Climate risk and firm-level cost structures: empirical analysis of Saudi listed companies

Climate change has evolved from being a purely environmental issue into a multidimensional challenge with profound implications for food security, public health and the stability of economic and financial systems. Increasingly frequent floods, heatwaves, droughts and extreme weather events are disrupting supply chains, damaging infrastructure and raising operational and financial risks for firms. In Saudi Arabia, these threats are particularly acute. The Kingdom’s already extreme desert climate is becoming more volatile, with rising average temperatures, recurrent heatwaves and sporadic heavy rainfall creating major challenges for sectors such as energy, construction, transportation and agriculture. Such climatic variability translates into higher operating expenses, increased financing costs and ultimately greater financial vulnerability for listed companies (Giang et al., 2021; Sun et al., 2020; Hamdouni and Smaoui, 2025).

Empirical evidence shows that climate risk negatively affects corporate performance across economies, reducing profitability, efficiency, sales, asset values and increasing financing costs (Huang et al., 2018; Giang et al., 2021; Sun et al., 2020; Zhang et al., 2023; Mawejje, 2024; Wu, 2025; Kling et al., 2021; Çepni et al., 2024). Recent studies also highlight rising operating costs due to environmental deterioration (Saieed, 2022; Srivastav et al., 2025; Ponticelli et al., 2023; Addoum et al., 2021; Xue et al., 2025; Zhao and Zhang, 2025; Liang et al., 2025). In Saudi Arabia, Hamdouni and Smaoui (2025) confirm significant climate effects on firm profitability and valuation, underscoring the importance of integrating climate risk into corporate strategy.

Despite the growing body of literature, significant research gaps remain. Existing studies have predominantly focused on profitability indicators such as return on assets (ROA) and return on equity (ROE), while the implications of climate variability for broader cost structures, including operating and financing costs, remain underexplored. This gap is particularly evident in arid economies, such as the Middle East and especially Saudi Arabia, where climate exposure is high and economic diversification is a central pillar of Vision 2030.

Cost-based indicators may provide a more direct measure of climate-risk transmission because climate shocks can first materialize through higher operating expenses, financing costs and total costs (TC) before their effects are fully reflected in aggregate profitability measures such as ROA and ROE. This distinction is particularly relevant in climate-vulnerable economies, where recurrent environmental shocks can directly increase firms’ resource, adaptation, insurance and financing burdens.

To address this limitation, this study adopts a cost-structure perspective that explicitly integrates operating cost dynamics and financing-cost transmission channels. Rather than focusing solely on profitability outcomes, we conceptualize climate risk as a multidimensional determinant of operating expenses, cost of capital (CC) and TC. This integrated framework strengthens the alignment between the research objective and the underlying economic mechanisms through which climate variability affects firm-level financial stability.

Against this background, the present study contributes to the understanding of how climate variability affects firm-level cost structures in Saudi Arabia. Specifically, the study uses a balanced panel data set comprising 80 publicly listed firms from seven climate-sensitive sectors – utilities, energy, materials, construction, transportation, agriculture and real estate – over the period 2010–2024. Using dynamic panel techniques (System GMM), this study examines the persistent effects of climate risk on firm-level costs, offering insights for Saudi firms seeking resilience under Vision 2030.

This paper proceeds as follows: Section 2 covers the theoretical background; Section 3 reviews the literature and develops hypotheses; Section 4 describes the data and methodology; Section 5 presents and discusses the empirical findings; and Section 6 concludes with implications, limitations and future research directions.

Climate change refers to long-term shifts in temperature, precipitation, humidity and wind speed compared to historical patterns, largely driven by greenhouse gas emissions from human activities. The Intergovernmental Panel on Climate Change (IPCC, 2021) stresses that the burning of fossil fuels such as coal, oil and natural gas releases high concentrations of carbon dioxide into the atmosphere, intensifying the greenhouse effect and leading to global warming. In addition, deforestation and agricultural activities emit large volumes of methane and nitrous oxide, further disrupting the natural carbon cycle and increasing climate instability (Giang et al., 2021; Guermazi et al., 2025).

These anthropogenic drivers of climate change have accelerated the occurrence of extreme weather events such as floods, heatwaves and droughts, all of which disrupt ecosystems and impose substantial economic costs (Huang et al., 2018). Empirical studies confirm that climate change does not only threatens the natural environment but also imposes financial and operational burdens on firms through higher energy consumption, lower productivity and increased maintenance costs (Srivastav et al., 2025; Zhao and Zhang, 2025).

