The effects of recent terrorist attacks on risk and return in commodity markets

Abstract

The event study methodology of Brown and Warner (1985) is adopted and augmented to evaluate the effect of the recent terrorist attacks on risk and returns in commodity markets. The methodology is supplemented with various techniques, such as the non-parametric ranking test and kernel regression, to find out if 20 terrorist attacks (including in Paris, Nice, Brussels, Stockholm, Manchester, Jakarta, London and Melbourne) generated abnormal returns. Asset pricing models are fitted with interaction variables while GARCH, TARCH, EGARCH and PARCH models are used to determine changes in systematic risk. Our results show abnormal returns and delayed reactions of up to 120 days in the commodity markets—evidence which defies the efficient market hypothesis. In addition, we identify sector and commodity effects in risk analysis.