Structural Change and Diagnosis of Nepalese Stock Market Volatility

Abstract

Purpose: This paper studies the volatility of the Nepalese stock market. It focuses on two aspects of volatility: persistence and leverage effects. It aims to evaluate whether the presence of deterministic structural shifts in volatility produces biased estimations of persistence and leverage parameters.

Design/Methodology: This study uses time-series econometric modeling. The family of asymmetric GARCH specifications has been used to capture the leverage effect and persistence in the longitudinal univariate time-series indices data. The study covers 13 subindices along with a composite NEPSE index covering the period from 2003 to 2023.

Findings: Findings suggested that the estimates of persistence and leverage parameters are biased when deterministic structural shifts are not considered. Result indicated the downward adjustment of the estimated parameters of leverage and persistence when such structural shifts are incorporated in the models.

Conclusions: Failure to incorporate deterministic structural shifts are expected to produce biased estimate of volatility attributes of high-frequency financial time series data. A downward shift in persistence parameters indicates that the impact of recent lagged conditional variance and information shocks have larger effect on expected volatility compared to that of distant conditiona variances. Decrease in leverage parameters in the presence of deterministic structural shifts suggested the reduction in asymmetry of the news impact (‘bad news’ vs ‘good news’).

Implications: Investors and researchers when analyzing the high-frequency univariate financial time-series shall incorporate deterministic structural shifts in their analysis to ensure more robust analysis. Understanding the adjusted persistence and leverage parameters allows decision-makers to make informed risk assessments and enables to recalibrate their strategies to respond more promptly to market changes.