Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The article presents the results of research analysing the impact of ESG ratings on financial performance (ROA and ROE) using stepwise linear regression and feedforward neural networks. 

This is a very interesting approach to ESG, especially since, as mentioned, there is a ‘lack of uniform standards for assessing and measuring impact.’

The theoretical part reviews a wide range of international research on ESG activities, presenting both their positive and negative impact on companies' financial performance.

The materials and methods are clearly described. The authors used not only traditional methods but also modern machine learning techniques in their research, which is an addedd value. The assumptions made in the methodological section are explained in a clear and transparent manner.

The article is written in an interesting and understandable manner. The research results are very appealing and contribute new content to the topic described. 

However, in my opinion, two issues should be improved:

 - the questions/hypotheses were not highlighted,

- lines 63-65 – ‘several studies’ were mentioned, but only one publication on the subject was cited.

Author Response

Thank you for your valuable feedback. We wrote our responses to the attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper contributes to a better understanding of the effects of ESG actions on companies’ financial performance and provides insights into how investors and firms can integrate ESG ratings into their assessments. The authors investigate this problem using various financial indicators alongside ESG scores, estimating the relationship through both linear and nonlinear approaches, including classical linear regression and neural networks combined with explainable AI techniques. Their results reveal interesting differences in the effects on ROA versus ROE.

While the study addresses a highly relevant and timely topic, substantial improvements are required in both the analysis and presentation. Addressing the points outlined below will significantly enhance the paper’s clarity, rigor, and overall contribution.

Specific Comments:

• On lines 341, 381, 441, and 442, references are missing. The error message "Reference source not found" appears instead. Please correct.
• The regression results currently presented as Figure 1 are difficult to read. Please insert them as a properly formatted table rather than as a figure. The same adjustment is needed for Figure 2. For readability, it is sufficient to report coefficient values and p-values; standard errors, t-statistics, and confidence intervals can be omitted. Highlighting statistically significant coefficients in bold would also improve clarity.
• Since the linear model includes variables that may be serially correlated, please also report p-values based on an autocorrelation-robust variance-covariance matrix for the coefficient estimators.
• In Figure 3, the label of the x-axis is not informative. It should explicitly refer to the ESG index. Please revise.
• The effect on ROA displayed in Figure 3 appears to be quadratic in the ESG index. Could this finding be translated into the regression framework by including a quadratic ESG score term? This would allow for a more robust econometric analysis of the effect and provide results that are more easily interpretable.
• The authors should explicitly discuss the limitations of their analysis and outline potential directions for future research in the conclusion.

Author Response

Thank you for your valuable feedback. We wrote our responses to the attached file.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

I would like to thank once again Sustainability for the opportunity to serve as a reviewer for this prestigious journal. Regarding the review of the manuscript entitled “The Impact of ESG on Business Performance: Comparison of Existing Literature and Empirical Analysis of NASDAQ–NYSE-listed Companies” (sustainability-3871334), although under normal conditions this work could well be rejected outright, I believe it deserves a chance given that it is in its first round of review. I hereby submit my decision:

Reconsider after major revision

• Ambiguity of the title: “The Impact of ESG on Business Performance: Comparison of Existing Literature and Empirical Analysis of NASDAQ–NYSE-listed Companies”. This work is either a comparison of existing literature or an empirical analysis conducted on companies listed on NASDAQ–NYSE, although it seems to be mainly the latter. Therefore, please radically change the title of this work, bearing in mind that what you call “Comparison of Existing Literature” is nothing more than a simple review of the scientific literature on the subject (“State of the Art”).
• Regarding the above, the literature review is somewhat limited, since the concept of “risk,” although mentioned, is practically absent. This is particularly relevant considering that ESG stocks often exhibit quite interesting relationships with risk—or rather, with several types of risk. In this regard, and for the sake of an obvious and balanced conceptualization, please include in the literature review at least the following articles:
  • Shushi, T. (2025). ESG-driven optimal portfolio selection for separated environmental, social, and governance preferences. Operational Research, 25(2). doi:10.1007/s12351-025-00907-3
  • Bertelli, B., & Torricelli, C. (2025). Sustainable optimal stock portfolios: What relationship between sustainability and performance? European Journal of Operational Research, 323(1), 323–340. doi:10.1016/j.ejor.2025.01.021
  • Korzeb, Z., Karkowska, R., Matysek-Jędrych, A., & Niedziółka, P. (2024). How do ESG challenges affect default risk? An empirical analysis from the global banking sector perspective. Studies in Economics and Finance, 42(1), 89–114. doi:10.1108/sef-09-2023-0540
  • Martín-Cervantes, P. A., & Valls Martínez, M. del C. (2023). Unraveling the relationship between betas and ESG scores through the Random Forests methodology. Risk Management, 25(3). doi:10.1057/s41283-023-00121-5
• I believe the authors have certain problems distinguishing between “tables” and “figures” in this manuscript. In this respect, Figures 1 and 2 are not figures at all, since they are actually tables. It seems that the authors have taken a screenshot from a statistical program and then considered it as a graphic. Never, under any circumstances. Strictly speaking, these are tables.
• Please, when you submit a document to a scientific journal, review it carefully: I refer, for example, to lines 441–442 of the document (Error! Reference source not found.4).
• Figures 3 and 4 are not such, but rather Figure 3 a) title of the subfigure and b) title of the subfigure (in this case, I understand the variables ROA and ROE, respectively). In this regard, please remove from the title of each subfigure the label “esgClass3 – Sensitivity analysis.”
• Please, very important, again with the figures that are actually tables. It is not enough to list a series of statistical tests; they must also be commented upon. I am referring specifically to the AIC, BIC, Log-Likelihood criteria, or the Durbin-Watson test. Regarding the Jarque-Bera test or the characterization of skew/kurtosis, I urge you to reflect. Are they important for your analysis?
• In the final conclusions, please specify what limitations you encountered when carrying out this research.

With my best wishes in your personal and academic life.
The Reviewer

Author Response

Thank you for your valuable feedback. We wrote our responses to the attached file.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have appropriately answered all my comments. I recommend the revised version for publication.

Author Response

We sincerely thank the reviewer for their positive feedback and recommendation for publication.