金融危機以後,風險管理的重要性大幅提升,因此需要一個能夠準確衡量風險的方法,並提供正確的風險評估與控管。Value at Risk(VaR)是由國際清算銀行與JP Morgan的RiskMetrics所推動,為現在風險控管中計算風險的方法,它是一種衡量風險的指標,簡單易懂且方便運用,可計算不同類型的金融商品風險值,在估算期間及假設之信心水準下,透過所計算之數值,顯示其風險曝露的最大程度。 本文的研究標的為國內股票型基金,研究方法使用資料不需常態分配的歷史模擬法與蒙地卡羅模擬法,並將資料波動性納入兩種模擬法;加入Hull與White(1998)提出的指數加權移動平均法修正歷史模擬法,其方法利用指數遞減的方式修正近期和遠期波動,也將捕捉波動叢聚的GARCH模型加入蒙地卡羅模擬法,估算四種風險值模型之VaR。風險值模型必須經過回溯測試才能確定其準確性,而模型的失敗率與假設之顯著水準愈接近,則表示其模型愈準確,本文使用回溯測試及非條件涵蓋比率檢定,評估風險值模型之績效。 實證結果指出,在信心水準99%的假設下,捕捉資料波動叢聚現象的GARCH模型可改善蒙地卡羅模擬法之績效,在信心水準95%的假設下,將歷史資料之波動值納入考慮之HW模型,能改善歷史模擬法估計風險值之績效;整體而言,將波動值納入計算之模擬法,績效優於原始模擬法,其結果顯示最適模型為修正波動值之模型。
After the financial crisis, the importance of the risk management has elevated significantly. Therefore, it is needed a method to measure the risk accurately and provide the risk assessment and control correctly. Value at risk (VaR) is a index for measurement of the risk. VaR is an estimate of maximum potential loss to be expected over a given period a certain percentage of the time. The main purpose is used Demestic Equity Fund as the samplefor the research. It is used historical simulation method, Hull and White's (1998) historical simulation method, Monte Carlo simulation method and GARCH model of Monte Carlo simulation to predict the VaR. The research is inquired into which of models have better prediction, in order for us to check the accuracy of the VaR model. The Backtesting and LR test of unconditional coverage are evaluated performance of the VaR model. The result shows that model with volatility as the most suitable model.