Balancing Carbon and Profitability in Aviation: A Risk and Policy Perspective

Abstract

This study examines the intricate relationship between carbon emissions and sustainable financial performance in the global airline industry, a sector increasingly scrutinized for its environmental impact. Building upon the win–win hypothesis, trade-off theory, and emerging perspectives on non-linear environmental–financial linkages, this study explores how firm profitability is influenced not only by emission intensity but also by contextual factors such as temperature anomalies and the adoption of Sustainable Aviation Fuel (SAF). Using panel data from 29 major airlines headquartered across seven global regions, the analysis reveals a curvilinear relationship: while increased emissions are initially linked to higher profitability, likely reflecting operational scale, excessive emissions may diminish financial returns. The findings also underscore the moderating role of temperature anomalies, which can intensify both the initial benefits and the subsequent costs of emissions. Furthermore, the adoption of SAF appears to mitigate the financial risks of emissions under heightened climate-related pressure. Although initially costly and negatively associated with profitability, SAF investment shows potential long-term benefits, suggesting a non-linear payoff structure. Overall, the findings suggest that firms in carbon-intensive industries must carefully calibrate environmental strategies and investments to achieve long-term financial resilience. The study offers new insight into how internal decisions and external pressures jointly shape the emissions–performance dynamic.

1. Introduction

Climate change has moved to the forefront of global policy agendas. Reuters (2025) reports that 2024 was the warmest year ever, with above 1.5 degrees Celsius (34.7 degrees Fahrenheit) increase in world temperatures over pre-industrial levels for the first time. Last year, global CO₂ emissions from energy hit a historic high. Reducing carbon emissions is now widely accepted as essential to avoiding the most detrimental consequences of climate change. This urgency is underscored by the fact that over 70% of global emissions originate from just 100 companies, with the top 20 responsible for more than one-third (Taylor & Watts, 2019).

In this context, carbon emissions are becoming more widely acknowledged as a strategic corporate issue in addition to an environmental one. As such, the relationship between financial performance (FP) and corporate environmental performance (CEP) has gained significant attention from academics over the last few decades, making it a key issue in business and sustainability study. A growing body of research has investigated this relationship, but the findings remain mixed. Some studies report that reducing carbon emissions enhances financial performance by improving efficiency, stimulating innovation, and generating reputational benefits (Porter & van der Linde, 1995; Jones, 1995; Guenther & Hoppe, 2014). In contrast, others argue that environmental investments can impose substantial short-term costs, thereby weakening profitability (Konar & Cohen, 2001; Lankoski, 2008). Still others suggest a more complex, non-linear relationship—particularly in carbon-intensive sectors—where both excessive and insufficient environmental initiatives may hinder firm performance. Because carbon emissions represent only one dimension of broader corporate environmental performance, the overall relationship between corporate environmental performance (CEP)—and especially carbon emissions—and financial performance remains inconclusive.

There is substantial support for the “win–win” perspective (Porter & van der Linde, 1995; Wagner & Schaltegger, 2004; Fujii et al., 2013), which argues that strong environmental performance can enhance long-term competitiveness. Additionally, effective carbon strategies can create intangible assets that yield sustainable financial gains, according to resource-based and instrumental stakeholder theory (Abban & Hasan, 2021). Carbon-efficient firms can reduce regulatory and reputational risks, thereby enhancing long-term financial stability. For instance, BlackRock’s 2023 annual letter, led by CEO Larry Fink, highlighted the firm’s stance that climate change poses significant financial risks, thereby reaffirming the importance of shifting toward a low-carbon economy. Corporate carbon performance (CCP)—which reflects measurable efforts to reduce greenhouse gas emissions and improve energy efficiency—has emerged as an essential dimension of corporate responsibility (Velte, 2021). However, the “trade-off” perspective remains influential, arguing that environmental investment costs, especially in the short term, may outweigh benefits (Esty & Porter, 1998; Fujii et al., 2013). Compliance with regulations such as the EU Emissions Trading System can impose significant burdens (Chan et al., 2013), and managerial opportunism may lead to overinvestment in ESG initiatives that do not align with shareholder value (King & Lenox, 2001). Empirical studies confirm that firms, particularly in high-emission industries like aviation, often experience declines in ROA and profitability during early ESG (Environmental, Social, and Governance) adoption (Damert et al., 2017; Ganda & Milondzo, 2018; Kuo et al., 2021; Desai et al., 2022).

