A two-step approach to investment portfolio selec‎tion using stock clustering an‎d multi-criteria decision making

3/21/2026 12:00:00 AM

Portfolio optimization has always been one of the fundamental challenges in finance, as investors must balance expected returns with potential risk. Despite classical models, there is still a need for approaches that can take into account market realities an‎d multiple decision-making criteria. In this study, a hybrid approach was presented to form a portfolio. First, by calculating the risk an‎d return of all Nasdaq 100 stocks, clustering was performed into four risk-return groups to examine choices under similar conditions. Then, in each cluster, six fundamental criteria were weighted using the CRITIC method, an‎d the stocks were ranked using the TOPSIS an‎d WASPAS methods. Subsequently, the optimal portfolio was formed using the stan‎dard deviation measure an‎d considering practical weighting constraints. The results showed that portfolios from high-return clusters generated positive returns, while low-return clusters were loss-making. In particular, the high-return-high-risk cluster recorded the highest profits an‎d the low-return-high-risk cluster recorded the highest losses. These findings indicate that the proposed approach can be effective in selec‎ting optimal stock combinations commensurate with the risk tolerance level of investors.

انتخاب سبد سهام , بهينه‌سازي سبد سهام , تصميم‌گيري چند‌معياره , خوشه‌بندي , بازار سهام , سرمايه‌گذاري

Portfolio selec‎tion , Portfolio optimization , multi-criteria decision making , Clustering , Stock market , Investment

Mohammad Kouros

Dr. Ahmad Makui