Extreme Us Stock Market Fluctuations in the Wake of 9/11

Abstract

It is common wisdom that the 9/11 terrorist attacks boosted political and financial uncertainty and resulted in severe stock market meltdowns in the months after the attacks. Taking a sectoral focus of the market for US common stock, we apply statistical extreme value analysis (EVT) to assess whether downside risk measures like Value-at-Risk (VaR) and extremal sector linkages were significantly altered by 9/11. We use a semi-parametric quantile estimator for the VaR in order to test whether downside risk increased after 9/11. We also estimate the co-exceedance probability of observing joint meltdowns in sectoral and market portfolio indices by employing a dependence measure from bivariate extreme value theory. These tail beta's may be thought of as measuring systematic risk during high volatility regimes. Taking 9/11 as the sample midpoint we do find that nearly all sectoral tail beta's have risen over time. This indicates that the potential for domestic portfolio diversification during crisis periods, i.e. when diversification is most needed, decreased compared to the pre-9/11 era. As for the pre-9/11 and post-9/11 VaR estimates, we do not find much support for a structural change in downside risk.

Keywords: Market Crashes, Bivariate Extreme Value Analysis, Extreme Co-movements, heavy tails, Value-at-Risk, tail beta, extreme dependence

JEL Classification: G1, F3, C49