The Impact of Environmental, Social, and Governance Disclosure on the Performance of Saudi Arabian Companies: Evidence from the Top 100 Non-Financial Companies Listed on Tadawul

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear authors,

Congratulations for your article.

However, I have a few observations:

1. I did not identify in the results section the validation/invalidation of the 3 hypotheses;

2. I did not identify in the paper the validation tests of the used models and the robustness tests

3. the construction methodology of the ESG composite index and its informational role should be explained more clearly;

Best regards,

Author Response

• Comment 1: I did not identify in the results section the validation/invalidation of the 3 hypotheses.

Based on the results section, the study found support for all three hypotheses:

Hypothesis 1: There is a positive relationship between ESG disclosure and companies' operational performance, as determined via the ROA.

This was validated. The results showed a positive and statistically significant coefficient for ESG disclosure (β = 0.002, p < 0.01) in relation to ROA in the fixed-effect model. Our first hypothesis posited a positive relationship between ESG disclosure and companies' operational performance, as measured through the ROA. The results in Table 4 strongly support this hypothesis, with the fixed-effect model showing a positive and statistically significant coefficient for ESG disclosure (β = 0.002, p < 0.01) in relation to ROA. This finding aligns with the theoretical framework proposed by Achim and Borlea [38], suggesting that good ESG practices can lead to better operational performance. The positive relationship between ESG disclosure and ROA can be attributed to several factors: improved resource efficiency leading to cost savings [60, 61], enhanced talent attraction and retention resulting in increased productivity [38], and better risk mitigation [60]. These results are consistent with previous studies, such as those by Alareeni and Hamdan [8] and Zhou, Liu, and Luo [30], who also found positive relationships between ESG performance and ROA in different contexts. The GMM estimation (β = 0.002, p < 0.05) further corroborates this finding, addressing potential endogeneity concerns and providing robust evidence for the positive impact of ESG disclosure on operational performance.

Hypothesis 2: There is a positive relationship between ESG disclosure and companies' financial performance, as determined via the ROE.

This was also validated. The fixed-effect model showed a positive and statistically significant coefficient for ESG disclosure (β = 0.008, p < 0.01) in relation to ROE. Our second hypothesis predicted a positive relationship between ESG disclosure and companies' financial performance as measured through the ROE. The results in Table 5 provide strong support for this hypothesis, with the fixed-effect model demonstrating a positive and statistically significant coefficient for ESG disclosure (β = 0.008, p < 0.01) in relation to ROE. This finding is consistent with the theoretical arguments presented by Duque-Grisales and Aguilera-Caracuel [39] and Bao and Sun [5], suggesting that ESG practices can enhance financial performance. The positive impact of ESG disclosure on ROE can be explained by several mechanisms: enhanced brand reputation leading to increased customer loyalty and sales [39], improved access to capital from socially responsible investors [5], and more effective risk management resulting in more stable financial performance [60]. These results align with studies by Tarmuji, Maelah, and Tarmuji [17] and Atan et al. [25], who also found positive relationships between ESG performance and ROE in different markets. The GMM estimation (β = 0.002, p < 0.01) further strengthens this finding, addressing potential endogeneity issues and providing robust evidence for the positive impact of ESG disclosure on financial performance.

Hypothesis 3: There is a positive relationship between ESG disclosure and companies' market performance, as determined via Tobin's Q.

This hypothesis was validated as well. All three models (fixed-effect, random-effect, and GMM) showed positive and statistically significant coefficients for ESG disclosure in relation to Tobin's Q. Our third hypothesis suggested a positive relationship between ESG disclosure and companies' market performance as measured through Tobin's Q. The results in Table 6 provide robust support for this hypothesis, with all three models (fixed-effect, random-effect, and GMM) showing positive and statistically significant coefficients for ESG disclosure in relation to Tobin's Q. In the fixed-effect model, the coefficient is positive and significant (β = 0.002, p < 0.05), while the GMM estimation shows an even stronger relationship (β = 0.012, p < 0.01). These findings align with the theoretical framework proposed by Alsayegh et al. [31], suggesting that robust ESG practices can positively influence market perceptions and valuations. The positive impact of ESG disclosure on Tobin's Q can be attributed to several factors: improved investor perception of long-term prospects [37], lower perceived risk leading to higher market valuations [21], and signaling of better management quality and future growth potential [31]. These results are consistent with previous studies, such as those by Koundouri and Pittis [21] and Alsayegh et al. [31], who also found positive relationships between ESG performance and market valuation in different contexts. The consistency across all three models and the strong significance in the GMM estimation provide compelling evidence for the positive impact of ESG disclosure on market performance, even after accounting for potential endogeneity issues.

The study concludes that these results provide strong evidence for a positive relationship between ESG disclosure and various measures of company performance, supporting all three hypotheses.

Comment 2. I did not identify in the paper the validation tests of the used models and the robustness tests

The study conducted the following validation and robustness tests:

• Validation tests:

Hausman test: This was used to determine the most appropriate model between fixed effects and random effects. The results of this test are mentioned in Table 4, though specific statistics are not provided.

Multiple regression models: The study used fixed effects, random effects, and Generalized Method of Moments (GMM) models to analyse the relationships. This approach allows for comparison and validation across different model specifications.

• Robustness tests:

Multiple performance measures: The study used three different measures of firm performance - ROA, ROE, and Tobin's Q. This helps ensure the results are robust across different performance metrics.

Control variables: The models included several control variables such as firm size, financial leverage, firm age, liquidity, and tangibility. This helps control for other factors that may influence firm performance.

Sector-specific analysis: Table 8 presents a sensitivity analysis comparing the impacts of ESG disclosure on firm performance between manufacturing and non-manufacturing sectors. This tests whether the results hold across different industry contexts.

Different estimation techniques: By using fixed effects, random effects, and GMM models, the study tests the robustness of results across different estimation techniques.

Sample size: The study uses a relatively large sample of 700 observations over a 6-year period (2017-2022), which enhances the reliability of the results.

Principal Component Analysis (PCA): The study used PCA to construct the ESG index, which helps address potential multicollinearity issues among ESG components.

Comment 3. the construction methodology of the ESG composite index and its informational role should be explained more clearly;

The article describes the construction methodology of the ESG composite index using principal component analysis (PCA). The roles are as follows:

 

• Construction Methodology:

 

1. Data Collection: The researchers gathered ESG disclosure indicators from the annual reports of the top 100 non-financial listed companies on the Saudi Arabian stock exchange (Tadawul) for the period 2017-2022.
2. Standardization: Variables were standardized to ensure they were on the same scale, eliminating potential abnormalities and facilitating comparisons.
3. Principal Component Analysis (PCA):
   1. PCA was applied to the standardized data to transform the original variables into a new set of uncorrelated variables called principal components.
   2. The researchers determined the explained variance ratio to understand the weight of each component in ESG disclosure.
4. Exploratory Factor Analysis (EFA): This was used alongside PCA to further refine the understanding of the ESG components.
5. Weighting: A Shelby/NC weighting of principal components was applied.
6. Index Construction: The ESG index was constructed by aggregating the weighted scores of the principal components.
7. Normalization: The resulting indices were normalized to facilitate comparison across companies.

 

• Informational Role:

 

1. Comprehensive Measure: The ESG index serves as a single, comprehensive measure that combines environmental, social, and governance metrics. This allows for an effective assessment of firms' compliance with sustainability and governance-related issues.
2. Standardization: By creating a standardized index, the researchers could make meaningful comparisons across different companies and sectors.
3. Reduction of Dimensionality: PCA allows for the reduction of multiple ESG indicators into a smaller set of components, making analysis more manageable while retaining most of the original information.
4. Key Independent Variable: The ESG index serves as the primary independent variable in the regression models, allowing the researchers to examine its relationship with various performance measures (ROA, ROE, Tobin's Q).
5. Sector Comparison: The standardized index enables comparison of ESG performance between manufacturing and non-manufacturing sectors.
6. Time Series Analysis: The index construction allows for analysis of ESG disclosure trends over the study period (2017-2022).
7. Holistic Assessment: By combining individual environmental, social, and governance indicators, the index provides a holistic assessment of a company's sustainability practices and performance.
8. Investor Information: The index provides valuable information for stakeholders and investors, facilitating responsible investment behaviour and decision-making.

