On the link between the exchange rates and interest rate differentials in China: evidence from an asymmetric wavelet analysis

This paper adopts wavelet analysis to explore the time-frequency comovement and causality between the exchange rates and the interest rate differentials in China (compared to the USA) over the period from February 1999 to March 2018. While the existing literature in general treats the relationship between the two variables as being symmetric, this paper attempts to detect possible asymmetric patterns by extending the standard wavelet analysis to an asymmetric analysis. In addition, given that the relationship might be affected by exchange rate expectations and foreign exchange interventions, we further employ the partial wavelet tools to filter out the effects of the two controlled variables. The results show that the general pattern of the comovement and causality does not change. The comovement is intensified at the low frequency after the exchange rate reform in 2005. Moreover, exchange rates are found to positively comove with interest rate differentials, and the former leads the latter. More importantly, we provide robust evidence of meaningful asymmetry and substantial time and frequency variations in the comovement and causality between the two variables. These findings provide important implications for improving the transmission of monetary policy in China.