This study aims to investigate the relationship between firms' nature dependence level and cost of capital (cost of equity, cost of debt and weighted average cost of capital). We also tested how firm-level earnings volatility and macro-level climate vulnerability moderate the relationship between firms' nature dependence and the cost of capital.
This study is based on panel data of 57,297 firm-year observations from 87 countries covering the period 2015–2023. Fixed-effects (FE) regression models are employed to estimate the effect of nature dependence on financing costs and to test the proposed moderating channels. The study also uses alternative measures of nature dependence and addresses potential endogeneity concerns through a two-stage least squares (2SLS) instrumental-variables approach.
The results show that firms with greater dependence on the natural environment face significantly higher financing costs, aligning with risk-based asset pricing theory and the natural resource-based view. Furthermore, the effect is stronger for firms with high earnings volatility, suggesting that nature-related risks intensify existing firm-level uncertainty. At the macro level, the impact of nature dependence on financing costs is weaker in countries with high climate vulnerability, where broader climate risks may already be priced by financial markets. In less climate-exposed countries, nature dependence emerges as a clearer and more influential risk signal for investors.
This is the first study that examines the relationship between nature dependence and different measures of firms' cost of capital. By identifying the moderating roles of earnings volatility and climate vulnerability, this paper offers new insights with important implications for corporate risk management, investment decisions and financial regulation.
We must treat the natural world as we would the economic world—protecting natural capital so that it can continue to provide benefits well into the future
David Attenborough, Natural historian and narrator of the Netflix/WWF documentary series Our Planet and Christine Lagarde, President of the European Central Bank (ECB), former IMF managing director [1].
1. Introduction
The global economy is inextricably linked to the health of the planet's ecosystems. Natural ecosystems provide raw materials and resources as well as climate stability, which constitute the foundational capital for firms' operations (Dasgupta, 2021). However, as biodiversity declines at unprecedented rates, the firm's nature dependence has transformed from a structural necessity into a significant financial liability. The extant literature has recognized the importance of nature-related risks on firms such as valuation (Bracking, 2012), innovation (Mohan and Morris, 2025) and creditworthiness (Vu and Vo, 2026). While the literature on climate-related financial risk is well-developed, the specific nexus between a firm's reliance on ecosystem services and its financing costs remains underexplored.
Nature-related risks, arising from ecological degradation and biodiversity loss, can directly threaten firms' revenues and operational stability, ultimately altering risk perceptions among lenders and investors. In 2020, the World Economic Forum identified biodiversity loss as one of the five most significant threats to the global economy. It is estimated that over 50% of global gross domestic product relies moderately or heavily on natural capital, making economic activity highly exposed to the risks associated with ecosystem degradation (World Economic Forum, 2020; World Economic Forum and PwC, 2020). Kempa et al. (2021) indicated that renewable energy firms tend to face higher borrowing costs in the early stages of technological and market development, but this disadvantage diminishes over time, eventually turning into a cost-of-debt advantage. Yang and Li's (2025) study on A-share enterprises from 2000 to 2023 discloses that biodiversity risk can exacerbate firms' short-term debt for long-term use by weakening corporate risk-taking capacity and asset turnover. Equity investors also demand a risk premium to compensate for the prospect and uncertainty of future biodiversity-related regulation and litigation (Garel et al., 2024). However, improvements in environmental, social and governance (ESG) performance can reduce financing costs by signaling superior risk management and improving corporate reputation (Zhao et al., 2025).
Thus, we explore the relationship between a firm's natural dependence and its financing costs, in terms of both cost of debt (COD) and cost of equity (COE). Using global evidence over the period 2015–2023, we find that firms that are more nature-dependent have higher financing costs, with the effect being more pronounced for the COE. These findings remain robust after considering alternative measures of firms' nature dependence and addressing the endogeneity issue.
In addition, a critical and often overlooked dimension of this relationship is the moderating impact of volatility factors. Volatility, whether in the form of financial risk or environmental uncertainty, acts as a primary mechanism through which nature dependence translates into higher costs of capital. This paper aims to bridge this gap by examining the moderating role of volatility, at both firm-level and macro-level uncertainty, on the association between nature dependence and firms' financing costs, with a specific focus on the moderating impacts of volatility factors. We document that nature dependence has a more pronounced effect on financing costs for firms with unstable earnings. In contrast, there is a negative moderating effect of climate vulnerability on the relationship between nature dependence and firms' cost of capital.
This study makes several important contributions to the emerging literature on nature-related financial risks and sustainable finance. First, it provides large-scale cross-country empirical evidence on the impact of firms' dependence on ecosystem services, demonstrating that nature dependence is directly priced in capital markets through higher costs of capital. By documenting equity and debt market responses, this study extends prior ESG and climate risk research on how investors price environmental risks (Chen and Gao, 2012; Kling et al., 2021; Lemma et al., 2019), showing that biodiversity loss and ecosystem degradation are distinct and economically significant sources of systematic risk, rather than merely environmental concerns. Second, our findings contribute to the nascent literature on nature dependence risk by identifying earnings volatility as a key amplification channel, revealing that nature-related risks function as a risk multiplier for financially unstable firms. This insight advances risk-based asset pricing theory (Chen and Gao, 2012; Merton, 1977) by linking environmental dependence with firm-level uncertainty. Third, by uncovering heterogeneous effects across countries with different levels of climate vulnerability (ND-GAIN, 2023), the study sheds light on the interaction between micro-level nature exposure and macro-level climate risks, suggesting that creditors and investors differentially price nature dependence among countries with different levels of climate vulnerability. Finally, our results carry important implications for corporate strategy and financial regulation, emphasizing that nature-related risks should be integrated into firms' risk management frameworks and considered by investors and policymakers seeking to improve the alignment between capital allocation and ecological sustainability.
