Behaviour of Johannesburg Stock Exchange All Share Index Returns - An Asymmetric GARCH and News Impact Effects Approach

Carl H. Korkpoe, Peterson Owusu Junior

Abstract


We investigated the behaviour of returns of the Johannesburg Stock Exchange All Share Index using asymmetrical exponential-GARCH(1,1) and GJR-GARCH(1,1) incorporating the market reactions to news. We noted the returns distribution is skewed and have fat-tails with respect to the normal distribution. Thus we chose the skewed student t-distribution asymmetry to model the behaviour of the tails and capture the asymmetry in the distribution of the returns. The GJR-GARCH(1,1) with the skewed student t-distribution was found to be appropriate in describing the data generation process for the returns. The market was shown to react to news unequally. Volatility spikes sharply when unexpected adverse news reaches the market while remaining unresponsive for a large part to positive news. For investors this has implications for trading strategies and risk management with respect to equity portfolio risk and returns on the stock exchange. Bad news reaching the market can destabilize their portfolios. Risk mitigating actions by way of 'hedging' against the noise in the news is warranted.

JEL Classification: C01, C5, C58

Keywords


Backtesting, E-GARCH, GJR-GARCH, Volatility, News Impact

Full Text:

PDF




η δικτυακή πύλη της ευρωπαϊκής ένωσης ψηφιακή ελλάδα ΕΣΠΑ 2007-2013