Gender Diversity, ESG Performance, and Investment Efficiency: Evidence from Saudi Arabia

Abstract

This study examines the relationship between environmental, social, and governance (ESG) performance and investment efficiency and investigates whether board gender diversity moderates this relationship among Saudi listed firms. Using a sample of non-financial companies listed on the Saudi Stock Exchange (Tadawul) with available ESG scores over the period 2015–2023, the study employs panel regression analysis to assess the impact of ESG performance on investment efficiency. The findings indicate that higher ESG performance is associated with a greater tendency toward overinvestment rather than improved investment efficiency. However, board gender diversity significantly weakens this relationship, indicating a moderating effect of female board representation. The results remain robust across alternative specifications. This study advances to the ESG and corporate governance literature by providing empirical evidence from Saudi Arabia, an emerging market undergoing significant institutional reforms under Vision 2030 and highlights the importance of board gender diversity in improving the effectiveness of firms’ sustainability strategies.

1. Introduction

Efficient investment decisions are critical to firm value creation, yet whether Environmental, Social, and Governance (ESG) activities enhance or hinder firms’ investment efficiency remains an important question in the sustainable finance literature. The growing importance of sustainable investing has increased the importance of ESG performance in corporate and investment decision-making. Bloomberg Intelligence estimates that global ESG assets surpassed USD 30 trillion in 2022 and are expected to exceed USD 40 trillion by 2030, accounting for more than one-quarter of the projected USD 140 trillion in global assets under management. The continued growth of ESG assets reflects the increasing incorporation of ESG considerations into capital allocation decisions and the rising preference for sustainable investment approaches among investors (Bloomberg, 2024). This rapid expansion reflects investors’ increasing demand for responsible investment strategies and the growing integration of ESG information into capital allocation decisions. Consequently, the academic debate over the financial implications of ESG has intensified, with extensive research examining whether ESG activities create or destroy firm value (Al-Hiyari et al., 2023; Alobaid et al., 2024; Wu et al., 2024). Firms face rising pressure to boost their operational effectiveness and financial performances while fulfilling the high expectations of diverse groups of stakeholders to move beyond the legally mandated level of ESG efforts (Bilyay-Erdogan et al., 2024; Lian & Weng, 2024; J. Lin et al., 2023). Thus, investors and stakeholders now understand that ESG factors do not only facilitate the creation of a favorable business profile but do play an essential part in enhancing firm value. Investment efficiency reflects the degree to which a firm’s actual investment corresponds to its optimal, value-maximizing level, whereas departures from this benchmark indicate either overinvestment or underinvestment (Alobaid et al., 2024; Biddle et al., 2009). However, whether these sustainability initiatives ultimately improve or impair firms’ investment efficiency remains an unresolved empirical question.

The obligation of business management to make optimal investment decisions is of utmost importance, as these decisions have a substantial impact on the future profitability of the firm. Two competing views can explain the relationship between ESG and investment efficiency. On one hand, stakeholder theory posits that by adopting ESG practices, firms can reduce information asymmetries and agency conflicts, thereby addressing not only shareholders’ interests but also the needs of other stakeholders. This approach enhances the firm’s reputation, fosters trust and relationships, improves firm-level capital allocation efficiency, and ultimately increases firm value through a commitment to ethical and moral behaviors (Benlemlih & Bitar, 2018; Cornell & Shapiro, 1987). In contrast, agency theory suggests that ESG initiatives may increase agency costs and divert organizational resources from value-maximizing investments, potentially reducing investment efficiency (Jensen, 1986; Al-Hiyari et al., 2023). By comparison, slack-resources theory posits that firms with greater financial slack are better positioned to invest in ESG initiatives, as excess resources enable firms to undertake social and environmental activities (Waddock & Graves, 1997). From a trade-off perspective, Preston and O’bannon (1997) argue that ESG practices can consume capital and other limited firm resources, potentially leading firms to forego certain investment opportunities in order to achieve their ESG objectives. Furthermore, the symbolic adoption of ESG practices may be employed by managers to maintain legitimacy within the business community. Consequently, may also use that self-interested managers might utilize ESG initiatives as a form of greenwashing, pursuing personal benefits at the expense of stakeholders (Al-Hiyari et al., 2023). In light of the inconclusive findings and divergent viewpoints, researchers have not yet reached an agreement on the correlation between ESG performance and investment efficiency.

Mirza et al. (2020) argue that gender diversity has critical influence on the efficiency of investment. This argument is founded on the hypothesis that having female directors offers superior monitoring, reduces agency problems, and facilitates efficient resource allocation through monitoring. Gender diversity on boards has the capacity to strengthen the investment efficiency-ESG practice association through diversified perspectives with improved decision-making and corporate governance (Alkhawaja et al., 2023). Gender diversity makes efficient monitoring of ESG projects possible, improved risk management, and verification of proper integration of the firm’s operation and alignment with firm’s ESG practice strategic objectives (Alkhawaja et al., 2023). Following increased concerns regarding gender parity in high-level management, governments across the globe have implemented policies mandating increased women presence on company boards. The Sustainable Development Agenda 2030 places high value on the critical importance of gender equality and women’s empowerment in attaining sustainable development according to its Global Goals (Nicolò et al., 2022). Gender equality has been considered an immediate necessity of society, according to the agenda, and of paramount importance, an essential variable in strengthening environmental sustainability. Thus, this research contributes original value to the existing published research on ESG and investment efficiency through an examination of the influence of ESG policies on firm investment efficiency with an examination of board gender diversity’s influence on the relationship.

This study investigates the impact of ESG practices on firm investment efficiency and the moderating role of gender diversity, utilizing a sample of Saudi-listed firms. The Saudi market is selected for analysis for four key reasons. Firstly, aligning with Vision 2030, which prioritizes increasing women’s labor market participation. The Kingdom’s commitment to promoting gender diversity has resulted in substantial legislative and regulatory reforms. The economic implications of these reforms have not been thoroughly examined. Second, as Saudi Arabia reshapes its economy under Vision 2030, ESG considerations are playing an increasingly pivotal role in guiding businesses toward a future that balances profitability with environmental stewardship, social responsibility, and ethical governance (Grant Thornton, 2024). Third, previous research has predominantly concentrated on the determinants and impacts of ESG practices within developed economies, while there is limited literature exploring the significance of ESG practices for corporate investment decisions in emerging markets. Finally, examining the impact of gender diversity on the ESG–investment efficiency relationship in Saudi firms provides important insights for other emerging markets implementing similar gender inclusion strategies.

