Mr Charles Bean, Deputy Governor for Monetary Policy and Member of the Monetary Policy Committee, Bank of England, delivered the Schumpeter Lecture at the Annual Congress of the European Economic Association, Barcelona, 25 August 2009.
Selections from the text of the lecture published by the Bank for International Settlements:
The first distortion…
Creditors – especially if they are households rather than
sophisticated financial market participants – may not even factor in the implications of higher leverage for the possibility of default. And even if they do, the debt may be partially or wholly underwritten by the state, with the cost of the insurance only imperfectly passed back to the bank. Similarly, the bank may be thought to be too important to be allowed to fail, in which case people might expect an injection of capital by the state to make good abnormal losses. In any of these cases, there is an incentive for the bank to raise leverage. Moreover, the lower is the perceived uncertainty associated with the loans, the more the bank can afford to leverage up, while maintaining the same uncertainty over the return on its capital. So the environment of the Great Moderation would have been particularly conducive to intermediaries increasing the leverage of their positions.
The second distortion…
Moreover, a considerable amount of the remaining risk was contained in institutions which, while not formally recognised as banks, engaged in exactly the same sort of maturity transformation, financing long-term assets by short-term debt instruments. These included entities such as conduits, which housed the securitised loans and then financed them by selling short-term paper. But in many cases these entities had back-up credit lines to the supporting bank, so that when funding difficulties arose, the securitised loans in effect came back onto the bank’s balance sheet. And even where there was no formal obligation to act as a lender of last resort, originators often chose to provide back-up finance in order to protect their name in funding markets.
The motive for setting up these off-balance-sheet entities was entirely one of regulatory arbitrage. Off-balance-sheet vehicles were not required to hold capital in the same way as a bank would if the loans were on their balance sheet. So it appeared to be a neat way to boost profits without having to raise more capital. The Banco d’España, the Spanish banking supervisor, insisted that Spanish banks would have to treat conduits and the like as on balance sheet for capital purposes. As a result, Spain did not see the mushrooming of these off-balance-sheet vehicles.
And the third distortion…
One unintended consequence of financial innovation was that it enabled clever traders to create positions with considerable embedded leverage – that is, portfolios requiring little payment up front, but whose returns amplified changes in the value of the underlying assets. Traders then had a natural incentive to gravitate towards these types of highly risky instruments.
A related problem is that it is extremely difficult for management to observe the risk being taken on by their traders, particularly when innovative financial instruments have unusual return distributions. Take, for example, a deeply out of the money option. This pays a steady income premium and has little variation in value when the underlying instrument is a long way from the strike price, but generates rapidly escalating losses in bad states of the world. In good times this looks like a high return, low risk instrument. Only in very bad states of the world do the true risks taken on become apparent.
Knightian uncertainty about CDO returns increased information problems:
A typical CDO comprises a large number and variety of RMBS, including a mix of prime and sub-prime mortgages from a variety of originators. On the face of it, this might seem like a good thing as it creates diversification. However, even more than with plain vanilla RMBS, it becomes impossible to monitor the evolution of the underlying risks – it is akin to trying to unpick the ingredients of a sausage. That may not matter too much when defaults are low and only the holders of the first, equity, tranche suffer any losses. Holders of the safer tranches can in that case sit back and relax – a case of rational inattention. But once defaults begin to rise materially, it matters a lot what such a security contains. And with highly non-linear payoffs, returns can be extremely sensitive to small changes in underlying conditions.
When defaults on some US sub-prime mortgages originated in 2006 and 2007 started turning out much higher than expected, there was a realisation that losses could be much greater on some of these securities than previously believed. And a growing realisation of the informational complexity of these securities made them difficult to price in an objective sense. Essentially, investors switched from believing that returns behaved according to a tight and well-behaved distribution to one in which they had very little idea about the likely distribution of returns – a state of virtual Knightian uncertainty (Caballero and Krishnamurthy, 2008).
So who’s to blame?
First, in my view it would be a mistake to look for a single guilty culprit. Underestimation of risk born of the Great Moderation, loose monetary policy in the United States and a perverse pattern of international capital flows together provided fertile territory for the emergence of a credit/asset-price bubble. The creation of an array of complex new assets that were supposed to spread risk more widely ended up destroying information about the scale and location of losses, which proved to be crucial when the market turned. And an array of distorted incentives led the financial system to build up excessive leverage, increasing the vulnerabilities when asset prices began to fall. As in Agatha Christie’s Murder on the Orient Express, everyone had a hand in it.