Applying the Clustering and UTADIS Models to form an Investment Portfolio

Abstract

Objective: The aim of this study is to propose a synthetic approach including Clustering and UTADIS models to form a profitable investment portfolio. Methods: In this research, securities are clustered using K-means method and the ideal number of clusters is determined through certain validation indexes. The results obtained from the Clustering method were used as the input data for the UTADIS model and the securities were classified by UTADIS. After solving the primary model, in order to achieve better resulst, a post-optimality analysis was performed and the classification validity test and the classification error tests were carried out. Results: After reviewing previous studies in this field and carrying out a survey of professionals from the financial industry, eight key attributes including capital return, beta coefficient, net profit margin, BV/MV, ROA, ROE, P/E and EPS were identified. The investment portfolio consists of Iran tele companies, Khark Petrochemical, Shazand Petrochemical, Fanavaran Petrochemical, Information services, Iran refract, Khouzestan steel, and Iran zinc mines. Conclusion: The results of study showed that the proposed framework has created a profitable portfolio and capital return is the most important attribute in stock portfolio selection.

Keywords

• multi-attribute decision-making
• utadis
• clustering
• stock portfolio
• investment