Climate change risk and corporate risk-taking: Do board gender diversity and environment, social and governance performance matter?

The climate change risk (CCR) is currently one of the most important issues attracting widespread attention. The significance of this issue is evidenced by the recent increase in the occurrence of natural disasters, pandemics and natural resource shortfalls. These threats are associated with increasing adverse financial and economic shocks in terms of stock returns and dividend payments (Boubaker et al., 2024), cost of equity (Cepni et al., 2024), market valuation (Berkman et al., 2024), financial policy uncertainty (Hunjra et al., 2024), debt financing costs (Zhao et al., 2024) and green finance (Banerjee et al., 2024), among others. Studies examining the relationship between CCR and financial aspects conclude that CCR places companies in difficult economic situations and has a significant impact on corporate risk-taking (CRT) (Ren et al., 2023). The logic behind this prediction is twofold. First, as CCR incurs higher uncertainty and damages, managers behave in a more risk-averse manner and undertake more cautious decisions. Risk-averse behavior leads to suboptimal risk-taking. Second, according to loss aversion theory (Tversky and Kahneman, 1991), managers are no longer overly risk-averse when they face higher losses. Therefore, losses induced by climate risks may encourage risk-averse managers to accept risky projects with positive net present values.

Although the effects of climate change on corporate policies have been widely explored, evidence on particular channels through which CCR shapes CRT is far from unanimous. For example, little is known about whether climate risk encourages or hinders business risk-taking. However, there is no consensus on how companies should deal with CCR. Assuming the real effect of CCR on CRT, and given the lack of clarity regarding the interaction between the CCR-CRT nexus and corporate governance attributes, this study aims to fill the gap in the literature by testing whether CCR affects CRT and how board gender diversity (BoGD) and environment, social and governance (ESG) performance may moderate the CCR-CRT nexus.

This study applies a two-step generalized method of moments ^[1] to a sample of 10,874 firm-year observations from 34 countries from 2010 to 2021 and shows that managers accept more risks and exhibit better commitment to ESG practices in response to heightened CCR. Specifically, an increase in CCR by 1 standard deviation indicates an increase in total risk, idiosyncratic risk (IR) and market β risk (Mβ) by almost 21.5%, 8.7% and 7.7%, respectively. This result confirms the value-enhancing view of CCR, as it encourages companies to take more risks when they do not take enough risks and to behave risk-averse when the risks become too high. That is, CCR motivates managers to take more risk. These findings also indicate that BoGD (ESG performance) mitigates (enhances) the impact of CCR on CRT. These results are important because they suggest that firms with higher BoGD and ESG performance could have better mitigation efforts toward the risks associated with climate change. Furthermore, the combined effect of BoGD and ESG performance seems negative, indicating that they act as substitutes. These findings remain unaltered regardless of the metrics or alternative statistical techniques used.

This study differs from previous studies dealing with the impact of CCR on CRT in several ways. First, although scholars have focused on the impact of CCR on capital (Yue et al., 2024) and portfolio-level risk (Li and Lu, 2025), this study examines the impact of CCR on systematic and IRs at the firm level. Second, it relies on Sautner et al.'s (2023) procedure to assess firm-level CCR. Sautner et al.’s CCR is a multi-dimensional firm-level measure that captures firm-specific CCR induced by different origins such as physical, regulatory and transitional aspects. It has been used in several studies, including Arthur (2025) and Chen et al. (2024). Third, while earlier research has mainly been based on American and Chinese data (Naseer et al., 2024; Zhao and Parhizgari, 2024) , this study provides a more extensive analysis by considering a diversified group of firms from 34 countries worldwide. Moreover, the sample’s study covers various industries (two-digit standard industrial classification (SIC) code), which reflects the global and cross-sectoral empirical nature of the research. Fourth, this study intends to understand how BoGD and ESG performance matter in the CCR-CRT nexus. Previous studies have focused on the direct effects of BoGD and/or ESG performance on CCR and/or CRT (Trinh et al., 2024; Yin et al., 2024; Chen et al., 2025; Teodósio et al., 2021; He et al., 2023). In this study, BoGD and ESG performance moderated the effect of CCR on CRT jointly through direct and indirect channels. Specifically, women on the board and ESG performance not only directly enhance CCR but also promote the effect of CCR on CRT. Therefore, this study is the first to examine the moderating effects of corporate policies on board structure and ESG practices in the CCR-CRT association. Broadening the CCR assessment through selected corporate policies (CRT, BoGD and ESG performance) provides a more in-depth view of the managerial practices underlying CCR. To the best of our knowledge, no prior research has examined whether and how BoGD and ESG performance affect the association between firm-level CCR and CRT.