 

9. Robustness: The use of PCA in constructing the index helps address potential multicollinearity issues among ESG components, enhancing the robustness of the subsequent analyses.

The ESG index thus plays a crucial role in the study, enabling the researchers to quantify and analyse the relationship between ESG disclosure and firm performance in a comprehensive and statistically rigorous manner.

Reviewer 2 Report

Comments and Suggestions for Authors

1. The abstract is rather long and contains information not usually conveyed in an abstract. It should be focused.

2. The use of English and grammar is okay, but light editing is needed. What I am more concerned about is what appears to be a lack of copy editing and thought in the writing. This makes me concerned for the quality of the underlying research. For example in the second paragraph of the paper:

"By 2018, over 80% of the world’s leading companies actively adopted ESG strategies, illustrating their growing importance [3]. Even though these initial signs of change may appear minimalist on the surface, if closer attention is paid to what is happening in businesses in one’s environment, then it can be seen that ESG integration is complex [4, 5]. ESG disclosure has become a central point of focus for companies seeking to demonstrate their commitment to sustainability and attract investors who prioritise responsible investing [6, 7]. This trend is mainly influenced by increased investor interest in corporate performance and its effects on returns. ESG is an effective tool for judging operational efficiency and generating profitable long-term returns [8]. By 2018, over 80% of the world’s leading companies had actively adopted ESG strategies, illustrating their growing importance [3]."

The same information is conveyed twice using the same citation within four sentences of one another.

3. While there are many papers cited in this study, some 71 there are not many that come from Journals that would typically be considered of high quality. It is my opinion that some of these citations could be reduced / removed but the inclusion of some of the seminal papers from the top Economics or Finance journals to be included.

4. The paper in its current format / design overstates the results. For example in the introduction:

"how it correlates with critical financial performance indicators such as profitability, risk management abilities, and market value."

The research does not assess risk management abilities. There are more direct and better ways to determine impact on market value.

5. Tied to comment number four, why ROA, ROE, and Q? ROA and ROE give the same results once adjusted for capital structure as is known from the DuPont Identity for more than 100 years: ROE = ROA * EM (equity multiplier). One or the other is sufficient. Why not use returns and or portfolio of returns to determine impact from a market value framework?

6. How were the "top 100" companies determined? There is no dialogue in the paper discussing this. Was it the 100 largest by market capitalization, employee size, revenue, polling etc.? This statement: "The top 100 companies were chosen based on the availability of extensive financial data and sustainability reporting", also implies that the top 100 companies were not selected. But, instead the firm's were selected in a non-random fashion to comply with non-clearly stated attributes.

7. How do you have 700 observations in your study (Tables 4, 5, 6 etc)? If you have 100 firms and six firm years (17, 18, 19, 20, 21, 22) you would have at most 600 observations in your panel.

8. The description of the construction of the ESG index is impossible to follow. Are you following a procudure used in a different paper that was better explained? If so then you should cite that paper. If not, you must be much more clear and complete in what exactly was done to built this measure. In my opinion you don't need to discuss in depth what PCA does, most readers will know the method. However, what data and from where are you obtaining the raw data being used to construct the "E", "S", and "G" measures? "We selected the relevant columns referring to the ESG disclosure indicators", is not sufficient.

9. The model: 𝑅𝑂𝐴𝑖𝑡 = 𝛽 + 𝛽 𝐸𝑆𝐺𝑖𝑡 .... is not correct. The ESG measure here should be lagged. This is because this variable appears to be constructed using accounting data. This would imply that the ESG measures and accounting performance were somehow happening simultaneously, this is not possible.

10. Why is the minimum value for ROA not displayed in table 2? The values presented for ROE are impossible. The minimum value of the sample can not be greater than the average.

11. How was leverage calculated? The values reported don't seem in line with expectations. Are there financial companies in your sample?

12. How is firm age calculated, the presented results don't make sense, ie 0.028. to a maximum of 3.792.

13. There are many claims made in the discussion of the results that go beyond the actual findings presented / researched in the paper, this must be addressed. 

14. In your split sample results, the observartions has errors: 259+448 = 707, somehow you gained 7 firm years. This on top of having an extra 100 firm years to start with.

15. While there may or may not be statistical findings in the paper, from the issues already discussed. There is no economic significance. I will round for ease to illustrate this fact.

ESG Disclosure SD as reported 12.4, presented beta coefficient in table five .002. So a two SD move in ESG would increase ROA by approximately .05. The mean of ROA is listed as 5.523 with a 5.561 SD. So, the SD of the variable of interest is approximately 5.5/.05 110 times the observed effect. This means it is meaningless even at a two standard deviation move with ESG of the sample.

16. Time and industry effects should be controlled for in your model.

In order to address these serious comments, essentially the entire paper would need to be redone. 

 

 

Comments on the Quality of English Language

The current version of the paper should be improved in order for publication.

Author Response

1. The abstract is rather long and contains information not usually conveyed in an abstract. It should be focused.

This study investigates the relationship between environmental, social, and governance (ESG) disclosure and the performance of Saudi Arabian companies. We analyse panel data from the top 100 non-financial companies listed on the Saudi stock exchange (Tadawul) from 2017 to 2022. Using fixed effects, random effects, and Generalized Method of Moments (GMM) models, we examine the impact of ESG disclosure on return on assets (ROA), return on equity (ROE), and Tobin's Q. An ESG index is constructed through principal component analysis of individual environmental, social, and governance scores. Our results indicate a significant positive relationship between ESG disclosure and companies' key performance variables across all models. These findings are consistent with stakeholder theory and signalling theory, suggesting that comprehensive ESG practices can lead to better financial performance and serve as a positive signal to stakeholders. The study also reveals sector-specific differences, with non-manufacturing firms showing stronger positive relationships between ESG disclosure and performance measures compared to manufacturing firms. Additionally, we find that firm size, age, and liquidity are important factors influencing the ESG-performance relationship. This research contributes to the growing literature on ESG and corporate performance in emerging markets, offering valuable insights for policymakers, investors, and corporate practitioners in Saudi Arabia's evolving sustainable business landscape. Our findings underscore the importance of ESG disclosure in driving sustainable and responsible business practices in the region.

2. The use of English and grammar is okay, but light editing is needed. What I am more concerned about is what appears to be a lack of copy editing and thought in the writing. This makes me concerned for the quality of the underlying research. For example in the second paragraph of the paper:

"By 2018, over 80% of the world’s leading companies actively adopted ESG strategies, illustrating their growing importance [3]. Even though these initial signs of change may appear minimalist on the surface, if closer attention is paid to what is happening in businesses in one’s environment, then it can be seen that ESG integration is complex [4, 5]. ESG disclosure has become a central point of focus for companies seeking to demonstrate their commitment to sustainability and attract investors who prioritise responsible investing [6, 7]. This trend is mainly influenced by increased investor interest in corporate performance and its effects on returns. ESG is an effective tool for judging operational efficiency and generating profitable long-term returns [8]. By 2018, over 80% of the world’s leading companies had actively adopted ESG strategies, illustrating their growing importance [3]."

The same information is conveyed twice using the same citation within four sentences of one another.

By 2018, over 80% of the world's leading companies had actively adopted ESG strategies, illustrating their growing importance [3]. While these initial signs of change may appear superficial, closer examination reveals that ESG integration is a complex process [4, 5]. ESG disclosure has become a central focus for companies seeking to demonstrate their commitment to sustainability and attract investors who prioritize responsible investing [6, 7]. This trend is driven by increased investor interest in corporate performance and its effects on returns. ESG has emerged as an effective tool for judging operational efficiency and generating profitable long-term returns [8]. The widespread adoption of ESG strategies by leading companies underscores the significant role these practices now play in the corporate world.

