Currency Markets
Magic numbers?

Reports in the financial press routinely describe exchange rate movements in terms of `psychological barriers', which academic economists tend to dismiss, since principles of rationality and market efficiency dictate that the number with which an exchange rate ends should be irrelevant. In Discussion Paper No. 621, Research Fellow Paul De Grauwe and Danny Decupere test the efficient markets hypothesis for currency markets. If agents attach special importance to the last trailing digits of a currency's price, they are initially hesitant to pass this point, but once the `barrier is broken', the exchange rate movement accelerates through overbuying or overselling.

Classifying daily quotations of the dollar/DM and the dollar/yen exchange rates and their inverses during 1980-90 into 100 groups according to the last two digits, the authors find that there are psychological barriers at exchange rates of 130, 140, ... yen per dollar, etc., although the inverted exchange rates trigger no such effects. For the dollar/DM market, they find weak evidence of barriers at 2 or 3 Deutschmarks per dollar and no evidence for rates of 1.4, 1.5, 1.6 ... etc. (again, the inverted quotations have no impact). Currency markets are therefore inefficient, since irrelevant information does appear to influence agents' decisions and they select one method of quoting exchange rates rather than its inverse.
Professor de Grauwe will present this paper at an April lunchtime meeting, to be reported in more detail in the next issue of this Bulletin.

Psychological Barriers in the Foreign Exchange Market
Paul De Grauwe and Danny Decupere

Discussion Paper No. 621, January 1992 (IM)