IMPROVED DETECTION OF RARE-EVENT RISK OF A PORTFOLIO WITH U.S. REITs

Abstract

In measuring the market risks of a portfolio, value-at-risk (VaR) is one of the most commonly used tools. In this paper, the copula-generalized autoregressive conditional heteroskedasticity (GARCH) method is used to determine whether it is a better alternative for estimating the VaR of portfolios containing U.S. real estate investment trusts (REITs). The FTSE NAREIT US Real Estate Index, all REITs and the S&P 500 index are used to construct a portfolio. In total, 2800 daily data covering the period of the subprime mortgage crisis of 2007–2009 are used in this paper. We used six constant and two time-varying copula models combined with two GARCH models to form sixteen copula-GARCH models to depict the joint distribution of the two assets in the portfolio. We then computed corresponding one-day VaRs. Compared with the traditional VaR models, our results showed that the time-varying symmetrized Joe–Clayton (SJC) copula model combined with the GARCH Student-t innovation (tvSJC-copula–GARCHt) performed the best, regardless of the market situation. Hence, it could be served as a better way of detecting rare-event risk.