本文結合金融機構股票報酬率的時間序列與橫斷面資料，利用極端值理論，並以Nelson-Siegel模型的水平移動(β0)、斜率變化(β1)與曲度變化(β2)做爲利率因子的替代變數，探討利率因子變化對於金融類股風險值的影響程度。實證結果顯示利率因子與金融機構風險值間呈現負向關係，代表利率下降會導致金融機構的風險值增加。對於不同業務內容的金融機構，其風險值與利率變化的關聯性亦有所不同。而金控業由於其業務多元化可以分散風險，故能有效降低長短期利率變化對於其風險值的影響程度。

A time-series and cross-sectional combined data is employed to investigate the relationship between the Value-at-Risk (VaR) for stock returns of financial institutions and three interest rates factors based on Nelson-Siegel (1987) model. More precisely, we examine the impact of interest rate factors on the estimation of Value-at-Risk obtained from the Extreme Value Theory (EVT). The empirical results indicate that it shows a negative relationship between the calculated VaRs and the interest rate factors. In addition, we conclude the outcome is quite different among different financial sector stocks. Since the financial holding companies could diversify their business risk, it can effectively reduce the impact of term structure movements on their VaR calculations.