2. Literature review and hypothesis development
2.1 Nature-related risks and firms' cost of capital
A growing concern in the literature is that large-scale degradation of natural ecosystems may have substantial implications for firms' operations. Deteriorating ecosystem conditions can expose firms to heightened regulatory scrutiny and impose constraints on business activities, thereby affecting firms' risk profiles and economic outcomes (Wagner, 2023). According to risk-based asset pricing theory, asset prices reflect both revisions in expectations about future cash flows and changes in the perceived risk of those cash flows (Chen and Gao, 2012; Merton, 1977). Furthermore, the natural resource-based view suggests that a firm's competitive advantage and vulnerability to risk are influenced by how effectively the firm manages its relationship with the natural environment (Hart, 1995). Therefore, investors require higher expected returns for holding assets that are more exposed to systematic and non-diversifiable risks. Nature-related risks represent such a source of systematic risk because ecosystem degradation, biodiversity loss and environmental regulation affect entire industries and regions simultaneously and cannot be diversified away (Lucey et al., 2025). Existing studies primarily focus on how investors incorporate climate-related risks into financial decision-making, such as asset pricing and firms' cost of capital (Chen and Gao, 2012; Kling et al., 2021; Lemma et al., 2019). In contrast, far less is known about how firms' dependence on natural ecosystems is priced by financial markets, despite the close conceptual links between climate risk and nature-related risk.
The growing relevance of nature-related risks is emphasized by the World Economic Forum (2025), which identifies biodiversity loss and ecosystem collapse as the second most significant long-term global risks. Given the high dependence of economic activity on natural systems, the financial consequences of corporate nature dependence have become an increasing concern for managers, investors and regulators. Firms with greater reliance on natural ecosystems are more exposed to environmental damage and resource overuse, which can elevate litigation risk, reputational losses and operational disruptions (Garel et al., 2026). Similar to climate risks, nature-related risks can affect firms both directly through their operations and indirectly through supply chains. These risks may arise from physical channels such as firms' reliance on ecosystem services and their exposure to ecosystem degradation, which can disrupt access to natural inputs and impair normal business operations. In addition, risks may emerge from transition channels including environmental protection requirements and regulatory compliance pressures that increase operating costs and affect firms' production and investment decisions (Gjerde et al., 2026). Recent evidence further suggests that biodiversity loss represents an emerging financial risk, as equity markets have recently begun to price biodiversity transition risk (Garel et al., 2024). This delayed pricing implies that nature-related risks have not yet been fully internalized by financial markets, highlighting their growing relevance for firms' financing conditions.
Empirical evidence on debt markets also supports the financial materiality of nature-related risks. Using data from the USA syndicated loan market, Canipek et al. (2025) show that lenders incorporate nature-related risks into loan pricing, with firms exhibiting greater dependence on ecosystem services facing higher loan spreads. Specifically, a one percent increase in firm-level nature dependence is associated with an approximately 0.32% increase in loan spreads. Consistent with these findings, survey evidence from portfolio companies indicates that more than half of firms view nature-related risks as financially relevant, while the companies also acknowledge both the challenges and opportunities associated with integrating such risks into financial analysis and investor engagement (Gjerde et al., 2026). Regarding nature-related risks and the COE, Cepni et al. (2024) find that firms with greater exposure to climate risk face a higher COE. The positive association is stronger for firms operating in environments with heightened public and stakeholder attention to climate issues, clearer evidence of climate impacts and tighter financing constraints. Similarly, Chen (2025) shows that corporate biodiversity risk significantly increases the COE capital. The study further indicates that government scrutiny amplifies financing costs for high-risk firms, whereas greater attention from individual investors is associated with lower equity financing costs.
Taken together, prior studies suggest that nature-related risks are relevant for both equity investors and debt holders although their significance may differ across these capital providers. Building on this literature, firms with greater dependence on natural ecosystems are therefore more exposed to adverse operational shocks, supply-chain disruptions and regulatory interventions arising from systematic environmental risks that increase the volatility and downside risk of their future cash flows. As a result, capital providers expect higher risk premia when financing these firms, which is reflected in a higher cost of capital. Accordingly, we propose the following hypothesis:
Nature dependence significantly and positively affects the firms' cost of capital.
2.2 Moderating impacts of cash flow volatility on nature-related risks and firms' cost of capital nexus
Cash flow volatility reflects fluctuations in firms' operating cash flows and serves as an important indicator of uncertainty regarding future cash flow stability (Chen and Gao, 2012). Firms with higher cash flow volatility face greater difficulty in generating predictable returns, which heightens perceived risk of both investors and lenders. Building upon this view, prior studies document that greater earnings instability and uncertainty are associated with higher financing costs, as capital providers require additional risk premia to compensate for increased information asymmetry and default risk (Ghasemzadeh et al., 2021; McInnis, 2010). By contrast, firms tend to borrow more when borrowing conditions are more favorable and associated risks are relatively low, such as when operating costs and earnings volatility are limited (Ginglinger and Moreau, 2023). This suggests that earnings volatility plays a central role in shaping how financial markets price corporate risk exposures. From a risk-pricing perspective, firms with high earnings volatility are viewed as riskier because their cash flows are more sensitive to changes in the overall economic environment, making them more exposed to systematic risk and increasing the risk premium required by capital providers.
In the context of nature-related risks, when earnings volatility is high, nature dependence is more likely to amplify concerns about firms' future cash flows, prompting investors to price nature-related exposures more strongly. This reasoning is consistent with evidence that markets demand higher risk compensation when environmental and transition-related risks interact with firm-level uncertainty (Bolton and Kacperczyk, 2023). Therefore, when earnings volatility is high, the positive association between nature-related dependence and firms' cost of capital becomes stronger, indicating that investors price nature-related exposures more intensively under conditions of higher earnings volatility. As a result, the effect of nature dependence on firms' cost of capital is expected to be stronger for firms with more volatile earnings. We expect that:
Earnings volatilities positively and significantly moderate the relationship between nature dependence and firms' cost of capital.
2.3 Moderating impacts of climate vulnerability on nature-related risks and firms' cost of capital nexus
Compared to nature-related risks, climate risks have received substantial attention in both academic research and financial markets over the past decade. Existing studies document that climate transition risk has been priced since at least the Paris Agreement in 2015 (Bolton and Kacperczyk, 2021, 2023), while climate physical risks appear to have been reflected in asset pricing for an even longer period (Kruttli et al., 2025). This suggests that investors have had sufficient time to form expectations and incorporate climate-related risks into firms' financing conditions. For example, prior research shows that climate-related physical risks can impose significant direct and indirect costs on firms, including damage to physical assets and disruptions to supply chains. These costs may increase firms' financing costs and, in turn, affect their expected profitability as well as investment and financing decisions (Sautner et al., 2023). In addition, climate transition risks such as more stringent environmental regulation and the adoption of renewable energy technologies can further raise firms' operating and compliance costs, thereby contributing to a higher cost of capital (Meneses Cerón et al., 2024).