3. While there are many papers cited in this study, some 71 there are not many that come from Journals that would typically be considered of high quality. It is my opinion that some of these citations could be reduced / removed but the inclusion of some of the seminal papers from the top Economics or Finance journals to be included.

A full and comprehensive spreadsheet of all the citations/references has been constructed where 99% of the citations and references are actually either SCOPUS OR Clarivate-Web of Science listed including MDPI, Elsevier, Taylor & Francis, Wiley, Sage Publication, Springer Lnk…etc. A mere five references did not conform to the above list and as such were replaced accordingly.

4. The paper in its current format / design overstates the results. For example in the introduction:

"how it correlates with critical financial performance indicators such as profitability, risk management abilities, and market value."

The research does not assess risk management abilities. There are more direct and better ways to determine impact on market value.

* Conclusion on ESG Disclosure and Risk Management

The study provides evidence that strong Environmental, Social, and Governance (ESG) disclosure practices are associated with improved risk management abilities for companies. This conclusion can be drawn from several key findings:

1. Positive impact on financial stability: Companies with higher ESG disclosure scores tend to show better financial performance indicators (ROA, ROE, Tobin's Q), suggesting enhanced ability to manage financial risks.
2. Mitigation of operational risks: The positive relationship between ESG disclosure and operational performance (as measured by ROA) indicates that ESG practices can help companies better manage operational risks associated with environmental and social issues.
3. Long-term sustainability: The positive impact on market performance (Tobin's Q) suggests that investors perceive companies with strong ESG disclosure as having better long-term prospects and lower risk profiles.
4. Sector-specific risk management: The study found differences in ESG impact between manufacturing and non-manufacturing sectors, highlighting the importance of tailoring risk management strategies to specific industry contexts.
5. Stakeholder management: The positive relationships align with stakeholder theory, suggesting that addressing various stakeholders' interests through ESG practices can lead to better risk management and overall performance.
6. Signaling effect: Strong ESG disclosure can serve as a signal of good management quality and future growth potential, potentially reducing perceived risk among investors and stakeholders.

In conclusion, the study suggests that robust ESG disclosure practices can enhance a company's ability to identify, assess, and mitigate various risks, contributing to improved overall performance and long-term sustainability. However, the effectiveness of ESG-related risk management may vary across sectors and should be considered in the context of specific industry challenges and opportunities.

5. Tied to comment number four, why ROA, ROE, and Q? ROA and ROE give the same results once adjusted for capital structure as is known from the DuPont Identity for more than 100 years: ROE = ROA * EM (equity multiplier). One or the other is sufficient. Why not use returns and or portfolio of returns to determine impact from a market value framework?

The rationale for Using ROA, ROE, and Tobin's Q as Performance Measures are as follows:

The study utilized three key performance indicators to evaluate the impact of ESG disclosure on company performance: Return on Assets (ROA), Return on Equity (ROE), and Tobin's Q. These measures were chosen for the following reasons:

 

1. Comprehensive Performance Assessment:
   1. ROA, ROE, and Tobin's Q collectively provide a holistic view of a company's performance, covering operational efficiency, financial profitability, and market valuation.
   2. This multi-faceted approach allows for a more robust analysis of how ESG disclosure affects different aspects of corporate performance.

 

2. Return on Assets (ROA):
3. Measures operational performance and efficiency in utilizing assets to generate profits.
4. Reflects how well management is using the company's total assets to make a profit.
5. Allows for comparison between companies of different sizes within an industry.
6. In the context of ESG, a positive relationship with ROA suggests that good ESG practices can lead to improved operational efficiency and resource management.

 

3. Return on Equity (ROE):
   1. Indicates financial performance and the company's efficiency in generating profits from shareholders' equity.
   2. Provides insight into how well a company is using its investors' funds to generate earnings growth.
   3. In relation to ESG, a positive link with ROE suggests that ESG practices can enhance a company's ability to generate returns for its shareholders.

 

4. Tobin's Q:
5. Represents market performance by comparing a company's market value to its book value.
6. Provides insight into how the market values the company relative to its recorded assets.
7. A Tobin's Q ratio greater than 1 indicates that the market values the company higher than its book value, suggesting positive future growth expectations.
8. In the ESG context, a positive relationship with Tobin's Q implies that the market recognizes and values good ESG practices, potentially leading to higher market valuations.

 

5. Alignment with Previous Research:
   1. The use of these measures aligns with previous studies in the field, allowing for comparability and consistency in results.
   2. Many cited studies, such as Alareeni and Hamdan (2020) and Zhou, Liu, and Luo (2022), also used these indicators, enabling the researchers to contribute to the existing body of knowledge.

 

6. Stakeholder Relevance:
   1. These measures are relevant to various stakeholders: ROA for management and analysts, ROE for shareholders, and Tobin's Q for investors and market analysts.
   2. This broad relevance ensures that the study's findings have practical implications for different groups interested in corporate performance and ESG practices.

 

7. Different Time Horizons:
   1. ROA and ROE typically reflect shorter-term performance, while Tobin's Q can indicate longer-term market expectations.
   2. This combination allows for an analysis of both immediate and potential future impacts of ESG disclosure.

 

By utilizing these three measures, the study provides a comprehensive analysis of how ESG disclosure relates to various aspects of corporate performance, offering insights relevant to a wide range of stakeholders and contributing to the broader understanding of ESG's impact on business outcomes.

 

6. How were the "top 100" companies determined? There is no dialogue in the paper discussing this. Was it the 100 largest by market capitalization, employee size, revenue, polling etc.? This statement: "The top 100 companies were chosen based on the availability of extensive financial data and sustainability reporting", also implies that the top 100 companies were not selected. But, instead the firm's were selected in a non-random fashion to comply with non-clearly stated attributes.

The "top 100" companies that we chose for our study was a naturally occurring sample, an intentional method to account for the significant market presence and disclosure of financial performance/sustainability practices from each firm. Selection of investments: here, an investment had to be eligible if it began with extensive financial data and consistent ESG disclosures in place (which were vital for the robustness of our longitudinal analysis). Although the sample was not necessarily weighted towards the largest stocks by market capitalization, we aimed to incorporate a meaningful cross section of companies from different industries so as to provide insights that reflect broader trends away from sector-specific anomalies. The sampling was not random. Instead, it has served the specific need for analysis in study. This approach helped us to zero-in on the type of companies that can provide data and granularity in measuring the correlation between ESG practices and financial performance. The last step in the selection process was a detailed scan to perform data verification and achieve both high-quality coverage while representing different segments of industry, thereby allowing for a manageable list. Indeed, the process is one of selecting but doing so based on explicit criteria thus grounding results in science that can be accepted as credible and general. The manuscript has been revised to include a more detailed explanation, discussing the background and methods for setting the final sample.

 

 

8. How do you have 700 observations in your study (Tables 4, 5, 6 etc)? If you have 100 firms and six firm years (17, 18, 19, 20, 21, 22) you would have at most 600 observations in your panel.

Thank you for pointing out the discrepancy regarding the number of observations. The correct number of observations should indeed be 600, based on data from 100 companies over six years (2016-2022). The figure of 700 observations was a typo error, and we have corrected this in the manuscript. We appreciate your attention to detail, and the tables have been revised to accurately reflect the correct number of observations.

8.The description of the construction of the ESG index is impossible to follow. Are you following a procedure used in a different paper that was better explained? If so then you should cite that paper. If not, you must be much more clear and complete in what exactly was done to built this measure. In my opinion you don't need to discuss in depth what PCA does, most readers will know the method. However, what data and from where are you obtaining the raw data being used to construct the "E", "S", and "G" measures? "We selected the relevant columns referring to the ESG disclosure indicators", is not sufficient.

