Covered interest parity: The untestable hypothesis

Although post Keynesian economists advocate the realistic bankers' view of the forward exchange rate, neoclassical economists formulate (CIP) as a testable hypothesis. In reality, CIP represents a formula used by bankers to calculate the forward rates they quote to their customers. This article provides arguments for the post Keynesian view of the forward exchange rate and suggests that CIP is not a theory, that it is a microeconomic relation, and that it is a hedging rather than an arbitrage condition. An empirical illustration shows that deviations from CIP are observed whenever published data are used, but these deviations disappear when transaction data are used instead. It is concluded that CIP is an untestable hypothesis.