Be grateful the global financial crisis didn't hit sooner

Author John Legge Published 17 August 2009

Two of the crucial assumptions behind conventional financial economics have been challenged by the Global Financial Crisis.

The Efficient Market Hypothesis—roughly summarised as the assertion that every transaction in a market occurs at the best possible price for both the parties to it given the current state of knowledge—has been seriously wounded. Even former chair of the US Federal Reserve Alan Greenspan no longer believes in it.

But the Efficient Market Hypothesis was what provided the justification for the deregulation of financial markets over the last 30 years. The theory went that if every market transaction provided the greatest possible benefit to each party then a regulator (or even a public record of transactions in the case of a derivatives market) was an unnecessary cost.

The proximate cause of the current crisis was the creation of synthetic Collateralised Debt Obligations (CDOs), which were marketed as a safe-as-houses way to collect repayments from domestic mortgages. And just in case some house buyers failed to make their repayments, the loss would be made up with a Credit Default Swap (CDS) underwritten by the world’s largest insurer, AIG.

Unfortunately, most of the mortgages underlying many CDOs had been made to NINJA borrowers (no income, job or assets). The CDSs that were supposed to cover this contingency had been written by the London subsidiary of AIG. The AIG group took advantage of the fact that British regulators were even more somnolent and their regulations even less strict than US ones.

When the NINJA borrowers started to default the mortgage-backed securities lost much of their value and their various owners asked for their CDSs to be honoured. It then turned out that the aggregate liabilities AIG had underwritten were more than ten times its total assets and its guarantees were worthless. Subsequently, the US government has paid out more than a trillion dollars to prevent AIG’s failure from destroying the financial system.

For Unit 406 students the key message is that the Security Market Line (SML) and the Capital Asset Pricing Model (CAPM) are not divinely inspired Truths, but approximations—sometimes wildly incorrect approximations. Managers who use their company’s share price as a measure of the success or otherwise of their initiatives will certainly not make the best possible decisions and they may make some appalling ones.

A corollary of the Efficient Market Hypothesis is the assertion that it is 'impossible to beat the market'—that since the price at any moment incorporates all available knowledge about the underlying asset, price movements are random except in hindsight. If both the direction and magnitude of a price change are random the price will follow a random walk of some kind; by convenient rather than a logically derived further assumption, asset prices will follow a geometric random walk leading to a lognormal distribution of prices around the mean or the trend.

This convenient assumption underlies the Black-Scholes formula for calculating the value of an option and by extension the Dixit formula for estimating an appropriate hurdle rate. The crucial problem is the inconvenient fact that the real distributions of asset prices have 'fat tails'—most of the time the lognormal assumption holds; but a statistically significant number of times the price of an asset will move further and faster than expected.

The Six Sigma methodology is based on the assumption that events more than six standard deviations from the mean can be ignored, given that their probability, assuming a normal distribution, is about one in a billion. If a security had an average annual movement of ten per cent and its fractional movements were normally distributed a 60 per cent movement in a year would be expected once in a billion years.

Chebyshev’s Inequality shows that, when you drop the assumption that the asset price follows a geometric random walk, a Six Sigma event can occur as often as once in 36 trials. The Great Depression started 80 years ago: rather than being surprised at the current global financial crisis perhaps we should be grateful that it has taken so long to arrive.