Evaluating the Risk-Return Profile of a Portfolio of ESG and Traditional Assets Using a Hybrid Optimisation Model
Evaluating the Risk-Return Profile of a Portfolio of ESG and Traditional Assets Using a Hybrid Optimisation Model
Author(s): Attila Banyai, Tibor Tatay, Gergő Thalmeiner, László PatakiSubject(s): Energy and Environmental Studies, Financial Markets, ICT Information and Communications Technologies
Published by: The London Academy of Science and Business
Keywords: portfolio rebalancing; sustainable investment; risk-adjusted returns; ESG and NonESG assets allocation; shape ratio; support vector machine algorithms;
Summary/Abstract: This article examines the risk-return dynamics of portfolios combining environmental, social, and governance (ESG) assets with traditional investment instruments. A hybrid optimisation framework is applied, uniting mean-variance principles with combinatorial selection and machine learning techniques. The study addresses two central questions: whether ESG funds provide diversification benefits, and whether they mitigate downside risk in periods of financial stress. The analysis draws on a dataset of five ESG and five non-ESG funds, spanning varied sectors and risk profiles, observed over a five-year horizon marked by diverse macroeconomic conditions. Portfolio performance is evaluated using the Sharpe ratio, with differential evolution and support vector machine algorithms employed to capture linear and non-linear aspects of risk-adjusted returns. The findings reveal a consistent positive association between ESG allocation and portfolio performance. Optimised portfolios frequently allocated 80-90 per cent of their weight to ESG assets, particularly GRID and ESGV. ESG holdings were shown to strengthen diversification, improve upside potential, and reduce downside exposure, especially during volatile market phases. Traditional assets contributed stability but played a weaker role in enhancing risk-adjusted returns. Statistical analysis confirmed both research hypotheses: portfolios integrating ESG investments achieved higher Sharpe ratios without excessive risk, and ESG funds demonstrated resilience under adverse conditions. Machine learning models further underscored the significance of non-linear patterns, which enhanced the explanatory power of ESG exposure in the optimisation process. In sum, the study contributes to growing evidence that ESG assets not only advance sustainability objectives but also deliver measurable financial benefits. The hybrid methodological approach illustrates the importance of balanced allocation constraints and robust optimisation in portfolio design. These results suggest that incorporating ESG assets can simultaneously reinforce financial stability and support long-term sustainable development.
Journal: Virtual Economics
- Issue Year: 8/2025
- Issue No: 1
- Page Range: 16-39
- Page Count: 24
- Language: English
