Time-Varying Volatility Dynamics of Dhaka Stock Exchange (DSE) Using GARCH-Type Models
Date
2016-06
Student ID
Research Supervisor
Editor
Journal Title
North South Business Review
Volume
06
Issue
2
Journal Title
Journal ISSN
Volume Title
Abstract
Our study in this paper examines the changes in volatility dynamics of Bangladesh stock market, due to stock market crash back in 2010 using Generalized Autoregressive Conditional Heteroscedasticity (GARCH) models. Both symmetric and asymmetric GARCH models were used to estimate the conditional volatility in daily returns of the main stock exchange of Bangladesh namely as the Dhaka Stock Exchange (DSE). Using the closing values of DSE General (DGEN) and DSE broad (DSEX) indices, we conduct our analysis by keeping the crash period of 2010 in focus and dividing the data series into three sub-periods namely as crash, pre-crash and post-crash periods. The data series exhibit evidences of skewness, kurtosis and deviations from normality as expected from Finance literature. The unit root test concludes that the data series is not stationary at level but become stationary when we take returns of the series into consideration. The conditional variance is found to be highly persistent with leverage and asymmetric effects. Our results also indicate that asymmetric GARCH models are better fitted to model volatility dynamics than the symmetric GARCH model for all sub-periods. Lastly, the comparison among various model parameters in different sub-periods in our study also exhibit significant change in volatility patterns of DSE from pre-crash to post-crash period as indicated by ARCH, GARCH, Leverage and Power coefficients.
Description
Keywords
Citation
Department Name
Publisher
North South University