This study has several important managerial, policy and theoretical implications. First, the positive association between CCR and CRT implies that the latter acts as an effective CCR-hedging mechanism. Intuitively, when CCR losses occur, managers undertake more risks, as the loss aversion theory predicts. By contrast, they accept a higher risk of profit in the long run. Therefore, CRT may benefit all stakeholders. Second, the obtained results explore how managers and regulators must adjust corporate policies toward board structure and ESG performance to hedge against CCR. On the one hand, findings display a valid reason for corporate leaders to justify their ESG engagement, as it helps to overcome the devastating nature of climate change. On the other hand, the positive effect of ESG performance on CCR could help shareholders understand the benefit of ESG practices instead of treating them as excessive investments that harm their wealth. Third, the results highlight that female directors should not merely be tokens, as it is found that women on boards play a crucial role in hedging CCR. Consequently, policymakers must endorse gender quota legislation and normalize gender diversity on boards. Finally, this study contributes to corporate governance literature. Although, CCR has an impact on CRT, the effectiveness of board oversight remains unclear. My evidence reveals that some forms of board control, including enhanced BoGD and ESG practices, provide effective mechanisms for corporate response to CCR.

The remainder of this paper is organized as follows. Section 2 reviews the literature and develops our hypotheses. Section 3 highlights the research design, including the empirical model, variables and data construction. Section 4 presents the empirical and robustness results, and Section 5 concludes the study.

Studies examining the relationship between corporate change risks and corporate risk appetite are generally based on two assumptions: value-enhancing vs value-destroying.

The value-enhancing view asserts that CCR promotes CRT. There are two reasons for this assumption. On the one hand, higher uncertainty caused by CCR is hypothesized to enhance managerial risk-aversion. As managers are thought to be risk-averse, they behave more risk-averse and are extremely cautious when making decisions. Under higher uncertainty, they would reject projects with positive net present values. Thus, the higher the uncertainty, the higher the managerial risk aversion and the suboptimal the risk-taking propensity. To correct for this sub-level of risk-taking, CRT incentives (e.g. stock options) are regarded as constructive strategies to prompt managers to accept additional risks. On the other hand, according to the theory of loss aversion (Tversky and Kahneman, 1991), managers are typically more sensitive to losses than gains and are more prone to take risks when they incur losses. Therefore, as climate risks undoubtedly cause losses, they are bold enough to adopt positive net present value projects in the hope of overcoming stressful economic conditions. On balance, companies are more likely to choose high-margin projects to avoid losses and thus take more risks.

The value-destroying assumption predicts that CCR mitigates CRT. Arguments supporting this assumption fall broadly into four theoretical backgrounds. The first argument is based on the agency theory (Jensen and Meckling, 1976), assuming that, for several reasons, managers are unwilling to take excessive risks, especially in periods of uncertainty. First, they suffer from job anxiety and career concerns, which heavily limit their motivation to take more risks. Second, managers are believed to be selfish, and they may turn down risky projects with positive net present values. Third, because these projects are costly in the short term and may produce uncertain long-term benefits, managers opt to postpone them to avoid accountability from shareholders.

The second argument relies on stakeholder theory’s assumption that stakeholder activism may increase managers’ awareness of climate risks. Specifically, environmental shareholders forces managers to pay more attention to a firm’s vulnerability to CCR (Flammer et al., 2021). Consequently, managers engage in environmentally friendly practices that have short-term financial returns at the expense of positive net present value projects that may benefit the company financially in the long run (Xu et al., 2022). Stakeholder theory’s predictions also echo risk management theory’s assumptions, asserting that managerial risk aversion provides “insurance-like protection” against adverse climate events. According to Benz et al. (2021), hazardous climate events are harmful to corporate assets. Pan et al. (2024) show that Chinese banks reduce their risk-taking during extreme air pollution days. Similarly, Xue et al. (2020) found evidence of a significant mitigation effect of environmental risk on CRT.

The third argument is based on real options theory, asserting that managers behave more cautiously, even extremely conservatively, when there is uncertainty. As climate changes are hazardous events, firms could under-invest in positive net present value “investments to hedge corporate” systematic risk. Additionally, managers’ investment policies may be selective. They could invest in certain projects with positive net present values – but not all – in highly uncertain circumstances.

