Do We Need Multi-Country Models to Explain Exchange Rate and Interest Rate Dynamics?

This paper examines characterizations of exchange rate and short-term interest rate dynamics, based on the implications of multi-country versions of the Cox, Ingersoll, and Ross (1985) class of term structure models. The countries considered are the US, Germany, Japan, and the UK. Our tests reveal that multi-country models are, in some cases, better able to explain the dynamics of exchange rates and interest rates than two-country and single-country models respectively. This is particularly true for the Japanese interest rate as well as the rate of appreciation of the Deutsche mark relative to the US dollar. Our inference is conducted using the small-sample distributions of test statistics, in addition to their asymptotic distributions