Intellectual errors at the heart of crisis21 April 2011
At its heart, the credit crunch was fuelled by a deep intellectual error. It was this error which encouraged banks to take such a naïve and flawed view of the double or triple credit boom of the last 15 years. Rather than being something that regulators knew was wrong but did not have the power to stop this error was not just shared by regulators but actively promulgated.
Worse still, as accountancy standards, national and then finally international regulatory standards committed the same error, regulatory agencies actively compounded the problem and all but forced all banks to go down the same, flawed, path with the same, flawed, models and assumptions. What was this error and how did it bring the world's financial system crashing down?
Over the last ten years the unprecedented supply of debt helped drive "yields" to historically low levels. So great was the supply that there was little price for risk. Banks responded by steadily increasing the degree of risk on their balance sheets. Visible as a trend from the 1990s, this rise became even more rapid from about 2001. This can be measured in leverage ratios or in other words the ratio between the equity that investors have invested in a bank and the volume of loans that banks were making to companies or individuals. Clearly the greater the leverage ratio, the greater the risk of equity holders losing their shirts should loans start to go bad. In 2001 the five largest investment banks all had leverage ratios of between 10 and 25 (i.e. they had lent or had trading book exposure to 10 to 25 times the size of the equity that investors had invested in their companies). By 2007-08 only two of the same banks had leverage ratios below 20 and others were between 30 and 100.
In fact, the rise in the level of risk that banks were taking was even greater than that. Banks were able to use loopholes in international accountancy rules to place some risk in quasi-separate legal entities commonly known as Special Investment Vehicles (or SIVs). Banks ultimately underwrote these entities but for reporting purposes in the heady days pre-2007 they were not visible on bank balance sheets. Loans held in SIVs rose from $100bn in 2003 to $300bn by 2007. In short, banks' search for returns in the low yield environment meant that they were taking even more risk than was visible in their annual returns.
Even worse than this were discrepancies in risk methodologies that meant that the capital held against risks taken by banks' newer trading business was starkly less than the capital held against risks taken by banks' more traditional corporate banking business. (Trading businesses buy, sell, short or hedge securities which are tradable on capital markets. Corporate banking is the business of advancing loans to companies to help them expand or meet their working capital needs). The Bank for International Settlements estimated that under the old regime a bank with 57 per cent of its risk on its trading book might only need to hold four per cent of its capital base against that risk.
This meant not just that banks were undercapitalising. It also meant that they were able to arbitrage their own balance sheets by using securitisation to recycle assets from their capital-heavy loan books to their capital-light trading books. This is the economics of the looking glass or the Cheshire cat. However in a way this is to detail the problem rather than explain it. Banks stuffed their businesses with an unprecedented degree of risk in order to maintain their income. But why were they so foolish? Why did boards and senior managers allow the risk on their balance sheets to build up to such historically high levels? Why were trading books permitted to require such shockingly small levels of capital? In part, no doubt, because Basel I told them that they could and investors therefore continued to have confidence in banks that were becoming riskier and riskier. Three years ago most banks held a pretty good cushion above the minimum capital required by the Basel accords. How could they not be safe? But again, this is narrative not explanation. Why were regulators just as foolish as bank directors?
The ultimate answer is intellectual. In the quarter century before 2007 the so-called efficient market hypothesis became the bedrock of the risk management approach that banks and regulators took to their industry. This held that the price of a financial asset would always reflect all pertinent information that was relevant to its value. Market prices and the volatility of those prices can therefore be used to understand the risk inherent in that financial asset. This theory says that if a corporate bond is expensive with a price that does not move around much then it must ipso facto be a low risk investment. Or that if the price of a share is too low, well-informed investors will all but immediately spot the opportunity and make a fast buck driving it up to its "correct" price.
Upon this approach were based the myriad of financial models and calculations that were used to estimate the price and required capital for a wide range of complex financial products. The best known was the value at risk framework (VAR) which calculated how much capital investors should set aside for holding complex assets.
Unfortunately this whole intellectual edifice is built upon a fundamental misunderstanding of free markets. Markets are the best, the most dynamic means of discovering information about a product or service. It does not follow, and does not need to follow, that they always and in all circumstances "know" the price or risk of a product. Markets are fluid. Things are discovered, forgotten, misunderstood. As recent research into the impact of behavioural psychology has redemonstrated, homo sapiens is categorically not, never has been and never will be a pure homo economicus. He is influenced by less rational fears, prejudices and hopes. He tends to extrapolate out too confidently from recent events - exaggerating booms and deepening busts. To assume that the market prices, therefore, are a sort of platonically perfect reflection of reality, and indeed the future, at almost any given time is to subvert the very concept of the market in a curiously statist fashion. Rather than being something perfect and controlled the market is a mess. It is the best mess we have to hand but that does not mean that it is something perfect which can be taken at any point in time to reflect a statistically confident view of future risks and price movements.
But that is precisely what bankers and risk professionals tried to do. The value at risk framework calculated capital requirements by assuming that future losses would be like losses in the past adjusted for the current price and that two independent investments could not have correlated losses. This meant that apparently readily-tradable assets (such as those bought by a bank trading desk) required in principle far less capital to protect than less "liquid" or tradable assets such as "traditional" loans to companies.
However, when the crunch came, none of these assumptions held good. Past losses were too low as they reflected the period of the cheap debt-fuelled credit boom. This proved to be no guide to the future. Nor did apparently unconnected assets remain uncorrelated at the moment of market stress when liquidity drained out of the system. This was particularly the case for re-tranched and "re-packaged" debt sold via securitisation where the underlying risk was too opaque.
Previously tradable assets ceased to be tradable at any price. People panicked and the market behaved, by the standards of the market-efficiency framework, in an unpredicted, overly-correlated and irrationally fear-led fashion. But that was reality. And it was a reality that a longer understanding of the past than the last ten quarters of expected loss figureswould have confidently predicted.
(Indeed it was a reality that an understanding of securitisation and ratings methodology would also have predicted. Many senior notes -- the least risky investments with the highest AAA ratings-- only obtained their low risk rating because managers were obliged to wind up a Special Investment Vehicle when losses started. This just caused a glut of sales from many SIVs at one time and a further collapse in prices.)
The culmination of this process was in the run up to 2007 when Basel II and the national rules which enforced it further institutionalised these errors. This gave banks the opportunity to replace the simple and rather arbitrary risk weightings of Basel I with finer calculations of risk capital based on the same, flawed, approach that the banks and ratings agencies were using. The message to banks was simple. Demonstrate that you can use the "latest" VAR-style capital allocation models and you can reduce your required capital. (And, as a consultant, I worked with several institutions worried that if they did not move fast enough to so-called Advanced Basel II they would be at a relative disadvantage compared to their more "sophisticated" competitors).
In short, real depth or breadth of thought were not just discouraged but actively disincentivised. The most up to date economic thinking (the behavioural economists who were questioning the "pure" efficient-markets hypothesis) was rendered irrelevant as banks were encouraged to engage in a lemming like race to subscribe to VAR style risk management and take down their capital levels - a race it must be said in which they were only too happy to engage.
Hardly surprisingly therefore, capital levels fell further and leverage ratios rose yet higher to ever giddier heights as Basel II or similar national systems were anticipated and then promulgated prior to 2007. Leverage ratios for almost all banks were never higher than in 2006 and early 2007 months or weeks before the crisis exploded. It is not that regulators would have stopped the madness if only they could have done. They could have done. But they not only made the same error as the bankers and the risk professionals. By institutionalising the flawed approach to risk management through the Basel II accords they compounded the problem and further encouraged bank directors in their insane confidence in their own ability.