在財務最佳化投資的應用上,常被應用找出標第物的最佳化投資權重方法為Mean-Variance 方法,此方法在無風險利率上的假設為常數,本篇要介紹的方法由HJB PDE 導出最佳化投資權重,在W. H. Fleming 及S. J. Sheu(2000)提出的方法也得到相同的結果。假設無風險利率為隨機情況下,所求的最佳化投資權重。在對數型效用函數且在允許買空賣空(投資權重允許為負值)的情況下,比較兩個方法所得期末報酬表現。在推導中, 將介紹利用HJB PDE 推導隨機利率的最佳化投資權重結果。評判兩個方法的標準為夏普比率之高低,實證分析中將採用美國金融市場歷史資料。 Mean-Variance portfolio optimization is the most commonly applied method to find the portfolio weight for risky assets. The interest rate is assumed to be a constant in the framework. We derive the optimal portfolio weight by Hamilton-Jacobi-Bellman (HJB) equation under log utility when the interest rate is stochastic. We compare the Sharpe ratio as a measure of performance of the two methods, allowing short sales. The empirical analysis on US historical data is conducted.