Sensitivity Analysis Using Garch Model: Evidence from Indian Stock Market

Abstract

The sensitivity of a financial market can be assessed by understanding the volatility in the stock returns. Volatility forecasting measures the riskiness of the investments. In a financial time series, there are periods where volatility is higher in comparison to other periods. Furthermore, the volatility in the stock movement tends to increase during economic disturbances such as recessions and financial crises due to compulsive selling and buying of stocks. The aim of the study is to conduct a sensitivity analysis of the Indian Stock Market across time and different frequencies. The S&P BSE 500 index has been selected to study the sensitivity of the Indian stock market as it represents 93% of market capitalisation of 20 major industries of the Indian economy. The daily returns, calculated using the closing value of the selected BSE index for a period of 19 years from 01-02-1999 to 31-08-2018, has been used as the variable for the study. The stationarity of time series data has been assessed using Augmented Dickey fuller test with breakpoints to identify the significant date in the time series. The Augmented Dickey Fuller test suggests that, at level difference, 18th May 2009 is a significant breakpoint date in the daily returns series, which coincides with the report on recession by National Bureau of Economic research. Hence, two time series, one before and one after 18th May 2009 were created. The normality of the two series has been tested using Jarque Bera Test, which suggests that the time series data are not normally distributed. The Autoregressive Conditional Heteroscedasticity (ARCH) model was applied to study the sensitivity of the Indian stock returns. The ARCH LM test highlights the significant existence of ARCH effect in the series before the breakpoint date, yet no ARCH effect was found in the series from 18-05-2009 to 31-8-2018. The results demonstrate that there is a significant decrease in the volatility in the daily returns after the recession period. The results of the study further suggest that volatility in daily returns existed before the period of recession, which was caused due to excessive leverage effect. However, the volatility in the daily returns has significantly reduced. This feature of the stock market movement can be attributed to increased financial literacy among investors and improved prudential norms of Securities Exchange Board of India. Hence, it can be concluded that minor fluctuation no longer cause panic amongst investor as they did before the 2009 market crash.

Keywords: GARCH, ARCH effect, S&P BSE 500 Index, Breakpoint Unit Root

View PDF