两阶段投资组合均值－变异数模式最佳化分析.pdf

經營管理論叢 Vol.6, No.1, 2010第 49-71頁 兩階段投資組合均值－變異數模式最佳化分析 Optimization Analysis of Two-stage Portfolio Mean-variance Model 1 2 許晉雄 鄒慶士 (Received: Dec. 1, 2009 ；First Revision: Jan. 20, 2010 ；Accepted: Feb. 24, 2010 ) 摘要 如何將有限的資金在眾多之投資標的中作有效的配置，是財務管理領域中重要的議 題之一，此問題稱為投資組合 選擇。在傳統的投資組合理論中，最著名的是馬可維茲 (1952)所提出的均值 -變異數模式，在過去處理此類的投資組合選擇問題大多是將此單階 段且具有衝突性的目標合併轉為單一目標的形式，並以單目標最佳化技術進行求解。本 文即針對當投資組合最佳化均值-變異數模型中包含一些實務上所需的限制條件，例如 基數限制或允許賣空情況，進行投資組合問題求解，主題探討分別在賣空與不允許賣空 和基數限制與無基數限制的條件下獲得效率之投資組合。在第二階段，有關於在投資組 合選擇問題中，再利用多屬性決策分析中的簡單加權法和 TOPSIS 法兩種方法，對投資 組合進行分析排序，提供投資者更多樣化之投資組合選擇，由研究結果可發現投資者在 選擇其投資組合時，應考慮更多全面性的績效指標。 關鍵詞： 投資組合、基數限制、簡單加權法、 TOPSIS 法 Abstract The allocation of limited capital to a variety of assets available is one of the important problems in financial management. This problem is called portfolio selection problem. The mean-variance (MV) portfolio model originally proposed by Markowitz (1952) has been playing a critical role in the portfolio selection theory. Most studies had been made to solve this single-stage problem with single-objective optimization techniques by aggregating conflicting and incommensurate objectives into a single one. In this paper, we discuss the MV portfolio model with some practical considerations, such as cardinality constraints and short sales cases, are incorporated into this portfolio optimization models. Finally, Multi-Attribute Decision Making (MADM) methods named Simple Additive Weighting (SAW) Method and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method are employed to outrank the efficient portfolio that decision makers satisfy most in the second stage.