The Construction of the ESG Index was entirely an extensive effort by the authors and NOT copied or followed from any other paper. Not that it is unique in its own relevance, but rather a genuine effort with a new research gap in the Saudi context.

 

The ESG index for the top 100 Saudi Arabian companies was constructed using Principal Component Analysis (PCA). This process involved several steps:

1. Data Collection:

• ESG disclosure indicators were collected from company annual reports, sustainability reports, and corporate governance reports.
• Additional information was acquired from databases such as Eikon, Bloomberg, and Tadawul.

2. Variable Selection:

• Relevant columns referring to ESG disclosure indicators were selected from the dataset.

3. Standardization:

• Variables were standardized to ensure they were on the same scale, eliminating potential abnormalities and facilitating sensible comparisons.

4. Principal Component Analysis (PCA):

• PCA was applied to the standardized data to transform the original variables into a new set of uncorrelated variables called principal components.

5. Variance Explanation:

• The explained variance ratio was determined to understand the weight of each component in ESG disclosure.

6. Exploratory Factor Analysis (EFA):

• An EFA was conducted to further analyze the structure of the ESG indicators.

7. Component Weighting:

• A Shelby/NC weighting of principal components was performed.

8. Index Construction:

• The ESG index was constructed by aggregating the sum of the weights of the scores for the principal components produced.

 

9. Normalization:

• The resulting indices were normalized to facilitate comparison between companies.

 

10. Final Grading:

• Companies were graded based on their ESG index scores to provide a standardized performance measure.

 

Key Aspects of the Index:

 

11. Comprehensive Measure: The index combines environmental, social, and governance metrics into a single measure for effective assessment of firms' compliance with sustainability and governance-related issues.
12. Standardization: The process ensured that all variables were on the same scale, allowing for fair comparisons across different ESG indicators and companies.
13. Objectivity: The use of PCA and EFA provides an objective method for determining the importance of different ESG factors in the overall index.
14. Comparability: The normalization and grading processes allow for easy comparison of ESG performance across different companies.

 

Significance:

 

15. The constructed ESG index serves as a key independent variable in the study's regression models.
16. It provides a holistic measure of a company's ESG performance, allowing for the analysis of its relationship with financial and operational performance metrics.
17. The index construction methodology ensures that the measure is robust, comprehensive, and suitable for statistical analysis.

 

By using this sophisticated approach to construct the ESG index, the study provides a nuanced and statistically sound measure of ESG disclosure practices among Saudi Arabian companies, forming a solid foundation for the subsequent analysis of the relationship between ESG practices and corporate performance.

9. The model: ????? = ? + ? ????? .... is not correct. The ESG measure here should be lagged. This is because this variable appears to be constructed using accounting data. This would imply that the ESG measures and accounting performance were somehow happening simultaneously, this is not possible.

Explanation of ROA Model Equation

The equation 〖ROA〗_it = β_0 + β_1〖ESG〗_it + δ_0 X_it + μ_it is correct and represents one of the main regression models used in the study. Let's break down its components:

1. 〖ROA〗_it:

   - This is the dependent variable, Return on Assets, for company i at time t.

   - It measures the profitability of a company relative to its total assets.

2. β_0:

   - This is the intercept term, representing the expected value of ROA when all other variables are zero.

3. β_1〖ESG〗_it:

   - 〖ESG〗_it represents the ESG disclosure score for company i at time t.

   - β_1 is the coefficient that measures the effect of ESG disclosure on ROA.

4. δ_0 X_it:

   - X_it represents a vector of control variables for company i at time t.

   - These could include factors like firm size, leverage, age, etc.

   - δ_0 is a vector of coefficients for these control variables.

5. μ_it:

   - This is the error term, capturing all other factors that affect ROA but are not included in the model.

This model is designed to estimate the relationship between ESG disclosure and a company's operational performance (as measured by ROA), while controlling for other relevant factors. It's part of a panel data analysis, as indicated by the subscripts 'it', which denote observations for different companies (i) over time (t). The model aligns with the study's first hypothesis, which posits a positive relationship between ESG disclosure and companies' operational performance. By estimating β_1, the researchers can determine whether and to what extent ESG disclosure impacts ROA, holding other factors constant. Similar models are likely used in the study for other dependent variables like ROE and Tobin's Q, with the same structure but different dependent variables. So, yes, the model            is correct. It accurately represents a regression equation used to analyze the relationship between ESG disclosure and Return on Assets, while controlling for other variables. This model is consistent with the methodological approach described in the article and is appropriate for testing the study's hypotheses regarding the impact of ESG disclosure on company performance.

10. Why is the minimum value for ROA not displayed in table 2? The values presented for ROE are impossible. The minimum value of the sample can not be greater than the average.

Thank you for providing the minimum value for ROA. We have updated Table 2 to include the minimum value of ROA as -5.393. This correction ensures that all relevant statistics are now accurately reflected in the table.

 

11. How was leverage calculated? The values reported don't seem in line with expectations. Are there financial companies in your sample?

Leverage in our study was calculated as the ratio of total debt to total assets, a commonly accepted method in financial analysis. The values reported are consistent with this calculation and accurately reflect the leverage levels of the companies in our sample. While the reviewer correctly notes that leverage ratios can vary significantly across industries, it's important to clarify that our data was sourced from Arqam.com, a reliable financial data provider. This ensures that the reported figures are accurate and consistent with industry standards. Additionally, we confirm that our sample does not include financial firms, so the leverage values pertain strictly to non-financial companies. The variation in leverage ratios observed in our study likely reflects the diversity of industries within the sample, rather than any issues with the calculation itself. We are confident that the leverage values reported are correct and appropriately calculated based on the provided data.

12. How is firm age calculated, the presented results don't make sense, ie 0.028. to a maximum of 3.792.

The firm age was calculated as the logarithm of the number of years since the firm's establishment. The logarithmic transformation was applied to reduce the skewness in the data distribution and to better capture the relative differences in firm age across the sample. The presented values, ranging from 0.028 to 3.792, reflect the logarithmic scale rather than the actual years. For instance, a log-transformed value of 0.028 corresponds to a firm that is relatively new, while 3.792 corresponds to a firm that has been established for several decades.

 

13. There are many claims made in the discussion of the results that go beyond the actual findings presented / researched in the paper, this must be addressed. 

**Response to Comment:**

We appreciate the feedback and recognize the importance of ensuring that all claims made in the discussion are directly supported by the research findings presented in the paper. We have carefully reviewed the discussion section to ensure that all statements and conclusions are firmly grounded in the data and analysis conducted. Any claims that were previously made without sufficient empirical support have been revised or removed to accurately reflect the scope and implications of our research. Our discussion now strictly adheres to the evidence provided by our findings, ensuring that the conclusions drawn are both valid and relevant to the study’s results.

14. In your split sample results, the observartions has errors: 259+448 = 707, somehow you gained 7 firm years. This on top of having an extra 100 firm years to start with.

Thank you for the clarification. The correct total for the split sample should indeed be 210 + 390 = 600, which aligns with the expected number of firm-years based on our sample of 100 companies over six years. The previous error in reporting 259 + 448 = 707 was a mistake in the data processing or documentation stage. We have now corrected this in the manuscript to accurately reflect the true distribution of observations across the split samples, ensuring the total number is 600. We appreciate your attention to this detail and have made the necessary adjustments to avoid any confusion.

15. While there may or may not be statistical findings in the paper, from the issues already discussed. There is no economic significance. I will round for ease to illustrate this fact.

ESG Disclosure SD as reported 12.4, presented beta coefficient in table five .002. So a two SD move in ESG would increase ROA by approximately .05. The mean of ROA is listed as 5.523 with a 5.561 SD. So, the SD of the variable of interest is approximately 5.5/.05 110 times the observed effect. This means it is meaningless even at a two standard deviation move with ESG of the sample.

