Modelling exchange rate volatility and tourism demand in Sri Lanka: A bivariate GARCH approach

Abstract

This study investigates the relationship between exchange rate volatility and tourism demand in Sri Lanka, focusing on India and the United Kingdom as major source countries over the period from 2015 to 2024. A thorough examination of the literature revealed that no similar research had been conducted in Sri Lanka. The main goal is to determine the extent to which volatility fluctuations between these two important economic variables occur and how changes in exchange rates affect the number of tourists. Sri Lanka’s economy depends heavily on tourism, and the performance is quite sensitive to external economic variables, especially changes in exchange rates. The study uses a comprehensive modelling framework that includes ARIMA, univariate GARCH, and bivariate DCC-GARCH models in order to capture both the deterministic patterns and the stochastic volatility in the data. Using the Augmented Dickey- Fuller test, the time series data for tourist arrivals and exchange rates (LKR/INR and LKR/GBP) were examined for stationarity and appropriate transformations were applied. In order to model short-term dynamics and trends, ARMA and ARIMA models were estimated for each countryspecific dataset. The adequacy and prediction accuracy of the fitted models were validated by diagnostic tests. The exchange rate series showed significant ARCH effects, which led to the modelling of volatility using GARCH (1,1) models. The exchange rate series showed persistent volatility clustering according to these univariate GARCH models, even if some of the variance coefficients were not statistically significant. However, these models served as a foundation for forming the bivariate DCC-GARCH (1,1) models, which successfully represented the dynamic conditional correlations between exchange rate volatility and tourism demand. The findings show that there are differences in the magnitude and direction of volatility fluctuations between India and the UK, which may be attributed to variations in tourist behaviour and economic sensitivity. The DCC-GARCH models improve forecasting and policymaking by offering insightful information on the co-movement of exchange rate volatility and tourism demand. The findings hold significant suggestions for policymakers, tourism planners, and businesses looking to reduce economic risks and enhance forecasting in the face of exchange rate volatility.