Testing for contagion under asymmetric dynamics: Evidence from the stock markets between US and Taiwan

This article is an attempt to test, through the use of forward forecasting test on dynamic conditional correlation (DCC), for contagion between Taiwan and US stocks under asymmetry. The process includes three steps. The first step uses the iterated cumulative sums of squares (ICSS) algorithm to detect the structural breaks of market return. The second step creates dummy variables for breaks, estimates EGARCH model of conditional generalized error distribution, and computes dynamic conditional correlation coefficients of DCC multivariate GARCH model. The third step employs one-step and N-step forecast test to check for contagion effect. The evidences prove the asymmetric leverage effect of Taiwan weighted stock index and New York-NYSE Composite Index. Interestingly, we discovered that there are two kinds of contagion, "positive" and "negative", between markets. (c) 2006 Elsevier B.V. All rights reserved.