We need to make the right sort of mistake

Why do markets make so many mistakes? It is one of the disagreeable questions that those of us who believe that markets are the best (or perhaps the least bad) method of allocating resources too rarely consider.

The question has been rammed home in the last few weeks with the collapse of several East Asian currencies and stock markets - a collapse on a scale that was perfectly easy to forecast with hindsight but that took the markets completely by surprise.

But, of course, that is just one example of many: there have been a host of other examples of market failure, from the over-valuing of Japanese equities and property at the end of the 1980s, to the undervaluing of sterling during the period between 1992 and this spring. The malignant impact of these errors varies. The overvaluation of Japanese assets was catastrophic and has hampered the recovery of the economy through the whole of this decade. The undervaluing of sterling mercifully turned out not only to be relatively benign, but compensated for the earlier error of joining the ERM (and hence having German interest rates imposed on Britain) by kick-starting the economy. But it was still a mistake, and imposed costs - including a billion-plus bill for the British taxpayer - which ought to have been unnecessary.

Like most apparently simple questions, though, it is really a multi-layered one. For there are different sorts of markets and different sorts of mistakes. At the moment, when everyone is aware that US equities are valued very highly by historical standards and therefore pondering whether there is a looming discontinuity there, it might be helpful to make some distinctions. But to give away the punch-line, I am going to argue that some mistakes are both necessary and healthy to the world economic system, but that some are not. In short, you have to have the right sort of mistake and not the wrong sort.

The starting point is surely to distinguish between market decisions on individual projects and big macro-economic financial flows.

The great advantage of market decisions on whether to back a project is that there are lots of different decision points. The newspaper you have in your hand is a good example of this, for the group which publishes it was not started by the decision of an executive committee of some giant publishing house that thought it had spotted a gap in the market. It was started by three individuals who then persuaded fund managers to take a risk. The money came from multiple decision points, each of which made an assessment of risk and reward - rightly, as it turned out, for when the ownership changed to its present status of being owned by two large newspaper groups, the original investors more than tripled their investment. This was judged as a decent, if unspectacular, return, because venture capital funds need winners to cover the inevitable cases where they lose their entire investment.

The point here is that the venture capital industry needs to make mistakes. Of course it doesn't like making them, but if it didn't make mistakes ventures that require capital would not get started. In any case, the system is self-purifying for people who consistently make mistakes are weeded out pretty soon.

When one moves from the particular to the general, the problems start. Markets, as practitioners always tell us, thrive on risk, and at the level of a project or an individual equity investment the string of mistakes is a sign that a market is doing its job.

Periodically, however, they make a general mispricing of that risk. When they realise this there is the sudden discontinuity, and that discontinuity can be damaging to the real economy. Why do they misprice? I suggest are three broad reasons: that the markets have been rigged and so have not been able to respond to signals; that investors simply get carried away by, in the words of Alan Greenspan last year, "irrational exuberance"; and (slightly different from the previous point) that the increased "professionalism" of the investment community is inhibiting the contrarians, so that markets become something of a one-way stream.

If markets are rigged there is an easy solution. Don't rig 'em. With hindsight it is easy to see that the surge in Japanese asset prices was the result of a rigged market, where valuations were pushed to extreme levels by financial manipulation. There have been some elements of this in East Asia, where banks have increased their lending without paying proper attention to the risks involved, expecting (rightly as it happens) that they would be bailed out by government if things went wrong. It looks as though Thailand's taxpayers may end up spending the equivalent of about 8 per cent of GNP on bailing out duff banks.

The second reason for mispricing is deeply rooted in human nature: history is littered with examples of investment manias, from the South Sea bubble, through the crash of 1929, to the one that everyone in Britain remembers - the house price boom of 1988. Because we do remember these things, we have some protection; wise central banks can lean against manias not just by warning against them but by jacking up the cost of money. One of the signs of health at the moment is the number of people who are comparing the present situation in US equities with the condition in 1987, though they are right also to note the differences.

We can live with reasons one and two; the one that concerns me more is number three, that markets have become too "professional", that no-one wants to step out of line. Have a look at the graphs. They show the latest IMF forecasts for world growth and inflation. They are a sensible, rational piece of work by competent, well-paid professionals. But they are going to be wrong, aren't they?

Take the growth forecast on the left. The obvious thing about the shaded area, showing growth of a bit over 4 per cent, is that it is completely different from the past experience where growth has swung from 1 per cent to nearly 7 per cent. It may well be that the amplitude of future swings will be lower than in the past, but to suggest that growth will be flat must be wrong.

Now look at the one on the right, on inflation. Again, this is what good professionals expect to happen, with inflation in the developed world perking up a bit and inflation in the developing world continuing to fall. That outcome is possible and certainly more likely than the flat growth projection on the left. But it is surely just as likely that the developed countries' inflation will push on down, following the secular trend since 1974, while that of the developing countries will perk up again.

In investment, as in forecasting, the same sort of consensus approach now dominates. Everybody does pretty much the same thing, for two reasons. The first is that everyone is judged against his or her peers rather than by absolute standards, so making the same mistake as everyone else is not a hanging offence, whereas making an individual mistake is seriously bad news. The second is there are strong fiduciary pressures on fund managers not to run unbalanced portfolios, so they could not get a fund out of line with the market, even if they wanted to.

And so there is a paradox. Because individually fund managers are constrained from taking a risk by betting against the market, collectively they take an even bigger risk because markets become one-way. We need more fund managers making mistakes and risking money in companies and indeed countries where others fear to tread. What we do not need is the systemic mistake where fund managers all pile into the same thing because there is apparent safety in numbers - which there is not.

Copyright: IOS & Bloomberg