The Importance of Interest Rate Volatility in Empirical Tests of Uncovered Interest Parity

Uncovered interest rate parity provides a crucial theoretical underpinning for many models in international finance and international monetary economics. Though theoretically sound, this concept has not been supported by the empirical evidence. Typically, econometric tests not only reject the null hypothesis, but also find significant slope coefficients with the wrong sign. Following the approach employed in Kool and Thornton (2004), we show that the empirical procedure conventionally used to test for UIP may produce biased slope coefficients if the true data-generating process slightly differs from the theoretically expected one. Using monthly data for ten industrial countries during the period W75-2004, we estimate the UIP relation for all possible bilateral country pairs for each of the six fiveyear sub-periods. The evidence supports the biasedness hypothesis: when the interest rate volatility of the anchor country is very high (very low), this estimation procedure reports significantly higher (lower) slope coefficients.