The evolving patterns in economic volatility: a comparative regime analysis of pre- and post-crisis period

Abstract

Economic volatility modeling is a very crucial part of statistical finance and computational economics, with significant implications for risk management and policy formulation. This study presents a comparative analysis of the Nigerian Stock Exchange (NSE) All-Share Index across two distinct phases: a pre-crisis era and a post-crisis period characterized by significant macroeconomic and global disruptions. Employing the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) families for volatility estimation and a Markov Regime Switching (MRS) framework for regime identification, the research captures both continuous volatility dynamics and discrete shifts between high- and low-volatility states. Data preprocessing involved return computation, stationarity testing, and heteroskedasticity diagnostics. Findings shows that, a low mean return in both the pre- and post-crisis periods, confirming the absence of substantial average daily gains. However, standard deviation increased from  (pre-crisis) to  (post-crisis), pointing to heightened volatility in the latter period. The post-crisis results for GARCH (1, 1) model show  and, implying slightly longer-lasting shocks compared to the pre-crisis values of  EGARCH results highlight asymmetric volatility effects, with  negative in the post-crisis period, indicating that negative shocks increase volatility more than positive ones. TGARCH results similarly confirm leverage effects. While, MRS shows that the transition probability matrix reveals that the probability of remaining in the low-volatility regime dropped from  (pre-crisis) to  (post-crisis), indicating heightened volatility persistence and extended high-volatility regimes in the post-crisis period, emphasizing the structural impacts of global shocks. The study contributes a replicable statistical-computational framework for volatility analysis in emerging markets.