Chaos theory of currencies

What makes currencies move? You would think that with so much experience, we would be quite good at understanding the answer to this question by now. But I've been sitting at a screen this week looking at financial news agency copy on the tumbling dollar, the "juggernaut" German mark and the "all-time lows" of the franc and pound, and I'm confused. The banal dealers' comments that agency writers use to fill endless pages reflect something strange about our knowledge of foreign exchange markets.

The most common failing is simply to use a redescription of something as an explanation for it. Some comments illustrated a division between dealers who believed the dollar problems were caused by a strong mark and others who believe the strong mark had its roots in a weak dollar. More interestingly, they are apparently inconsistent with the most fundamental principle of economic analysis of markets. For economists it is "irrational" to believe you can predict a change in prices without believing it will be in response to new information.

For the economist, the Japanese financial year-end could not have an impact on the value of the yen, as many have argued. If if did, someone less intelligent than Nick Leeson could anticipate the coming end of the year, and buy yen in advance.The Kobe earthquake, in contrast, could have an effect on the price of the yen, but that would be because only someone a great deal more informed than Nick Leeson could ever have anticipated it.

The market participants, however, have never taken economists too seriously. They see there are such things as psychologically important levels. They know that when the franc falls below 3.55, instead of thinking "how cheap francs are, I should buy some more," people might easily think the reverse.

Economists long ago conceded that markets do not operate exactly as efficient markets theory suggests. They acknowledged that there could be speculative bubbles in markets, in which more or less rational market participants could knowingly continue to trade while prices deviated from "fundamental value". You could have "pyramid selling", near worthless items selling for high prices, merely because the buyer believes someone else will buy at an even higher price.

Bubble-like behaviour - which is seen as rational by some, and irrational by others - is invoked to explain crashes and volatility in markets. And it certainly seems to be doing its fair share of work this week. When the plethora of explanations offered in the media for different market events is so confused, and thin of substance, you can be sure that it is not the fundamental values which are driving the changes. Changes in the fundamentals are normally relatively easy to agree upon.

It is when the fundamentals are not so obvious, the task becomes harder: to predict what others are thinking. In that event, dealers grapple for a standard: an argument that is useful not because it describes reality, but because everybody else accepts it. It is similar to battles we observe in consumer electronics, when a product is useful not so much because it is better than its rivals, but because everybody else uses it (so you can get the software).

We know that standards evolve in unsettling jumps rather than incrementally. We lurch from vinyl records to CD; we don't stop at in-between stations as we once did in the evolution of motor cars. That is because it is hard to displace a standard once it is in operation.

One way of looking at recent market instability is as a collapse of an old standard - the old relative values of the mark, dollar and other European currencies, accompanied by an unsettling jump towards a new standard. The old rates looked wrong; once the perception became widespread, they dismantled themselves. The new trading ranges have yet to settle.

The analogy does not offer a clear vision of which way markets will turn. One view is to argue that if bubble theory is relevant, and the concept of standards is useful in understanding the changing views of the dealer community, then the markets have probably overreacted in recent days; bubble positions (in the mark) have accumulated. They will fall back as fundamentals reassert themselves. But, equally, they will not fall back to where they were. The old range has passed.

At the risk of sounding like a dealer, poised to be quoted in some Reuter copy, in my view the worst may not be over, but an improvement in dollar fortunes - at least relative to the mark - will probably be forthcoming before long.