Exchange rate volatility and economic indicators: applications of a regime-switching linear regression model in the Kenyan context

Abstract

This study examined the relationship between USD/KES exchange rate volatility and key economic indicators—GDP growth, inflation, trade balance, and the central bank rate—in Kenya from 2010 to 2023 using a regime-switching linear regression model. It explored how these relationships varied across low, medium, and high volatility regimes, providing a more detailed analysis than traditional models. Results showed that in low-volatility periods, relationships were weak; in medium-volatility periods, only the central bank rate was significant; and in high-volatility periods, inflation had the strongest influence, with the central bank rate stabilizing the economy. The 3-regime model outperformed the single- and two-regime models in terms of AIC, BIC, and R-squared. Despite limitations such as fewer observations per regime, the study highlighted the importance of regime-specific policies for economic stability and provided valuable insights for policymakers and researchers.