THE EFFICIENCY OF FINANCIAL MARKETS

The efficient market hypothesis has been around for some time. There are various forms of it, weak, semi-strong and strong, but it basically says that the price of an asset reflects all information available about that asset. Put another way, unless you know something that the markets don't know, there is no point expecting to achieve above-average returns in the long run from your investment. The best strategy would be to put your money into a fund which simply tracked the market, and went up and down with it.

There have been modifications on the theme. Rudiger Dornbusch's famous "overshooting model" (once voted by the Economist as one of the ten most influential economics papers of all time) demonstrated that the foreign exchange market could overshoot, even with perfect information, instantaneous adjustment and rational expectations, if other prices (in his case, domestic price inflation) were sticky. Olivier Blanchard demonstrated much the same thing, this time with a domestic stock market and domestic price inflation. And Stanley Fischer got the same result with wages and domestic price inflation, if wages were constrained within 2-year contracts but prices were not. However, none of these disowned the hypothesis as such; what they said - in essence - was that markets are related, so inefficiencies in one can spill over into, and affect, ones which a priori are efficient.

More recent research has tended to concentrate on the informational aspects. Whatever the theory, it is a fact that some people know more about some things than others. Knowing that, people adjust for it (in the same way as they adjust the prices of second-hand goods in order to take account of the fact that they might be buying a "lemon"). However, while for second-hand cars this informational aysmmetry might be large, decisive even, the consensus seems to be that for big, publicly traded markets, it is much smaller. Indeed, it may well be reducing, owing to the power of the internet, and increased disclosure requirements. Notwithstanding the likes of Enron and Madoff, the hypothesis still holds.

But the financial fall-out from the B.P. oil spill is putting the hypothesis severely under strain, in my view. Nobody knows what the eventual costs of the clean-up will be. The Economist looked at worst-case scenarios and came up with a figure of US$20 billion for costs and damages, and US$17 billion for fines. US$37 billion is a very large number. But what is often forgotten is that B.P. is a very, very large company; and to use a trite mathematical fact, very, very large is very large in comparison with very large. In an average year the company generates US$20-30 billion in cash, so we are talking about a couple of years' cashflow max. Yet since the well erupted, the value of B.P. in the stock market has fallen by US$89 billion; even allowing for general market falls, and the fact that B.P. only had a 65% stake in the well, its value has still fallen by about US$65 billion. Much, much more than the worst-case scenarios.

Why? All of the figures above are known to the world (or, at least, that part of the world who deals with this sort of thing). Shouldn't every fund manager worth his salt therefore be ploughing into B.P. shares as fast as they can, on the grounds that the stock is ridiculously cheap? After all, that is what the efficient market hypothesis would suggest should happen. But it isn't happening.

I think the answer lies in the informational asymmetry outlined above. In this particular case, the asymmetry refers to what the U.S. Government and Congress will do. B.P. is currently America's most hated company, and it is clear that a lot of people, from the President down, want to hurt it. Nobody outside of the Government and Congress yet knows what that hurt will be, though there is a lot of talk about retrospective fines for economic loss (currently capped at US$75m by the legislation introduced post the Exxon Valdez disaster). It is significant that on the day that B.P. visited the White House and announced an agreement to set up a US$20 billion escrow account to cover the costs of the clean-up, an account moreover which would be administered by the Government rather than B.P. itself, the company's share price rose by 7%. Better the devil you know (US$20 billion) than the devil you don't (a much higher, and unknown, figure).

So, our efficient market hypothesis can survive; B.P. is a special case. The problem I have with this analysis is that in today's straightened times, nearly all publicly traded assets are, or could be, affected by Government actions; indeed, they are increasingly being so affected. Banks are nationalised instead of going bust; the price of money is kept artificially low through quantitative easing; wage and pension entitlements are being changed; special levies and corporate taxes are being imposed on financial and mining companies; energy policies are being adapted in the face of climate change. And so on and so forth.

If every company is, or could be, subject to aysmetrical information constraints such as those above, then what is the point of the efficient market hypothesis? You could be forgiven for saying "not much".

Walter Blotscher