In contrast, nature-related risks remain at a relatively early stage of recognition in financial markets. Recent evidence indicates that biodiversity-related risks have only recently begun to be priced by investors, suggesting that perceptions of nature risk are still evolving (Canipek et al., 2025; Garel et al., 2024). This difference in the timing and maturity of risk pricing implies that climate risks and nature risks may interact in shaping firms' financing conditions. Nature dependence represents a firm-specific source of risk through firms' reliance on ecosystem services. On one hand, investors and creditors might demand higher capital costs in highly climate-vulnerable contexts due to elevated risk exposure (Cepni et al., 2024). Cepni et al. (2024) show that a key mechanism influencing financing costs is climate transition risk, which arises from uncertainty surrounding the emergence of new business opportunities. On the other hand, when nature-dependent firms operate in highly climate-vulnerable environments, the positive effect of nature dependence on firms' cost of capital may be weaker, as investors already account for significant climate-related uncertainty at the country level or the firms have engaged in pro-environmental policies (Yildiz and Temiz, 2024). In other words, climate vulnerability may attenuate the pricing effect of firm-level nature dependence on firms' cost of capital. Accordingly, we propose the following hypothesis.
Climate vulnerability negatively moderates the relationship between nature dependence and firms' cost of capital.
3. Methodology and data
3.1 Methodology
Based on prior literature that investigates how firm-level sustainability attributes, such as ESG performance, influence the cost of capital (e.g. Priem and Gabellone, 2024; Koutoupis et al., 2026), we develop an empirical model to assess the effect of firm-level nature dependence on financing costs. The model is specified as follows:
where, is firm i's financing cost or cost of capital in year t; is the firm-level nature dependence of firm i in year t; represent a vector of Nth control variable; is the constant; and are coefficients; and are the industry-, country- and year-fixed effects and is the error term.
The cost of capital is proxied using three measures: the firm's weighted average cost of capital (WACC), the COE and the COD. To capture nature dependence, we employ the recently developed firm-level nature dependence dataset of Garel et al. (2026). Accordingly, the measure is constructed by integrating firm-level, segment-level revenue data with dependency materiality ratings from the Exploring Natural Capital Opportunities, Risks and Exposure (ENCORE) database. These ratings assess the reliance of economic activities on 25 ecosystem services grouped into 3 categories: (1) provisioning services (e.g. biomass provisioning and water supply); (2) maintenance and regulation services (e.g. soil quality and pollination) and (3) cultural services. For each ecosystem service, dependence is evaluated based on the degree to which an activity's functioning would be impaired by ecosystem disruption and the financial cost required to adapt to such disruptions.
Consistent with prior studies (Chen, 2025; Koutoupis et al., 2026) that use panel data to test the relationship between ESG and cost of capital, we include a set of control variables: firm size (TA), leverage (LEV), capital expenditure ratio (Capex), revenue growth (Salesgrowth), asset tangibility (Tangibility), profitability (return on assets (ROA)) and firm value (TOBINQ).
To examine how firm-level earnings volatility and macro-level climate vulnerability moderate the relationship between nature dependence and the cost of capital, we estimate the following extended models:
where, represents earnings volatility, which is the standard deviation of operating cash flows computed over a rolling three-year window of firm i in year t. represents climate vulnerability of human societies in a given country to adverse effects from climate-related events.
Definition and data sources of variables in Models (1)–(3) are presented in Table 1.
Definition and data sources of variables
| Variables | Definitions | Data source |
|---|---|---|
| WACC | Weighted average cost of capital | Refinitiv Eikon |
| COE | Weighted average cost of equity, calculated using the Capital Asset Pricing Model (CAPM) | Refinitiv Eikon |
| COD | After-tax weighted average cost of debt, calculated as interest expense scaled by total debt | Refinitiv Eikon |
| Nature_Dep | Firm-level nature dependence. Computed as the revenue-weighted average dependence of each firm on non-cultural ecosystem services. Only ecosystem services with a segment-level dependence score of at least 1 are included in the calculation | Garel et al. (2026) |
| Nature_High | Alternative measure of firm-level nature dependence. It captures the number of non-cultural ecosystem services for which a firm's revenue-weighted dependence score is equal to or greater than 4 | Garel et al. (2026) |
| Nature_Average | Alternative measure of firm-level nature dependence. It is defined as the annual average value of Nature_Dep across all firms in the sample. It captures the overall level of firms' revenue-weighted dependence on non-cultural ecosystem services in a given year | Authors' calculation |
| Size | Firm size, measured as Natural logarithm of total assets | Refinitiv Eikon |
| ROA | Return on asset, measured as Total net income divided by average total assets | Refinitiv Eikon |
| Lev | Leverage, measured as Total debts (short-term and long-term) divided by total assets | Refinitiv Eikon |
| Capex | Capital expenditure, measured as Capital expenditure divided by total assets | Refinitiv Eikon |
| Salesgrowth | Sale growth, measured as Annual percentage change in sales revenue | Refinitiv Eikon |
| Tangibility | Firm tangibility, measured as Total fixed assets scaled by total assets | Refinitiv Eikon |
| Tobinsq | Firm value, measured as Ratio of the firm's market value (market capitalization plus total debt) to the book value of total assets | Refinitiv Eikon |
| GDP | Economic development, measured as Log of GDP per capita for the firm's country | World Bank |
| CFVolatility | Earnings volatility, measured as Standard deviation of operating cash flows computed over a rolling three-year window for each firm | Authors' calculation |
| Vulnerability | The climate vulnerability of human societies in a given country to adverse effects from climate-related events | The Notre Dame Global Adaptation Initiative Index (https://gain.nd.edu/our-work/country-index/) |
