“Green Finance, Policy Uncertainty, and Corporate Environmental Performance: Institutional Pathways in Developed Markets”

3.1. Green finance and corporate environmental performance (CEP)

The Resource-Based View (RBV) suggests that firms with access to unique strategic resources such as long-term capital can develop rare and inimitable environmental capabilities, leading to superior competitive advantages (Barney, 1991). Meanwhile, according to Stakeholder Theory, firms act on the expectations of regulators, investors and customers enacting environmentally friendly activities in order to protect their legitimacy (Freeman 1984).

Empirical evidence of the developed market reveals that the firms who engage in green bonds issuing or sustain-linked financing are able to raise ESG ratings and lower down pollution intensity (Flammer, 2021; Sun et al., 2023; Zhao et al., 2024). Hence:

3.2. The moderating role of economic policy uncertainty (EPU)

Economic policy uncertainty (EPU) represents the uncertainty in fiscal, monetary or regulation environments, whichwould increase firms’ perceived risk and discourage long-term investment (Baker et al., 2016; Bloom et al., 2008).

Based on Real Options Theory (Dixit & Pindyck, 1994), under high uncertainty firms delay or reduce the scale ofirreversible investments like low-carbon transition. In developed markets also, fluctuations in the form of carbon prices, tax policies or climate-policy declarations have been shown to dampen connection between green capital and CEP (Pastor & Veronesi, 2013; Wang, 2024).

Emerging studies (Zhang et al., 2021; Dong & Zhang, 2024) find that EPU stifles firm investments into sustainability and green innovation. Accordingly:

3.3. Institutional buffers in developed markets (TCFD/CSRD Disclosure)

Institutional Theory, for example, specifically considers how regulatory and governance regimes might influence company strategy (North, 1990; Scott, 1995). Disclosure requirements such as the EU Corporate Sustainability Reporting Directive (CSRD) and the Task Force on Climate-related Financial Disclosures (TCFD) help to reduce information asymmetries and to increase investor confidence in more developed markets (Krueger et al., 2020). Firms with more-better-quality disclosures have better climate-risk management capabilities, thus being better able of turning governance mechanisms into environmental performance.

3.4. Commitment mechanisms and long-term performance (SBTi Adoption)

Such commitment tools such as the SBTi signal credible long-term decarbonization paths (Sullivan & Gouldson, 2021). These commitments reduce the risk of perceived greenwashing, and foster the trust of stakeholders. As a result of greater alignment with global climate goals, the companies that set SBTi targets are able to use GF for more substantialenvironmental improvements (Abbasi et al., 2022).

3.5. Triple interaction: gf, epu, and institutional buffers

Institutional quality and disclosure readiness may serve as buffers in extremely uncertain environments. While weaker firms are more susceptible to uncertainty shocks, firms with strong TCFD/CSRD disclosures or SBTi adoption may maintain CEP improvements despite elevated EPU. This dynamic is a novel contribution because it has not been empirically tested in developed-market contexts.

3.6 Conceptual framework explanation ( Figure 1)

Figure 1. Theoretical Framework.

The conceptual model demonstrates the pathways through which green finance influences corporate environmental performance under varying levels of economic policy uncertainty. Consistent with the resource-based view and stakeholder theory, access to green financial resources enables firms to invest in low-carbon technologies, renewable energy, and emission-reduction strategies, thereby improving environmental performance. However, economic policy uncertainty can weaken this relationship by increasing investment risk and discouraging firms from undertaking irreversible sustainability investments, as predicted by real options theory. Institutional governance mechanisms, including high-quality climate disclosure and science-based emission reduction commitments, can mitigate these negative effects by strengthening transparency, improving policy credibility, and supporting long-term environmental investment strategies. The framework also incorporates firm-level control variables such as carbon price exposure, green institutional ownership, and renewable energy share, which capture contextual factors influencing firms’ environmental strategies in developed economies.

Overall, the conceptual framework highlights that the environmental effectiveness of green finance depends not only on financial resources but also on policy stability and institutional governance. Perera (2021) found that microfinance significantly contributes to women’s economic empowerment in rural Sri Lanka. (Perera, 2021).