The findings provide new evidence of a significant and positive association between ESG score and investment inefficiency, suggesting that ESG performance at higher levels is associated with higher investment inefficiency. This finding indicates that while ESG initiatives can enhance the social legitimacy of company, they may simultaneously forward the focus of managers and resources away from value-maximizing investments, aligning with the view that excessive ESG involvement can lead to overinvestment or increased agency costs (Biddle et al., 2009). This result in line with prior studies (Ma & Ma, 2025), which find that ESG significantly reduce overall firm investment efficiency in the U.S. and Chinese contexts, respectively, where the cost of ESG involvement could decrease investment efficiency for the current period (Keisse & Jaafar, 2025). However, we find that the noted of the positive association is contingent by the extent of gender diversity among board members. In particular, we observe that the positive association between investments in ESGS and investment inefficiency becomes more prominent when a company has more female member in the board. This suggests that firms with greater diversity in boards are more likely to integrate ESG initiatives with strategy-driven investment objectives, thereby enhancing efficiency. Our finding when using alternative measures of ESGS and gender diversity, as well as when employing a two-stage least squares (2SLS) estimation approach.

2. Literature Review and Hypotheses Development

2.1. ESG and Investment Efficiency

ESG performance has become an increasingly important component of corporate strategy and investment evaluation. It reflects a firm’s ability to direct capital toward value-creating investments while avoiding both overinvestment and underinvestment. From a theoretical perspective, the relationship between ESG performance and investment efficiency remains inconclusive. Stakeholder theory argues that ESG practices enhance transparency, strengthen stakeholder relationships, reduce information asymmetry, and improve corporate governance, thereby facilitating more efficient investment decisions (Benlemlih & Bitar, 2018; Cornell & Shapiro, 1987). In contrast, agency theory suggests that ESG initiatives may increase agency costs and divert managerial attention from value-maximizing investments, thereby reducing investment efficiency (Al-Hiyari et al., 2023; Jensen, 1986). Meanwhile, slack resources theory argues that firms with greater organizational resources are better positioned to invest in ESG initiatives without compromising investment efficiency (Waddock & Graves, 1997).

Despite these competing theoretical perspectives, the majority of empirical studies support the view that ESG contributes to more efficient capital allocation. Early evidence demonstrates that firms with stronger corporate social responsibility (CSR) engagement tend to achieve greater investment efficiency by reducing both overinvestment and underinvestment (Benlemlih & Bitar, 2018). Similarly, Safi et al. (2023) and Afrin and Rahman (2024) show that socially responsible practices improve investment efficiency by mitigating agency conflicts, strengthening stakeholder relationships, and enhancing managerial decision-making. Extending this line of research, studies focusing on broader ESG performance consistently report that firms with superior ESG practices allocate capital more efficiently through enhanced transparency, lower information asymmetry, and improved market monitoring (Bilyay-Erdogan et al., 2024; Githaiga, 2025; Lian & Weng, 2024). Furthermore, Desai (2025) demonstrates that mandatory ESG disclosure regulations significantly improve investment efficiency, suggesting that regulatory enforcement enhances the credibility and usefulness of sustainability information for investment decisions. However, emerging evidence indicates that the economic consequences of ESG are not uniformly positive. Recent evidence suggests that ESG-related risks are associated with lower investment efficiency. Ma and Ma (2025) find that higher ESG performance is associated with lower investment efficiency among Chinese listed firms.

These studies suggest that ESG initiatives may not always improve capital allocation, particularly when the costs of ESG engagement outweigh their governance and informational benefits. Although stakeholder and slack resources theories suggest that ESG activities may improve investment efficiency by strengthening stakeholder relationships and enabling firms with abundant resources to undertake sustainability initiatives, the agency theory and resource-diversion perspective argue that extensive ESG engagement may divert managerial attention and corporate resources away from value-maximizing investments. This argument is particularly relevant in Saudi Arabia where the high ownership concentration, the substantial family and state influence, and an evolving ESG disclosure environment may constrain external monitoring and increase managerial discretion over sustainability-related expenditures. In this setting, ESG investments could be pursued beyond value maximizing amounts for legitimacy or reputational reasons, resulting in agency cost and resource-diversion mechanisms becoming predominant. Accordingly, Saudi-specific institutional features and recent governance reforms may shape the ESG–investment efficiency relationship and strengthen the monitoring role of gender-diverse boards. Therefore, the following hypothesis is proposed:

H1. 

ESG performance is positively associated with firms’ investment inefficiency.

2.2. Gender Diversity, ESG and Investment Efficiency

Recent research suggests that the relationship between ESG performance and investment efficiency is contingent upon firms’ governance structures. Rather than viewing ESG as a universally beneficial mechanism, scholars increasingly argue that board characteristics determine whether sustainability initiatives are effectively translated into corporate investment outcomes. For example, Al-Hiyari et al. (2023) find that board cultural diversity strengthens the positive relationship between ESG performance and investment efficiency by enhancing board monitoring and strategic decision-making. Conversely, Abbas et al. (2025) report that larger boards weaken this relationship, suggesting that ineffective governance structures may reduce the benefits derived from ESG initiatives. Collectively, these findings indicate that governance quality plays a critical role in shaping the effectiveness of ESG implementation.

Among governance mechanisms, board gender diversity has received increasing attention due to its influence on board effectiveness and corporate sustainability. Stakeholder and agency theories suggest that female directors contribute to stronger board monitoring, greater ethical awareness, enhanced stakeholder orientation, and more transparent decision-making. These attributes improve the board’s ability to oversee managerial actions, reduce agency conflicts, and ensure that ESG initiatives are aligned with firms’ long-term strategic objectives rather than managerial self-interest.

Empirical evidence generally supports this view. Studies consistently show that gender-diverse boards improve ESG disclosure and sustainability performance by enhancing transparency and accountability (Alkhawaja et al., 2023; Nicolò et al., 2022; Odriozola et al., 2024). Female directors have also been found to improve the quality of environmental disclosure (Abbasi et al., 2024) and strengthen investment decision-making through more effective monitoring and governance (Mirza et al., 2020). More directly, Keisse and Jaafar (2025) find that board gender diversity enhances the positive association between ESG and investment efficiency among European listed firms.

Despite these advances, important gaps remain. Existing evidence have focused on developed markets, especially the U.S., and on China, with relatively little evidence from Saudi Arabia and the Gulf Cooperation Council (GCC). Furthermore, although previous studies have examined several governance moderators—including board cultural diversity, board size, carbon sensitivity, and corporate reputation—the moderating role of board gender diversity in the ESG–investment efficiency relationship remains largely unexplored. Despite the recent evidence provided by Keisse and Jaafar (2025) from European firms, it remains unclear whether the moderating role of board gender diversity extends to the Saudi institutional environment, where differences in corporate governance, ownership structure, and regulatory development may influence ESG-related investment decisions. This gap is particularly relevant in Saudi Arabia, where Vision 2030 has substantially increased female participation in corporate leadership while simultaneously encouraging firms to strengthen ESG practices. Therefore, board gender diversity is expected to enhance the effectiveness of ESG initiatives by improving governance quality and strengthening the alignment between sustainability practices and corporate investment decisions.