Saudi Arabia is among the countries most exposed to climate variability, with rising average and extreme temperatures, shifting rainfall patterns and frequent extreme events. Evidence shows that climate variability in the Kingdom has already imposed significant risks on corporate financial outcomes. Hamdouni and Smaoui (2025) found that precipitation and specific humidity negatively affect ROA and ROE, while wind speed strongly reduces both profitability and market valuation. This confirms that climate risks are not only environmental challenges but also financial threats in the Saudi market.

Beyond profitability, climate risks are closely tied to firm-level costs. Studies from the Middle East and North Africa (MENA) region (Saieed, 2022) demonstrate that reduced precipitation, soil moisture deficits and severe droughts directly raise operating expenses, especially in resource-dependent sectors such as agriculture, construction and transportation. These findings are echoed by global evidence, where extreme heat waves and rising temperatures have been shown to elevate energy expenditures, reduce labor productivity and accelerate equipment depreciation, all contributing to higher firm-level costs (Addoum et al., 2021; Xue et al., 2025).

In the Saudi context, these risks translate into higher operational costs for listed firms, greater volatility in financial results and increased vulnerability of corporate strategies to external shocks. As highlighted in the national Vision 2030 framework, integrating climate risk into corporate and financial decision-making is no longer optional but a necessity to ensure resilience and long-term sustainability.

Climate change has become a multidimensional threat to economic and financial stability, as extreme weather events disrupt infrastructure and supply chains, causing financial losses. Accordingly, research increasingly explores the impact of physical and transitional climate risks on corporate performance.

Evidence from multiple countries generally supports a negative association between climate risk and financial outcomes. Huang et al. (2018), using the Global Climate Risk Index across 55 countries, found that higher climate risk correlates with lower and more volatile earnings and cash flows. Similarly, Giang et al. (2021) reported that among Vietnamese manufacturing firms, humidity risk significantly reduced ROA, ROE and return on sales (ROS), even when temperature, rainfall and sunshine had no significant effects. Sun et al. (2020) reached similar conclusions for Chinese mining firms, showing that both physical and transitional climate risks diminished ROA and ROE. In a broader cross-country setting, Zhang et al. (2023) observed that climate risk had a negative, though statistically insignificant, impact on ROA and cash flow from operations (CFO), but led firms to adopt more conservative financing policies by increasing long-term debt.

The adverse impact of climate variability is also evident in studies on developing economies. Mawejje (2024) found that in Uganda, weather and temperature shocks significantly harmed sales, profits and capacity utilization, with micro and small enterprises, especially in agriculture and industry, most vulnerable. Wu (2025) showed that in China, extreme temperatures above 30°C or below −10°C reduced asset values, particularly in capital- and labor-intensive sectors. Griffin et al. (2025) highlighted a hump-shaped relationship in the European Union (EU) and United Kingdom (UK): moderate warming benefited firms in cooler climates, but performance declined sharply beyond ∼23°C, without notable improvements in environmental, social, and governance (ESG) practices or emissions reductions.

Ginglinger and Moreau (2023) show that greater exposure to physical climate risk reduces firms’ leverage, reflecting precautionary debt reduction and higher lending costs. The effect strengthens after 2015, underscoring the rising impact of climate awareness and regulation on financing decisions.

Saieed (2022) finds that climate factors significantly affect firms’ operating costs in the MENA region. Reduced precipitation and soil moisture, higher drought severity and increased vapor pressure raise operating expenses, particularly in resource-dependent sectors such as agriculture, construction and transportation, while technology-oriented sectors show weaker effects. Wind speed is negatively related to operating costs.

These studies converge on the finding that climate variability and extreme weather conditions exert adverse effects on firm performance across different economic contexts. However, they also reveal substantial heterogeneity in the magnitude and transmission channels of these effects. While evidence from developing and emerging economies emphasizes losses in sales, asset values and operational efficiency, studies in advanced economies point to nonlinear and threshold-based responses to temperature changes. Moreover, although recent contributions have begun to examine financing and operating cost channels, the existing literature remains fragmented, with limited integration of cost-based mechanisms, particularly in climate-vulnerable regions.