Figure 1 illustrates the conceptual framework used to examine the relationship between green finance and corporate environmental performance in developed markets. In this framework, green finance (GF) represents the main explanatory variable, while corporate environmental performance (CEP) is the dependent variable. The model proposes that access to green financial resources enables firms to invest in environmentally sustainable technologies and practices, thereby improving their environmental performance. Economic policy uncertainty (EPU) is introduced as a moderating variable that may weaken the positive impact of green finance on corporate environmental performance. When policy environments become unstable, firms may delay or reduce long-term environmental investments due to regulatory and economic uncertainty.

In addition, institutional governance mechanisms including climate disclosure quality under frameworks such as the Task Force on Climate-related Financial Disclosures (TCFD) and the Corporate Sustainability Reporting Directive (CSRD), as well as corporate commitments to the Science-Based Targets initiative (SBTi) are expected to strengthen the effectiveness of green finance and mitigate the negative influence of policy uncertainty. These mechanisms enhance transparency, improve investor confidence, and support more stable environmental investment strategies.

The conceptual framework therefore captures both direct and interaction effects between financial resources, policy environments, and institutional governance mechanisms in shaping corporate environmental performance in developed markets.

Hypothesized Relationships.

Path 1: GF → CEP (direct positive effect; H1).

Path 2: GF × EPU → CEP (negative moderating effect; H2).

Path 3: GF × TCFD/CSRD → CEP (positive institutional reinforcement; H3).

Path 4: GF × SBTi → CEP (positive commitment effect; H4).

Path 5: GF × EPU × Institutional mechanisms → CEP (buffering triple interaction; H5)

4.1. Research framework

This paper explores the impact of Green Finance (GF) on Corporate Environmental Performance (CEP) in developed markets, noting the moderating role of Economic Policy Uncertainty (EPU) as well as buffer effects from institutional arrangements including TCFD/CSRD disclosure quality and decision to join Science-Based Targets initiative (SBTi).

Equal representation across different types of industries with high environmental exposure (energy, manufacturing), medium environmental exposure (transportation), and low environmental exposure (technology & services) was sustainedthrough a stratified sampling strategy. This method also increases the external validity and assures that results are representative of both high- and low-carbon sectors, which are prominently included in decarbonization policies forseveral developed economies.

To account for the intricacies of institutional and firm-level mechanisms, a number of other new control variables were added including: Carbon Price Exposure (CPE), SBTi Commitment, Green Institutional Ownership (Green_IO%), Renewable Energy Share/RE100 membership, and TCFD/CSRD Disclosure Quality.

4.2. Population, sampling, and sample size

The population includes all listed non-financial firms in developed economies with advanced ESG disclosure regulations. A two-stage stratified sampling design was implemented:

Sample Size:

Approximately 1,200–1,500 firms were analyzed, yielding 12,000–15,000 firm-year observations (2012–2024). This longitudinal scope covers critical milestones such as the Paris Agreement (2015) and the EU’s CSRD introduction, ensuring adequate statistical power and capturing structural policy shifts.

Justification:

The sampling design ensures (1) wide coverage across countries and industries, (2) temporal depth to observe policy transitions, and (3) focus on environmentally sensitive sectors most influenced by EPU and green finance availability.

4.3. Sample composition

The final dataset reflects the following approximate distribution of firms by sector and geographic region. The allocation ensures coverage of carbon-intensive industries and variation across institutional contexts in developed markets.

The sectoral distribution follows proportional weighting: 20% Energy & Utilities, 30% Manufacturing & Heavy Industry, 15% Transportation & Logistics, 35% Technology & Services., The geographic distribution mirrors the relative weight of green finance markets globally: 40% North America, 35% Europe, 25% Asia-Pacific., Final dataset covers 1,280 firms (12,000–15,000 firm-year observations) for the period 2012–2024, ensuring statistical power for panel data econometric models.

4.4. Data sources

The study uses multiple reputable international databases to construct the dataset and ensure reliable measurement of financial, environmental, and institutional variables. Firm-level financial data, including green bonds, loans, equity structures, and firm fundamentals, were obtained from Refinitiv Eikon, Bloomberg, and Compustat Global. Corporate environmental performance (CEP) was measured using environmental pillar scores from Refinitiv ESG datasets, MSCIESG ratings, and climate disclosure scores from CDP. Economic policy uncertainty was captured using the widely used indices developed by Scott R. Baker, Nicholas Bloom, and Steven J. Davis. Carbon price exposure data were collected from the European Union Emissions Trading System, the UK Emissions Trading Scheme, and the World Bank Carbon Pricing Dashboard. Additional information on corporate climate commitments, institutional ownership, and renewable energy adoption was obtained from the Science Based Targets initiative database, Morningstar sustainable finance datasets, the Principles for Responsible Investment, and the RE100 initiative, complemented by energy transition data from BloombergNEF.