| Variables | Definitions | Data source |
|---|---|---|
| WACC | Weighted average cost of capital | Refinitiv Eikon |
| COE | Weighted average cost of equity, calculated using the Capital Asset Pricing Model (CAPM) | Refinitiv Eikon |
| COD | After-tax weighted average cost of debt, calculated as interest expense scaled by total debt | Refinitiv Eikon |
| Nature_Dep | Firm-level nature dependence. Computed as the revenue-weighted average dependence of each firm on non-cultural ecosystem services. Only ecosystem services with a segment-level dependence score of at least 1 are included in the calculation | |
| Nature_High | Alternative measure of firm-level nature dependence. It captures the number of non-cultural ecosystem services for which a firm's revenue-weighted dependence score is equal to or greater than 4 | |
| Nature_Average | Alternative measure of firm-level nature dependence. It is defined as the annual average value of Nature_Dep across all firms in the sample. It captures the overall level of firms' revenue-weighted dependence on non-cultural ecosystem services in a given year | Authors' calculation |
| Size | Firm size, measured as Natural logarithm of total assets | Refinitiv Eikon |
| ROA | Return on asset, measured as Total net income divided by average total assets | Refinitiv Eikon |
| Lev | Leverage, measured as Total debts (short-term and long-term) divided by total assets | Refinitiv Eikon |
| Capex | Capital expenditure, measured as Capital expenditure divided by total assets | Refinitiv Eikon |
| Salesgrowth | Sale growth, measured as Annual percentage change in sales revenue | Refinitiv Eikon |
| Tangibility | Firm tangibility, measured as Total fixed assets scaled by total assets | Refinitiv Eikon |
| Tobinsq | Firm value, measured as Ratio of the firm's market value (market capitalization plus total debt) to the book value of total assets | Refinitiv Eikon |
| GDP | Economic development, measured as Log of GDP per capita for the firm's country | World Bank |
| CFVolatility | Earnings volatility, measured as Standard deviation of operating cash flows computed over a rolling three-year window for each firm | Authors' calculation |
| Vulnerability | The climate vulnerability of human societies in a given country to adverse effects from climate-related events | The Notre Dame Global Adaptation Initiative Index ( |
Following empirical literature, this study employs a firm-level panel regression with fixed effects (FE) (Iqbal et al., 2020). The FE specification assumes a common underlying effect across the estimated models, with variation in observed coefficients attributable to sampling variability (Borenstein et al., 2010). To account for potential heterogeneity, we incorporate year dummies and cluster standard errors at the firm level, thus ensuring robust statistical inference in the presence of heteroskedasticity and within-firm correlation (Stock and Watson, 2008; Anh et al., 2025).
3.2 Sample and data
Our sample includes publicly listed firms from 87 countries and spans the period from 2015 to 2023. The choice of this sample period follows data availability of nature dependence and cost of capital measures, while ensuring a broad cross-country coverage to capture substantial variation in firms' dependence on the ecosystem. Financial data are obtained from the Refinitiv Eikon database, while nature dependence scores are drawn from Garel et al. (2026) and made publicly available at https://osf.io/d85e7/overview. To reduce the influence of extreme observations, all financial variables are winsorized at the 1st and 99th percentiles. After addressing missing values, the final panel dataset comprises 57,297 firm-year observations.
4. Results and discussions
4.1 Descriptive statistics
Table 2 reports the summary statistics for the variables used in Models (1)–(3). The average cost of capital (WACC) in the sample was 7.24%. The mean COE is 8.91%, which is approximately four times higher than the mean COD (2.44%), indicating that equity financing is substantially more expensive than debt financing across firms and countries in the dataset. The maximum COE reaches 24%, compared with a maximum COD of 8.85%. The mean LEV ratio is 22.5%, suggesting that most firms in the sample maintain relatively low borrowing levels. Nonetheless, the maximum LEV value is 1.00, reflecting that some firms operate under very high financial risk. Firms exhibit an average sales growth rate of 10.8%, indicating solid growth over the estimation period. Regarding nature dependence, the average dependence score is 2.83, ranging from 1.95 to 3.83. This distribution shows meaningful variation in firms' reliance on natural capital across the sample. The correlation matrix in Table A1 (Online Appendices) indicates that there are no multicollinearity issues in our regression models.
Descriptive statistics of variables
| Variables | N | Mean | S.D | p50 | Min | Max |
|---|---|---|---|---|---|---|
| WACC | 57,297 | 0.072 | 0.034 | 0.069 | 0 | 0.205 |
| COE | 57,297 | 0.089 | 0.041 | 0.087 | 0 | 0.240 |
| COD | 57,297 | 0.024 | 0.018 | 0.023 | 0 | 0.089 |
| ROA | 57,297 | 0.026 | 0.100 | 0.034 | −1.278 | 0.313 |
| Lev | 57,297 | 0.225 | 0.178 | 0.203 | 0 | 1.000 |
| Capex | 57,297 | 0.037 | 0.039 | 0.026 | <0.001 | 0.293 |
| Salesgrowth | 57,297 | 0.108 | 0.428 | 0.052 | −0.952 | 3.769 |
| Size | 57,297 | 8.758 | 2.801 | 8.466 | −2.129 | 15.420 |
| Tangibility | 57,297 | 0.284 | 0.200 | 0.248 | 0 | 0.896 |
| Nature_Dep | 57,297 | 2.828 | 0.409 | 2.890 | 1.945 | 3.833 |
| Nature_High | 57,297 | 2.104 | 2.261 | 1.000 | 0 | 7.000 |
| Tobinsq | 57,297 | 1.717 | 3.093 | 0.959 | 0.167 | 26.920 |
| CFVolatility | 57,297 | 5,492 | 22,194 | 216.800 | 0 | 271,482 |
| Vulnerability | 57,297 | 0.371 | 0.0440 | 0.381 | 0.251 | 0.569 |
| Variables | N | Mean | S.D | p50 | Min | Max |
|---|---|---|---|---|---|---|
| WACC | 57,297 | 0.072 | 0.034 | 0.069 | 0 | 0.205 |
| COE | 57,297 | 0.089 | 0.041 | 0.087 | 0 | 0.240 |
| COD | 57,297 | 0.024 | 0.018 | 0.023 | 0 | 0.089 |
| ROA | 57,297 | 0.026 | 0.100 | 0.034 | −1.278 | 0.313 |
| Lev | 57,297 | 0.225 | 0.178 | 0.203 | 0 | 1.000 |
| Capex | 57,297 | 0.037 | 0.039 | 0.026 | <0.001 | 0.293 |
| Salesgrowth | 57,297 | 0.108 | 0.428 | 0.052 | −0.952 | 3.769 |
| Size | 57,297 | 8.758 | 2.801 | 8.466 | −2.129 | 15.420 |
| Tangibility | 57,297 | 0.284 | 0.200 | 0.248 | 0 | 0.896 |
