The Yen/Dollar Carry Trade: Carrying the UIP Condition Too Far?

Abstract

The yen carry trade between the US and Japan has existed as a feasible investment strategy in direct violation of the uncovered interest parity (UIP) condition. Using yen-dollar spot and forward exchange rate data from 1993 thorough 2007, I demonstrate that the forward exchange rate quoted in the market in this period was based on the UIP formulation, and as such the investment strategy was based on favorable variations in the spot exchange rate. Empirical analysis using OLS regressions verified that the forward rate is not an unbiased estimator of the future spot rate, and demonstrated the failure of the UIP condition in this specific case. Adapting models from Frankel and Froot (1989), empirical analyses suggested the insufficiency of peso problems (systematically irrational expectations in the market) and time-varying risk premiums in explaining the failure of UIP. A comprehensive model incorporating monetary policy variables (interest rate smoothing, resisting rapid changes in exchange rates, and random policy influences), adapted from McCallum (1994), demonstrated that the UIP condition is actually not violated, and the observed bias of the forward rate as a predictor may be attributed to monetary policy response strategies by the US and Japanese governments. I find that the subsequent gap between predicted and actual future spot rates accounts for the existence of the yen carry trade as a viable arbitrage opportunity.