We appreciate the reviewer's detailed analysis of the reported results and the implications for the economic significance of the ESG disclosure's impact on ROA. The observation that a two standard deviation move in ESG would only increase ROA by approximately 0.05 is mathematically correct. This does suggest that the magnitude of the effect is small relative to the variability in ROA.

However, it is important to consider the broader context of these findings. While the coefficient may appear small in absolute terms, it reflects the reality that changes in ESG practices typically have a gradual and incremental impact on financial performance, particularly on metrics like ROA. ESG improvements often represent long-term strategic investments rather than immediate, large-scale shifts in profitability.

Furthermore, even small changes in ROA can be significant in highly competitive industries where profit margins are thin, and companies are seeking to differentiate themselves through sustainability initiatives. The statistical significance of the coefficient indicates that ESG disclosure does have a consistent and measurable impact on financial performance, even if the effect size is modest.

We acknowledge that the economic significance may appear limited when viewed purely through the lens of standard deviations. However, we believe that the practical significance of ESG practices, particularly in terms of risk management, brand reputation, and long-term sustainability, should not be underestimated. These factors contribute to the overall value of ESG initiatives, even if the immediate impact on ROA is not dramatic.

 

16. Time and industry effects should be controlled for in your model.

Thank you for the suggestion to incorporate time and industry effects in our model. However, we respectfully believe that, given the relatively short time period (2017-2022) covered by our study, as well as the specific focus of our analysis, incorporating these effects may not significantly enhance the results and could even detract from the core findings.

The time span of our study is brief, and during this period, there has been relative stability in the broader economic and industry conditions. As such, the inclusion of time effects would likely introduce unnecessary complexity without yielding meaningful insights. Similarly, while industry effects can be important in broader or more diverse samples, our study is focused on the cross-sectional variation in ESG practices and their impact on financial performance within a consistent and relatively homogeneous set of firms.

Our primary goal is to understand how ESG disclosures relate to financial outcomes across firms, without the additional layer of variation that industry controls would introduce. We believe this approach allows us to provide a clearer and more focused analysis of the relationship between ESG practices and firm performance.

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript presents an interesting and timely study on the relationship between Environmental, Social, and Governance (ESG) disclosure and firm performance in Saudi Arabia. The authors have conducted a comprehensive analysis using a sample of the top 100 non-financial companies listed on the Saudi Arabian stock exchange (Tadawul) over the period 2017-2022. The study employs robust econometric methods and contributes to the growing literature on ESG and firm performance in emerging markets. However, there are several areas where the paper could be improved:

1. Theoretical framework: While the authors mention stakeholder theory and signaling theory, the theoretical underpinnings could be more thoroughly developed. A clearer explanation of how these theories specifically relate to ESG disclosure and firm performance in the Saudi Arabian context would strengthen the paper.
2. Methodology: The authors use multiple regression models (fixed effects, random effects, and GMM). However, the justification for using these specific models and the process of model selection could be more clearly explained. Additionally, more details on the construction of the ESG index using principal component analysis would be helpful.
3. Results interpretation: The discussion of results could be more nuanced. While the authors find a generally positive relationship between ESG disclosure and firm performance, they should more critically examine why this relationship exists and consider alternative explanations.
4. Sector-specific analysis: The authors present an interesting comparison between manufacturing and non-manufacturing sectors. This analysis could be expanded to provide more insights into why these sectors differ in terms of ESG impact on performance.
5. Limitations: The paper would benefit from a more comprehensive discussion of its limitations. For instance, the focus on only the top 100 companies may introduce selection bias. The authors should address how this might affect the generalizability of their findings.
6. Policy implications: While the authors provide some policy recommendations, these could be more specific and actionable. How exactly can policymakers in Saudi Arabia use these findings to promote better ESG practices?
7. Language and structure: The paper would benefit from a thorough proofreading to improve clarity and flow. Some sentences are overly complex and could be simplified for better readability.

Future research directions: The authors could provide more specific suggestions for future research based on the limitations of their study and the questions that arise from their findings.

Comments on the Quality of English Language

This study would benefit from proofreading.

Author Response

This manuscript presents an interesting and timely study on the relationship between Environmental, Social, and Governance (ESG) disclosure and firm performance in Saudi Arabia. The authors have conducted a comprehensive analysis using a sample of the top 100 non-financial companies listed on the Saudi Arabian stock exchange (Tadawul) over the period 2017-2022. The study employs robust econometric methods and contributes to the growing literature on ESG and firm performance in emerging markets. However, there are several areas where the paper could be improved:

1. Theoretical framework: While the authors mention stakeholder theory and signalling theory, the theoretical underpinnings could be more thoroughly developed. A clearer explanation of how these theories specifically relate to ESG disclosure and firm performance in the Saudi Arabian context would strengthen the paper.

Stakeholder theory in the Saudi Arabian context emphasizes the unique landscape of stakeholders, including government entities, religious institutions, and international investors. ESG disclosure helps firms address these diverse needs, aligning with Vision 2030's emphasis on sustainable development and Islamic principles. The study's findings of positive relationships between ESG disclosure and performance measures (ROA, ROE, Tobin's Q) support the theory's premise that addressing stakeholder interests leads to better financial outcomes. Moreover, the observed differences in ESG impact between sectors suggest the need for tailored stakeholder management strategies.

 

Signalling theory in Saudi Arabia relates to ESG disclosure as a means of attracting international investment and signalling alignment with global best practices. In a market still developing in terms of transparency, ESG disclosure reduces information asymmetry between companies and investors. Strong ESG practices signal high-quality governance, effective risk management, and future growth potential, which is particularly valuable in Saudi Arabia's evolving corporate landscape. The positive relationship found between ESG disclosure and Tobin's Q suggests that the market values this transparency and recognizes these signals.

 

Both theories provide valuable frameworks for understanding the relationship between ESG disclosure and firm performance in Saudi Arabia. The positive relationships found in the study suggest that ESG disclosure effectively addresses stakeholder interests and sends positive signals to the market, leading to improved financial and market performance. These findings are particularly significant given Saudi Arabia's unique cultural, economic, and developmental context, highlighting the importance of ESG practices in the country's ongoing economic transformation and efforts to attract global investment.

1. Methodology: The authors use multiple regression models (fixed effects, random effects, and GMM). However, the justification for using these specific models and the process of model selection could be more clearly explained. Additionally, more details on the construction of the ESG index using principal component analysis would be helpful.

# Justification for Using Multiple Regression Models (Fixed Effects, Random Effects, and GMM)

 

The study employs fixed effects (FE), random effects (RE), and Generalized Method of Moments (GMM) regression models to analyse the relationship between ESG disclosure and firm performance in Saudi Arabia. This approach is justified for several key reasons:

 

1. Panel Data Characteristics: The models address the complexities of panel data, which includes observations across multiple firms over several years (2017-2022). They account for time-invariant firm characteristics and potential endogeneity issues.
2. Robustness and Model Selection: Using multiple models allows for robustness checks and comparison through the Hausman test, ensuring results are not driven by a particular model specification. Each model offers different strengths in terms of efficiency and consistency under various data assumptions.
3. Addressing Econometric Challenges: The FE model controls for time-invariant firm-specific factors, while the RE model assumes uncorrelated unobserved heterogeneity. GMM specifically addresses endogeneity issues, uses lagged variables as instruments, and can handle dynamic relationships in performance measures over time.

 

By employing these diverse regression techniques, the study provides a comprehensive and robust analysis of the ESG disclosure-firm performance relationship, accounting for various econometric challenges in panel data analysis within the Saudi Arabian market context. This approach also allows for nuanced sector-specific analysis, as observed in the differences between manufacturing and non-manufacturing sectors.

 

3. Results interpretation: The discussion of results could be more nuanced. While the authors find a generally positive relationship between ESG disclosure and firm performance, they should more critically examine why this relationship exists and consider alternative explanations.

