The puzzle of efficiency is one of the unresolved but still challenging issues in financial economics: this should come as no surprise since inefficiency implies the existence of unexploited profits in capital markets. Today there is growing evidence that foreign exchange markets have, to a large extent, acted inefficiently. It is claimed that forward rates contain very little, if any, useful information about the future course of spot exchange rates.

The efficiency hypothesis in the foreign exchange market is re-examined by Research Fellow Nicos Christodoulakis, Sarantis Kalyvitis and Nicos Karamouzis in Discussion Paper No. 1016. The traditional efficiency testing equations are reviewed and a model is developed that incorporates Bayesian revisions in the form of devaluation expectations. A number of propositions regarding the pattern of the coefficients in efficiency testing equations are established. The results are confirmed by the empirical estimation of the model for the Greek foreign exchange market. In each period the monetary authority is supposed to set a target for the exchange rate which is (directly or indirectly) known to agents. The results indicate that past devaluations play an important role in the formation of expectations. Therefore, simple efficiency tests may reject the efficient market hypothesis. It is suggested that when such a long memory process is dominant in the market, past devaluations should be included in efficiency testing equations.

Efficiency and Expectations Revisited: A Foreign Exchange Market with Bayesian Players
Nicos M Christodoulakis, Sarantis C Kalyvitis and Nicos V Karamouzis

Discussion Paper No. 1016, September 1994 (IM)