H2. 

Board gender diversity moderates the relationship between ESG performance and investment efficiency.

3. Research Design

3.1. Sample Selection

To empirically examine the proposed hypotheses, this study concentrates on non-financial firms listed on the Tadawul Stock Exchange with available ESG scores over the period 2015 to 2023. The sample begins in 2015, providing a pre-Vision 2030 baseline against which subsequent developments following the launch of Saudi Vision 2030 in 2016 can be considered. This national agenda is likely to have stimulated an expansion of non-financial activities among Saudi firms. This assertion is supported by (Alharbi, 2021), who documented a notable increase in corporate social responsibility disclosures following the announcement of Vision 2030. The year 2023 was selected because it was the most recent fiscal year for which complete ESG data were available at the time of data collection.

Data for this study were collected from Refinitiv database and companies’ annual reports. ESG and financial data were extracted from the Refinitiv database, while corporate governance data and data missing from Refinitiv were manually collected from the companies’ annual reports. Newly listed firms were included from the year in which the required data became available, while delisted, insolvent, merged, or otherwise exiting firms were included for all years prior to their exit for which complete data were available. Additionally, to mitigate the impact of outliers on the regression results, all continuous variables were winsorized at the 1st and 99th percentiles. The unbalanced panel comprises all Saudi-listed firms that disclosed ESG information, resulting in a total of 342 firm-year observations from 14 industries.

3.2. Measurement of Variables

Investment Efficiency (INVEFF) is the dependent variable in this study and is measured following (Al-Hiyari et al., 2023; Alobaid et al., 2024; Biddle et al., 2009). Expected investment is estimated cross-sectionally for each year by regressing firms’ investment (INVEST), measured as capital expenditures scaled by lagged total assets, on sales growth (SGROWTH), measured as the percentage change in sales from year t − 2 to year t. The investment model is estimated as follows:

$${INVEST}_{it}={\beta }_0+{\beta }_1{SGROWTH}_{it-1}+{\epsilon }_{it}$$

(1)

where INVEST represents the investment of firm i in year t, calculated as capital expenditures divided by total assets in year t − 1, and SGROWTH denotes the percentage change in sales from year t − 2 to year t. The model is estimated cross-sectionally for each year in the sample period.

The residuals (ε) from Model (1) represent deviations from expected investment and serve as the proxy for investment inefficiency. Positive residuals indicate overinvestment, whereas negative residuals indicate underinvestment (Ismail & Abdul Wahab, 2024). Accordingly, in the subsequent regression analyses, a positive coefficient on an explanatory variable indicates a tendency toward greater overinvestment, whereas a negative coefficient indicates a tendency toward greater underinvestment. Coefficients that move the residuals closer to zero imply improvements in investment efficiency.

In terms of interest independent variables, this study has two primary independent test variables. The first is the ESG score (ESGS), obtained from Refinitiv Eikon, which serves as the main independent variable. It reflects a firm’s overall ESG performance, calculated annually as a weighted average of over 500 ESG metrics, with scores ranging from 0 to 100, where higher values indicate stronger ESG practices (Al-Hiyari et al., 2023; Samet & Jarboui, 2017). Based on stakeholder theory and prior empirical evidence, stronger ESG performance is expected to reduce deviations from expected investment, thereby improving investment efficiency. Accordingly, ESGS is expected to move investment residuals closer to zero rather than consistently in either a positive or negative direction.

The second variable is the percentage of female directors on the board (FBOD) (Ismail & Abdul Wahab, 2024), which is expected to strengthen the relationship between ESG performance and investment efficiency. The key variable of interest is the interaction term ESGS*FBOD. A significant positive (or negative) coefficient would imply that female board representation influences the strength and direction of the relationship between ESGS and investment efficiency.

Following prior studies (Al-Hiyari et al., 2023; Alobaid et al., 2024; Biddle et al., 2009; Ismail & Abdul Wahab, 2024; Samet & Jarboui, 2017), this study incorporates several control firm-level control variables to account for factors that may influence investment efficiency. These include firm size (FSIZE), net cash flows (NCFASST), profitability (ROE), market-to-book value (MTBV), sales growth (SGROWTH), leverage (LEVERAGE), and the ratio of property, plant, and equipment to total assets (PPEASST).

We also account for firms’ governance characteristics, including board size (BODSIZE), board independence (BODINDP), board expertise (BODEXP), audit committee independence (ACINDP), the presence of a nomination committee (PNOMCOMM), a compensation committee (PCOMPCOMM), a governance committee (PCGCOMM), and whether the CEO also serves as the board chair (CEODUAL). Industry and year fixed effects are incorporated into the empirical models to account for unobserved industry heterogeneity and macroeconomic conditions that may influence investment efficiency.

3.3. Research Models

To examine the relationship between ESG score and investment efficiency (H₁), Equation (2) is estimated using a random effect panel regression model.

$$\begin{gathered} {INVINEFF}_{it}={\beta }_0+{\beta }_1{ESGS}_{it}+{\beta }_2{FBOD}_{it}+{\beta }_3{FSIZE}_{it}+{\beta }_4{NCFASST}_{it}+{\beta }_5{ROE}_{it} \\ +{\beta }_6{MTBV}_{it}+{\beta }_7{SGROWTH}_{it}+{\beta }_8{LEVERAGE}_{it}+{\beta }_9{PPEASST}_{it} \\ +{\beta }_{10}{ACINDP}_{it}+{\beta }_{11}{BODEXP}_{it}+{\beta }_{12}{BODSIZE}_{it}+{\beta }_{13}{BODIND}_{it} \\ +{\beta }_{14}{PNOMCOMM}_{it}+{\beta }_{15}{PCOMPCOMM}_{it}+{\beta }_{16}{PCGCOMM}_{it} \\ +{\beta }_{17}{CEODUAL}_{it}+{\beta }_{18}{BIG4}_{it}+{\beta }_{19–31}{INDUSDUM}_{it} \\ +{\beta }_{32–38}{YEARDUM}_{it}+{\epsilon }_{it} \end{gathered}$$

(2)

Next, to test H₂, Equation (3) is estimated by incorporating the interaction term (ESGS*FBOD), which captures the moderating role of board gender diversity on the ESGS–investment efficiency relationship.

