Forming An Optimal Portfolio by Applying the Data Envelopment Analysis (DEA) Approach Based on The Single Index Model

Abstract

The study aims to determine the optimal portfolio formation using the Data Envelopment Analysis (DEA) approach based on the Single Index Model. The significance of this research lies in providing investors with an effective tool for maximizing returns while minimizing risk through portfolio diversification. The DEA method is applied to measure the efficiency of selected stocks listed on the Indonesian Stock Exchange between 2020 and 2023. By utilizing technical efficiency, the study evaluates stocks to identify efficient Decision-Making Units (DMUs) for portfolio candidates. The Single Index Model, developed by William Sharpe, is used to calculate expected returns and associated risks. The research uses purposive sampling, focusing on companies in the industrial sector. Monthly stock prices, dividends, and macroeconomic data such as the Indonesian Stock Exchange Index (IHSG) and Bank Indonesia Certificate (SBI) interest rates serve as inputs. Using DEA-CCR and DEA-BCC models, efficient stocks are selected based on efficiency scores. The results highlight 11 stocks as optimal portfolio candidates, including Astra International (ASII) and Perma Plasindo (BINO). The portfolio provides an expected monthly return of 0.0211213 with a standard deviation of 0.026621, indicating potential outperformance relative to the market index.