The Interval Estimation of Value at Risk on Stocks with Different Volatilities

[[abstract]]摘要 風險值是一種用來評估風險的方法，已被廣泛用來衡量資產價格變動之風 險。惟風險值評估的結果本身亦具有不確定性，然於應用上卻較少針對此一不確 定性加以探討。本文之目的擬採用極值理論來估計台灣股價加權指數之風險值和 建構其信賴區間，並進一步探討在台灣股市價格限制之規定下，比較不同波動性 之個股風險值的不確定性估計，並探討此一主管機關所訂定之價格限制政策的合 理性。實證結果顯示使用極值理論所計算出之區間間距較小，這代表模型的穩定 性較高；在加權股價指數方面，發現多頭部位的風險大過於空頭部位；個股之部 分則顯示，波動度高低、類股之不同與多空頭不同的情況下，所計算出之風險值 差異性很大，在投資人避險策略的方面應針對不同的特性訂定不同的標準，以達 到更加的風險控管效果。[[abstract]]Abstract Value at Risk (VaR) is a way to estimate risk and widely used to evaluate the market risk However we should notice the uncertainty of VaR But few researches focus on the uncertainty of VaR This paper is aimed to apply the extreme value theory (EVT) to estimate VaR of Taiwan weighted stock index and construct the confidence interval of VaR Furthermore under the price limit policy we consider the uncertainty of VaR with stocks in different volatilities and to inspect whether the price limit policy is rational or not The empirical results show that confidence interval computed by EVT is much small This represent EVT model is stable In Taiwan weighted stock index we can see the long position’s VaR is bigger than the short position In stocks the results show that we should legislate different hedge strategy for stocks with different volatilitie