SRI LANKAN STOCK MARKET VOLATILITY ANALYSIS: AN ARMA- GARCH APPROACH

Abstract

Beyond its role in capital generation, a stock market is emulated as a facet in economic development indication. Sri Lankan stock market, the Colombo Stock Exchange (CSE) is a languidly developing market known for instability and periodical fluctuations increasing the volatility risk for the investors. Toward market development, it is imperative in attracting and retaining long term investors. Thus, the study aimed to identify the dynamics of the CSE through volatility estimation of the Sri Lankan stock market during a high volatile period. Further, the use of ARMA-GARCH models aims to contribute to the local empirical studies on the applicability of ARMA-GARCH models in the Sri Lankan context.

The study used the daily closing prices of the All-Share-Price Index (ASPI) from January 2018 to December 2022 in log return volatility. Owing to the non-normality and serial dependence conditions inherent in the data, the study developed an ARMA (2,2) mean equation and separate volatility equation applying symmetric models of GARCH, and TGARCH and asymmetric GARCH models of EGARCH, and GJR-GARCH.

The study findings identified that asymmetric GARCH models are more reliable in volatility estimation and forecasting. Further, ASPI indicated a leverage effect where negative information caused idiosyncratic volatility.

Keywords: ARMA-GARCH, GARCH, Leverage Effect, Stock Markets, Volatility