| Nature_Dep | 57,297 | 2.828 | 0.409 | 2.890 | 1.945 | 3.833 |
| Nature_High | 57,297 | 2.104 | 2.261 | 1.000 | 0 | 7.000 |
| Tobinsq | 57,297 | 1.717 | 3.093 | 0.959 | 0.167 | 26.920 |
| CFVolatility | 57,297 | 5,492 | 22,194 | 216.800 | 0 | 271,482 |
| Vulnerability | 57,297 | 0.371 | 0.0440 | 0.381 | 0.251 | 0.569 |
4.2 Baseline regression results
The baseline results reported in Table 3 indicate that firm-level nature dependence (Nature_Dep) is positively and statistically associated with financing costs, thus supporting Hypothesis H1. Specifically, nature dependence significantly increases the WACC at the 10% level, with the effect being more pronounced for the COE (at the 1% level of significance) while remaining statistically significant for the COD (at the 5% level). Our findings are consistent with prior literature, which suggests that nature-related risks are considerable for both equity investors and debt holders, although their significance may differ across these capital providers (Garel et al., 2024; Canipek et al., 2025; Gjerde et al., 2026). In particular, the effect is strongest for the COE, indicating that equity investors require a higher return to compensate for uncertainties associated with firms' reliance on ecosystem services and potential nature-related disruptions (Cepni et al., 2024). The COD is also positively affected, but at a lower level of significance, implying that creditors incorporate nature-related risks into their lending decisions, though to a lesser extent than equity investors, due to the fixed-income nature of the lending. Overall, these results highlight that both equity and debt markets are increasingly pricing nature-related risks, and firms with greater dependence on natural resources may face higher financing costs if they do not effectively manage or disclose such exposures. Our findings support recent studies that document that firms' dependence on ecosystem services can translate into substantial financial risks (Garel et al., 2026; Giglio et al., 2024). The heightened sensitivity of capital markets to environmental risk is closely linked to the observed financing effects of nature dependence. According to the findings of Bolton and Kacperczyk (2021), Pástor et al. (2021) and Cai (2025), investors increasingly recognize that nature-related risks can materially affect firms' cash flows, risk profiles and long-term performance. The investors' advanced capacity to assess environmental vulnerabilities strengthens market discipline, reinforcing the association between nature dependence and higher financing costs and ultimately shaping firm performance through higher capital costs and tighter financing conditions (Cai, 2025).
Baseline regression results: nature dependence and cost of capital
| Variables | (1) WACC | (2) COE | (3) COD |
|---|---|---|---|
| Nature_Dep | 0.001* | 0.003*** | 0.001** |
| (1.702) | (3.978) | (2.576) | |
| Size | <−0.001** | 0.002*** | <0.001 |
| (−2.192) | (10.235) | (1.518) | |
| ROA | −0.016*** | −0.024*** | −0.015*** |
| (−6.940) | (−8.613) | (−15.619) | |
| Lev | −0.038*** | 0.013*** | 0.024*** |
| (−29.548) | (8.183) | (43.760) | |
| Capex | 0.032*** | −0.002 | 0.003* |
| (6.414) | (−0.374) | (1.720) | |
| Salesgrowth | 0.003*** | 0.002*** | <0.001 |
| (7.679) | (4.186) | (1.368) | |
| Tangibility | −0.008*** | −0.008*** | <0.001 |
| (−6.479) | (−5.142) | (0.854) | |
| Tobinsq | 0.001*** | <0.001*** | −0.000*** |
| (5.682) | (3.119) | (−6.707) | |
| GDP | −0.047*** | −0.065*** | 0.004*** |
| (−17.860) | (−19.762) | (3.444) | |
| Constant | 0.555*** | 0.729*** | −0.018* |
| (20.828) | (21.672) | (−1.776) | |
| Observations | 57,144 | 57,144 | 57,144 |
| R-squared | 0.372 | 0.323 | 0.532 |
| Industry FE | Yes | Yes | Yes |
| Country FE | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes |
| Variables | (1) WACC | (2) COE | (3) COD |
|---|---|---|---|
| Nature_Dep | 0.001* | 0.003*** | 0.001** |
| (1.702) | (3.978) | (2.576) | |
| Size | <−0.001** | 0.002*** | <0.001 |
| (−2.192) | (10.235) | (1.518) | |
| ROA | −0.016*** | −0.024*** | −0.015*** |
| (−6.940) | (−8.613) | (−15.619) | |
| Lev | −0.038*** | 0.013*** | 0.024*** |
| (−29.548) | (8.183) | (43.760) | |
| Capex | 0.032*** | −0.002 | 0.003* |
| (6.414) | (−0.374) | (1.720) | |
| Salesgrowth | 0.003*** | 0.002*** | <0.001 |
| (7.679) | (4.186) | (1.368) | |
| Tangibility | −0.008*** | −0.008*** | <0.001 |
| (−6.479) | (−5.142) | (0.854) | |
| Tobinsq | 0.001*** | <0.001*** | −0.000*** |
| (5.682) | (3.119) | (−6.707) | |
| GDP | −0.047*** | −0.065*** | 0.004*** |
| (−17.860) | (−19.762) | (3.444) | |
| Constant | 0.555*** | 0.729*** | −0.018* |
| (20.828) | (21.672) | (−1.776) | |
| Observations | 57,144 | 57,144 | 57,144 |
| R-squared | 0.372 | 0.323 | 0.532 |
| Industry FE | Yes | Yes | Yes |
| Country FE | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes |
Note(s): This table presents the fixed-effects regression results of Equation (1). Dependent variables are Weighted average cost of capital (WACC), Cost of Equity (COE) and Cost of Debt (COD), reported in columns (1)–(3), correspondingly. Nature_dep captures firm-level exposure to nature dependence. All specifications include industry, country, and year fixed effects. Standard errors, reported in parentheses, are heteroskedasticity-robust and clustered at the firm level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively
Regarding the control variables, the estimates for LEV in Table 3 reveal heterogeneous effects across different components of the cost of capital. While higher LEV is significantly associated with a lower WACC, it is linked to significantly higher costs of equity and debt when these components are examined separately, with statistical significance at the 1% level. This finding is consistent with the notion that greater indebtedness heightens perceived risk of both creditors and equity holders, leading them to demand higher required returns. This finding is consistent with Modigliani and Miller (1958), indicating that higher LEV leads to a higher COE, and aligns with the results of Fandella et al. (2023). In contrast, firm profitability (ROA) and asset tangibility are negatively related to both equity and debt financing costs. These results are consistent with findings of Ge and Liu (2015) and Fandella et al. (2023), suggesting that markets reward firms with stronger operating performance and a higher proportion of fixed assets, as such characteristics are perceived to mitigate risk and consequently reduce the cost of external financing.
