VaR夏普法則的應用限制性

• Main Authors: SHU-CHUAN LIAO, 廖淑娟
• Other Authors: 郭震坤
• Format: Others
• Language: zh-TW
• Published: 2001
• Online Access: http://ndltd.ncl.edu.tw/handle/66681979011160357380

碩士 === 國立臺灣大學 === 國際企業學研究所 === 89 === VaR Sharpe Ratio is first proposed by Dowd (1999). He replaces the standard deviation inside the original Sharpe Ratio by VaR. He also applies this kind of Sharpe Ratio on portfolio management. When the returns of assets is normal distribution, VaR Sharpe Ratio will reduce to the original Sharpe Ratio. Noted that when calculating VaR, according to the distribution of assets return being either normal distribution or not, we should use different methods. Once the skewness of the distribution is positive (negative), if we still calculate VaR by the way based on normal distribution, VaR will be over-estimate (under-estimate). Although the method based on normality assumption is easy to use, it will lead a investor to make a wrong decision. In this paper, we show that when a investor faces some assets with non-normal return distribution, he should applies VaR Sharpe Ratio on portfolio management very carefully. We use the data of warrants in Taiwan as the source of distribution with positive skewness and at the same time, we manufacture a set of data artificially as the source of distribution with negative skewness. We find that if the investor uses the normality assumption to derive the VaR Sharpe Ratios of these assets and chooses portfolio based on these VaR Sharpe Ratios, he may has possibility to make a wrong decision.