Exchange Rates
Do Markets Use News?

A market is said to be "efficient' if prices reflect all the information available to participants in that market. This hypothesis has been used to explain the volatility of exchange rates over the last decade. If foreign exchange markets are efficient, then these fluctuations are merely the market's response to the arrival of new information concerning the economy or government policies.

How can the "efficient' markets hypothesis be tested? One commonly applied test is based on the behaviour of the forward exchange rate, the rate at which foreign currency can be bought or sold today for delivery in the future. Suppose we consider the market for sterling to be delivered one month hence. If the market is efficient, the forward rate today and the spot rate one month from today should differ only because of new information which has become available to the market in the intervening month. If this is genuinely new information, it cannot be forecasted on the basis of today's information. The future spot rate will therefore equal the forward rate plus a random and unpredictable "shock'.

Empirical investigations of the efficient market hypothesis using this approach generally lead to a weak rejection of the hypothesis. These tests based on the forward rate are only strictly applicable if market participants are "risk-neutral'; that is, they are not too concerned about risk. We may therefore be unsure why the efficient markets hypothesis has been rejected; it may be because agents are risk-averse or because markets do not make full use of available information.

In Discussion Paper No. 53 Research Fellow Charles Bean attempts to assess the relative importance of each factor in the rejection of efficient markets. Bean demonstrates that the difference between the forward rate and the realised future spot rate can be decomposed into two components - a risk premium and a term due solely to new information. He also shows that it is possible to place a lower bound on the exchange rate variation due to movements in the risk premium. The analysis suggests that changes in the risk premium are at least as important as new information in explaining movements in the sterling/dollar, franc/dollar and lira/dollar rates. Only for the DM/dollar rate does "news' explain more than half the exchange rate variation, and even in this case nearly forty per cent of the variation could be attributed to the risk premium. Bean concludes that new information may play a smaller role in explaining exchange rate movements than was previously thought, and that departures from the efficient markets hypothesis are quantitatively significant.


Exchange Rates, Risk Premia and New Information: A Note
Charles R Bean

Discussion Paper No. 53, February 1985 (IM)