ABSTRACT
In C hapter 10 the unpredictable nature of exchange rate changes was highlighted. Hence even if the extant evidence on the EM H had unambiguously dem onstrated that the forward rate is an unbiased forecast of the future spot rate, it is unlikely to be a particularly good forecast of the future spot rate in periods in which new information is important: although the EM H implies that antici pated changes in the exchange rate will be orthogonal to the forecast error, unanticipated changes in the determinants of exchange rates will be correlated with e, in
that is, the ‘news’ represents the update of agents’ expectations. The ‘news’ view of the determ ination of foreign exchange rates
would seem to have wide appeal. For example, the financial columns of the daily press abound with headlines such as ‘unexpectedly good money supply figures result in an appreciation of the exchange ra te’ and ‘an unexpected deterioration in the current account led to an exchange rate depreciation’. If new information is important in foreign exchange markets then it is perhaps more appropriate to empirically implement exchange rate models, such as the m onetary and portfolio approaches, in a news context rather than regressing the exchange rate on the levels of, for example, relative money supplies. As was demonstrated in C hapter 9 this latter approach has not been particularly successful due mainly to problems of misspecification.