4.5. Variable definitions

Classic firm-level controls: Firm size (lnAssets), Leverage (Debt/Assets), Profitability (ROE), Industry dummies, and Year dummies.

4.6. Econometric specification

Baseline Model

Where:

Difference-in-Differences (DiD) extension

To capture the causal impact of green finance adoption, a Difference-in-Differences (DiD) framework compares treated firms (with GF exposure) and control firms (without GF exposure) before and after key policy events (e.g., Paris Agreement 2015, TCFD introduction 2017):

The coefficient δ3δ3 estimates the treatment effect of GF policies on CEP, validating the robustness of the main FE model.

Extended Interaction Models

These interaction terms clarify how institutional and governance quality modify uncertainty’s influence on environmental outcomes.

4.7. Estimation strategy, using software stata and python

All estimations were conducted using STATA 18 and Python 3.10 to ensure methodological robustness and reproducibility. The primary analysis employs a two-way fixed-effects panel model controlling for firm-specific and year-specific heterogeneity. Standard errors are clustered at the firm level to address heteroskedasticity and serial correlation. To address potential endogeneity concerns, additional robustness checks were conducted using instrumental variable (IV–2SLS) estimation and propensity score matching combined with difference-in-differences (PSM–DiD). Furthermore, dynamic panel estimations and placebo tests were implemented to validate the stability of the results. Perera et al. (2026) demonstrate that digital green finance significantly improves corporate environmental performance. (Perera et al., 2026).

The complete econometric design of how green finance effects CEP in the presence of different economic policy uncertainty levels (EPU) in developed financial markets is shown in Figure 2.

Figure 2. Empirical Strategy Overview.

Figure 2 showsThe methodology starts with a two-way fixed-effects (FE) panel estimation to control for firm and year heterogeneity (along the lines of Eq.) before moving on to instrument-variable (IV) regressions due to concern for endogeneity, and propensity-score-matching (PSM) methods to further reinforce causal inference. We then use a Difference-in-Differences (DiD) approach to compare treated (green financed) and control firms around key sustainability policy events (e.g., Paris Agreement, TCFD introduction). Protractinteraction models contain both moderator and buffer channel GF × EPU, GF × TCFD/CSRD, GF × SBTi, and GF × EPU × Disclosure/SBTi to examine the impacts of institutions and governance on CEP. Finally, robustness is compliedwith based on placebo, nonlinear and dynamic GMM methods to form a stringent, multiple-tiered method to verify the hypothesized linkage of green finance–uncertainty–environmental performance relationships for advanced economies.

This research employs Carbon Price Exposure, SBTi commitments, Green Institutional Ownership, Renewable Energy Share and TCFD/CSRD Disclosure Quality to focus on institutional, operational and investorimposedcharacteristics of advanced markets which have not been considered in prior work. Disclosure–Uncertainty Interaction: Examines whether the high-quality disclosures (TCFD/CSRD) attenuate the detrimental impact of EPU on green finance effectiveness. Cross-Country Developed Market Panel: This is the first study that compared across multiple developed markets using robust econometric techniques (FE, IV, PSM) and new disclosure and investor variables.

Figure 3. Green finance, epu, and moderation effects, developed vs emerging markets.

Figure 3 differences in the coefficients for GF, EPU, and the interaction of GF × EPU between developed and emerging markets are exhibited in Figure 02: GF effect: relatively positive and stronger in developed markets (β ≈ 0.42) than in emerging markets (β ≈ 0.29). That is because stronger institutions facilitate transforming green finance into environmental performance as a result. EPU Effect: Negative for both, but much larger in emerging markets (β ≈ − 0.21) than in developed countries (β ≈ − 0.12), consistent with the notion that weaker policy environments increase the size of uncertainty shocks. GF × EPU Moderation: Both significant, but the interaction effect is more negative in emerging markets, as it should be. H2, and H3. and EPU is more negative for the effect Green finance on Green finance onDestroywindows More importantly, in supporting H1, H2, and H3 this figure suggests that the channel. It really demonstrates how the institutional differences connect to the aims of your research, and it’s well articulated.