$$\begin{gathered} {INVINEFF}_{it}={\beta }_0+{\beta }_1{ESGS}_{it}+{\beta }_2{FBOD}_{it}+{\beta }_3{ESGS\ast FBOD}_{it}+{\beta }_4{FSIZE}_{it} \\ +{\beta }_5{NCFASST}_{it}+{\beta }_6{ROE}_{it}+{\beta }_7{MTBV}_{it}+{\beta }_8{SGROWTH}_{it} \\ +{\beta }_9{LEVERAGE}_{it}+{\beta }_{10}{PPEASST}_{it}+{\beta }_{11}{ACINDP}_{it}+{\beta }_{12}{BODEXP}_{it} \\ +{\beta }_{13}{BODSIZE}_{it}+{\beta }_{14}{BODIND}_{it}+{\beta }_{15}{\mathrm{P}\mathrm{N}\mathrm{O}\mathrm{M}\mathrm{C}\mathrm{O}\mathrm{M}\mathrm{M}}_{it} \\ +{\beta }_{16}{\mathrm{P}\mathrm{C}\mathrm{O}\mathrm{M}\mathrm{P}\mathrm{C}\mathrm{O}\mathrm{M}\mathrm{M}}_{it}+{\beta }_{17}{\mathrm{P}\mathrm{C}\mathrm{G}\mathrm{C}\mathrm{O}\mathrm{M}\mathrm{M}}_{it}+{\beta }_{18}{\mathrm{C}\mathrm{E}\mathrm{O}\mathrm{D}\mathrm{U}\mathrm{A}\mathrm{L}}_{it} \\ +{\beta }_{19}{\mathrm{B}\mathrm{I}\mathrm{G}4}_{it}+{\beta }_{20–32}{\mathrm{I}\mathrm{N}\mathrm{D}\mathrm{U}\mathrm{S}\mathrm{D}\mathrm{U}\mathrm{M}}_{it}+{\beta }_{33–39}{\mathrm{Y}\mathrm{E}\mathrm{A}\mathrm{R}\mathrm{D}\mathrm{U}\mathrm{M}}_{it}+{\epsilon }_{it} \end{gathered}$$

(3)

Investment inefficiency (INVINEFF) serves as the dependent variable, while ESG performance is measured using the ESG score (ESGS). Board gender diversity is represented by the proportion of female directors (FBOD). The operational definitions of all control variables are provided in

Table 1. Definition for variables.

Variables	Definitions	Sources	Used in
INVINEFF	Investment inefficiency	Calculated by author using data from Refinitiv database	(Alobaid et al., 2024; Biddle et al., 2009; Ismail & Abdul Wahab, 2024)
ESGS	The Refinitiv rating assesses a firm’s environmental, social, and governance (ESG) performance on a scale ranging from 0 to 100,	Refinitiv database	(Al-Hiyari et al., 2023)
FBOD	The proportion of female directors serving on a company’s board of directors, expressed as a percentage,	Refinitiv database/Annual report	(Ghaleb et al., 2024; Ismail & Abdul Wahab, 2024)
FSIZE	The natural logarithm of a firm’s total assets, used as a proxy for firm size,	Refinitiv database	(Ismail & Abdul Wahab, 2024)
NCFASST	Net cash flow from operating activities scaled by lag total assets, serving as a measure of operational cash-generating efficiency,	Refinitiv database	(Ismail & Abdul Wahab, 2024)
ROE	Return on equity, used as an indicator of a firm’s profitability by measuring the return generated on shareholders’ equity,	Refinitiv database	(Alobaid et al., 2024)
MTBV	Market to book value	Refinitiv database	(Al-Hiyari et al., 2023)
SGROWTH	Sales growth	Refinitiv database
LEVERAGE	Leverage, calculated as the ratio of total debt to total assets, indicating the firm’s reliance on debt financing,	Refinitiv database	(Al-Hiyari et al., 2023)
PPEASST	Property, Plant, and Equipment scaled by total assets,	Refinitiv database	(Al-Hiyari et al., 2023)
ACINDP	Percentage of independent directors in audit committee, measured as the number of independent directors divided by the total number of audit committee members.	Refinitiv database/Annual report	(Ismail & Abdul Wahab, 2024)
BODSIZE	Board size, measured by the total number of directors serving on the board.	Refinitiv database/Annual report	(Ismail & Abdul Wahab, 2024)
BODEXP	Percentage of directors with financial expertise, calculated as the number of financially expert directors divided by the total number of board members.	Refinitiv database/Annual report	(Ismail & Abdul Wahab, 2024)
BODINDP	Proportion of independent directors, measured as the number of independent directors divided by the total number of board members.	Refinitiv database/Annual report	(Ismail & Abdul Wahab, 2024)
PNOMCOMM	Presence of a nomination committee, represented by a dummy variable coded as “1” if the firm has a nomination committee and “0” otherwise.	Refinitiv database/Annual report	(Med Bechir & Jouirou, 2024)
PCOMPCOMM	Presence compensation committee, indicated by a dummy variable coded as “1” if the firm has a compensation committee and “0” otherwise.	Refinitiv database/Annual report	(Med Bechir & Jouirou, 2024)
PCGCOMM	Presence of a corporate governance committee, represented by a dummy variable coded as “1” if the firm has such a committee and “0” otherwise.	Refinitiv database/Annual report	(Med Bechir & Jouirou, 2024)
CEODUAL	CEO duality, indicating whether the CEO also serves as the Board Chairman, represented by a dummy variable coded as “1” if the roles are combined and “0” if they are separated.	Refinitiv database/Annual report	(Ullah et al., 2025)
BIG4	Audit quality, represented by a dummy variable coded as “1” if the company is audited by a Big 4 audit firm and “0” otherwise.	Refinitiv database/Annual report	(Al-Qadasi et al., 2023)
INDUSDUM	Indicator variables used to control for industry-specific effects.	Refinitiv database	(Ismail & Abdul Wahab, 2024)
YAERDUM	Indicator variables used to control for time-specific effects.	Refinitiv database	(Ismail & Abdul Wahab, 2024)

.

4. Results and Discussion

4.1. Descriptive Analysis and Pairwise Correlation of Study Variables

Table 2. Descriptive statistics.