Finally, from a macroeconomic perspective, consequences of climate change are largely detrimental. For example, rising sea levels may damage the assets of firms located in coastal areas. Extreme weather would indeed severely lead to business interruptions or shutdowns. Climate change is commonly supposed to be severe in industries (e.g. utility, recreation and tourism, insurance, etc.) and harms firms’ investment in sustainability (Gao et al., 2024).

Based on the above analysis, I hypothesize the following:

Empirical investigations of the value-enhancing and value-destroying views addressed in the previous section recognize numerous determinants of the CCR as well as the CRT. Among these investigations, corporate governance attributes are thought to impact the CCR-CRT nexus. What is still unclear is how this nexus is shaped by the interaction between specific corporate governance features. For example, while scholars have confirmed that BoGD and ESG performance influence business risk, there is a gap in the existing literature on the jointly causal effect of BoGD and ESG performance on the CCR-CRT link. I aim, in the following, to fill this gap.

Several scholars have revealed that the BoGD, CCR and CRT are connected. On the one hand, a negative causal effect was found between BoGD and CCR. This assertion stems from different theoretical arguments. First, behavioral theory (Cyert and March, 1963) states that women are innately more risk-averse and more sensitive to losses. Second, agency theory predicts that female directors improve board oversight and effectiveness toward social and environmental tasks because they exhibit higher awareness of the community than their male counterparts do. Third, legitimacy theory argues that by being diverse, boards can better legitimize their green credentials and gain more trust from a broad range of stakeholders. Fan et al. (2023) posit that, relative to their male-dominated counterparts, boards with more female members are better implied in addressing climatic issues such as greenhouse gas (GHG) and carbon dioxide (CO₂) emissions. Tingbani et al. (2020) provided evidence of a strong positive association between GHG intensity and gender diversity. Altunbas et al. (2022) found that BoGD enhances business efforts to curb CO₂ emissions. Nuber and Velte (2021) provide evidence that female directors improve European firms’ ESG performance by curbing their total carbon emissions. Altogether, these arguments contend that BoGD plays an effective role in addressing firm-level CCR.

However, research has suggested that BoGD impacts CRT differently. A systematic review by Teodósio et al. (2021) showed that boardroom gender diversity has different effects on CRT. On the one hand, board and TMT gender diversity has a universal effect on CRT, as it attenuates corporate risks, including litigation, failure and operational risks. On the other hand, this effect seems contingent on IRs and systematic risks. These effects rely mainly on three lenses. First, behavioral theory suggested that men are more likely to take risks than are women because of distinct biological, psychological and social considerations. Second, agency theory hypothesized that BoGD promotes boardrooms’ oversight and independence. Third, resource dependency theory (Pfeffer and Salancik, 1978) asserted that female directors favor higher corporate legitimacy, better decision-making and moderate risk-taking.

In light of these findings, I propose the following hypothesis:

Previous research has shown potential correlations between ESG performance, CCR and CRT. The relationship between ESG performance and CCR has been examined by numerous empirical investigations. For example, Yin et al. (2024) showed that ESG performance mitigates a corporation’s exposure to climate risks. Gao et al. (2024) find that environmental investment significantly enhances a firm’s climate resilience. Tang et al. (2024) pointed out that companies’ ESG performance is significantly better during climate shocks. Fang (2024) reported that Chinese companies pursue green strategies when exposed to CCR. Ozkan et al. (2023) revealed that firms the incur greater climate risks behave more socially responsibly. Using branch-level data on American commercial banks’ exposure to physical climate risk, Erhemjamts et al. (2024) asserted that banks react to climate disasters by boosting ESG investments. These findings support risk-management theory, assuming that firms with higher ESG performance express better resilience to CCR. Their assets are less vulnerable to detrimental effects of climate change. Indeed, they are more likely to maintain their financial position.

However, the CCR has been shown to restrict a firm’s ESG engagements. According to Li et al. (2024), CCR significantly inhibits corporate ESG performance. Chen et al. (2024) found that injuries caused by climate changes alleviate the ESG performance of A-share Chinese firms. Similarly, He et al. (2023) found that abnormal temperature – as a proxy for CCR – significantly reduces corporate green innovation. Mohy‐ud‐Din et al. (2024) supported the role of corporate social responsibility in bolstering firm resilience against threats of climate change and climate policy uncertainties. Consistent with this view, Jia and Li (2020) argued that the uncertainty underlying climate changes harms corporate sustainability.