5.8. Summary statistics and multicollinearity analysis

5.8.1 Establish sample representativeness and institutional context

The descriptive evidence indicates that companies in developed markets have high basic levels of CEP (ie, mean = 64%), indicates that environmental performance is already embedded in the companies’ strategies. The percentage that Occupies in financing amount is 18.6%, GFs firm financing function is increasing, but it has been incomplete entirely (Flammer, 2021). That the EPU is not significant is interested as the average level (102) for EPU indicate that political uncertainty is not an irrelevant issue, even for developed markets, and that this could potentially affect investment decisions (Baker, Bloom, & Davis, 2016).

The five new control variables are the institutional preparation and market dynamic: CARBON PRICE EXPOSURE (CPE) (average = 14.5) indicates that the firms are under direct carbon-pricing regimes, hence the higher pay-offs to companies to have GF practices (World Bank, 2023). SBTi targets (46% of firms) as serious decarbonization pathways, intended to join global net zero benchmarks (Sullivan & Gouldson, 2021). This is also evidence of a strong ESG investor pressure in western courts (Krueger et al., 2020). LVR: even impressive Green Institutional Ownership (29%) suggest a strong ESG investor pressure in the west court (Krueger et al., 2020). Share of Renewable Energy (42%) is real operation decarbonization of RE100 and clean power purchasing (CDP, 2022). TCFD/CSRD disclosure scores (mean = 2.9/5) also suggest some progress in terms of climate-related transparency, however it varies between markets (EC, 2022). Perera et al. (2025) argue that green finance research has expanded significantly under policy uncertainty conditions. (Perera et al., 2025).

5.8.2. Bivariate associations: correlation matrix

Besides testing for multicollinearity, the correlation matrix allows for a descriptive index of interdependence betweenvariables. This is useful for conservation purposes, to establish the preliminary relationships and channels through whichgreen finance may influence CEP.

Table 4 correlation matrix indicates that there is a positively correlation between GF and CEP (r = 0.32, p < 0.01), thusproviding support for the baseline hypothesis. EPU is negatively related to CEP (r = −0.19, p < 0.05) which imply that uncertainty may erode environmental performance according to real-options theory (Dixit & Pindyck, 1994). The institutional (CPE, SBTi, Green IO%, Renewable Share, TCFD/CSRD disclosure) enablers of sustainability are positively associated with CEP (Krueger et al., 2020; EC, 2022). Significantly, no relationship is greater than 0.70, suggesting no collinearity issues (Hair et al., 2019).

Figure 4. - GF, EPU, and GF *EPU effects across robustness variants (Developed markets).

Figure 4 shows how GF, EPU, and GF × EPU affect different types of robustness in developed markets. The bars show the coefficient estimates (β) from two-way fixed-effects models for the full sample, the high-EPU subsample, the low-EPU subsample, and a non-COVID sample (not including 2020–2021). Green Finance (GF) is always positive and bigger in low-EPU settings, while Economic Policy Uncertainty (EPU) is negative, and this effect is stronger in high-EPU settings. The interaction GF × EPU is most negative in high-EPU settings and stays negative even when COVID-era years are left out. This shows that the moderating effect is structural and not caused by a crisis. There were fixed effects for the firm and the year, and the robust SE was clustered at the firm level.

Overall, the empirical results consistently demonstrate that green finance improves corporate environmental performance, while economic policy uncertainty weakens this relationship. Institutional governance mechanisms such as disclosure quality and climate commitments help mitigate these negative effects, highlighting the importance of stable policy environments and strong governance frameworks for effective green finance.

Direct effect of green finance on CEP To test the direct effect of green finance on CEP, the baseline RE and FE regression models are estimated (H1). This step answers the question of whether GF is a root cause of CEP in developed markets without including moderation (EPU) or institutional interaction terms. We present the results in Table 04, which reports coefficients, clustered standard errors and model fit statistics.

Table 5 The baseline regression results strongly support Hypothesis 1. Green finance has a positive and statistically significant effect on corporate environmental performance (β = 0.352, p < 0.01), indicating that firms with greater access to green financial instruments achieve higher environmental performance. This finding supports the resource-based view and stakeholder theory, suggesting that access to sustainable financial resources enables firms to invest in cleaner technologies and environmentally responsible practices.

