DCC-GARCH model for market and firm-level dynamic correlation in S&P 500

Understanding the dynamic correlations among asset returns is essential for ascertaining the behavior of asset prices and their comovements. It also has important implications for portfolio diversification and risk management. In this chapter, we apply the DCCGARCH model pioneered by Engle (2001) and Engle and Sheppard (2002) to investigate the dynamics of correlations among S&P 500 stocks during the sub-prime crisis. Using the daily data of stocks in the S&P 500 index, we document strong evidence of persistent dynamic correlations among the returns of the index component stocks. Conditional correlations between S&P 500 index and the component stocks increase substantially during the period of sub-prime crisis, showing strong evidence of contagion. In addition, stock return variance is time-varying and peaks at the crest of financial crisis. The results show that the DCC-GARCH model is a powerful tool for forecasting return correlations and performing value-at-risk portfolio analysis.