Volatility Analysis: An Empirical Investigation from South Asian Stock Markets Using the GARCH Model
Keywords:
Rist, Return, Volatility, ARCH, GARCH, Stock Returns, South Asian Stock MarketsAbstract
This study examines the risk, return, and volatility of selected South Asian stock markets. It helps to measure the volatility patterns and compare the stability and risk characteristics of these markets. Major South Asian stock exchanges (PSX, BSE, and DSE). The main objective of this research is to use sophisticated GARCH models to assess volatility persistence over time and provide empirical support for regional diversification strategies within the SAARC block. This study uses secondary data collected from sources including Yahoo Finance, Investing.com, and the official websites of the PSX, BSE, DSE, and CSE. Closing prices used to compute returns in time series data. The data frequency is daily from January 2022 to April 2026. Key variables are risk, return, and volatility. The Univariate GARCH model is applied to the data. KSE has the highest average return of 0.12% with a risk of 1.2%, whereas it has the lowest risk per $ of average return (10.44), followed by the Bombay and Dhaka stock markets, respectively. The results show that ARCH and GARCH coefficients for all markets are significant at the 1% significance level. Mean reversion is computed as the sum of the ARCH and GARCH coefficients; a value closer to 1 indicates slower mean reversion (i.e., higher volatility). The Bombay Stock Exchange is the most volatile, followed by the Karachi and Dhaka stock exchanges. Half-life indicates the number of days it takes for markets to revert to their mean positions. Bombay, Karachi, and Dhaka markets take 40.58, 20.8, and 6.37 days, respectively, to revert to their mean positions. KSE offers better returns with moderate risk, which is the best option for investors.
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 International Journal of Emerging Business and Economic Trends

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