To test Hypothesis 2 (H2), we further extended the baseline model by including the interaction term GF × EPU. This is how we test the hypothesis that EPU may weaken the positive effect of green finance on corporate environmental performance (CEP) in developed countries in this study. The model further includes firm-level mechanisms (CPE, SBTi, Green_IO%, RE_Share, TCFD/CSRD disclosure) along with firm and year fixed-effects. The results are presented in Table 05 on the full sample of 1,280 firms (6,200 firm-year observations).

Table 6 Findings provided strong support for H2. The coefficient of green financing is still positive and significant (β = 0.428, p < 0.01 established in developed markets. Nevertheless, the interaction term (GF × EPU) is negative and significant (β = −0.0042, p < 0.05), meaning that higher degree of economic policy uncertainty weakens the positive impacts of GF on environmental performance. This result is also in line with real options theory, in that, in the face of uncertainty, firms tend to postpone and adjustirreversible decisions (e.g. investments in low-carbon technologies) downwards (Dixit & Pindyck, 1994; Pastor & Veronesi, 2013). This dynamic is first validated by empirical studies that reveal that EPU negatively influences firms’ innovation1 and the strength of their abatement investment (e.g., Zhang et al., 2021; Wang, 2024).

5.8.2 Estimation strategy

To set up a reference point of empirically observed linkage, we start with a two-way fixed effects (FE) model to account for both firm-specific and year-specific unobserved heterogeneity. In this design, it is impossible for unobservable time-varying factors (e.g., corporate culture and sectoral regulations) and macroeconomic shocks shared by all firms within the given year to confound the estimated link between green finance (GF), economic policy uncertainty (EPU), and corporate environmental performance (CEP). As is usual in the econometric literature, we cluster standard errors at the firm level to allow for potential correlation, across time, within firms (Wooldridge, 2019).

The model incorporates the five new institutional and firm-level control variables Carbon Price Exposure (CPE), SBTi commitments, Green Institutional Ownership (Green_IO%), Renewable Energy Share (RE_Share), and TCFD/CSRD disclosure quality to pin down the direct effect of GF and EPU on CEP. The estimate is made for the sample of developed market 1,280 firms, and 5,500 firm-year names (2012–2024).

The empirical evidence in Table 7 strongly supports both H1 and H2. First, GF has a positive and very significant impact on CEP (β = 0.362, p < 0.01), which supports that in developed markets, firms with more GF achieve higher environmental performance. The result is in line with the stakeholder theory (Freeman, 1984) and the resource-based view (Barney, 1991) prediction that access to sustainability-related financial resources increase firms’ capacity to adopt cleaner technologies and legitimize themselves toward their stakeholders. This appears to be confirmed by empirical evidence: Flammer (2021) demonstrates that U.S. firms issuing green bonds have improvements in environmental ratings and Tan (2025) finds similar effects for technology upgrading.

5.8.3 Robustness check i: instrumental variable (IV/2SLS) estimation

While the 2-way fixed effect (FE) baseline controls for observed heterogeneity endogeneity is still a concern. Particularlyfirms with higher CEP may be more likely to receive green finance (reverse causality) or omitted variables (e.g., unobserved policy shocks) might bias estimates346. To control for this non-randomness, we use instrument variable (IV) estimation based on two-stage least squares (2SLS). 3.2 Instrument selection: Consistent with the extant sustainable finance literature (Flammer, 2021; Tang & Zhang, 2020), we instrument firm-level GF with country-level lagged green bond market development and share of global green fund inflows. These are associated with access to green finance, but considered to be exogenous to the CEP of a specific firm atany given point in time.

Table 8 The IV/2SLS test result suggests that GF has a significant positive impact on CEP (β = 0.389, p < 0.01), which is in linewith the baseline FE model but a little bit higher. This indicates that any residual effect of reverse causality (stronger CEP luring GF) was not forcing the baseline results. The first-stage F-statistic of 19.8 is above the standard rule of thumb of 10, suggesting that our instruments are strong and valid (Staiger & Stock, 1997). Economic policy uncertainty still exhibits a significantly negative effect (β = −0.152, p < 0.05), supporting H2: graterpolicy volatility is still undercutting GF’s environmental benefits. These results support the intuition of real options theory (Dixit & Pindyck, 1994), and attest the importance of stable institutional architecture to secure the efficiency of the capital allocation.

