Multi-choice goal programming approach for multi-objective portfolio optimization problem

In real-world investment scenarios, investors tend to make decisions with regard to all the criteria they want including return, risk, and liquidity all together in simple terms to enhance portfolio optimization. In more complex decision-making contexts, they may also have multiple aspirations or preferences associated with these various goals. Thus, the aim of this study is to apply Multi-Choice Goal Programming (MCGP) approach for solving a multi-objective portfolio optimization (MOPO) problem. The model considers three objectives which are maximizing return, minimizing risk and maximizing liquidity. The model also incorporates environmental, social and governance (ESG) constraints alongside other practical constraints, including cardinality, sector, floor and ceiling constraints to reflect real-world investment conditions. Computational experiments were performed to verify and validate the model, using a dataset of 30 companies listed on Bursa Malaysia, which includes monthly return and turnover rates over the period from January 2020 to December 2023. The data was prepared in Excel before being processed in MATLAB2018a. The model was tested across four scenarios to examine the impact of changing priority weight on the solutions. The results show that the MCGP model could generate optimal portfolios that meet the specified target goals. The results also highlight the effectiveness of the MCGP model in providing investors with the flexibility to define their target goals as interval values.