Recent empirical literature highlights that climate change directly increases operating costs for firms through multiple channels. Srivastav et al. (2025) demonstrate that rising climate risks alter firms’ energy consumption behavior, thereby raising energy-related operating expenses. Similarly, Ponticelli et al. (2023) show that higher temperatures increase energy demand for cooling and reduce the efficiency of industrial equipment, which elevates production costs and affects local industry concentration. Addoum et al. (2021) find that extreme heat waves depress earnings in climate-sensitive sectors by driving up operating costs in areas such as energy and logistics. Supporting this view, Xue et al. (2025) confirm that the operational impacts of climate change manifest primarily through three channels: higher energy use, reduced human capital productivity and accelerated equipment depreciation, all of which contribute to rising costs. Along the same lines, Zhao and Zhang (2025) show that extreme weather events generate financial losses by increasing repair, maintenance and energy expenditures. Liang et al. (2025) provide further evidence by demonstrating that heatwaves in China trigger recurrent power outages, forcing firms to incur additional costs to maintain operations and mitigate production losses.

Hamdouni and Smaoui (2025) show that climate risks significantly affect the financial performance of Saudi listed firms (2010–2022). Precipitation and specific humidity reduce ROA and ROE, while relative humidity and wet-bulb temperature have positive effects. Wind speed strongly lowers profitability and market value. The study highlights the need to integrate climate risk management into corporate strategies in line with Vision 2030. Despite these important insights, empirical evidence on the direct relationship between climate variability and firms’ cost structures in Saudi Arabia remains scarce. Existing studies largely emphasize profitability and market valuation, while the cost-based transmission channels, such as operating and financing costs, have received limited attention. This lack of evidence is particularly striking given Saudi Arabia’s arid climate conditions and high exposure to climate-related risks, highlighting the need for focused empirical analysis on climate–cost mechanisms in this context.

In a related firm-level study, Hamdouni (2026) shows that multidimensional climate risks significantly shape corporate green innovation among Saudi firms, underscoring that climate exposure influences not only profitability but also firms’ strategic and operational responses.

Kling et al. (2025) show that climate vulnerability significantly worsens firms’ financing conditions by increasing the cost of debt and restricting access to credit. Using a large cross-country panel, they find that firms in more climate-exposed environments face systematically higher borrowing costs, while the effect on the cost of equity remains weak. Similarly, Çepni et al. (2024) document that climate change exposure significantly increases firms’ cost of equity, as investors demand higher risk premiums, particularly for firms facing stronger financing constraints or operating in highly climate-exposed regions.

Recent empirical research increasingly conceptualizes climate risk as a cost-transmission mechanism that affects firms not only through profitability outcomes but also through operating expenses and financing costs. At the operational level, Addoum et al. (2020) show that temperature shocks significantly reduce establishment-level sales and productivity, particularly in climate-sensitive industries. Their findings indicate that extreme temperatures impair operational efficiency and disrupt production processes. Such productivity losses imply higher per-unit production costs, increased adjustment expenses and reduced operational efficiency, thereby linking climate variability directly to firms’ operating cost structures rather than solely to revenue fluctuations.

From a financing perspective, climate exposure generates a measurable risk premium in capital markets. Chava (2014) demonstrates that firms with higher environmental risk face significantly higher costs of debt and equity, reflecting investor and creditor re-pricing of environmental externalities. Extending this argument to climate vulnerability, Kling et al. (2021) find that firms operating in more climate-exposed environments incur higher borrowing costs and experience restricted access to credit. More recently, Cang and Li (2024) provide direct bond-market evidence showing that firms with greater exposure to corporate climate risk face significantly wider bond credit spreads, confirming that climate-related uncertainty is incorporated into debt pricing. This evidence establishes a clear financing-cost transmission channel through which climate risk increases firms’ cost of debt and ultimately their overall CC.

In conclusion, the existing literature demonstrates that the impact of climate change on firms has been examined across several interrelated dimensions. A substantial body of research has concentrated on financial performance, consistently showing that climate risks, such as temperature fluctuations, humidity and extreme weather events, tend to reduce profitability and operational efficiency. Another strand of literature focuses on financing conditions, showing that climate vulnerability raises the cost of debt and restricts credit access (Kling et al., 2021) and more recently raises the cost of equity as investors demand higher returns (Çepni et al., 2024). A smaller but growing body of research examines operating cost dynamics, showing that environmental deterioration raises expenditures on energy, maintenance and productivity losses (Saieed, 2022; Srivastav et al., 2025; Zhao and Zhang, 2025). Despite this expanding body of evidence, much of the empirical literature remains centered on profitability-based indicators, particularly return measures such as ROA and ROE, while a comprehensive analysis of how climate risk reshapes firms’ overall cost structures remains limited. Specifically, limited attention has been given to the joint examination of operating expenses, CC and TC within an integrated empirical framework. This gap matters most for climate-vulnerable emerging economies, where environmental shocks can simultaneously affect operational efficiency and financing conditions. The present study addresses this gap by examining the impact of climate variability on operating costs, CC and TC for listed firms in Saudi Arabia – a country highly exposed to climate variability and undergoing significant economic transformation under Vision 2030. By moving beyond profitability metrics and focusing explicitly on cost-structure dynamics, the study provides new empirical evidence on the transmission channels through which climate risks influence firm-level financial stability. In doing so, it contributes to the literature by integrating operating and financing cost perspectives within a unified framework and offers policy-relevant insights for corporate decision-makers and regulators seeking to enhance climate resilience in line with the Kingdom’s sustainability objectives.