5.8.4 Robustness check II: Propensity score matching (PSM) with DID

To enhance causal reasoning, we also add a propensity score matching (PSM) in combination with the difference-in-differences (DID) estimator to the FE and IV/2SLS specifications. PSM generates a balanced control group of non green finance accessing firms which are selected on the basis of size, leverage, profitability, and sectoral exposure to avoidselection bias (Rosenbaum & Rubin, 1983). By combining matching with DID, we compare the pre- vs. post-treatment changes in CEP between recipients of GF (treated firms) and matched non-recipients (control firms).

The coefficient is estimated using the sample of 1,280 firms that originate from developed markets (5,500 firm-year observations, 2012–2024), with treatment being receipt of nontrivial GF inflows (green bonds, sustainability loans or ESG-linked credit).

Table 9 The PSM-DID results validate the validity of the main findings. First, GF remains significant and positive for CEP (β = 0.371, p < 0.01), which supports H1 and suggests that access to GF leads to environmental gains that are, for the most part, observable in the environment after controlling for selfselection bias. Second, the coefficient of EPU is negative and significant, β = −0.139, p < 0.05), providing evidence for H2-policy uncertainty indeed suppresses CEP effects. Finally, the interaction term (GF × EPU) is once more negative and statistically significant (β = −0.0038, p < 0.05), meaning that uncertainty reduces the benefits GF.

These results are the in line with previous studies on sustainability finance, the in which quasi-experimental an methods are used to address the issue of selection bias (Flammer, 2021; Sun et al., 2023). Moreover, the post-matching diagnosis demonstrates that the average covariate bias is 3.4%, below the 10% threshold (Caliendo & Kopeinig, 2008), which means that the treated and untreated groups are similar.

5.13. Robustness check III: Alternative CEP measure (Emission Intensity Reduction)

In order to check the robustness of our results, we conduct the baseline regression with secondary measures of CEP. Inbaseline models, ESG environmental scores may be indicative of disclosure/reporting practices rather than actual operational effects. Hence, to reinforce their validity we include the following two other measures of performance: (i) emissions intensity, as a direct operational measure (the log of scope 1 + 2 emissions divided by output), and (ii) CDP climate scores, as a disclosure- and governance-based measure of environmental performance. These multi-specifications strategy enables us to test the parallel effects of GF and EPU on CEP on the basis of rating-based, operational and disclosure CEP dimensions.

Dependent variables by column: (1) ESG Environmental Score (0–100), (2) Emissions Intensity (Scope 1 + 2, log), (3) CDP Climate Score (A = 4 … D = 1).

Table 10 For ESG Environmental Scores (Model 1), GF continues its significant influence at a positive level (β = 0.354, p < 0.01), and EPU shows a significantly negative effect (β = −0.142, p < 0.05). The coefficient for the interaction term (GF × EPU) is negative (β = −0.0041, p < 0.05), which strongly supports H1 and H2. These results have confirmed that green finance enhances firms’ environmental performance; however, the effectiveness of green finance weakens with increasing level of policy uncertainty.

On the other hand, the findings for Emissions Intensity (Model 2) find no impact of either GF and EPU. This implies that real emission cuts need long time horizons and could lag behind financial mobilization, as supported by previous findings that decarbonization returns come only after continued capital expenditure and technology diffusion (CDP, 2022; Cui, 2023).

Stared whether the 25 For CDP Climate Score (Model 3), GF is still both significant and positive (b = 0.209, p < .01), while EPU stays negative and significant (b = −0.088, p < .05). The interaction term is also negative (β = −0.0022, p < 0.05), suggesting that policy volatility negatively affects disclosure-based climate performance. These findings the suggest that, despite accruing transparency and governance benefits from green finance, firms are not immune to policy uncertainty exactly.

5.8.5 Robustness checks

We test if the EPU moderation is different across levels of policy stability within the developed market by separating the sample into high and low EPU countries (where high and low EPU countries are countries with EPU value above and below the EPU index median). This sub-sample analysis sheds light on within-developed heterogeneity, which constitutesa theory-based test of the GF-CEP link in different institutional certainty conditions.

Table 11 tests H2 and H3 more directly by splitting the sample into high- vs. low-EPU developed economies. In high-EPU situations, the GF coefficient is small and not important, while the EPU coefficient is very negative and the GF × EPU interaction is very negative The findings demonstrate that GF is insignificant in high-EPU conditions, but positive and significant in low-EPU conditions, and that EPU and the interaction term are negative only in the high-EPU sub-sample. This result supports the view that the same remains for a stable policy that is a prerequisite for the effectiveness of GF, as evidenced in real options (Dixit & Pindyck, 1994) and institutional theory (North, 1990).