Variables	Mean	Median	Sd. Dv.	p25	p75
INVINEFF	0.000	−0.014	0.040	−0.022	0.019
ESGS	33.253	33.345	18.936	16.610	45.850
FBOD	2.356	0.000	5.387	0.000	0.000
FSIZE	17.016	16.977	1.896	15.271	18.524
NCFASST	0.085	0.064	0.089	0.023	0.142
ROE	11.539	11.485	16.324	4.350	17.440
MTBV	2.809	2.000	2.280	1.340	3.560
SGROWTH	−0.492	−0.639	0.504	−0.910	−0.296
LEVERAGE	21.408	17.655	18.132	4.500	33.920
PPEASST	0.345	0.340	0.297	0.016	0.607
ACINDP	65.969	66.670	27.507	50.000	100.000
BODSIZE	3.902	3.820	2.289	2.580	5.410
BODEXP	9.526	9.000	2.335	9.000	11.000
BODINDP	43.244	42.860	12.353	33.330	50.000
PNOMCOMM	0.982	1.000	0.131	1.000	1.000
PCOMPCOMM	0.988	1.000	0.108	1.000	1.000
PCGCOMM	0.208	0.000	0.406	0.000	0.000
CEODUAL	0.035	0.000	0.184	0.000	0.000
Sample	342 firm-year observations

presents the descriptive statistics for all variables used in this study. The mean investment inefficiency residual (INVINEFF) is 0.000, with a standard deviation of 0.040 and an interquartile range from −0.022 to 0.019, indicating variation in firms’ deviations from expected investment, which is consistent with the findings reported by (Alobaid et al., 2024). The ESGS, which captures firms’ performance on environmental, social, and governance dimensions, has a mean value of 33.25, with substantial variation across firms (SD = 18.94), and an interquartile range between 16.61 and 45.85. Board diversity (FBOD), measured by the proportion of female directors serving on a company’s board of directors, expressed as a percentage, has a mean of 2.36 but a median of zero, indicating that a large proportion of firms in the sample do not have any female directors. This finding is consistent with the broader context of gender representation in corporate governance within the Saudi market.

In terms of control variables, FSIZE is normally distributed with a mean of 17.02, and ROE has a mean of 11.54 percent. The average NCFASST is 0.0845, while SGROWTH is negative on average (−0.49), possibly reflecting market contraction in some years. The leverage has a mean of 21.41% and a median of 17.66%, with a standard deviation of 18.13% and an interquartile range from 4.50% to 33.92%, suggesting a right-skewed distribution and considerable variation across firms, while the ratio PPEASST shows a mean of 0.345, a median of 0.340, a standard deviation of 0.297, and an interquartile range from 0.016 to 0.607, reflecting substantial heterogeneity in asset structure among firms. In terms of corporate governance attributes, ACINDP averages 66 percent, and boards tend to be relatively small (mean = 3.90 members). Most firms have nomination, compensation, and governance committees, although only about 20.8% of firms have a dedicated corporate governance committee). Finally, the CEO–Chairman separation variable indicates that the roles are combined in most cases, with separation occurring in only about 3.5% of firms.

Table 3. Pairwise Correlation Matrix and Variance Inflation Factors.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)	(17)	(18)	VIF
INVINEFF (1)	1.000
ESGS (2)	0.029	1.000																	1.53
FBOD (3)	0.032	0.189 ***	1.000																1.44
FSIZE (4)	−0.033	0.439 ***	0.020	1.000															2.06
NCFTASS (5)	0.312 ***	0.069	0.161 ***	−0.099	1.000														3.15
ROE (6)	0.066	0.092 *	0.181 ***	0.041	0.634 ***	1.000													2.46
MTBV (7)	0.157 ***	0.003	0.156 ***	−0.297 ***	0.518 ***	0.489 ***	1.000												1.97
SGROWTH (8)	0.000	−0.044	0.040	−0.420 ***	0.477 ***	0.490 ***	0.595 ***	1.000											2.34
LEVERAGE (9)	0.268 ***	−0.011	−0.069	0.050	−0.083	−0.179 ***	−0.060	−0.083	1.000										1.72
PPETASS (10)	0.552 ***	−0.098 *	−0.009	−0.101 *	0.435 ***	0.038	0.137 **	0.110 **	0.514 ***	1.000									2.59
ACINDP (11)	−0.129 **	0.184 ***	0.129 **	−0.056	0.056	0.054	0.145 ***	0.140 ***	−0.212 ***	−0.179 ***	1.000								1.33
BODEXP (12)	0.002	0.119 **	0.051	0.095 *	−0.078	−0.029	−0.152 ***	−0.135 **	−0.011	−0.045	−0.122 **	1.000							1.13
BODSIZE (13)	−0.073	0.043	0.004	0.223 ***	−0.061	−0.069	−0.105 *	−0.141 ***	−0.056	−0.077	−0.188 ***	−0.012	1.000						1.16
BODINDP (14)	−0.063	−0.008	−0.047	−0.179 ***	0.028	−0.066	−0.005	−0.015	−0.121 **	−0.139 ***	0.229 ***	0.073	0.009	1.000					1.24
PNOMCOMM (15)	0.061	0.014	0.059	0.055	−0.026	−0.160 ***	−0.156 ***	−0.156 ***	0.073	0.057	−0.003	0.069	0.021	0.040	1.000				3.52
PCOMCOMM (16)	0.035	0.023	0.048	0.011	0.057	−0.039	0.008	0.035	0.079	0.079	0.063	0.066	−0.010	0.024	0.814 ***	1.000			3.44
PCGCOMM (17)	−0.247 ***	0.260 ***	−0.135 **	0.202 ***	−0.125 **	0.088	−0.008	0.008	−0.055	−0.316 ***	0.068	−0.066	0.175 ***	0.141 ***	−0.041	0.056	1.000		1.5
CEODUAL (18)	−0.037	−0.012	0.024	0.086	−0.016	0.000	−0.054	−0.035	0.014	−0.043	0.016	0.001	−0.002	−0.079	0.026	0.021	0.098 *	1.000	1.11

Note(s): *** p < 0.01, ** p < 0.05, * p < 0.1.

presents the Pearson correlation matrix for the study variables. INVINEFF exhibits a positive and statistically significant correlation with NCFASST (r = 0.312, p < 0.01) and PPEASST (r = 0.552, p < 0.01), indicating that firms with stronger internal financing capacity and greater capital intensity tend to exhibit higher levels of investment inefficiency. A positive correlation is also observed between INVINEFF and ROE (r = 0.066). Interestingly, INVINEFF shows a weak and insignificant positive correlation with both ESGS (r = 0.029, p = 0.599) and FBOD (r = 0.032, p = 0.562), indicating that the potential effect of gender diversity and ESGS on investment efficiency may operate through indirect channels or require further multivariate analysis to reveal.

FBOAD is moderately and significantly correlated with ESGS (r = 0.189, p < 0.01), providing initial support for our expectation that gender-diverse boards may promote sustainability practices. Further, ESGS is strongly correlated with FSIZE (r = 0.439, p < 0.01), highlighting that larger firms tend to engage more in ESG initiatives. Several governance variables also demonstrate meaningful correlations. For example, BODINDP is negatively and significantly correlated with FSIZE (r = −0.179, p < 0.01), indicating that larger firms in the sample tend to have a lower proportion of independent directors. ACINDP is significantly associated with ESGS (r = 0.184, p < 0.01) and FBOD (r = 0.129, p < 0.05), reinforcing the role of governance quality in ESG outcomes. Additionally, the presence of a corporate governance committee is negatively correlated with investment inefficiency (r = −0.247, p < 0.01), suggesting that formal governance structures are more prevalent in firms exhibiting greater investment efficiency or reduced agency concerns.

