Value at Risk of Multivariate Financial Time Series Model

• Main Authors: Chieh-Hsiang Hsu, 許傑翔
• Other Authors: Yi-Ping Chang
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/24770209990745404733

碩士 === 東吳大學 === 商用數學系 === 92 === Value at Risk(VaR) models developed in recent years has been well accepted in the field of risk management. Then, how to estimate VaR accurately is an important key to avoid crisis. While calculating VaR we need a model which can effectively describe the volatility dynamic and the distribution of asset’s return. In this paper, we use the univariate GARCH model and the multivariate CCC model together with several different error distributions to calculate volatility and VaR. Through a portfolio consisting of two currencies, our empirical research shows that the multivariate CCC model together with excess kurtosis or fat tail provides a better representation in VaR framework.