Concerning the link between ESG performance and CRT, prior contributions supported opposite directions. On the one hand, according to the risk management and stakeholder views, firms with high ESG performance better mitigate their riskiness, especially during high uncertainty. Managers behave more cautiously and conservatively when incurring uncertainty. On the other hand, according to real options theory, they heavily act in ESG practices to improve the firm’s nonfinancial ratings and deflect the market’s attention from their sub-optimal financial equilibrium. Thus, firms with high ESG standards perform better during a crisis. In other words, investing in ESG provides firms insurance-like benefits to manage high uncertainties. In accordance with this prediction, Azimli (2023) founds that the insurance-like benefits of ESG in the form of reduced risk are larger for firms that face a higher degree of uncertainty. Eratalay and Cortés Ángel (2022) revealed that firms highly ranked in ESG performance display up to 7.3% less systemic risk.

Drawing on the above arguments, I propose the following hypothesis:

To emphasize the effect of CCR on CRT, I develop the following model:

Where, CCR_it is a firm-level climate risk, and CRT is the total risk, IR or market beta risk (Mβ). BoGD and ESG are proxies for boardroom gender diversity and ESG performance, respectively. X denotes a set of firm controls including size, R&D expenditures, financial leverage, market-to-book ratio and tangible assets. μ_i (μt) represents the firm (year) fixed effect. ε_it stands for the random error term. β refers to the core regression coefficients. The subscripts i, t and k indicate the number of firms, years and the control variables, respectively.

To identify how BoGD and ESG performance affect the relationship between CCR and CRT, the following model is developed:

Where, γ₁ (γ₂) determines the validity of H2 (3). γ₃ checks the possible substitution/complementarity effect between the BoGD and ESG performance. To address endogeneity concerns, Models (1) and (2) are tested by running a two-step generalized method of moments as in Blundell and Bond (1998). Variables are winsorized at the 1st and 99th percentiles to reduce the impact of potential outliers.

To gauge the CCR at the firm level and in the cross-country sample, I exploit the CCR index developed by Sautner et al. (2023). This index is assessed using the total frequency of climate change bigrams scaled by the total number of bigrams. The latter are computed using the frequency of keywords such as “risks,” “uncertainties” and their synonyms in surrounding sentences that discuss climate change issues occurring in the earnings conference calls transcript. The CCR index captures three basic dimensions of firm-level climate risk (physical, regulatory and transitional risks). I employ the aggregate CCR index as in the yearly version of 2021-Q4. For better readability, the CCR index was multiplied by 10,000. As in prior studies, CRT is weighted by the firm’s total risk, IR and M𝛽.

Data on BoGD (% female) were obtained from the DataStream database. The firm’s ESG performance is assessed using ESG disclosure scores from the Bloomberg Data Service. Xk represent firm-level controls gathered from the WorldScope database. The definitions and details of all variables are summarized in the  Appendix.

The initial sample was based on the entire comprehensive list of ISIN codes for firm-level CCR as assessed by Sautner et al. (2023). This consists of 7,328 firms around 34 international equity markets with 16,642 firm-year observations ^[2] over the 2010–2021 period. After excluding observations from financial firms, observations with missing values and those not covered by Bloomberg database, the final sample corresponds to 10,874 firm-year observations from 2,146 nonfinancial firms.

Table 1 reports the summary descriptive statistics including means and standard deviations (SD). The mean value and standard deviation of the CCR variable are 0.0317 and 0.1622, respectively. They are in line with Sautner et al. (2023). The average (SD) firm’s total risk is 0.0238 (0.0134) which is in line with the values reported in previous studies (e.g., Xue et al., 2020). The IRs and Mβs have slightly lower values. The mean (SD) for IR is 0.0147 (0.0178) and that for Mβ is 0.0116 (0.0095). The mean value of the ESG disclosure score is 33.541 (out of a potential 100) and the standard deviation is 20.315, which is similar the corresponding mean (34.374) and standard deviation (14.497) of Temiz and Acar’s (2023) study of 14,185 firm-year observations from 43 countries during 2010–2019. On average, there are 22.51% women on corporate boards (SD of 11.73%).

Focusing on the control variables, the firm’s average size is 17.3, whereas the mean of R&D expenditure over total assets is nearly 2%. The sample’s average leverage and market-to-book ratio are, respectively, 53.21% and 6.137. Tangible assets represent on average 21.59% of total assets. These statistics are nearest to those reported by Naseer et al. (2024) based on a panel of 1,529 multinational listed firms over the 2012–2021 period.

To test for multicollinearity, bivariate correlations among the variables are summarized in Columns (2)–(11) in Table 1. Interestingly, correlations seem fairly low, indicating that multicollinearity issues are nonexistent ^[3]. Additionally, the lower variance inflation factors, shown on the right side of Table 1, indicate that there is no multicollinearity problem between variables.

