MODELING EXCHANGE RATE VOLATILITY IN PAKISTAN: AN APPLICATION OF ARCH, GARCH, GARCH-M AND EGARCH

Abstract

This study examines the dynamics of exchange rate volatility in Pakistan by employing autoregressive conditional heteroscedasticity (ARCH) and its extensions generalized ARCH (GARCH), GARCH-in-Mean (GARCH-M), and Exponential GARCH (EGARCH). Using monthly data from January 2001 to September 2025, the research investigates the persistence, clustering, and asymmetric behavior of volatility in the Pakistani foreign exchange market. The results from ARCH and GARCH models confirm the presence of volatility clustering, where current volatility is significantly influenced by past shocks and variances. Findings from the GARCH-M model highlight that volatility itself negatively affects exchange rate returns, suggesting a risk-return trade-off. Moreover, EGARCH results demonstrate significant size and sign effects, revealing that large shocks contribute disproportionately to volatility and that good news impacts exchange rate stability more strongly than bad news. Overall, the study provides robust evidence of persistent and asymmetric volatility in Pakistan’s exchange rate market, carrying important implications for policymakers, investors, and risk managers in designing effective hedging and stabilization strategies.

Keywords: Exchange rate volatility, Autoregressive Conditional Heteroskedasticity, ARCH, GARCH, GARCH-M, EGARCH