This research aims to explore the risk profile of the value-oriented environmental, social, and governance (ESG) portfolio in the United States stock market using explainable artificial intelligence (XAI) approaches. To achieve this, we use a range of key market, risk, and uncertainty indicators. The results indicate that the value ESG portfolio is significantly influenced not only by the stock market but also by other factors—particularly cash flow–related indicators such as default risk and bond-market conditions. We observe an approximately linear relationship between the value ESG portfolio and the most significant factors. Specifically, higher (lower) values in the stock and bond markets are associated with higher (lower) values of the value ESG portfolio, whereas the opposite is true for the default spread. The interpretability provided by the XAI-based approach offers valuable insights into developing effective portfolio investment strategies.
Baris Kocaarslan (Sat,) studied this question.