Estimation of Conditional Volatility Models for Indian Stock Market
Natchimuthu N.*, Dr. Prakasam Karthigai**
*Research Scholar, Bharathiar University, Coimbatore, India. firstname.lastname@example.org
**Associate Professor, Department of Commerce, Christ University, Bengaluru, India. email@example.com
Online published on 31 May, 2018.
The leverage effect or asymmetric effect is defined as the unequal influence of negative shock on volatility in comparison with positive shock. The negative shocks or the bad news seems to have larger influence on next period volatility. The objective of this study is to test the presence of volatility clustering and long-term memory features in Indian capital market and to compare the symmetric and asymmetric conditional models. The presence of asymmetric effect was tested with the help of conditional volatility models like TGARCH, EGARCH and PGARCH. These volatility models were applied to NSE broad indices and sectoral indices. The data for a period of 18 years was collected from PROWESS database. The results suggest that there is volatility clustering and long-term memory features in Indian capital market. Asymmetric volatility models performed better on all the indices tested. In selecting the best volatility models, AIC information criterion and log likelihood parameters were used. AIC selects PGARCH and TGARCH models as best performing model and log likelihood selects PGARCH model as the best model. Hence, it was concluded that PGARCH model performs better than other asymmetrical models for Indian capital market.
Asymmetric conditional volatility models, GARCH, TGARCH, EGARCH and PGARCH.