Application of Operations Research in Portfolio Optimization in Finance

The analysis covers 761 of the 765 records retrieved from Scopus on portfolio optimization between 2017 and 2026. Publications rose from 35 in 2017 to 162 in 2026, an average annual growth of 18.56 percent, though the path was uneven. For RQ1, output dropped 22 percent in 2019, recovered through 2024, and jumped 63.9 percent in 2025 to 159 papers. The years 2025 and 2026 together hold 42.2 percent of the corpus. 2026 is incomplete; 22 of its papers remain in press. Mean citations fall from 35.1 per paper in 2017 to 0.9 in 2026 because recent papers have had less time to be cited. The median is 3 citations. 161 papers (21.2 percent) have none, so impact is concentrated in a smaller set of older work. Of the 761 papers, 170 do not contain “portfolio” in the title or author keywords. One of the three most-cited items is a 2023 healthcare paper on synthetic data (370 citations) that is not about portfolio optimization. The other two are a 2019 review of quantum computing in finance (612 citations) and a 2020 stock-market study (268 citations). Both touch related topics but are not central to portfolio selection. Citation totals therefore describe the retrieved set, not portfolio optimization alone. Comparisons with studies that used narrower searches should be made with that difference in mind. For RQ2, the work is spread across many countries yet concentrated in a few. The set contains 2,039 authors and 400 sources, with 2.98 authors per paper on average. Only 14.1 percent of papers are single-authored; 30.75 percent have international co-authors. China accounts for 300 of 1,708 author-affiliation appearances (17.6 percent), the United States for 185 (10.8 percent), and India for 135 (7.9 percent). These three countries make up 36.3 percent of appearances; the top ten make up 57.5 percent. Li Shuai (eight papers), Cao Xinwei, and Qi Yue (seven each) appear in recurring international teams. Multi-country papers (237, or 31.1 percent) average 22.5 citations against 9.7 for single-country papers, a statistically significant difference. Reviews average more citations (55.6 versus 11.7), but that difference does not fully account for the collaboration effect. For RQ3, keyword co-occurrence at minimum frequencies of 5, 7, and 9 produced similar themes, with the number of clusters falling from seven to five as the threshold rose. The largest clusters center on portfolio optimization, finance, risk management, and sustainable finance; on operations-research methods such as constrained optimization, genetic algorithms, and quantum computing; and on financial markets, electronic trading, and machine learning. Smaller clusters cover decision-making, artificial intelligence, multi-objective optimization, risk measurement, and investment strategies. These map onto the literature review: mean-variance optimization remains the base, with later work adding optimization under uncertainty, metaheuristics, reinforcement learning, and quantum computing. Quantum computing still forms a distinct cluster at the stricter thresholds but links to risk and mean-variance topics at the broader ones (Cui et al., 2022; Soler-Dominguez et al., 2018; Orús et al., 2019; Singh et al., 2022). “Decentralized finance” appears as an index keyword in 100 papers, 99 of them from 2024 to 2026. Only 10 of those papers mention DeFi, blockchain, or cryptocurrency in the title or author keywords. Part of the recent rise in that term may therefore come from indexing rather than from a shift in research focus. For RQ4, the main limits come from the export. All cited-reference fields are empty, so co-citation analysis and bibliographic coupling cannot be done. A re-export that includes references would open those methods. Index keywords are missing for 388 papers (51 percent), so the keyword maps cover only about half the set. The broad search also means the results do not map cleanly onto portfolio optimization. Restricting the analysis to author keywords and dropping the main search term would tighten the focus. The treatment of incomplete years is inconsistent: four 2027 records were removed, yet all 162 papers from 2026 were kept, including 22 still in press. Papers from 2025 and 2026 have had less time to accumulate citations, so their lower counts do not by themselves indicate lower quality. The counts and maps still show how the literature under this search has grown, which countries and authors contribute most, and how the keyword clusters track the methodological shifts described in the literature review. Later bibliometric work on the topic would gain from a narrower query, an export that includes cited references, and conclusions that stay within what the chosen data can support. The present study acknowledges these limits.

1. Aboussalah, A. M., Xu, Z., & Lee, C.-G. (2022). What is the value of the cross-sectional approach to deep reinforcement learning? Quantitative Finance, 22(6), 1091–1111. https://doi.org/10.1080/14697688.2021.2001032 2. Abu-Alkheil, A., Khan, W. A., Parikh, B., & Mohanty, S. K. (2017). Dynamic co-integration and portfolio diversification of Islamic and conventional indices: Global evidence. Quarterly Review of Economics and Finance, 66, 212–224. https://doi.org/10.1016/j.qref.2017.02.005 3. Akhtaruzzaman, M., Banerjee, A. K., Boubaker, S., & Moussa, F. (2023). Does green improve portfolio optimisation? Energy Economics, 124, 106831. https://doi.org/10.1016/j.eneco.2023.106831 4. Alcazar, J., Leyton-Ortega, V., & Perdomo-Ortiz, A. (2020). Classical versus quantum models in machine learning: Insights from a finance application. Machine Learning: Science and Technology, 1(3). https://doi.org/10.1088/2632-2153/AB9009 5. Ali, F., Khurram, M. U., Sensoy, A., & Vo, X. V. (2024). Green cryptocurrencies and portfolio diversification in the era of greener paths. Renewable and Sustainable Energy Reviews, 191, 114137. https://doi.org/10.1016/j.rser.2023.114137 6. Anquetin, T., Coqueret, G., Tavin, B., & Welgryn, L. (2022). Scopes of carbon emissions and their impact on green portfolios. Economic Modelling, 115, 105951. https://doi.org/10.1016/j.econmod.2022.105951 7. Aouni, B., Doumpos, M., Pérez-Gladish, B., & Steuer, R. E. (2018). On the increasing importance of multiple criteria decision aid methods for portfolio selection. Journal of the Operational Research Society, 69(10), 1525–1542. https://doi.org/10.1080/01605682.2018.1475118 8. Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007 9. Aria, M., & Cuccurullo, C. (2026). Science mapping analysis: A primer with Biblioshiny. McGraw-Hill Education. ISBN 978-88-386-2297-7. 10. Aria, M., Cuccurullo, C., D’Aniello, L., & Spano, M. (2026). Biblioshiny and the SAAS workflow: An integrated framework for transparent and reproducible science mapping. A demonstration through the replication of a study. Journal of Informetrics. https://doi.org/10.1016/j.joi.2026.101837