Cointegration of Bombay Stock Exchange with Major Asian Markets—A Copula Approach

Abstract

This article examines the level of cointegration of the weekly returns of Bombay Stock Exchange (BSE), the representative index of India, with those of other major Asian markets of China (Shanghai), Hong Kong (Hangseng), Japan (Nikkei) and Taiwan using the vector autoregression (VAR) for long-term dependencies and copulas for tail dependencies. A linear relationship is observed among all these markets, but the upper and lower tail dependencies of the BSE with Nikkei and Hangseng are found to be much stronger. The Clayton copula was used to study the lower tail dependency and the Gumbel copula for the upper tail dependencies of the returns. These results will be helpful for international portfolio diversification and the investors may find India, Japan and Hong Kong markets attractive for such diversifications. In addition, the GARCH(1, 1) methodology was also employed for studying the long-term variances of each of the market and it is observed that the weekly volatility of Nikkei is much higher as compared to other markets, while that of the BSE is moderate. Apart from this, all these markets show excess kurtosis, pointing towards heavy-tailedness of the return series apart from the non-normality. The VAR results show that BSE weekly returns are affected by up to two lags of the Hong Kong stock market (HANG). At the same time, the NIKKEI and the SHANG are affected by up to 1 week and 2 week lags of the BSE, respectively.