The reported correlation coefficients are all below the conventional threshold of 0.80, with the exception of the correlation between the PNOMCOMM and PCOMCOMMP. To address this, the main model was re-estimated after excluding PCOMCOMM, and the results remained consistent, indicating that multicollinearity is unlikely to pose a concern a concern (Gujarati & Porter, 2009). The final column of Table 3 reports the variance inflation factors (VIFs), which range from 1.11 to 3.52, with a mean of 1.91. As all VIF values are below the conventional threshold of 5, multicollinearity is unlikely to materially affect the regression estimates. ESGS and FBOD were mean-centered before constructing the interaction term to reduce nonessential multicollinearity.

4.2. Regression Results

Table 4. Results of the main models.

Variables	Coef.	z	Coef.	z
ESGS	0.00033	2.170 **	0.00048	3.060 ***
FBOD	0.00006	0.140	0.00282	2.840 ***
ESGS * FBOD	-	-	−0.00006	−3.060 ***
FSIZE	−0.00418	−1.570	−0.00368	−1.380
NCFASST	−0.04903	−1.430	−0.05073	−1.500
ROE	0.00009	0.550	0.00002	0.160
MTBV	0.00131	0.990	0.00164	1.260
SGROWTH	−0.00619	−0.720	−0.00424	−0.500
LEVERAGE	0.00024	1.270	0.00024	1.260
PPEASST	0.08225	4.260 ***	0.07955	4.150 ***
ACINDP	−0.00016	−2.020 **	−0.00016	−2.050 **
BODSIZE	−0.00127	−1.690 *	−0.00100	−1.340
BODEXP	0.00046	0.420	0.00073	0.680
BODINDP	−0.00007	−0.460	−0.00007	−0.470
PNOMCOMM	0.00124	0.060	−0.00206	−0.110
PCOMPCOMM	−0.00801	−0.330	−0.00521	−0.220
PCGCOMM	−0.00115	−0.160	−0.00275	−0.400
BIG4	0.01093	1.030	0.00989	0.930
CEODUAL	0.00937	0.800	0.00838	0.720
Year and Sectors	Included
No. Obs.	342		342
R-squared	0.364		0.374
Wald chi2(36)	97.360 ***		108.710 ***

Note(s): *** p < 0.01, ** p < 0.05, * p < 0.10.

shows the regression result for the impact of ESG on investment efficiency. The Breusch–Pagan Lagrangian Multiplier (LM) test rejects the null hypothesis that the variance of the panel-specific effect is zero, indicating that a panel estimator (random effects or fixed effects) is preferred to pooled OLS. The output for the LM test indicates that the chibar² value is 112.10 while the p-value is 0.0000, rejecting the assumption that the variance for the unobserved individual effects is equal to zero. This suggests the presence and statistical significance of unobserved firm-level heterogeneity. As such, the hypothesis that pooled Ordinary Least Squares (OLS) is sufficient is rejected. These results justify the application of panel data for the estimation of the model over pooled OLS since they enable the inclusion of firm-specific effects that could otherwise compromise the validity of the results. In addition, identify the most appropriate panel data estimator, both fixed effects and random effects models were estimated and compared using the Hausman test. The test produced a chi-square statistic of 16.26 with a p-value of 0.505, suggests that the random effects estimator is appropriate for the analysis.

To examine the determinants of investment efficiency, a random-effects panel regression model was estimated. As reported in Table 4, the model is statistically significant (Wald χ² = 97.36, p < 0.01). The within R² of 0.364. The coefficient on ESGS is positive and statistically significant (β > 0, z = 2.170), providing support for H1.

To examine the moderating effect of female board representation on the relationship between ESG performance and investment efficiency, the interaction term (ESGS × FBOD) was incorporated into the random-effects regression model. The results, reported in Table 4, show that the model remains statistically significant (Wald χ² = 108.71, p < 0.01), with a slight improvement in explanatory power compared with the baseline model. Consistent with Model 1, ESGS retains a positive and statistically significant coefficient, indicating that higher ESG performance is associated with greater investment inefficiency. In contrast, the interaction term (ESGS × FBOD) is negative and statistically significant (z = −3.060, p < 0.01), providing support for H2.

To better explain the interaction, marginal effects were calculated at realistic levels of female board representation. ESGS has a positive and significant marginal association with the signed investment residual when FBOD is 0% (dy/dx = 0.000484, p = 0.002; 95% confidence interval [0.000174, 0.000793]), but the association becomes negative and insignificant when FBOD is 10% (dy/dx = −0.000091, p = 0.654; 95% confidence interval [−0.000491, 0.000308]). The difference between the two marginal effects (ΔME = −0.000575, p = 0.002) further validates the positive ESGS–investment residual association significantly decreases with increasing female board representation.

4.3. Discussion

The results indicate that firms with stronger ESG performance is associated with a greater tendency toward overinvestment rather than improved investment efficiency. The result is consistent with prior studies (Ma & Ma, 2025), which document a negative association between ESG performance and investment efficiency, and supports the argument that ESG initiatives do not necessarily translate into more efficient investment decisions in the short run (Alobaid et al., 2024; Keisse & Jaafar, 2025).

Among the control variables, PPETSS exhibits a positive and statistically significant coefficient (β = 0.082, z = 4.260), indicating that firms with greater asset tangibility tend to experience higher investment inefficiency. Regarding governance characteristics, ACINDP has a negative and statistically significant coefficient (z = −2.020), indicating a negative association with the signed investment residual. Board size (BODSIZE) also exhibits a weak negative association (z = −1.690); however, this effect is only marginally significant and becomes insignificant after including the interaction term. In contrast, the remaining governance variables, including board independence and the existence of nomination and compensation committees, are not statistically significant. Likewise, firm size, return on equity, and sales growth do not show significant associations with investment inefficiency. An additional analysis controlling for the COVID-19 period confirmed that the main findings remain unchanged, with ESG performance retaining a positive and significant association with investment efficiency (β = 0.00033, p = 0.030).