The CCR index was positively and highly significantly correlated with all CRT variables providing initial support of H1. This result corroborates the findings of Xu et al. (2022) who show that managers adopt risky projects when companies are exposed to drastic climate risks. Additionally, a highly positive correlation exists between CCR risk and ESG performance variables, indicating that climate risk involves a better firm’s commitment to environmental and social concerns. Erhemjamts et al. (2024) obtained comparable results for a sample of American commercial banks. Furthermore, Wang and Li (2023) revealed that climate-related risks successfully predict the volatility of the CSI 300 ESG index. However, Chen et al. (2025) and Li et al. (2024) reported a negative association. CCR and BoGD were negatively correlated (at the 1% level). This result suggests that BoGD inhibits CCR at the firm level and corroborates the results of Altunbas et al. (2022) which support a negative relationship between the percentage of female directors and the level of CO₂ emissions. Furthermore, BoGD is negatively correlated with all CRT metrics at higher levels of significance indicating that board diversity mitigates CRT. This result is in line with that of Alzayed et al. (2024), indicating that BoGD has a minor impact on firm risk. Similarly, ESG performance seems to have a detriment effect on CRT as the correlations between these variables are negative. This result is similar to that of Menla Ali et al. (2024) who showed that ESG performance curbs CRT. Besides, the positive correlation found between BoGD and ESG variables supports predictions from sustainability and corporate social responsibility literature that BoGD is associated with greater social and environmental duty (Bannò et al., 2023; Chang et al., 2024). The bivariate correlations between remaining variables were similar to those shown by prior related studies.

To display the direct effect of CCR on CRT, a two-step generalized method of moments was applied to Model (1). Before that, I accessed the validity of the model. As shown in Table 2, there was no evidence of serious aserial correlations. First, insignificant Sargan tests imply that the instruments are exogenous, which is a critical assumption for the validity of the GMM estimates. Second, an insignificant Blundell–Bond second-order autocorrelation (AR2) shows that the lagged variables are not correlated with the error term confirming the validity of the model used in our estimation process. Additionally, Wald statistics for the coefficients and time dummies further confirm our model choice and the use of time dummies. Therefore, the coefficients in Regressions (1) to (5) are appropriate.

In Columns (1) to (3), only the core explanatory variables of CRT are added. The results show that the regression coefficients of CRT are positive and statistically significant at the 1% level, indicating that CCR significantly promotes CRT as the value-enhancing view predicts.

The estimated coefficient of the annual total risk variable is 0.2153, implying that a 1-standard-deviation increase in the CCR index is associated with an approximately 12.12% increase in CRT relative to its sample mean. Furthermore, 12.12% is calculated as 0.2153 × 0.0134 /0.0238, where 0.2248 is the estimated coefficient of TR in Column (1), and 0.0134 and 0.0238 are, respectively, the standard deviation and the mean of TR from Table 1. Similarly, a 1-standard-deviation increase in annual CCR implies 10.48% and 6.33% increases in IR and M𝛽, respectively. These results support the value-enhancing assumption that CCR promotes CRT. They support findings of Xu et al. (2022). Indeed, they are consistent with Li et al. (2024) who asserted that firms excessively invest in risky projects as a risk management strategy to hedge against CCR. These results also support Ali and Gao’s (2023) findings of a positive linear effect of a firm’s exposure to abnormal climate events and CRT^[4]. Furthermore, results support those of Chen et al. (2024) showing that carbon pressure significantly promotes risk-taking in Chinese A-share listed energy firms. A recent study by Yue et al. (2024) also showed that the CRT level increases during climate clusters. Altogether, the results from Columns (1) to (3) support H1.

In Column (4), I gradually controlled for the effects of CCR on CRT metrics, BoGD and ESG performance. The most notable findings are as detailed in the following section.

First, the results show that the coefficients of total risk, IRs and M𝛽s are still positive, indicating that CCR significantly encourages CRT. Moreover, the coefficient of BoGD is negative (−0.1128) and significant at the higher level. This means that an added 1 standard deviation of the percentage of female directors significantly inhibits CCR by nearly 5.88%. This result is in agreement with Altunbas et al. (2022) who asserted that the higher the number of women holding managerial positions, the lower the firm-level CO₂ risk. Moreover, it confirms the findings of Trinh et al. (2024) that firms with boards dominated by women are likely to incur lower CCR. Arguments supporting this finding rely on the extant theoretical literature including the resource-based view, resource dependency, legitimacy and noninstitutional theories. Commonly, these theories support that, as women have innately stronger moral standards and are socially more sensitive, emotional and empathic, gender-diversified boards act more efficiently in addressing climate change-related issues (Teodósio et al., 2021; Trinh et al., 2024).

