Tests of the uncovered interest parity

Abstract

This paper carries out empirical testing of the Uncovered Interest Parity for US-Mexico, US-Brazil and US-Japan using general OLS and GARCH from monthly data. Similar to numerous other studies UIP failed to hold empirically. I also test if deviations from UIP are in any way effected by business cycles but did not find any supporting evidence. In contrast to a number of other studies my slope coefficient was significantly different from unity. The coefficient also showed a negative sign for one of the economies. Additionally, there were presence of ARCH and GARCH effects in UIP deviations. Finally, no evidence was found for UIP to hold better for developed nations like Japan and not for emerging markets like Mexico and Brazil.