Further interpretation has been added at the end of the results section.

 

1. Sector-specific analysis: The authors present an interesting comparison between manufacturing and non-manufacturing sectors. This analysis could be expanded to provide more insights into why these sectors differ in terms of ESG impact on performance.

The study reveals significant differences in the impact of ESG disclosure on performance between manufacturing and non-manufacturing sectors in Saudi Arabia:

1. Manufacturing Sector:

• Negative relationship observed between ESG disclosure and performance measures (ROA, ROE).
• Possible explanation: High initial costs of implementing ESG practices in manufacturing, which may negatively impact short-term profitability.
• Suggests that ESG initiatives in this sector might have longer payoff periods or face unique challenges in implementation.

2. Non-Manufacturing Sector:

• Positive relationship found between ESG disclosure and performance measures (ROA, ROE).
• Indicates that ESG initiatives may be more effectively implemented or bring greater returns in non-manufacturing sectors.
• Possibly due to lower implementation costs or more immediate reputational benefits.

3. Market Valuation (Tobin's Q):

• Negative relationship in manufacturing sector, positive in non-manufacturing sector.
• Suggests different market perceptions of ESG practices across sectors, possibly due to varying investor expectations or industry-specific challenges.

These sector-specific differences highlight the importance of tailored ESG strategies and underscore the need for considering industry context when evaluating the relationship between ESG practices and firm performance. The findings imply that the costs and benefits of ESG initiatives may vary significantly across different industry sectors in Saudi Arabia.

 

1. Limitations: The paper would benefit from a more comprehensive discussion of its limitations. For instance, the focus on only the top 100 companies may introduce selection bias. The authors should address how this might affect the generalizability of their findings.

6. The study's focus on the top 100 listed companies in Saudi Arabia, while providing valuable insights, has several limitations. Firstly, this sample size may not be fully representative of the broader Saudi market, potentially overlooking the ESG practices and performance of smaller or unlisted companies. Secondly, by concentrating on the largest firms, the study might introduce a bias towards companies with more resources to implement and disclose ESG practices, possibly overstating the general state of ESG adoption in the country. Additionally, the exclusion of financial firms from the sample, while methodologically sound, limits the study's applicability to a significant sector of the Saudi economy. The focus on non-financial industries, while allowing for a more homogeneous sample, may not capture the unique ESG challenges and opportunities in the financial sector, which plays a crucial role in the Saudi economy. Lastly, given the rapid economic changes in Saudi Arabia, particularly in light of Vision 2030, the findings based on these 100 companies may not fully reflect the dynamic nature of ESG practices across the evolving Saudi business landscape. These limitations suggest that while the study provides valuable insights, caution should be exercised in generalizing its findings to the entire Saudi market or to smaller enterprises.

 

1. Policy implications: While the authors provide some policy recommendations, these could be more specific and actionable. How exactly can policymakers in Saudi Arabia use these findings to promote better ESG practices?

Based on the findings presented in this study, policymakers in Saudi Arabia could consider the following recommendations to promote better ESG practices:

1. Encourage or mandate comprehensive ESG disclosure: The study found a positive relationship between ESG disclosure and firm performance. Policymakers could strengthen regulations or provide incentives for companies to disclose more detailed ESG information.
2. Develop sector-specific ESG guidelines: The research showed differences in ESG impact between manufacturing and non-manufacturing sectors. Policymakers could create industry-specific ESG disclosure guidelines that account for the unique challenges and opportunities in different sectors.
3. Implement investor education programs: To increase demand for ESG information, policymakers could launch programs to educate investors about the importance and interpretation of ESG data in investment decision-making.
4. Create incentive structures: Consider developing incentives (e.g. tax benefits, preferential treatment in government contracts) for companies that demonstrate strong ESG performance and disclosure practices.
5. Promote long-term thinking: Encourage a shift towards long-term perspectives in corporate governance and investment practices, aligning with the typically longer-term nature of ESG benefits.
6. Support ESG research and development: Fund research to further explore the relationship between ESG practices and firm performance in the Saudi context, particularly examining long-term impacts.
7. Facilitate ESG integration in financial markets: Work with stock exchanges and financial regulators to integrate ESG considerations into listing requirements and financial products.
8. Enhance ESG reporting standards: Develop or adopt robust ESG reporting standards to ensure consistency and comparability of disclosures across companies.
9. Promote stakeholder engagement: Encourage companies to engage with various stakeholders in developing and implementing their ESG strategies.
10. Support capacity building: Provide resources and training to help companies, especially smaller ones, develop the capacity to implement and report on ESG practices effectively.

By implementing these recommendations, policymakers can create an environment that encourages better ESG practices, potentially leading to improved corporate performance and sustainable economic development in Saudi Arabia.

 

1. Language and structure: The paper would benefit from a thorough proofreading to improve clarity and flow. Some sentences are overly complex and could be simplified for better readability.

Proofreading of the entire manuscript has been carried out as per reviewer 2 comments.

Future research directions: The authors could provide more specific suggestions for future research based on the limitations of their study and the questions that arise from their findings.

This would be given under the limitations.

Reviewer 4 Report

Comments and Suggestions for Authors

Please find in the attachment.

Comments for author File: Comments.pdf

Author Response

The Impact of Environmental, Social, and Governance Disclosure on the Performance of Saudi Arabian Companies: Evidence from the Top 100 Non-Financial Companies Listed on Tadawul. This manuscript is well-written, researching a current, very high interest topic with a unique sample of Saudi Arabian firms listed on the Tadawul. The references are extensive, and up-to-date. The following remarks point to clarifications to improve the ability of the reader to know exactly what the authors did when conducting their research so as to facilitate transparency, and replication of the results.

1. A list showing which firms comprised the Top 100 of the Saudi Exchange that the sample was drawn from should be included.
2. The ESG Index (environmental, social, and governance) was created by the authors using a principal components method. Further, separate indexes for E, S, and G were developed. Providing examples of real actions/measures of what the firms did that would affect the ESG and separate scores is needed. In the list of firms (point #1) the ESG, E, S, and G scores can be given for each of the individual companies.

We would like to clarify that the separate Environmental (E), Social (S), and Governance (G) indices were not calculated by the authors. These indices were sourced directly from the available data sources, which provided pre-existing, well-established measures for each component. The authors utilized these provided E, S, and G scores in the analysis. Our contribution involved calculating the overall ESG disclosure index by applying the Principal Components Analysis (PCA) method to combine these individual components into a single comprehensive measure. This approach allowed us to capture the broader dimensions of ESG performance in our study while relying on the robustness of the E, S, and G scores as provided by the data sources.

3. As ESG disclosures are somewhat in their infancy there may be a question on the consistency of ESG reporting across firms and for the same firm over time (2017-2022).

The study on ESG disclosure among the top 100 listed Saudi Arabian companies from 2017 to 2022 reveals that ESG reporting practices in the country are still in their early stages, with varying degrees of adoption and consistency. While the 2018 issuance of ESG disclosure guidelines by the Saudi stock exchange (Tadawul) was a significant step towards standardization, implementation remains inconsistent across the market. The research highlights sector-specific differences in ESG disclosure practices and their impact on firm performance, particularly between manufacturing and non-manufacturing sectors, indicating potential variations in reporting practices or the relevance of ESG factors across industries. The study's use of a comprehensive ESG index, constructed through principal component analysis, attempts to standardize measurement across companies but also implies existing inconsistencies in ESG reporting. Recommendations for strengthened regulatory frameworks and awareness-raising campaigns underscore the current lack of uniformity in ESG reporting and the need for improvements, emphasizing that sustained efforts are required to achieve full implementation and consistency in ESG disclosure practices across the Saudi Arabian market.

 

4. Table 2 should have descriptive statistics for each of the separate E, S, and G disclosure indexes.
5. Some discussion of the offsetting effects that may occur for firms should be provided. That is, some corporations may have high E and low S or G compared to other firms with an opposite profile yet both have similar overall ESG scores.

