Using Utility Additive Discriminates Analysis for Stock Evaluation

• Main Authors: Wei-Jen Wang, 王偉任
• Other Authors: Wen-Wu Chang
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/83233067201558944486

碩士 === 真理大學 === 管理科學研究所 === 91 === This study presents a real application of a multicriteria decision aid (MCDA) approach to stock evaluation based on preference disaggregation, using ordinal regression and linear programming(UTADIS method; UTilités Additives DIScriminantes). The procedure is applied on data concerning common stocks of MSCI components. We evaluated the sample stocks under consideration and determined a trichotomous classification scheme, according to the stock cumulative returns of each share. And using 15 criteria and three kind of stock cumulative returns formation period. The additive utility function that are derived through this approach have the extrapolation ability that any new share can be correctly classified into one of three user-predefined groups. Through the application of preference disaggregation analysis in portfolio selection, it enables the portfolio manager to classify a large number of stocks into a few categories according to their financial and stock market performance. Thus, instead of considering simultaneously hundreds of stocks traded in a stock exchange, a portfolio manager can determine a limited number of stocks upon which to base portfolio construction. The results show that the additive utility function can classify our sample into predefined stock evaluative rating more correctly, when the periods of criteria selection as the same as the stock cumulative returns formation period. Besides, the professional institution index owns the highest weights of all criteria, especially for “fund share-holding ratio”, “recently three months changes in fund share-holding”, “foreign share-holding ratio”, “recently three months changes in foreign share-holding”. It appears that the change of stock holding of institutional investors reflects stock cumulative returns and stock evaluative rating in Taiwan’s security market. Furthermore, the comparison with multiple discriminant analysis illustrates the superiority of the proposed methodology over a well-known multivariate statistical technique that has been extensively used to study financial decision-making problems.