Prior studies have highlighted that climate risk can increase business costs in various ways, although its effects differ across sectors. Therefore, we propose the following hypotheses to examine the influence of climate factors on firm-level costs.

Soil Moisture (GW):

Precipitation (PRE):

Specific humidity (QV2M):

Relative humidity (RH2M):

Wet-bulb temperature (T2MWET):

Wind speed (WS2M):

Leverage (Lev):

Firm size (Size):

Growth rate (G):

This study uses a balanced panel of 80 Saudi listed firms from seven climate-sensitive sectors – utilities, energy, materials, construction, transportation, agriculture and real estate – over 2010–2024. These industries are structurally exposed to climate variability due to dependence on natural resources, energy intensity and infrastructure-based operations.

Sector selection follows an explicit climate-exposure classification framework used in the climate-finance literature (Addoum et al., 2021; Kling et al., 2021; Ponticelli et al., 2023). Sectors are considered climate-sensitive if they meet at least one of three criteria: (1) reliance on weather-driven inputs, (2) temperature and energy sensitivity or (3) infrastructure and logistics exposure to extreme conditions.

To ensure representativeness, the selected sectors account for approximately 70% of nonfinancial listed market capitalization and about 65%–75% of Saudi non-oil gross domestic product (GDP), indicating that the sample reflects the core climate-exposed structure of the Saudi economy.

Firms are included only if continuous financial and climate data are available throughout the period to maintain a balanced panel suitable for dynamic GMM estimation. Financial data are obtained from Tadawul, company reports and Bloomberg. Climate variables are sourced from NASA MERRA-2 and include soil moisture (GW), precipitation (PRE), specific humidity (QV2M), relative humidity (RH2M), wet-bulb temperature (T2MWET) and wind speed (WS2M), matched to firms by geographic location.

The merged data set provides a representative climate-relevant panel for analyzing long-term cost effects of climate variability. Variable definitions are reported in Table 1.

To ground the empirical specification economically, the study adopts a cost-transmission framework in which climate variables affect firm costs through two channels: operating costs and financing costs.

Physical climate conditions (temperature, humidity, soil moisture and wind) alter production environments. Heat and humidity increase cooling demand, electricity use, productivity losses and equipment depreciation, while wind and moisture raise maintenance and logistics expenses. These effects increase operating expenses and TC. In arid economies such as Saudi Arabia, precipitation relaxes water constraints and reduces operating costs. Climate variables therefore represent physical productivity shocks affecting firms’ cost functions.

Climate exposure also influences financial risk. Persistent climate stress increases cash-flow volatility and perceived default risk, leading investors and lenders to require higher risk premia. Consequently, climate conditions affect the CC by acting as systematic risk signals even without immediate physical damage.

This study uses a dynamic panel specification estimated with the two-step System GMM estimator to analyze the impact of climate variables on firm costs. The approach controls for endogeneity, autocorrelation and unobserved heterogeneity and allows estimation of short- and persistent climate effects on operating expenses (OE), CC and TC.

The lagged dependent variable captures partial adjustment, as firm costs cannot respond instantaneously to shocks due to contractual and financing rigidities. This specification distinguishes short-run responses from persistent cost effects.

Lagged dependent variables are treated as endogenous, firm controls (e.g. leverage and growth) as predetermined and climate variables as exogenous since firms cannot influence regional weather conditions. This classification addresses simultaneity and reverse causality. This assumption is supported both conceptually and empirically: individual firms are price-takers with respect to climate and cannot plausibly alter regional conditions, and the climate variables are drawn from MERRA-2 data at a coarse spatial grid resolution well above the firm level, leaving no plausible channel for firm-level feedback.

