Essays on international parity conditions

This thesis brings together a collection of essays on parity conditions in international economics: covered interest parity: covered interest parity: purchasing power parity and real interest parity. While each essay is an independent study of a particular problem area, there exists a common theme in that the set of parity conditions chosen for analysis is thought to be important in determining the short and long-nm behaviour of exchange rates. The justification for the study arises from two related issues. Firstly, as it is often assumed that exchange rates are determined in efficient markets, an analysis of international parity conditions may help us comment on the efficient markets hypothesis. We define efficiency according to Fama (1970), Where the market is said to be efficient if prices 'fully reflect' all currently available information. Secondly, models of exchange rate determination, within which the above parity conditions play a fundamental role, have exhibited a poor empirical performance in the recent past. An examination of the foundations of such models may therefore be helpful in allocating 'blame'. Of the four problem areas analysed ally covered interest parity was unconditionally accepted as a plausible assumption. Fran a possible 6330 potential arbitrage opportunities observed during the months of August and September 1987, only eight would have been profitable. Agents were efficient in terms of ensuring the forward exchange premium equalled the relevant interest rate differentials, subject to transaction costs. Some evidence however was found for the existence of a risk premium in the forward exchange rate during the 1920s, but attempts to model the premium as both a function of past forecast errors and as a latent variable, had limited success. We were therefore unable to verify the existence of risk averse speculative agents in foreign exchange markets. Purchasing power parity, analysed in terms of a theory of arbitrage for the period 1975 to 1980, using a recently developed econometric technique - cointegration - was rejected. This would imply that commodity arbitrage may be inefficient. A direct test of real interest parity using the bivariate vector autoregression approach, was also decisively rejected for the period 1979 to 1986. The observation that real interest rates do not fully reflect all currently held information suggests that the long term credibility of the European Monetary System may be suspect and that governments can influence national investment/saving decisions by intervention in domestic financial markets