5.8.6 Robustness check IV: Post-COVID exclusion analysis (2012–2019)

To verify the results are not dominated by the peculiar dynamics of the COVID-19 era (2020–2021), we re-estimate the baseline specification with pre-COVID samples only (2012–2019). This period gives us a cleaner environment by isolating global pandemic shocks, COVID-19 fiscal packages and unusual regulatory moves which may have transitedboth green finance (GF) flows and coporate environmental performance (CEP). The estimate includes 1,280 firms and 4,000 firm-year observations (2014–2019).

Dependent Variable: Corporate Environmental Performance (CEP).

Table 12 The exclusion of COVID-19 years validates the strength of the main results. Green finance remains to exert a positive and significant influence on CEP (β = 0.341, p < 0.01), supporting H1. Likewise, EPU continues with its negative influenceand statistically significant (β = −0.121, p < 0.05), and the interaction term GF × EPU is still negative and significant (β = −0.0036, p < 0.05), supporting again H2.

The findings imply that the relationships identified are not caused by transient pandemic-induced shocks, but insteadreveal more fundamental dynamics across the levels of GF, EPU, and CEP in developed markets. This is in line with previous studies that suggest the COVID years entail unusual volatility and government interventions that could affect the estimates if not accounted for (Demir & Danisman, 2021; OECD, 2021).

5.8.7 Robustness check V: sectoral heterogeneity

To check whether the effectiveness of GF differs between industries, we divided the sample into high-emitting sectors (energy, heavy industry, and transportation), and low-emitting sectors (services, high-tech, and retail). This enables us to examine whether the GF–CEP connection and the EPU moderating impact are more pronounced in emissionheavysectors, in which sustainability transitions become capital- and policy-bound. The estimation again employs firm- and year-fixed effects with the earlier described controls on 1,280 firms; 5,500 firm-year observations (2012–2024).

The results demonstrate that the GF–CEP association is more robust in high-emission sectors (β = 0.421, p < 0.01) than the low-emission sectors (β = 0.189, p < 0.05). Similarly, the moderating effect of EPU is only significant in high-emission sectors (β = −0.0062, p < 0.05). In the low-emission industries, both EPU and the interaction term are not statistically significant.

This heterogeneity indicates that energy, the manufacturing, and transport firms are particularly exposed more to policy uncertainty, as their environmental strategies involve substantial irreversible investments (Dixit & Pindyck, 1994). In contrast, it would be relatively costly for GF to achieve regulatory independence from EPU, so the degree of exposure to regulatory volatility is lower for firms in services and technology industries.

6.1 Policy implications for developed economies

We found as results of the study provide important policy guidance to governments, regulators and financial institutions in developed economies that want to speed up a shift towards low-carbon and climate-resilient economies. First, policymakers should focus on stable and credible climate policy frameworks. Long-term policy frameworks (consistent carbon pricing systems, multi-year climate strategies and clear regulatory standards) are necessary to reduce investment risk and promote large scale environmental investments by firms.

Stable policy environments ensure effectivetranslation of green financial resources into real environmental changes. Secondly, further improving the depth of climate disclosure frameworks would significantly strengthen green finance. Expanding and enforcing international disclosure standards, such as the Task Force on Climate-related Financial Disclosures (TCFD) and Corporate Sustainability Reporting Directive (CSRD), may rely critically on improvingtransparency and reducing information asymmetries between firms and investors. The increased disclosure helps to restore market confidence as well as further reduce the threats of greenwashing and misallocation of sustainable finance.

Third, governments and financial regulators should encourage the use of risk-sharing mechanisms that allow financial institutions to better manage uncertainty regarding green investments. Such instruments including public guarantees, blended finance structures and climate investment insurance schemes can mitigate perceived risks and incentivizebanks and institutional investors to increase their green lending and investing activities. Fourth, targeted transition policies are all the more important for high-emission industries. Industry, transport and energy sectors require large capital expenditure as they work to decarbonise. Policy instruments such as contracts-for-difference, tax breaks, accelerated depreciation for clean technologies and targeted transition subsidies could help to decarbonisethese sectors. Lastly, banks should incorporate dynamic risk assessment in green financial products. Issuing green loans and bonds linked to performance-based sustainability indicators, climate transition plans, and science-based emission reduction targets can also help ensure that companies maintain their environmental performance when political uncertainty is high.

