. In paragraph 39, we are told that supervisors should actively question the assumptions underlying bank risk models. In fact, the chief assumption that needs questioning is that those models can give a robust measure of risk.
The truth is that the risk modelling paradigm makes claims far outside its proper domain. It fails to take seriously Keynes's distinction between risk and uncertainty. That is very important for all clear thinking about both the causes of the crisis and what changes in the structure of the banking system are needed. What will average house prices be in 10 years' time? Is any noble Lord willing to bet on a number? Is any other noble Lord willing to insure him against the bet being wrong? Our risk management models are constantly making the assumption that we can attach precise numbers to those contingencies.
It is for those reasons that George Soros has called for credit default swaps to be banned. The report does not go that far; it simply suggests that the credit default swap market should be made more robust. The main problem is not that those trades are opaque, but that the insurance offered was phoney, because it was not based on actuarial principles. We should not allow firms to offer insurance against uninsurable events. That is a species of fraud.
In their response to the report, the Government wanted more effective stress testing,
"which allows firms to assess the impact of more extreme events ... not captured by traditional risk management models".
However, that also presupposes the possibility of giving numbers to extreme events, and that is just not correct.
Large consequences for bank supervision and regulation follow from taking uncertainty seriously. As noble Lords will know, there is a big debate going on, which finds an echo in this report, on whether countering financial fragility requires the separation of different forms of banking along the lines of the Glass-Steagal Act of 1933, or whether it can be accomplished solely by what is called macro-prudential supervision. The risk management school naturally favours the second. It proposes increasing the buffers of capital and liquidity to guard against high-risk events. Those might include contracyclical capital ratios—the requirement that capital ratios go up in good times and down in bad times.
Of course those measures will do good, they will reduce the fragility in the system, but they are designed to leave the structure of banks untouched and to give them greater capital cover for the same set of activities, without restricting their right to engage in those activities. What is more, they assume that the size of the buffers and the timing of their variants can be measured, whereas uncertainty is unmeasurable.
For example, they assume that the regulators will be able to discover exactly at what point in the cycle they are. Does anyone really believe that? First, it presupposes the existence of the cycle. Secondly, it presupposes knowledge about the length of the cycle. Thirdly, it presupposes knowledge about where we are in the cycle. All that is supposed to be gathered in some central regulatory agency and then instructions issued to the banks on the basis of the figures. That is a completely utopian view. What is more, banks will easily be able to game such regulations. They are very clever at doing that.
Anyone who takes uncertainty seriously is bound to favour some substantial break-up of the banking system into different sectors. The logical line of division should be between commercial banks and investment banks. The noble Lord, Lord MacGregor, quite rightly said that the report states that we should think about these things seriously and not just dismiss them out of hand as being impossible to achieve in a global economy. Many arguments in the report cry out for this kind of separation. For example, conglomerate banks, which are very big banks, have become too complex to manage. Everyone wants better non-executive directors but, as was also said by the noble Lord, Lord MacGregor, the complexity of the business makes it difficult for boards to understand all its aspects. Then let us take pre-funded insurance levied on the size of contributions. That runs up against the difficulty of measuring the riskiness of different classes of loan business. There are many problems with the strategy of simply leaving the structure of the banks unchanged and making them hold more capital and liquidity. This deserves serious thought.
I shall end on this note. The City of London performs a valuable service for the economy in this country. It is also its most powerful economic lobby, and like all lobbies, it tends to rate the value of its services higher than they are worth, a point recently made by the noble Lord, Lord Turner. This is a problem of long standing. It may amuse your Lordships to realise that Keynes wondered, way back in 1913, so before the First World War,
"how long it will be found necessary to pay City men so entirely out of proportion to what other servants of society commonly receive for performing social services not less useful or difficult".
Plus ça change, plus c'est la même chose. At least a time of crisis offers the Government a chance to reform the City in the light of Winston Churchill's remark, made when he was Chancellor of the Exchequer in 1925, that he,
"would rather see finance less proud and industry more content".