Determinants of Exchange Rate: Evidence from Sri Lanka

Abstract

This present research investigated the modelling of exchange rate volatility of USD/LKR and analyses whether macroeconomic factors could influence the exchange rate. A combination of Autoregressive Integrated Moving Average (ARIMA) and Autoregressive Conditional Heteroskedasticity (ARCH) family models were used to model the exchange rate volatility. The most suitable model was determined based on the maximum likelihood criterion. To explore the presence of dynamic short-run and long-run relationships, and the impact of macroeconomic variables on the exchange rate was analyzed using the ARDL model. The empirical findings indicate that the most appropriate statistically significant model for volatility is ARIMA (1,0,0)- ARCH (1). The ARDL model suggested that a long-run relationship between the macroeconomic variables and the exchange rate does not exist. In contrast, a short-run relationship exists between exchange rate lag one, exchange rate lag two, inflation, Merchandising trade balance. Thereby, we suggest that improving the merchandising trade balance and minimizing inflation could minimize volatility in the exchange rate. The practical implications inferred from this study could influence all stakeholders exposed to foreign exchange volatility, including policymakers, importers, exporters, and financial institutions. The contribution of this research considered the most recent economic phenomena of Sri Lanka and used Gross official reserve as a variable that was not used earlier in Sri Lankan studies.