4.3 Alternative measures of nature dependence
Garel et al. (2026) developed two alternative aggregate measures of firms' nature dependence. In addition to the baseline indicator used in our main analysis, Nature_Dep, which reflects the revenue-weighted average dependence of a firm's activities on economically relevant ecosystem services, the authors introduce Nature_High to capture the concentration of nature-related risks. Specifically, Nature_High measures the number of ecosystem services on which a firm exhibits high dependence after revenue-weighting dependencies at the activity level. By construction, this measure avoids averaging high and low dependencies and is therefore better suited to capturing tail risks arising from strong reliance on a limited set of critical ecosystem services.
To assess the robustness of our findings to alternative measurements, we replaced Nature_Dep with Nature_High in the regression framework. The results, reported in Table A-2 (Online Appendices), are consistent with the baseline estimates in Table 3. Nature_High is positively and statistically significantly associated with all three measures of the cost of capital, including WACC, COE and COD, indicating that nature-related risks matter for both equity investors and debt holders, though their impact may vary across these groups (Garel et al., 2024; Canipek et al., 2025; Gjerde et al., 2026). In addition, we replace the firm-level Nature_Dep with its annual average value across all firms in the sample (Nature_Average) to capture the overall level of nature dependence in a given year. The results are presented in Table A-3 (Online Appendices). These consistent results across conceptually distinct measures of nature dependence strengthen our conclusion that firms' exposure to nature-related risks is systematically priced by capital markets and support Hypothesis H1.
4.4 The moderating effects of volatility factors
Table 4 reports the moderating effects of volatility-related factors on the relationship between nature dependence and firms' cost of capital (WACC). We account for both firm-level volatility, proxied by earnings volatility, and macro-level uncertainty, measured by country-level climate vulnerability.
Moderating effects of volatility factors on nature dependence and WACC nexus
| Variables | (1) WACC | (2) WACC |
|---|---|---|
| Nature_Dep | 0.001* | 0.009** |
| (1.657) | (2.015) | |
| Size | −0.001*** | −0.001*** |
| (−6.688) | (−7.251) | |
| ROA | 0.004** | 0.005** |
| (2.003) | (2.214) | |
| Capex | 0.031*** | 0.030*** |
| (5.983) | (5.894) | |
| Salesgrowth | 0.003*** | 0.003*** |
| (6.762) | (6.629) | |
| Tangibility | −0.018*** | −0.018*** |
| (−14.515) | (−14.451) | |
| Tobinsq | 0.001*** | 0.001*** |
| (4.744) | (4.787) | |
| GDP | −0.046*** | −0.047*** |
| (−17.454) | (−18.090) | |
| CFVolatility | <−0.001*** | |
| (−3.118) | ||
| Nature_Dep × CFVolatility | <0.001*** | |
| (2.959) | ||
| Vulnerability | 0.496*** | |
| (7.793) | ||
| Nature_Dep × Vulnerability | −0.022* | |
| (−1.708) | ||
| Constant | 0.550*** | 0.378*** |
| (20.369) | (10.804) | |
| Observations | 57,202 | 57,297 |
| R-squared | 0.345 | 0.346 |
| Industry FE | Yes | Yes |
| Country FE | Yes | Yes |
| Year FE | Yes | Yes |
| Variables | (1) WACC | (2) WACC |
|---|---|---|
| Nature_Dep | 0.001* | 0.009** |
| (1.657) | (2.015) | |
| Size | −0.001*** | −0.001*** |
| (−6.688) | (−7.251) | |
| ROA | 0.004** | 0.005** |
| (2.003) | (2.214) | |
| Capex | 0.031*** | 0.030*** |
| (5.983) | (5.894) | |
| Salesgrowth | 0.003*** | 0.003*** |
| (6.762) | (6.629) | |
| Tangibility | −0.018*** | −0.018*** |
| (−14.515) | (−14.451) | |
| Tobinsq | 0.001*** | 0.001*** |
| (4.744) | (4.787) | |
| GDP | −0.046*** | −0.047*** |
| (−17.454) | (−18.090) | |
| CFVolatility | <−0.001*** | |
| (−3.118) | ||
| Nature_Dep × CFVolatility | <0.001*** | |
| (2.959) | ||
| Vulnerability | 0.496*** | |
| (7.793) | ||
| Nature_Dep × Vulnerability | −0.022* | |
| (−1.708) | ||
| Constant | 0.550*** | 0.378*** |
| (20.369) | (10.804) | |
| Observations | 57,202 | 57,297 |
| R-squared | 0.345 | 0.346 |
| Industry FE | Yes | Yes |
| Country FE | Yes | Yes |
| Year FE | Yes | Yes |
Note(s): This table reports the fixed-effects regression results of Equations (2) and (3). The dependent variable is WACC. Columns (1) and (2) differ in terms of volatility measures, which are the firm-level earnings volatility (CFVolatility) and the country-level climate vulnerability (Vulnerability), respectively. Industry, country, and time fixed-effects are considered in all regressions. Heteroscedasticity-robust standard errors are clustered at the firm level in parentheses. *, **, *** indicate the significance levels at 10%, 5% and 1%, respectively
The results in Table 4 show that the interaction of nature dependence with firm-level cash flow volatility is positive at a 1% level of significance, indicating that nature dependence has a more pronounced effect on financing costs for firms with unstable earnings. This finding provides empirical support for Hypothesis H2 and suggests that when earnings volatility, a firm-level financial characteristic, is unstable, nature-related risks exacerbate uncertainty about future cash flows, prompting investors to demand higher risk compensation. Our result aligns with evidence that nature-related policy and market risks increase operating instability and earnings volatility (Lin and Zheng, 2026; Guo et al., 2025). This finding also supports studies demonstrating that investors price environmental and transition risks more strongly when these risks interact with firm-level uncertainty (Bolton and Kacperczyk, 2023).
