本文以Delta-Normal法,歷史模擬法及蒙地卡羅模擬法衡量多種股價指數與貨幣投資組合之風險值,期問從97/3/3至98/2/28,本文發現在Delta-Normal法中,包括日圓或馬克的投資組合之風險較不包含日圓或馬克的組合之風險低,在歷史模擬法中,香港恆生指數與馬幣的投責組合風險最大,此點解釋為何馬來西亞政府自1998/9/2起,將馬幣由浮動匯率改採固定匯率,蒙地卡羅模擬法發現只要包括德國馬克的投資組合,其風險小於不包括德國馬克的組合,本研究也發現兩個不同金融商品組成的投資組合風險低於兩徊相同商品組成的投資組合,證明分散風險的重要性,同時本研究也支持沒有一個衡量 VaR 的方法在各衡量角度下均為最佳,並建議當一個地區發生金融風暴時,投資人最好不要將該地區的貨幣與股票包含於投資組合內,以降低風險。
This study uses three commonly-used VaR methods to measure the financial risk of various portfolios from March 3, 1997 to Feb. 28, 1998. Using all three methods, portfolios including Japanese yen or German marks have less risk than those excluding Japanese yen or German marks. The Delta-Normal method requires an accurate estimation of the model’s parameters because different confidence intervals or observation period variations affect VaR numbers. According to historical simulations, a portfolio consisting of Malaysian ringgit and the Heng Seng Stock Index has the highest risk among all portfolios, and is the 2(superscript nd) most risky portfolio under the Delta-Normal method. This finding accounts for why the Malaysian government has abandoned the floating exchange rate system since Sep.2, 1998 and Hong Kong government sacrifized the stability of stock market to control the exchange rate of Hong Kong dollars. This study also confirms the diversification effect, as the VaR value of a portfolio consisting of two different instruments is less than that of a portfolio including two identical instruments. However, no single VaR method is superior by every measure. This study suggests that when a financial crisis occurs in one region, conservative investors should divest the currencies and stocks from that region from their portfolios.