Instruments use lags t − 2 to t − 4 and are collapsed following (Roodman, 2009), resulting in 24–26 instruments, below the number of firms (n = 80). This restriction prevents instrument proliferation and supports Hansen test reliability (see Table 8).

Within this framework, endogenous variables are instrumented using their own lagged levels and differences, while selected controls are treated as predetermined.

The general dynamic panel model is specified as follows:

where:

• $yi,t$⁠: Dependent variable;

• $yi,t-1$⁠: Lagged dependent variable;

• $Xi,t$⁠: Vector of climate variables;

• $Zi,t$⁠: Control variables;

• $ηi$⁠: Unobserved firm effects; and

• $εi,t$⁠: Idiosyncratic error term.

This study applies the Arellano-Bover/Blundell-Bond two-step System GMM estimator to examine the impact of climate variables on firm-level costs. The dynamic specification, with a lagged dependent variable, controls for endogeneity, autocorrelation and unobserved heterogeneity.

To avoid instrument proliferation, instruments are collapsed and restricted, resulting in 24–26 instruments across models, well below the number of firms (n = 80). Time-fixed effects were tested via joint significance tests of year dummies and found statistically significant. They are therefore included in all specifications to control for macroeconomic conditions, regulatory changes and common climate- or policy-related shocks. In the Saudi context, these period effects capture economy-wide cost pressures such as post-2016 energy price reforms, the COVID-19 economic disruption (2020–2021) and sustainability and climate policy initiatives under Vision 2030, which simultaneously influenced energy expenses, financing conditions and operational efficiency across sectors. This methodology ensures consistent and efficient estimation of both short-and long-term effects while satisfying key diagnostic criteria such as the Hansen J-test for instrument validity and the Arellano–Bond test for autocorrelation.

To ensure the robustness of the model estimates, Arellano–Bond AR(1) and AR(2) tests are conducted to assess serial correlation in the first-differenced residuals, while Hansen and Sargan tests are used to evaluate the validity of the overidentifying restrictions. The absence of second-order serial correlation and the failure to reject the null hypothesis of the Hansen test confirm the appropriateness of the instrument set and provide further evidence that the results are not affected by instrument proliferation or overfitting.

Table 2 reports summary statistics. Operating expenses (OE) average 0.577 (median 0.407), indicating high-cost intensity. The CC averages 7.38%, while TC vary widely across firms. Climate exposure differs across variables: soil moisture is stable; precipitation is highly variable and humidity and temperature indicators show moderate dispersion. Control variables also vary, with leverage slightly above one, wide firm-size dispersion and heterogeneous growth rates.

Table 3 reports Pearson correlations among cost measures, climate variables and firm characteristics. Operating expenses (OE) are strongly correlated with TC (r = 0.62). Climate variables show weak direct correlations with OE but strong intercorrelations, notably between specific humidity (QV2M) and relative humidity (RH2M) (r = 0.91), indicating multicollinearity. Therefore, principal component analysis (PCA) is used as a robustness check to summarize shared climatic variation and stabilize the estimates.

Table 4 reports Levin–Lin–Chu (LLC) and Im–Pesaran–Shin tests showing all variables are stationary at levels (I(0)), rejecting the unit-root null. This supports dynamic panel estimation without differencing.

Panel cross-sectional dependence tests assess whether firms are affected by common shocks (Basak and Das, 2018; De Hoyos and Sarafidis, 2006). Table 5 shows all statistics are significant (p < 0.05), confirming cross-sectional dependence across the 80 firms (2010–2024) and justifying its treatment in the econometric specification for reliable inference.

Table 6 reports the Pesaran and Yamagata (2008) slope homogeneity test for OE, CC and TC. All Delta statistics are significant (p < 0.05), rejecting homogeneity and indicating cross-firm heterogeneity. This supports using System GMM, which accounts for unobserved firm effects and heterogeneous relationships.

Table 7 reports VIF diagnostics for the OE, CC and TC models. All values are below 10, indicating no serious multicollinearity. Skewness and kurtosis show mild nonnormality typical of economic panel data, but this does not affect reliability, especially as System GMM is robust to such deviations.

Table 8 reports the GMM estimates for operating expenses (OE), CC and TC. The lagged dependent variables are significant across all models, confirming strong persistence in firm cost structures. Taken together, the persistence and significance patterns indicate that climate variables operate through a transmission mechanism rather than isolated shocks: physical climate conditions first alter operational efficiency and subsequently influence financing conditions.