At the country level, using the ND-GAIN climate vulnerability index, we document a negative moderating effect of climate vulnerability on the relationship between nature dependence and firms' cost of capital, consistent with Hypothesis H3. We uncover that in countries that are already highly exposed to climate risks, financing costs are generally elevated, leaving limited room for additional risk pricing attributable to firms' nature dependence. Also, firms in those countries have engaged in pro-environmental policies to mitigate the risk (Yildiz and Temiz, 2024). In contrast, in less climate-vulnerable countries, nature dependence is more salient to investors and is therefore associated with a higher incremental risk premium. Beyond its direct effects on the COD and equity, the key components of the cost of capital and high national exposure to climate risk may also increase firms' financial constraints. Our results align with this channel, indicating that climate vulnerability raises the COD both directly and indirectly by tightening firms' access to external finance (Kling et al., 2021).
4.5 The 2SLS model to address endogeneity issue
To mitigate potential endogeneity issues, we adopt an instrumental variables approach based on two-stage least squares (2SLS) estimation (Sargan, 1958; Murray, 2006), which is widely used in the literature to address endogeneity concerns (Jean et al., 2016). Specifically, we employ an instrument variable (Nature_Dep_iv) – the average nature dependence of industry peers in the same country and year, excluding the focal firm. Firms operating within the same country-industry context share similar natural resource conditions and regulatory environments; this ensures a strong correlation between the peer mean and the focal firm's nature dependence. Meanwhile, the peer average is unlikely to directly affect an individual firm's cost of capital through channels other than its own nature-dependent ones. Columns (1) and (2) of Table 5 report the first- and second-stage results, respectively, with the peer-mean instrument entering positively and significantly (Kleibergen-Paap F = 24.75, exceeding the Stock-Yogo 10% critical value of 16.38). In addition, column (3) reports the results based on propensity score matching (PSM) (Rosenbaum and Rubin, 1983). In this approach, firms with high levels of nature dependence are matched with firms exhibiting lower dependence but similar firm characteristics using nearest-neighbor matching. The matching procedure helps ensure that the treated and control firms are comparable in terms of observable characteristics before examining differences in financing costs. Overall, the results from Table 5 are consistent with our baseline model that nature dependence is positively and significantly associated with firms' cost of capital at the 1% level.
Endogeneity issue – 2SLS model and propensity score matching
| (1) First stage | (2) Second stage | (3) PSM | |
|---|---|---|---|
| Variables | Nature_Dep | WACC | WACC |
| Nature_Dep | 0.047*** | 0.002*** | |
| (3.598) | (2.671) | ||
| Nature_Dep_iv | 0.185*** | ||
| (4.975) | |||
| Size | 0.016*** | −0.001*** | −0.000 |
| (7.239) | (−4.000) | (−1.025) | |
| ROA | 0.011 | −0.016*** | −0.014*** |
| (0.402) | (−6.013) | (−4.491) | |
| Lev | −0.052** | −0.036*** | −0.039*** |
| (−2.463) | (−19.852) | (−19.816) | |
| Capex | 0.252*** | 0.020*** | 0.021*** |
| (3.444) | (2.965) | (2.621) | |
| Salesgrowth | −0.016*** | 0.004*** | 0.003*** |
| (−3.916) | (7.510) | (5.848) | |
| Tangibility | 0.259*** | −0.020*** | −0.007*** |
| (11.635) | (−5.360) | (−3.699) | |
| Tobinsq | 0.002 | 0.001*** | 0.001*** |
| (1.433) | (3.857) | (4.327) | |
| GDP | −0.046* | −0.047*** | −0.037*** |
| (−1.932) | (−15.894) | (−6.948) | |
| Constant | 2.536*** | 0.468*** | 0.454*** |
| (8.427) | (8.944) | (8.519) | |
| Observations | 56,123 | 56,123 | 45,120 |
| R-squared | 0.253 | 0.138 | 0.393 |
| Industry FE | Yes | Yes | Yes |
| Country FE | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes |
| First stage F-statistic of excluded instrument | 24.75*** | ||
| Anderson-Rubin 95% confidence interval | [0.0214, 0.0728] | ||
| Kleibergen-Paap rk Lagrange Multiplier (LM) statistics | 26.905*** |
| (1) First stage | (2) Second stage | (3) PSM | |
|---|---|---|---|
| Variables | Nature_Dep | WACC | WACC |
| Nature_Dep | 0.047*** | 0.002*** | |
| (3.598) | (2.671) | ||
| Nature_Dep_iv | 0.185*** | ||
| (4.975) | |||
| Size | 0.016*** | −0.001*** | −0.000 |
| (7.239) | (−4.000) | (−1.025) | |
| ROA | 0.011 | −0.016*** | −0.014*** |
| (0.402) | (−6.013) | (−4.491) | |
| Lev | −0.052** | −0.036*** | −0.039*** |
| (−2.463) | (−19.852) | (−19.816) | |
| Capex | 0.252*** | 0.020*** | 0.021*** |
| (3.444) | (2.965) | (2.621) | |
| Salesgrowth | −0.016*** | 0.004*** | 0.003*** |
| (−3.916) | (7.510) | (5.848) | |
| Tangibility | 0.259*** | −0.020*** | −0.007*** |
| (11.635) | (−5.360) | (−3.699) | |
| Tobinsq | 0.002 | 0.001*** | 0.001*** |
| (1.433) | (3.857) | (4.327) | |
| GDP | −0.046* | −0.047*** | −0.037*** |
| (−1.932) | (−15.894) | (−6.948) | |
| Constant | 2.536*** | 0.468*** | 0.454*** |
| (8.427) | (8.944) | (8.519) | |
| Observations | 56,123 | 56,123 | 45,120 |
| R-squared | 0.253 | 0.138 | 0.393 |
| Industry FE | Yes | Yes | Yes |
| Country FE | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes |
| First stage F-statistic of excluded instrument | 24.75*** | ||
| Anderson-Rubin 95% confidence interval | [0.0214, 0.0728] | ||
| Kleibergen-Paap rk Lagrange Multiplier (LM) statistics | 26.905*** |
Note(s): This table reports the fixed-effects regression results addressing potential endogeneity concerns using a two-stage least squares (2SLS) approach and propensity score matching (PSM). In columns (1) and (2), the dependent variable in the first stage is Nature_Dep, while WACC is used in the second-stage regression. The instrument variable (Nature_Dep_iv) is the average nature dependence of industry peers in the same country and year, excluding the focal firm. Column (3) reports the results based on a matched sample obtained from the PSM approach. Industry, country, and time fixed-effects are considered in all regressions. Heteroscedasticity-robust standard errors are clustered at the firm level in parentheses. *, **, *** indicate the significance levels at 10%, 5% and 1%, respectively
