Estimation of Value at Risk - The application of GARJI model - the Case of Down Jones Industry Index and S&P 500 Index

• Main Authors: Ting-Huei Liao, 廖丁輝
• Other Authors: Chien-Liang Chiu
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/55669470962688961883

碩士 === 淡江大學 === 財務金融學系碩士在職專班 === 92 === In this thesis, we employ RiskMetrics, GARCH and GARJI model to estimate the VaR of DownJones Industry Index and S&P 500 Index. Due to the heavier tails, skewness and leptokurtosis of financial assets returns, the behavior of price movement is discontinuous. Therefore, we use GARJI model to capture the discontinuous status and incorporate the status into the VaR calculations. Additionally, we also adjusted the percentile for the heavier tail of the financial assets returns to enhance the calculations of VaR more efficiently. The results of our finding are as follows： (1) GARJI model can solve the problem of irregular jump and capture the fat tails and leptokurtosis of financial asset returns. Therefore, using the GARJI model is more precise to evaluate the VaR than theGARCH model, and the RiskMetrics model is a medial level. (2) As to the measures of accuracy, GARJI model is better than the GARCH model under the same confidence level. That is, GARJI can cover the most risk effectively and capture the volatility. However, both of them have the ability to estimate the Value at Risk. (3) As to the measures of efficiency, there is a trade-off between efficiency and accuracy. That is, GARJI model is more accuracy but is less efficiency than GARCH model. Keywords︰GARJI model、GARCH model、RiskMetrics model、 Value at Risk