The finding suggests that the association between ESG performance and the signed investment residual varies with the level of female board representation. In other words, while ESG performance alone is associated with higher investment inefficiency, this adverse relationship becomes less pronounced in firms with greater female board representation. This finding is consistent with the view that female directors enhance board monitoring, improve strategic decision-making, and facilitate the effective integration of ESG initiatives into firms’ investment decisions. Furthermore, the positive and statistically significant coefficient on FBOD indicates that female board representation has a direct association with investment inefficiency after accounting for the interaction effect. However, the significant interaction term suggests that its primary contribution lies in moderating the relationship between ESG performance and investment efficiency rather than acting solely as an independent determinant.

4.4. Robustness Tests

4.4.1. Exploring Alternative Measurement of ESG

As a robustness check, the ESGS variable was decomposed into the three constituent pillars, environment pillar score (EPS), social pillar score (SPS), and governance pillar score (GPS) to examine their respective impacts on investment efficiency (Al-Hiyari et al., 2023). The random-effects regression results in

Table 5. Results based on alternative measurements of ESGS and board diversity.

	1		2		3
Variables	Coefficient	z	Coefficient	z	Coefficient	z
ESGS			0.000	2.480 **	0.001	3.470 ***
EPS	0.000	−3.410 ***	-	-	-	-
SPS	0.000	2.960 ***	-	-	-	-
GPS	0.000	2.010 **	-	-	-	-
FBOD	0.000	0.500	0.005	0.850	0.037	3.100 ***
ESGS * FBOD	-	-	-	-	−0.001	−3.080 ***
FSIZE	−0.001	−0.430	−0.003	−1.270	−0.003	−1.070
NCFASST	−0.028	−0.840	−0.034	−1.000	−0.036	−1.100
ROE	0.000	0.630	0.000	0.530	0.000	0.280
MTBV	0.001	1.180	0.002	1.480	0.002	1.750 *
SGROWTH	−0.006	−0.800	−0.008	−1.040	−0.007	−0.950
LEVERAGE	0.000	1.560	0.000	1.400	0.000	1.350
PPEASST	0.085	5.870 ***	0.078	5.520 ***	0.076	5.440 ***
ACINDP	0.000	−1.590	0.000	−1.770 *	0.000	−1.670 *
BODSIZE	−0.001	−1.530	−0.001	−1.810 *	−0.001	−1.520
BODEXP	0.000	0.070	0.000	0.180	0.000	0.330
BODINDP	0.000	−1.200	0.000	−0.710	0.000	−0.710
PNOMCOMM	−0.012	−0.620	0.001	0.040	0.000	0.000
PCOMPCOMM	0.002	0.100	−0.007	−0.280	−0.006	−0.270
PCGCOMM	−0.002	−0.240	−0.002	−0.320	−0.004	−0.640
BIG4	0.008	0.820	0.009	1.020	0.008	0.880
CEODUAL	0.010	0.840	0.010	0.820	0.007	0.590
_cons	0.006	0.130	0.038	0.900	0.024	0.570
Year and Sector	Included
No. Obs.	342.000		342.000		342.000
Wald chi2(28)	112.630		92.740		104.420
Prob > chi2	0.000		0.000		0.000
R-squared	0.328		0.320		0.325

Note(s): *** p < 0.01, ** p < 0.05, * p < 0.10.

Colum 1 indicate that the EPS is negative and significantly associated with the signed investment residual (β = −0.000, p = 0.001), suggesting that environmental strength is the driving factor for investment efficiency enhancement. The finding suggests that companies with robust environmental practices are more efficient in allocating capital, possibly because such practices reduce regulatory risk, improve resource utilization, and enhance stakeholder trust. On the other hand, the SPS and GPS indicate positive and significant associations with the signed investment residual (p = 0.003 & 0.045, respectively), suggesting that an increase in social and governance activities might be due to overinvestment or malfunctioning capital allocation. In the case of social and governance programs, agency problems may become more pronounced because managers, capitulating to external pressures for compliance or high-expenditure programs, may resort to overinvestments by committing excessive resources. Such over commitments may drain productive opportunities by bringing about misplacement of resources and lower efficiency of investments (Chen et al., 2024; Ma & Ma, 2025). Environmental demands at the same time also tend to bring about strengthened corporate governance exposure, thereby heightening investors’ expectations for greater transparency (Di & Li, 2023). In sum, these findings suggest that that the aggregate ESG relationship varies across its constituent pillars: EPS is associated with improved investment efficiency, whereas SPS and GPS are associated with a greater tendency toward overinvestment.

4.4.2. Exploring Alternative Measurement of Board Diversity

To examine the robustness of the main findings, board gender diversity was remeasured using the approach proposed by Ghaleb et al. (2024). Specifically, a dummy variable is constructed, taking the value of “1” if the firm’s BOD includes at least one female member, and “0” otherwise. Subsequently, we re-estimate the models 2 & 3 using the FBOD dummy variable and interaction variable (Table 5, Columns 2–3). The findings obtained in Columns 2 and 3 consistently mirror the main results presented in Table 4, indicating significant positive coefficients on ESGS and investment inefficiency, thereby suggesting that higher levels of ESGS are associated with greater investment inefficiency. Further, the results for the interaction term (FBOD*ESGS) tabulated in Columns 5 and 6 are negative and significant, consistently replicating the main results. That is, board diversity moderates the connection by reducing the inefficiency that is typical of high ESG engagement. Hence, the evidence supports that diverse boards are more effective at monitoring and integrating ESG activities with strategy-based investment goals, improving efficiency.

4.4.3. Endogeneity Issue

It is argued that the association between sustainability practices and corporate investment efficiency is likely to suffer from endogeneity problems (Al-Hiyari et al., 2023; Benlemlih & Bitar, 2018; H. L. Lin & Yen, 2022). This endogeneity could be an outcome of reverse causality (efficiency firms could have more sources for enhancing ESG) or of omitted variables (e.g., unobserved quality of corporate governance). To address this potential problem, we employed a two-stage least squares (2SLS) regression using two alternative instrumental variables, CSR Strategy Score (CSRSS) and Community Score (CS) based on their high theoretical as well as empirical relevance (Bilyay-Erdogan et al., 2024; Orazalin & Baydauletov, 2020). The CS measures a firm’s involvement in community-related issues such as philanthropy, local development issues, as well as stakeholder interactions. Though such a dimension provides a measure of a firm’s overall ESG commitment and thus strongly correlates with the overall ESG score (relevance condition), its direct effect on efficiency of investment is unlikely. The CSRSS measures the existence as well as quality of a company’s policies as well as strategic commitments towards corporate social responsibility. This score shows structural as well as procedural embedding of ESG principles into corporate strategy that strongly corresponds with ESG performance. However, by itself, a company’s announced CSR policies or formal frameworks are likely not to hinder efficiency of investment directly, particularly if they have not yet affected actual decisions of operations by fulfilling the condition of exogeneity. Both of these are conceptually linked with ESGS but are unlikely by themselves to hinder the firm’s efficiency of capital investment.