5. Conclusions and implications
5.1 Summary of findings
This study investigates the relationship between firms' nature dependence level and cost of capital. Using a sample of 57,297 firm-year observations from 87 countries over the period from 2015 to 2023, we demonstrate that firms more reliant on the natural environment face significantly higher financing costs. This effect is particularly prominent in equity markets, suggesting that investors increasingly view ecosystem degradation and biodiversity loss as systemic threats to long-term valuation and support recent findings about nature dependence risk (Garel et al., 2024; Canipek et al., 2025; Gjerde et al., 2026). Our findings highlight that nature dependence is no longer an external and irrelevant factor but an obvious financial burden. Firms with high nature dependency, particularly those in less climate-vulnerable countries, must prioritize the integration of nature-related strategies into their core business models to mitigate the COE and debt. Stabilizing earnings through better resource management may also help dampen the penalizing effects of nature risk on financing.
Our findings remain robust after accounting for alternative measurements of nature dependence and addressing potential endogeneity through a 2SLS instrumental variables approach. Furthermore, our analysis reveals that this “nature-related risk premium” is not uniform across uncertain environments. Specifically, the impact of nature dependence on financing costs is amplified for firms with high earnings volatility, indicating that nature-related risks act as a risk multiplier for already unstable firms (Lin and Zheng, 2026; Guo et al., 2025). For macro-level climate vulnerability, the pricing of nature dependence is dampened in countries with high climate exposure, likely because the broader climate risk is already priced in, whereas in less vulnerable regions, nature dependence serves as a more distinct and salient risk signal for investors (Kling et al., 2021).
5.2 Implications
The study highlights important implications for firms, investors and policymakers. Since firms' costs of capital increase with higher nature dependence, particularly under conditions of elevated earnings volatility, firms should take proactive steps to manage nature-related risks and reduce excessive reliance on natural resources where feasible. In practice, firms can strengthen risk management by systematically identifying and mapping their dependencies and impacts on ecosystems across operations and supply chains, given their exposure to nature dependence (Hoang and Vu, 2026). Integrating these assessments into enterprise risk management and financial planning can help firms anticipate operational disruptions linked to ecosystem degradation and biodiversity loss. Firms may also reduce exposure by diversifying suppliers and sourcing regions, investing in resource-efficient production technologies, and substituting scarce natural inputs with more sustainable alternatives where possible. In addition, strengthening ecosystem stewardship initiatives, such as watershed protection, sustainable land use and biodiversity restoration, can help safeguard the natural assets on which firms depend (Dasgupta, 2021). Implementing robust monitoring systems and aligning disclosures with emerging frameworks such as the taskforce on nature-related financial disclosures can further enhance transparency and signal stronger environmental risk governance to capital markets. These measures can help stabilize earnings, reduce perceived environmental risk and ultimately mitigate higher capital costs, particularly among equity investors who appear most sensitive to nature dependence.
The findings also suggest that investors should incorporate firm-level nature exposure into valuation models, portfolio construction and risk assessment frameworks. Investors should treat nature dependence as a risk factor that interacts with both firm-level characteristics and macro-environmental uncertainty. Integrating indicators of ecosystem dependency, supply-chain exposure to biodiversity loss and firms' mitigation strategies can improve risk-adjusted return assessments and support more informed stewardship and engagement strategies. In countries with lower climate vulnerability, nature dependence appears to provide a clearer signal of firm-specific risk and may therefore warrant greater weight in investment analysis.
Lastly, policymakers and regulators should consider strengthening nature-related policies not only to support ecological sustainability but also to improve capital allocation by enabling markets to more accurately price firms' exposure to ecosystem degradation. Policymakers should recognize that nature-related financial risks are not priced uniformly across firms and countries. In less climate-vulnerable economies, where nature dependence appears to be more strongly reflected in capital costs, clear reporting standards and consistent disclosure requirements can improve market transparency and pricing efficiency. In more climate-exposed countries, where broader climate risks may already dominate market perceptions, regulators should ensure that firm-level nature dependencies remain visible and measurable in corporate reporting. Strengthening disclosure frameworks, supporting standardized metrics for nature-related risk and encouraging corporate transition strategies that reduce unsustainable resource dependence can help financial markets better identify and manage nature-related financial risks.
Note
IMF (2019), “The Greatest Balancing Act”, retrieved from: https://www.imf.org/en/publications/fandd/issues/2019/12/nature-climate-and-the-global-economy-lagarde-attenborough.
The supplementary material for this article can be found online.

