| This paper investigates the issue regarding the financial contagion effect among stock markets after the crises. A financial crisis in one area may expand to other regions of the world and such a phenomenon is known as "Financial Contagion". Several stock markets tend to move together after the financial crisis breaking out in some area and this comovement of markets can be seen as an effect of financial contagion. Therefore, testing the correlation change among the stock markets can be treated as a tool to test the contagion effect. However, the correlation change is asymmetric since a prosperous stock market does not spread like the crisis and cause other markets prosperity. Moreover, the comovement effect among stock markets during the financial crisis is often much stronger than during the prosperous time.Copula models have advantages in describing the correlation between variables. Since they can handle the non-linearity and asymmetric properties of the correlation structure which cannot be characterized by the Pearson product-moment correlation coefficient, the copula is a good choice in the empirical research through this paper. Time-varying copula is a development of the constant coefficient copula models and is more suitable to cover the problem considered here.Time-varying copula mentioned by Patton is the main empirical method in this paper. Time series on S&P500, FTSE100 and TWII are used to investigate the financial contagion among the US, UK and Taiwan stock markets. Three crises are considered and samples are divided into three parts according to the corresponding time intervals. Financial contagion is finally detected through the correlation change before and after the crises. Empirical results show that time-varying copula is more suitable than constant coefficient copula models when describing the dynamic correlations. Moreover, the correlation between US and Taiwan stock markets behaves more volatile than that between US and UK markets. |
PDF Full Text Request