The GMM results indicate that operating expenses exhibits strong persistence, as reflected by the highly significant coefficient on the lagged dependent variable (0.45***). This persistence is consistent with cost-adjustment theory, which suggests that firms face adjustment frictions that prevent rapid changes in operational cost structures over time. Climate-related variables – including profile soil moisture (GW), specific humidity (QV2M), wet bulb temperature (T2MWET) and wind speed (WS2M) – are statistically significant and positively associated with operating costs, indicating that climate-induced stress systematically increases firms’ operational expenditures through higher energy use, maintenance requirements and resource management costs rather than transitory shocks. In contrast, precipitation is negatively related to operating costs, suggesting that improved water availability may reduce irrigation, cooling and input-related expenses. Relative humidity (RH2M) does not exhibit a statistically significant effect. Among firm-level controls, leverage increases operating costs, while firm size and revenue growth are associated with lower cost ratios, consistent with economies of scale and learning effects. This positive leverage effect in the baseline model reflects higher financing costs, debt servicing obligations and financial fragility under climate stress, particularly when individual climate variables are entered separately.

The magnitude and consistency of these effects suggest structural exposure rather than firm-specific sensitivity. Industries characterized by continuous production, outdoor activity or temperature-dependent infrastructure are therefore more vulnerable to climate shocks, indicating heterogeneous cost transmission across climate-sensitive sectors.

Economically, a one-unit increase in wet-bulb temperature is associated with roughly a 6% rise in operating cost ratios, showing that even moderate heat stress generates meaningful cost pressures for climate-sensitive firms.

Financing costs show strong persistence over time. Climate variables such as soil moisture, humidity, wet-bulb temperature and wind speed increase the CC, consistent with higher risk premia under climate exposure, while precipitation reduces financing costs. These results indicate that capital markets increasingly price climate risk rather than treating it as a purely operational concern.

This finding links the physical and financial dimensions of climate risk: operational stress increases expected cash-flow volatility, which is then incorporated into risk premia by lenders and investors, confirming the sequential transmission from physical exposure to financial risk pricing.

Total cost estimates show strong persistence, indicating that aggregate expenditures adjust slowly over time. Climate variables have positive and significant effects on TC, suggesting that climate variability operates through multiple channels, including production inefficiencies, infrastructure damage and higher energy and maintenance expenses. Precipitation remains negatively associated with TC, reflecting water-related savings. Leverage increases TC, whereas firm size and growth reduce cost ratios.

The simultaneous significance across operating, financing and total cost models indicates that climate exposure acts as a systemic cost driver rather than a single-channel disturbance, affecting production efficiency, capital pricing and expenditure planning jointly.

The consistent insignificance of relative humidity across all cost measures is theoretically intuitive. Relative humidity primarily affects perceived thermal comfort rather than directly influencing production efficiency, capital intensity or infrastructure performance once absolute humidity and heat stress (captured by QV2M and T2MWET) are controlled for. Moreover, the strong correlation between relative humidity and specific humidity suggests that moisture-related economic effects are already captured by absolute humidity measures, rendering RH2M redundant in the presence of more physically relevant climate indicators.

Overall, the results demonstrate a coherent climate-driven cost transmission process: environmental conditions first alter operational efficiency, then affect financing risk and ultimately accumulate into total expenditure adjustments. Therefore, the empirical evidence supports the study’s conceptual framework in which climate exposure represents a structural economic factor shaping firm cost structures in climate-sensitive industries rather than merely a set of statistically significant correlations.

Overall, the results indicate that climate variables influence firm costs through operational adjustment and financial risk channels across operating, financing and total cost measures.

The reliability of the System GMM estimates is supported by a restricted instrument strategy in which lagged dependent variables are instrumented using their own lags and the instrument matrix is collapsed to avoid proliferation (24–26 instruments, below n = 80). Hansen and AR(2) tests confirm instrument validity and correct specification, indicating that the estimates are not driven by weak instruments or overfitting.

To ensure the robustness of our empirical findings, we conduct a supplementary analysis using PCA. PCA is a widely used dimensionality reduction technique that transforms a set of correlated explanatory variables into a smaller number of uncorrelated principal components, capturing the majority of the variance in the original data (Jolliffe and Cadima, 2016). By applying PCA, we mitigate concerns regarding multicollinearity and potential measurement noise among the independent variables, thereby improving the stability and interpretability of our regression estimates. Following standard retention criteria, components with eigenvalues greater than one and cumulative explained variance exceeding 60% are considered economically informative and statistically sufficient to summarize climate conditions. The extracted principal components are subsequently used in our regression framework to verify whether the core relationships observed in the main analysis hold when the underlying data structure is simplified. Consistency of the results across both the original and PCA-transformed models strengthens the validity of our conclusions and demonstrates the robustness of our empirical strategy.