The present study adds to the literature by providing an integrated theoretical framework to understand the nexus between green finance, policy uncertainty, and corporate environmental performance. The first type of contribution is that the findings align with the resource-based view (RBV), showing how access to sustainable financial resources allows firms to develop environmental capabilities and achieve competitive advantages (Barney, 1991). Second, in line with stakeholder theory, the results show that firms respond to pressures from investors, regulators and society, by adopting environmentally-responsible practices that are underpinned by green finance (Freeman 1984).

Third, the findings verify the predictions of real options theory by demonstrating that policy uncertainty (1) decreases the expected return of long-term environmental investments and consequently deters firms from incurring irreversible sustainability projects (Dixit & Pindyck, 1994). Notably, the study also emphasizes institutional theory, showing how governance mechanisms like climate disclosure frameworks and emission reduction commitments can reduce adverse impacts driven by policy uncertainty (North, 1990; Scott, 1995). This study constructs a comprehensive framework to explain how financial systems, policy environment and governance structure interact with each other to determine corporate environmental outcomes examining both dual interaction effects (GF × EPU, GF × Disclosure) and triple interactions (GF × EPU × Institutional mechanisms). The analysis also shows significant cross-country and sectoral differences in the effectiveness of green finance.

Yet in countries with lower policy uncertainty and stronger institutional framework, including the United Kingdom, Germany and Sweden, the positive relationship between green finance and corporate environmental performance is more pronounced. These economies are characterized by stable climate policies, transparent disclosure systems and mature financial markets that facilitate such long-term sustainable investments. On the contrary, we find that green finance effects are weaker in countries with more economic policy uncertainty. Fluctuations in fiscal policy and regulatory frameworks, including commitments to climate policy targets, within evenadvanced economies can deter firms from making large-scale environmental investments. Sectoral differences are also evident. High-emission sectors like energy, heavy manufacturing and transportationexhibit astronger response to green finance stemming from the capital intensive requirements of decarbonizing these emissions-heavy industries. But these sectors are also more exposed to policy uncertainty because their investment decisions rely on regulatory set-ups such as carbon pricing and environmental standards. On the other hand, low-emission sectors like services and technology show weaker yet more stable relationships between green finance and environmental performance because of their lower exposure to climate transition risks. The findings underscore the need for policy coordination, sector-specific transition strategies and cross-country harmonization of green finance frameworks.

6.2 Limitations and future research

Although there are some contributions of this study, it does also have several limitations that should be discussed. First, part of the empirical analysis concerns listed non-financial firms in nine developed economies which may restrictthe reach of the findings to other contexts (e.g. emerging markets, SMEs or state-owned enterprises under alternativeinstitutional settings).

Second, corporate environmental performance is measured based on primarily ESG environmental scores, alongsideemissions intensity and CDP climate ratings. While this approach enhances the overall validity of measurement, these indicators are likely influenced at least in part by disclosure practices rather than purely operational environmental outcomes. Third, while the study employs rigorous econometric techniques including fixed effects, instrumental variable estimation and propensity score matching residual endogeneity and unobserved time-varying shocks cannot be completely excluded. There are several avenues for future research building on this work. First, studies may use more specific measures of firm-level uncertainty, for example variables from textual data such as risk disclosures in 10-K filings or sentiment in media articles to derive conceptual definitions about uncertainty. Second, future works may examine non-linear or threshold effects of policy uncertainty on environmental investment behavior (e.g., possible inverted-U relationships). investors and green bond markets to better understand the way in which green finance is channelled through financial systems. Last, newtechnological innovations from the world of finance like FinTech-enabled green finance, blockchain-based sustainability tracking schemes and digital climate-investment platforms merit attention as potential pathways for future research between financialization, innovation and environmental performance.

Third, more work could be done on financial intermediaries such as banks, institutional In general, the results accord with previous evidence showing green finance is instrumental to enhancing corporate environmental performance but its effectiveness is contingent on stable policy environments and robust institutional governance. Strengthening climate disclosure systems, reducing policy uncertainty, and promoting credible climate commitments are hence crucial to unlocking the full environmental potential of green finance in developed markets.