Second, the coefficient of ESG variable is positive (0.1103) and significant at the 1% level, implying that CCR boosts ESG performance. Explicitly, a 1-standard-deviation increase in annual CCR involves a 6.68% growth in the firm’s annual ESG performance. This result confirms the finding of Yin et al. (2024) that firm-level climate risk enhances ESG performance. The authors’ results are based on a sample of 31,664 firm-year observations from Chinese listed firms over the 2010–2021 period using a similar CCR proxy; the CCR time-varying measure was proposed by Sautner et al. (2023). The positive effect of CCR on ESG performance is also in line with Erhemjamts et al. (2024) who concluded that materiality and physical climate risk are positively associated with banks’ ESG performance. It corroborates moreover, the result of Tang et al. (2024), indicating that corporate risk exposure encourages environmental practices significantly. However, this result contradicts Li et al.’s, (2024) findings that physical climate risk mitigates ESG performance; an annual average temperature deviation of 1 unit results in a 0.04 unit decrease in ESG performance.

Column (5) reports the regression results for all variables included. Overall, the results correspond with those of similar recent studies. CCR suppresses a firm’s market-to-book value and R&D expenses. Firm size has a significant and negative impact on CCR, and firms with higher tangible assets experiences serious CCR.

The results of assessing the moderating effects of BoGD and ESG performance on the CCR-CRT nexus are summarized in Table 3. As shown in the above discussion, estimators of CRT metrics remained positive at a higher statistical significance level. Columns (1)–(3) display that coefficients of the interaction terms between total risk, IRs, market 𝛽 and BoGD are significantly negative. This indicates that BoGD plays a mitigating moderating role; companies that have more women in boards can effectively alleviate the excessive risk-taking caused by CCR. This result is important because it suggests that organizations may endorse better mitigating strategies against CCR by increasing BoGD. That is, the better women prevail in boards, the lesser CCR is harmful at the firm-level. This result confirms previous findings that women in boards enhance corporates’ effectiveness toward CCR mitigation (Trinh et al., 2024; Altunbas et al., 2022). Based on these findings, H2 is accepted.

Regarding the effect of ESG performance on the CCR-CRT nexus, the interaction term between CCR and ESG performance was positive and significant at the 1% level. Thus, it can be concluded that ESG performance strengthened the positive effect of CCR on CRT. This means that the higher the firm’s ESG performance, the greater the positive impact of CCR on CRT. Better ESG performance may enhance the impact of CCR on corporate risk through two channels. On the one hand, higher ESG performance indicates that the company is inclined to reach environmental and social goals that can make it more resilient to CCR (Bagh et al., 2024). On the other hand, a good ESG ranking mitigates a firm’s value decline due to adverse events. Consequently, it prevents companies from undervaluation because investors are less likely to penalize firms exposed to such events. Based on these findings, H3 is accepted, supporting, therefore, predictions from corporate social responsibility and risk-management theories. Taken together, results on the interaction terms indicate that BoGD and ESG performance are, respectively, positioned as holistic risk-mitigating and risk-enhancing strategies by effectively moderating the association between CCR and CRT.

Concerning the combined influence of BoGD and ESG performance on the CCR-CRT association, the results show that the interaction term between these moderators is negative and statistically significant across Regressions (1)–(4), indicating that, after controlling for other factors, the CCR-CRT nexus is restricted to firms with stronger ESG performance and female dominated boards. In terms of economic significance, within firms with higher BoGD and ESG performance, a 1 standard deviation rise in CCR induces a 7.98%, 32.6% and 3.37% decline in firm’s total risk, IR and Mβ, respectively ^[5]. All variables included [Column (4)], the interaction term BoGD × ESG remains notably negative and statistically significant at the higher economic level. These results suggest the substitutive relationship between BoGD and ESG as moderators of the CCR-CRT connection. Therefore, it is possible to defend that these variables replace each other when they work inversely, in the sense that higher levels of one of them reduce the effect of the other. Effectively, their joint impact is evidently lower than the sum of their individual effects [−0.1631 < (−0.103 + 0.1051)]. Within the corporate governance literature, numerous empirical studies recognized the substitutive view by showing that the prevalence of one governance mechanism leads to the lack of other(s) (Lan and Zhou, 2024; García‐Sánchez et al., 2023). This study is the first to delve into the substitutive effect between BoGD and ESG performance with reference to the CCR-CRT causal relationship. The logic behind the substitutive relationship between BoGD and ESG performance echoes agency and sustainable finance theories. In one side, as these theories have highlighted the dual good side of signals sent by management to investors and community in the form of higher BoGD and ESG commitment, and as environmental and social engagements are associated with an effective cost that business incurs which is lower than that associated with women appointment within boards, firms must decide between the cost and the benefits expected from BoGD and ESG performance. In other side, as CCR causes higher uncertainty and induces additional agency costs that may damage firm’s ESG performance, women in boards may act as “insurance-like protection” against business exposure to CCR (Azimli, 2023; Altunbas et al., 2022).