Applying this criterion, three principal components were retained, jointly explaining approximately 96% of the total variance in the six climate variables (71% for PC1, 19% for PC2 and 6% for PC3). These three components then replace the original climate variables as regressors in the GMM model, as shown in the Revised GMM Model specification below.

PCA on the six climate variables shows that three components summarize most climate variation. PC1 explains about 71% of total variance, PC2 about 19% and PC3 about 6%, together capturing over 96% of total variation. Economically, PC1 represents an aggregate climate stress index affecting operational efficiency and production conditions. PC2 captures atmospheric instability driven by wind variability, influencing operational risk and logistics costs. PC3 reflects interaction effects between soil moisture and air conditions that affect maintenance and infrastructure expenditures. This dimensionality reduction mitigates multicollinearity and improves model interpretability (Figure 1, Tables 9 and 10).

The Revised GMM Model (Using PCA) is:

where:

• $yi,t$⁠: Dependent variable;

• $yi,t-1$⁠: Lagged dependent variable;

• $PC1i,t$⁠, $PC2i,t$⁠, $PC3i,t$⁠: The first three principal components summarizing the variance in the six climate variables;

• $Zi,t$⁠: Control variables;

• $ηi$⁠: Unobserved firm effects; and

• $εi,t$⁠: Idiosyncratic error term.

In the regression framework, the principal components are interpreted as synthetic climate risk channels rather than isolated meteorological variables, allowing the estimated coefficients to reflect the impact of overall climatic pressure on firm cost behavior.

Table 11 reports the GMM estimations using the three principal components derived from climate variables. The lagged dependent variables are significant across all models, confirming strong persistence in operating costs, CC and TC. The principal components are all statistically significant across the models, highlighting the substantial role of climate factors in influencing firm costs. PC1, the humidity–temperature–moisture factor, is positively associated with OE and TC but negatively affects CC, indicating that combined hydro-thermal conditions exert mixed financial pressures across operating, financing and total-cost channels. PC2, dominated by wind speed (WS2M), shows a strong positive effect on CC while negatively impacting TC, indicating that wind-related climatic variability affects financing costs and expenditures differently. PC3, capturing interaction effects between soil moisture and wind dynamics, consistently exhibits a positive relationship across all dependent variables, implying that secondary climate interaction mechanisms uniformly elevate firm costs.

Control variables behave consistently: larger firms exhibit lower operating and TC, while leverage and growth reflect firm-specific financial adjustments. Diagnostic tests (AR(2), Hansen) confirm valid instruments and correct model specification. Overall, the PCA specification corroborates that aggregated climate exposure significantly affects firm costs through dynamic operational and financial channels.

Baseline GMM results (Table 8) show that soil moisture, precipitation, specific humidity, wet-bulb temperature and wind speed significantly affect operating efficiency, CC and TC, while relative humidity is weak. Physical climate exposure first affects operating efficiency through energy demand, cooling requirements and infrastructure stress, then increases perceived firm risk and financing costs, confirming the cost-transmission framework in which operating costs react immediately and capital costs adjust gradually.

Using PCA (Table 11), the six variables are summarized into three orthogonal components that remain significant across OE, CC and TC, indicating that firm costs are driven by joint climatic conditions rather than isolated indicators. The stronger persistence of lagged costs after PCA reflects clearer dynamic adjustment once multicollinearity is reduced, and diagnostic tests confirm model validity.

Hypothesis tests align with prior research. Humidity and soil-moisture factors increase costs (reject H01–H03; Saieed, 2022; Srivastav et al., 2025), precipitation reduces costs (support H12; Zhao and Zhang, 2025), and wet-bulb temperature and wind raise costs (support H15–H16; Addoum et al., 2021; Huang et al., 2018), while relative humidity is insignificant (accept H04). These findings reflect the Saudi context where cooling demand and infrastructure exposure dominate cost dynamics. Taken together, these results indicate that climate-sensitive industries are affected through interconnected operational and financial channels rather than isolated cost categories. Firms exposed to heat stress and atmospheric variability face immediate increases in operating expenditures and subsequently higher financing costs due to increased perceived risk, demonstrating compounded rather than independent cost pressures.