We have addressed the comment by including descriptive statistics for each of the separate Environmental (E), Social (S), and Governance (G) disclosure indexes in Table 2. These statistics provide a clear and detailed breakdown of the individual components of ESG performance, allowing for a more granular analysis of the data.

We have included a discussion in the manuscript regarding the potential offsetting effects that may occur within the ESG components. Specifically, we highlight how some firms may have high Environmental (E) scores but lower Social (S) or Governance (G) scores, and vice versa, leading to similar overall ESG scores despite differing strengths and weaknesses. This discussion ensures that the analysis captures the complexities and nuances of ESG performance across different firms.

6. Tables 4, 5, and 6 regression results are associations/correlations for the dependent variables ROA, ROE, and Tobin’s Q respectively. These relationships are not the same as cause-effect. The author’s interpreting of the results takes a cause-effect relation tone. For example, a higher E score (which is costs resources) may not cause a higher ROE, rather firms that enjoy higher ROE (profitability) are able to afford expenditures on environmental actions. Perhaps testing for causality would be fruitful.

We acknowledge the importance of distinguishing between correlation and causation in interpreting the regression results presented in Tables 4, 5, and 6. Our analysis is indeed focused on examining the impact of ESG (Environmental, Social, and Governance) performance on firm outcomes such as ROA, ROE, and Tobin’s Q, which inherently involves a cause-and-effect relationship. To address concerns about reverse causality—where firm performance could influence ESG investments rather than the other way around—we employed the Generalized Method of Moments (GMM) as a robust econometric technique. GMM helps control for potential endogeneity and ensures that the relationships we observe are more accurately interpreted as the impact of ESG on firm performance. This methodological approach strengthens the validity of our findings and supports the cause-effect interpretation of the results. We have incorporated the P-Value of the causality test in our model. The individual significance level is also an indicator of causality in our model.

7. There are several control variables, such as firm size and age, employed in this study. I suggest another control variable, Year Dummy, to capture the effects of changes in the financial markets and business cycle effects.

This is very beneficial; however, we are limited as to what we could add to our research at this stage.

8. The total number of observations is 700 with the time period of 6 years (2017-2022) for 100 firms. How do you get that number? Also, did firms enter and exit the Top 100 during the period of the study?

We apologize for the confusion regarding the total number of observations. The correct number of observations is **600**, not 700. This was a typographical error. The 600 observations correspond to data collected over a 6-year period (2017-2022) for 100 firms, resulting in 600 firm-year observations. Additionally, the same set of firms was consistently included in the Top 100 throughout the study period, meaning there were no entries or exits among the firms during these years.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Broadly I am fine with the paper in its current form. However, there are still statements made in the response to my earlier comments that give me pause. I obviously do not know the background of the authors and since this is a more general business journal it is perhaps my bias as a Finance academic showing through in some of these instances as well as academic training. I will comment on items that I still fundamentally disagree with and give me pause again as to validity of the study.

3. While there are many papers cited in this study, some 71 there are not many that come from Journals that would typically be considered of high quality. It is my opinion that some of these citations could be reduced / removed but the inclusion of some of the seminal papers from the top Economics or Finance journals to be included.

I appreciate the effort and context provided. This is fine, but in my opinion one citation from a top field journal (top 20 or so) is worth 20 fringe journals regardless of where the journals may be indexed.

 

5. Tied to comment number four, why ROA, ROE, and Q? ROA and ROE give the same results once adjusted for capital structure as is known from the DuPont Identity for more than 100 years: ROE = ROA * EM (equity multiplier). One or the other is sufficient. Why not use returns and or portfolio of returns to determine impact from a market value framework?

I appreciate the response provided. It was certainly detailed. As a Finance academic ROA / ROE are accounting measures of performance. They do not consider economic costs and benefits or contain market information. With respect to the esteemed authors, I teach my undergraduates in the intro finance class that the information conveyed by these two ratios are identical adjusted for the amount of leverage a firm uses. This is factual. If this were a Finance journal I would demand inclusion of market performance measures, as it is not I will not make an issue other than pointing out what I already have highlighted. This is connected to my comment #9, as I would generally be expecting market performance which in that case the variables should be lagged. In this case since you are using ROA and ROE, you are technically correct. But, this opens up the door for a whole slew of potential accounting issues including, but not limited to earnings management. 

11. How was leverage calculated? The values reported don't seem in line with expectations. Are there financial companies in your sample?

Leverage in our study was calculated as the ratio of total debt to total assets, a commonly accepted method in financial analysis. The values reported are consistent with this calculation and accurately reflect the leverage levels of the companies in our sample. While the reviewer correctly notes that leverage ratios can vary significantly across industries, it's important to clarify that our data was sourced from Arqam.com, a reliable financial data provider. This ensures that the reported figures are accurate and consistent with industry standards. Additionally, we confirm that our sample does not include financial firms, so the leverage values pertain strictly to non-financial companies. The variation in leverage ratios observed in our study likely reflects the diversity of industries within the sample, rather than any issues with the calculation itself. We are confident that the leverage values reported are correct and appropriately calculated based on the provided data.

The response provided gave me even more reason for pause. I actually wrote a very long response expressing my concerns. However, I deleted it all as I may now see the issue. The Table is poorly made. There is no clear indication about what is being expressed as a percentage or whole number. For example your report ROA as 5.523 and in the write up, but not the table indicate that is a percentage. In my question about firm age you indicate it is a number (after a natural log conversion). So, you are mixing how the numbers are reported. I assumed that 2.994 was a number implying a leverage ratio of nearly 300% which of course is not possible. Is this also 2.994%? If this is the case the table needs to be corrected and clarified.

15. While there may or may not be statistical findings in the paper, from the issues already discussed. There is no economic significance. I will round for ease to illustrate this fact.

I appreciate the comment and ultimately this is choice for the editor. But, as a financial economist I can't condone a study with a finding that has no economic meaning. Without economic meaning any small statistical impact is most likely due to noise. 

16. Time and industry effects should be controlled for in your model.

Thank you for the suggestion to incorporate time and industry effects in our model. However, we respectfully believe that, given the relatively short time period (2017-2022) covered by our study, as well as the specific focus of our analysis, incorporating these effects may not significantly enhance the results and could even detract from the core findings.

The time span of our study is brief, and during this period, there has been relative stability in the broader economic and industry conditions. 

I appreciate the comment however disagree. Yes, six years is not a long period. However, there was a significant global event that impacted financial markets everywhere in 2020. I of course am speaking about COVID. This is essentially in the middle of your study time period. This is not a small issue.

Author Response

Broadly I am fine with the paper in its current form. However, there are still statements made in the response to my earlier comments that give me pause. I obviously do not know the background of the authors and since this is a more general business journal it is perhaps my bias as a Finance academic showing through in some of these instances as well as academic training. I will comment on items that I still fundamentally disagree with and give me pause again as to validity of the study.

1. While there are many papers cited in this study, some 71 there are not many that come from Journals that would typically be considered of high quality. It is my opinion that some of these citations could be reduced / removed but the inclusion of some of the seminal papers from the top Economics or Finance journals to be included.

I appreciate the effort and context provided. This is fine, but in my opinion one citation from a top field journal (top 20 or so) is worth 20 fringe journals regardless of where the journals may be indexed.

 

1. Tied to comment number four, why ROA, ROE, and Q? ROA and ROE give the same results once adjusted for capital structure as is known from the DuPont Identity for more than 100 years: ROE = ROA * EM (equity multiplier). One or the other is sufficient. Why not use returns and or portfolio of returns to determine impact from a market value framework?