Table 6. Results of the 2SLS regression.

Variables	First Stage		Second Stage		First Stage		Second Stage
	Coefficient	t	Coefficient	z	Coefficient	t	Coefficient	z
ESGS			0.001	3.790 ***			0.001	3.320 ***
FBOD	−1.166	−4.290 ***	0.002	2.070 **	−0.424	−1.730 *	0.001	1.800 *
ESGS * FBOD	0.034	6.010 ***	0.000	−2.620 ***	0.018	3.640 ***	0.000	−2.280 **
FSIZE	1.901	4.060 ***	−0.003	−2.090 **	3.212	8.270 ***	−0.002	−1.660 *
NCFASST	10.625	0.910	0.022	0.640	17.665	1.730 *	0.028	0.820
ROE	−0.099	−1.710 *	0.000	0.860	−0.119	−2.330 **	0.000	0.670
MTBV	−0.207	−0.560	0.003	2.880 ***	−0.718	−2.210 **	0.003	2.850 ***
SGROWTH	3.126	1.590	−0.020	−3.430 ***	6.373	3.730 ***	−0.019	−3.250 ***
LEVERAGE	−0.151	−3.430 ***	0.000	0.420	0.005	0.140	0.000	0.430
PPEASST	−1.218	−0.380	0.062	6.620 ***	−7.103	−2.520 **	0.062	6.620 ***
ACINDP	0.049	2.020 **	0.000	−1.550	0.081	3.780 ***	0.000	−1.400
BODSIZE	−0.615	−2.210 **	−0.001	−0.790	0.045	0.190	−0.001	−0.820
BODEXP	−0.050	−0.180	0.000	−0.470	0.139	0.580	0.000	−0.290
BODINDP	0.056	1.050	0.000	0.590	0.013	0.280	0.000	0.570
PNOMCOMM	15.728	1.850 *	0.021	0.820	−0.383	−0.050	0.021	0.870
PCOMPCOMM	−17.382	−1.690 *	−0.019	−0.620	0.445	0.050	−0.020	−0.670
PCGCOMM	5.435	3.120 ***	−0.016	−2.900 ***	5.360	3.510 ***	−0.014	−2.640 ***
BIG4	4.414	2.460 **	0.003	0.600	1.425	0.900	0.004	0.730
CEODUAL	3.584	1.050	0.008	0.800	−3.676	−1.240	0.007	0.680
CSRSS	0.353	15.420 ***	-	-	-	-	-	-
CS	-	-	-	-	0.431	19.920 ***	-	-
_cons	−13.842	−1.260	0.020	0.610	−47.466	−5.130 ***	0.009	0.270
Year and Sector	Included
No. Obs.	342.000		342.000		342.000		342.000
F(27)/chi2(27)	27.330 ***		205.550 ***		38.540 ***		205.300 ***
R-squared	0.702		0.362		0.768		0.372

Note(s): *** p < 0.01, ** p < 0.05, * p < 0.10.

presents the first-stage regression results, indicating that the estimated coefficients of both instruments, CSRSS and CS, are statistically significant. This finding confirms their relevance, as the instruments are strongly associated with the potentially endogenous variable. The second-stage results are consistent with the main findings reported in Table 4, indicating a positive association with the signed investment residual. In addition, the interaction term (FBOD * ESGS) exhibits a statistically significant coefficient, suggesting that the association between ESG performance and the signed investment residual varies with the level of female board representation.

To mitigate potential reverse causality, we further re-estimated the model using investment outcomes measured in year t and one-year-lagged explanatory variables measured in year t − 1. The untabulated results show that lagged ESG performance remains positively and significantly associated with investment outcomes (β = 0.000428, p = 0.025), consistent with the main findings.

5. Conclusions

This study examines the impact of ESG scores on investment efficiency among Saudi listed companies and the moderating role of board gender diversity. The findings present significant insights into how sustainability initiative relates with corporate governance mechanism to form investment outcomes. Utilizing a random-effects panel regression model, the analysis reveals that higher level of ESG performance is associated with a positive shift in the signed investment residual, indicating a greater tendency toward overinvestment rather than improved investment efficiency. This suggests that higher ESG engagement is associated with investment above the model-predicted level, potentially reflecting implementation costs, agency conflicts, or resource diversion, however, these mechanisms are not directly tested in this study. This interpretation is consistent with prior studies (e.g., Ma & Ma, 2025), which report that higher ESG scores are associated with lower investment efficiency in both the U.S. and Chinese contexts. These findings imply that, despite their broader sustainability benefits, ESG initiatives may not immediately translate into more efficient investment decisions in the short term.

An important contribution of current study lies in detecting the moderating role of board gender diversity that weakens the positive association between ESG performance and overinvestment. This finding emphasizes that governance, particularly inclusive board composition, is associated with how sustainability practices relate to investment efficiency. This finding is particularly relevant in Saudi Arabia, where Vision 2030’s gender reforms are fostering greater female participation in corporate governance, amplifying the transformative potential of inclusive leadership.

Theoretically, this study enriches the ESG literature by documenting a context-specific association in an emerging market, highlighting that higher ESG performance is associated with a greater tendency toward overinvestment and that this relationship varies with board gender diversity. The findings contribute to agency and stakeholder theories by suggesting that board gender diversity is associated with differences in the relationship between ESG performance and investment outcomes, although the underlying mechanisms are not directly examined in this study. From a practical standpoint, these results may inform ongoing discussions among policymakers, regulators, and corporate leaders in Saudi Arabia and similar emerging markets regarding board composition and ESG implementation. As Saudi Arabia advances its Vision 2030 agenda, these findings suggest that board gender diversity may be a relevant governance characteristic when considering ESG-related investment decisions. However, the reported relationships should be interpreted as associations rather than causal effects, and further research using research designs that better support causal inference is needed before drawing policy recommendations.

Despite its contributions, the study has limitations. The focus on Saudi listed firms may limit generalizability to private or smaller enterprises, and reliance on aggregate ESG scores may obscure further pillar-specific dynamics. While endogeneity was addressed through 2SLS, potential unobserved governance factors suggest caution in causal interpretations. The findings should also be interpreted in light of potential ESG coverage bias, survivorship bias, and the observational nature of the study, which may limit causal inference. In addition, the study focuses primarily on statistical significance rather than economic significance. Future research could estimate marginal effects and predicted changes associated with realistic variations in ESG performance and board gender diversity to provide a more comprehensive assessment of the practical significance of the findings. Future research could also employ longer panel data, dynamic models, or qualitative approaches to further examine the mechanisms through which board gender diversity moderates the ESG–investment relationship.