Firm controls support H17–H19: leverage raises costs in the baseline model, while size and growth reduce them. In the PCA model, leverage becomes negative because reduced multicollinearity allows it to capture financial discipline and scale efficiency rather than climate-correlated stress. This sign change does not alter the main conclusion, climate exposure structurally increases firm costs through interconnected operational and financial channels, but shows that the leverage effect depends on how climate risk is measured.

The results clearly demonstrate that climate variables, particularly soil moisture, specific humidity, wet-bulb temperature and wind speed, significantly increase operating costs, CC and TC, while precipitation helps reduce them. These findings confirm that climate risks are not only environmental challenges but also material financial drivers for Saudi-listed firms. Importantly, the results indicate a sequential cost transmission process in which physical climate exposure first alters operational efficiency and subsequently translates into financial risk pricing, ultimately shaping total firm expenditures. Firm-level factors reinforce this pattern: leverage amplifies costs in the baseline specification, while size and growth provide protective scale advantages. Under PCA, however, leverage instead reflects financial discipline rather than climate-related financial stress.

Because climate variables affect operating and financing costs through distinct channels, the results offer decision-relevant guidance: firms exposed to heat and humidity should prioritize operational adaptation (cooling efficiency and maintenance planning), while those facing wind-driven instability should prioritize financial risk management (liquidity buffers, hedging and debt maturity adjustment).

In the Saudi context, these findings have concrete implications for firms, regulators and investors under Vision 2030. For firms, climate risk should be integrated into operational budgeting and financial planning, with greater investment in energy-efficient technologies, cooling and water systems and diversified financing to reduce climate-related cost pressures. For regulators, the findings support strengthening climate-risk disclosure, stress testing and sector-specific reporting requirements, while aligning climate-resilience incentives with Vision 2030’s sustainability and economic diversification objectives. For investors and lenders, the evidence that climate exposure increases financing costs highlights the need to incorporate sector-specific climate-risk indicators into portfolio allocation, credit assessment and pricing decisions. Such measures can improve climate resilience while supporting the financial stability and sustainable transformation targeted by Vision 2030.

The findings highlight practical climate-risk management strategies. Firms can ease cost pressures by investing in energy-efficient and climate-resilient technologies, diversifying energy sources toward renewables and improving cooling and water systems. Financially, diversifying funding sources, extending debt maturities and using green finance can help mitigate climate-related financing constraints. These measures directly interrupt the transmission channel identified in the empirical results by reducing operational exposure and stabilizing financial risk simultaneously. At the policy level, incentive schemes such as tax credits, green subsidies and preferential financing for climate-resilient investments can accelerate private-sector adaptation and lower systemic climate-related financial risk. Overall, the empirical evidence therefore supports differentiated adaptation, financing and supervisory strategies derived directly from the estimated climate cost channels rather than general sustainability compliance measures.

This study has several limitations. The sample includes only 80 listed firms from seven climate-sensitive sectors, which may limit generalization to other industries or private firms. The 2010–2024 period may not capture longer structural climate shifts. Although GMM and PCA mitigate endogeneity and multicollinearity, measurement error in climate and cost proxies cannot be fully excluded. The analysis also focuses on six climate variables and does not incorporate extreme events (e.g. floods, sandstorms and heatwaves). Moreover, spatial heterogeneity across Saudi regions may not be fully captured by the grid-based climate data, potentially limiting the precision of firm-level climate exposure measures. Sectoral differences in cost structures may also generate heterogeneous climate-cost transmission mechanisms across industries.

Finally, results reflect the specific institutional and climatic conditions of Saudi Arabia.

Future research can extend the implications of this study in several ways. Sectoral heterogeneity should be examined to distinguish operational exposure from financial risk exposure and improve capital allocation decisions. Studies could also evaluate firm-level adaptation strategies (green technologies, monitoring systems, resilient supply chains and disclosure) to determine whether they reduce operating volatility or financing premia. Incorporating energy and carbon intensity would further integrate climate–finance and corporate risk-management theory. Expanding samples to small and medium-sized enterprises (SMEs) and cross-country settings would clarify regulatory and disclosure implications and strengthen policy relevance.

The authors sincerely thank the Editor and the anonymous reviewers for their constructive comments and suggestions, which helped improve the quality and clarity of this manuscript.

This research received no external funding.

The datasets and code used in this study are available from the corresponding author upon request.