Long - Horizon Exchange Rate Predictability?

The importance to applied economists of having statistical tools that can reliably test for the presence of long-run predictability in time-series data can hardly be overstated. One such technique has gained prominence because of its apparent success in uncovering long-run relationships in international financial data. Recently, Mark (1995), Chinn and Meese (1995), and Bauer (1995), applied the long-horizon regression approach to investigate whether economic fundamentals have predictive power for exchange rates. The regressions employed by Mark and Chinn and Meese contain an error-correction term for spot rates and monetary fundamentals. Their analysis thus assumes that spot rates and fundamentals cointegrate. This paper uses the bivariate error-correction model that is implied by this assumption to show that little may be gained from long-horizon regressions. In particular, if the slope coefficient of the one-period-ahead regression is zero, then the slope coefficients must be zero for all horizons; on the contrary, if the estimated one-period-ahead coefficient is not zero, then estimated coefficients increase as the horizon increases. This paper further explores the conjecture that the existing evidence of long-run exchange rate predictability may be an artifact of the statistical technique. It shows that if spot exchange rates were independent of economic fundamentals, then long-horizon regressions would behave like spurious regressions. As a result, the finding of increasingly strong relationships at long horizons may not be evidence of an economic relationship. Several simulation experiments are conducted in which two independent time series are generated, one modeled after quarterly exchange rates and the other after monetary fundamentals as in Mark and Chinn and Meese. The results confirm that several diagnostic statistics, such as t-statistics and , are increasingly biased away frm zero for longer horizons of interest. Finally, the paper finds that even the graphical evidence of strong predictability over long horizons is misleading.