Three levels of robustness testing are performed to assess the sensitivity of the above insights.

To ensure the robustness of the variable measurement, I employ the followings approaches. First, the Climate Physical Risk Index (CPRI) developed by Guo et al. (2024) ^[6] was used as an alternative measure of CCR. Second, similar to Xu et al. (2022), the return on asset (ROA) volatility (RISK1) and ROA volatility adjusted by country and industry (RISK2) are used as alternative CRT metrics. Third, the Blau index was used as an alternative proxy for BoGD ^[7]. Fourth, to check for a possible nonlinear influence of BoGD on CCR (e.g. effects peaking at certain levels), the critical mass of at least three female directors is used as a substitute measure of BoGD (Kanter, 1977). Finally, the possible sensitive effects of CCR traits (physical, regulatory and transitional) were checked. Table 4 summarizes the results. As shown in Panel (A), the coefficients for the alternative proxies of CRT (BoGD) are still significantly positive (negative) and similar to those obtained based on the basic metrics (Tables 2 and 3). Panel (B) displays a robustness check of BoGD’s effect on CCR. In correspondence with the critical mass theory’s view, the binary variable (≥3 female directors) is negative, revealing that the BoGD could not be influential until the women’s seats reached a critical mass of three. In other words, boards with more than three women are more powerful in inhibiting a firm’s exposure to CCR. This finding confirms the results of the baseline analysis; there is a robust linear and negative relationship between BoGD and CCR. It highlights the U-shaped nature of this relationship. Regarding the sensitive effects of CCR traits (physical, regulatory, transitional), the results from Panel (B) display the obtained results for each CCR dimensions. These findings are similar to the corresponding estimates of the aggregate CCR measure.

To check the sensitivity of the obtained results to potential sample and time selections bias, I proceeded as follows. First, I excluded firms from the USA, China and Japan, as they dominate the sample (accounting for 17.52%, 15.35% and 10.45% of the sample, respectively). Second, I account for the possible impacts of external shocks such as the 2015 Paris Agreement and the COVID-19 pandemic on CCR. These events are assessed by the indicator variables POST-2015 and POST-2019 which are equal to 1 for the period from 2015 to 2019, and 0 otherwise.

The results are reported in Tables 5 and 6. On the one hand, Table 5 indicates that the significantly positive CCR impact on all business’ risk measures still holds, implying that the above insights are robust enough to be influenced by the USA, China or Japan firms’ selection. On the other hand, Table 6 shows that the coefficient of the binary POST-2015 variable is significantly negative. This result indicates that the CCR at the firm-level decreased after the enactment of the Paris Agreement. Firms become better aware of climate issues and take preemptive and precautionary actions in response to the ratification of this agreement. However, it seems that the COVID-19 outbreak has no effect on CCR. The coefficients of the dummy POST-19 variable are insignificant. Notably, M𝛽 is the most significant risk measure influenced by the COVID-19 pandemic.

Variations in CCR between sectors may mask possible heterogeneity within each sector. To investigate whether this possibility could impact the CCR-CRT nexus, I first ranked sectors (two-digit SIC code) by the CCR average values. Then, I re-run Model (1) using a propensity score matched analysis of industries in which those industries are defined as “highly exposed to CCR” (top-five industries). The propensity score matched analysis allowed us to match each firm in the top-five treatment group with that of the nearly identical features in the control group (slightly exposed to CCR). The results reported in Table 7 show that firms that are highly exposed to CCR behave in a less risk-averse manner. Their propensity to take risks was higher than that of their peers within the control group. Moreover, the magnitude of the causal effect between CCR and CRT is higher in the treatment group than that in the control group. Moreover, it exceeded that reported in Table 2. These results echo the findings of Sautner et al. (2023) who supported a large firm-level CCR within industries. They reinforced acceptance of H1 predicting the positive direction of the CCR-CRT nexus.