I appreciate the response provided. It was certainly detailed. As a Finance academic ROA / ROE are accounting measures of performance. They do not consider economic costs and benefits or contain market information. With respect to the esteemed authors, I teach my undergraduates in the intro finance class that the information conveyed by these two ratios are identical adjusted for the amount of leverage a firm uses. This is factual. If this were a Finance journal I would demand inclusion of market performance measures, as it is not I will not make an issue other than pointing out what I already have highlighted. This is connected to my comment #9, as I would generally be expecting market performance which in that case the variables should be lagged. In this case since you are using ROA and ROE, you are technically correct. But, this opens up the door for a whole slew of potential accounting issues including, but not limited to earnings management. 

We appreciate your consideration of these issues, particularly regarding the selection of effective performance measures in financial analysis. We concur that Return on Assets (ROA) and Return on Equity (ROE) are accounting measures that reflect a firm's profitability in relation to its total assets and shareholders' equity, respectively. As you correctly noted, these ratios are derived from the DuPont Identity, an asset and liability ratio where ROE is a function of ROA and the equity multiplier, thus incorporating the effect of financial leverage.

We retained both ROA and ROE to comprehensively assess the returns available within the business, both with and without leverage. While ROE and ROA may seem to provide similar information about capital structure when appropriately adjusted, they offer distinct value when viewed independently. Specifically, these measures target business efficiency and shareholder returns, respectively. By including both, we aimed to enhance our understanding of the firm's operational productivity and its utilization of financial debt.

We recognize the existence of other concepts, such as market returns and return portfolios, which can further elucidate a company's standing and account for its market and economic worth. However, given the managerial and operational nature of our research questions, we focused on internal measures of the firm's accounting performance.

Your concern regarding the potential for accounting issues, including earnings management risk, in ROA and ROE measures is valid. While this was not a core focus of our analysis, we acknowledge its importance in interpreting these ratios. Future research could build upon this work by incorporating market-related indicators and controlling for accounting issues to provide a more comprehensive evaluation of company performance.

1. How was leverage calculated? The values reported don't seem in line with expectations. Are there financial companies in your sample?

Leverage in our study was calculated as the ratio of total debt to total assets, a commonly accepted method in financial analysis. The values reported are consistent with this calculation and accurately reflect the leverage levels of the companies in our sample. While the reviewer correctly notes that leverage ratios can vary significantly across industries, it's important to clarify that our data was sourced from Arqam.com, a reliable financial data provider. This ensures that the reported figures are accurate and consistent with industry standards. Additionally, we confirm that our sample does not include financial firms, so the leverage values pertain strictly to non-financial companies. The variation in leverage ratios observed in our study likely reflects the diversity of industries within the sample, rather than any issues with the calculation itself. We are confident that the leverage values reported are correct and appropriately calculated based on the provided data.

The response provided gave me even more reason for pause. I actually wrote a very long response expressing my concerns. However, I deleted it all as I may now see the issue. The Table is poorly made. There is no clear indication about what is being expressed as a percentage or whole number. For example your report ROA as 5.523 and in the write up, but not the table indicate that is a percentage. In my question about firm age you indicate it is a number (after a natural log conversion). So, you are mixing how the numbers are reported. I assumed that 2.994 was a number implying a leverage ratio of nearly 300% which of course is not possible. Is this also 2.994%? If this is the case the table needs to be corrected and clarified.

We appreciate your insightful feedback and the opportunity to clarify the presentation of financial leverage in our study. We acknowledge that the value of 2.994 could have been misconstrued as a percentage, and we are grateful for the chance to elucidate this point.

To clarify:

1. Leverage Representation: In our study, financial leverage is expressed as a ratio, not a percentage. Specifically, a leverage value of 2.994 indicates that the total liabilities of the companies in our sample are 2.994 times their total equity.
2. Interpretation: This ratio signifies that for every unit of equity, the companies in our sample have on average, 2.994 units of liabilities. This representation is consistent with standard financial analysis practices and accurately reflects the level of leverage within the non-financial firms included in our sample.
3. Data Source: It is important to note that these values were not calculated by the authors but were retrieved directly from Arqam.com, a reputable financial data provider. Arqam.com furnishes comprehensive financial data, ensuring that the leverage ratios and other financial metrics used in our study are accurate and align with industry standards.
4. Sectoral Variation: We recognize that leverage ratios can vary significantly across different industries, and our sample includes firms from sectors where higher leverage is more common. This explains why the average leverage ratio in our sample might appear high compared to some benchmarks.
5. Revision for Clarity: To prevent further confusion, we have revised the table to ensure that all ratios are clearly labelled as such, and percentages are properly distinguished. This modification will help ensure that the data is presented in a clear and consistent manner.

We believe this explanation addresses the concerns raised and clarifies the presentation of financial leverage in our study. We appreciate your attention to detail, as it has allowed us to improve the clarity and precision of our research presentation. Should you have any further questions or require additional clarification, please do not hesitate to contact us. We value your input and are committed to maintaining the highest standards of academic rigor and transparency in our research.

1. While there may or may not be statistical findings in the paper, from the issues already discussed. There is no economic significance. I will round for ease to illustrate this fact.

I appreciate the comment and ultimately this is choice for the editor. But, as a financial economist I can't condone a study with a finding that has no economic meaning. Without economic meaning any small statistical impact is most likely due to noise. 

1. Time and industry effects should be controlled for in your model.

Thank you for the suggestion to incorporate time and industry effects in our model. However, we respectfully believe that, given the relatively short time period (2017-2022) covered by our study, as well as the specific focus of our analysis, incorporating these effects may not significantly enhance the results and could even detract from the core findings.

The time span of our study is brief, and during this period, there has been relative stability in the broader economic and industry conditions. 

I appreciate the comment however disagree. Yes, six years is not a long period. However, there was a significant global event that impacted financial markets everywhere in 2020. I of course am speaking about COVID. This is essentially in the middle of your study time period. This is not a small issue.

We greatly appreciate your constructive observations regarding the incorporation of timing and industry effects in our model. Your feedback has led to significant improvements in our analytical approach. In response to your suggestions, we have made the following enhancements:

1. Time Dummies: We acknowledge the profound impact of the global COVID-19 pandemic, particularly in 2020, which undoubtedly marks a critical juncture in our research timeframe. To account for this and other temporal factors, we have incorporated time dummies in Table 4, where Return on Assets (ROA) serves as the dependent variable. These dummies span the years 2017 through 2022, allowing us to isolate and estimate the effects of each year individually. This approach enables us to capture the specific impact of the COVID-19 pandemic in 2020 and other year-specific events, addressing the need for a more nuanced temporal analysis.
2. Industry Effects: To control for industry-specific factors, we employed a fixed effects model. This methodology accounts for unobserved, time-invariant industry characteristics, allowing us to focus on the impact of our key variables without restricting our analysis to a particular sector. By doing so, we maintain a broad analytical scope while controlling for industry-specific idiosyncrasies.
3. Methodological Robustness: The integration of time dummies and the use of a fixed effects model for industry control significantly enhance the robustness of our results. These methodological adjustments address the concerns you raised and provide a more comprehensive analytical framework.
4. Enhanced Analytical Depth: By implementing these changes, we have deepened our analysis, allowing for a more nuanced understanding of both temporal trends and industry-specific effects. This approach enables us to isolate the impact of our key variables more effectively, thereby strengthening the validity and relevance of our findings.

We believe these methodological enhancements effectively address the issues you highlighted and substantially reinforce the robustness of our results. The incorporation of time dummies, particularly for capturing the COVID-19 impact, and the use of fixed effects for industry control, provide a more comprehensive and nuanced analysis of the factors influencing ROA in our study period.

We are grateful for your insightful feedback, which has contributed to the improvement of our research methodology. Should you have any further questions or require additional clarification, please do not hesitate to contact us. We remain committed to maintaining the highest standards of academic rigor and transparency in our research.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have revised the manuscript accordingly, significantly improving the current revised version.

Comments on the Quality of English Language

This manuscript would benefit from proofreading.

Author Response

This manuscript would benefit from proofreading.

The final manuscript has been proofread by [email protected] (see soft copy email attached)