Motivated by an emerging strand of climate change literature delving into the effects of CCR on firms’ outcomes and strategies, this study empirically investigated whether firm-level CCR is associated with CRT and how BoGD and ESG performance moderate this association. Based on a sample of 10,874 company observations from 2,146 listed nonfinancial companies in around 34 countries in the 2010–2021 period and using a novel CCR at the company level developed by Sautner et al. (2023), the results reveal a significant positive impact of CCRs on CRT, supporting the value enhancement premise. By examining the moderating effects of BoGD and ESG performance, I reveal that excess risk-taking derived from stressful climate events seems to be mitigated in the presence of highly gender-diversified boards. However, ESG performance plays a meaningful role in CRT with increasing climate risks. Furthermore, the dual effect of BoGD and ESG performance on the relationship between climate and CCR-CRT shows that they act as substitutes: a higher level of the former reduces the effect of the latter. Together, these results prove that firms may gain advantages by integrating ESG practices and BoGD into their climate risk management strategies. These findings are robust to several robustness checks.

The results of this study have several important implications. On the managerial side, managers aiming to effectively alleviate CCRs should be conscious of the role of incentives for risk-taking. In times of climate hazards, corporate policies that encourage risk-taking could help firms’ commitment to serve stakeholders through the acceptance of projects with positive net present values. The latter helps firms to achieve long-term returns as well as managers building their careers and compensation. This could be achieved through pay-to-performance rewards such as stock-options or tax breaks for CEOs demonstrating high levels of risk-taking. Furthermore, the findings are beneficial for investors in making the best generative investments decisions. The positive effect of CCR on CRT highlights the financial implications of extreme climatic events. Therefore, investors may incredibly appreciate CRT as part of their investment strategies during climate hazards. In addition, this study’s insights encourage firms to adjust their strategies toward BoGD and ESG practices. The negative (positive) moderation effect of BoGD (ESG performance) on the CCR-CRT nexus suggests that improving women’s presence in boardrooms (ESG commitment) mitigates CCRs.

On the academic side, this study contributes to the current body of knowledge in two ways. First, it advances our understanding of the relationship between CCR and CRT. To the best of our knowledge, except for Xu et al. (2022), no study has delved into this nexus. Furthermore, while previous studies have separately investigated the effects of BoGD and ESG performance on CCR and CRT, this is the first study to investigate the dual effects of these variables on the CCR-CRT nexus. Additionally, this study contributes to the debate on the significance of good corporate governance practices in CCR management. The results show that higher BoGD and better ESG commitment result in better corporate immunity against incidences of harmful climate changes. Second, in contrast to the extant literature which is concentrated in the USA and China, this study provides a more extensive analysis by considering a diversified group of firms from 34 economies around the world. The results achieved are indeed valuable for policymakers who are called upon to develop appropriate organizational mechanisms that encourage managers to actively promote more environmentally friendly activities (e.g. reducing CO2 emissions and using clean and renewable energy).

This study had inherent limitations that merit additional attention. First, the sample was exclusively limited to the 34 economies covered by Sautner et al. (2023). This limitation prevents the study results from being applied more broadly in the global context. Hence, further studies are needed to explore the CCR-CRT nexus in different settings worldwide. Second, the study mainly recognizes BoGD and ESG performance as moderators of the CCR-CRT causal effect which may limit the comprehensiveness of the real attributes of this effect. Subsequent studies are called to examine other possible channels through which CCR promotes CRT. In addition, addressing other aspects of boardroom diversity and/or alternative ESG scopes may upgrade knowledge around CCR antecedents. Third, since ESG encompasses environmental (E), social (S) and governance (G) dimensions, scholars are called to analyze them separately to explore their independent and interactive effects on the CCR-CRT tie, thereby identifying the most influential ESG practices. Fourth, scholars may verify the robustness of my results by scoring using alternative ESG standards and other sources (e.g. Refinitiv or Morgan Stanley capital international). Furthermore, scholars are encouraged to explicitly analyze how BoGD and ESG performance affect different types of stakeholders (e.g. shareholders, creditors or employees). Indeed, beyond BoGD and ESG performance, firms may adopt other governance or strategic mechanisms (e.g. cross-departmental coordination and supply chain management) to address climate risks. Researchers can provide additional insights into how these mechanisms affect climate change mitigation. Last but not least, future research may enrich the contextual discussion of the “value-enhancing” versus “value-destroying” hypotheses by assessing the influence of diverse institutional settings, regulatory frameworks and cultural differences on firm‐level CCR-CRT nexus.

The author is grateful to the associate editor and anonymous reviewers for their constructive comments, which have significantly contributed to the refinement of this study.