To reconcile these opposing findings, a curvilinear or non-linear relationship has been proposed (Barnett & Salomon, 2006; Trumpp & Guenther, 2017). While initial environmental efforts may lower performance, gains may emerge as firms accumulate capabilities and stakeholder support. Trumpp and Guenther (2017) introduce the “Too-Little-of-a-Good-Thing” (TLGT) effect, warning against minimal engagement, and the “Too-Much-of-a-Good-Thing” (TMGT) effect, suggesting diminishing returns from excessive efforts. These perspectives converge on the notion that there exists an optimal level of environmental investment.

The issue of carbon intensity and the challenges of decarbonization become particularly salient in the aviation industry, which is widely recognized for its high emissions per passenger-kilometer and structural difficulty in transitioning to low-carbon alternatives. According to Earth.Org (2024), a single long-haul flight can emit more CO₂ in just a few hours than the average individual in 56 countries produces in an entire year.

Among full-service carriers, Deutsche Lufthansa AG recorded the highest CO₂ emissions in 2024, totaling 10 MtCO₂—a 6% increase from the previous year. It was followed by British Airways with 8.9 MtCO₂ (+3%) and Air France with 8.2 MtCO₂, which saw a slight decrease of 1%. Emirates and KLM each emitted 5.5 MtCO₂ and United Airlines emitted 3.6 MtCO₂ (+2%) (Transport & Environment, 2025). The airline industry, while vital to global economic growth, is under growing scrutiny for its environmental impact—accounting for around 2.5% of global CO₂ emissions. As sustainability becomes essential to corporate legitimacy and resilience, airlines face increasing pressure to reduce emissions and transition toward greener operations. Though accounting for only 2.5% of global CO₂ emissions, aviation’s net climate impact is estimated at approximately 4% (Ritchie, 2024).

Sustainable aviation fuel (SAF) is a viable substitute for lowering CO₂ emissions in the aviation sector. Sustainable aviation fuel (SAF) is derived from a variety of renewable sources, including waste materials. Unlike traditional fossil fuels, SAF recycles carbon already present in the atmosphere, thereby contributing to a meaningful reduction in net CO₂ emissions. Reaching net-zero by 2050 remains a complex challenge for airlines due to high costs and technological limits of sustainable aviation fuel (SAF) (Earth.Org, 2024). The adoption of sustainable aviation fuel (SAF)—a key mitigation innovation—offers the potential to reduce emissions but involves considerable cost, which may not be easily transferred to consumers due to demand elasticity (Watson et al., 2024; Jaume et al., 2024). Although SAF can cut emissions by up to 27%, it typically costs at least 120% more than conventional fuel, making its financial impact ambiguous. Though, industries like aviation, energy, and manufacturing exhibit unique emission-performance patterns (Andrew & Cortese, 2011; Iwata & Okada, 2011; Miah et al., 2021; Oestreich & Tsiakas, 2024; Włodarczyk et al., 2024). For firms in such sectors, aligning environmental responsibility with profitability remains a core dilemma. Yet, some industry leaders show that environmental and financial goals can move forward together (Business World, 2025).

Carbon disclosure, meanwhile, plays a crucial role in signaling a company’s climate accountability. However, the lack of standardized reporting frameworks limits data comparability and benchmarking (Iwata & Okada, 2011; Schaltegger & Csutora, 2012). Some airlines report detailed emission figures, while others omit them entirely or present them in incomparable formats. This lack of standardization—such as variation in units, scope definitions, or reporting frequency—posed significant challenges in compiling a reliable dataset for this study. Carbon emission data were therefore manually extracted from annual and sustainability reports, acknowledging these limitations.

Additional complexity arises from heterogeneity in climate across regions. Corporations are embedded in distinct environmental and regulatory settings that shape their exposure to climate-related risks. Due to climate heterogeneity—captured by regional temperature anomalies—firms may experience varying degrees of public and policy pressure regarding emissions. In this study, we incorporate temperature anomalies from the continent in which each airline is headquartered as a proxy for localized climate intensity. While acknowledging the limitation that airlines operate globally, the home-region anomaly serves as an indicator of the broader environmental context and regulatory pressure the firm faces. This approach allows us to explore whether the financial consequences of carbon emissions differ depending on the severity of regional climate deviations.

Given these intricacies, a number of important problems arise. First, especially in carbon-intensive industries like aviation, it is still uncertain whether the link between carbon emissions and financial success is best defined as linear or follows a more complicated, curved pattern—particularly in carbon-intensive sectors like aviation. Second, regional climate conditions, such as temperature anomalies, may influence the profitability of emissions by affecting public and regulatory responses. Third, the adoption of sustainable aviation fuel (SAF) may play a moderating role in this relationship, potentially offsetting the financial risks associated with high emission levels. Building on these considerations, this study sets out to examine the financial implications of carbon emissions in the global airline industry, with particular attention to the role of regional climate conditions and policy responses. Specifically, we investigate whether the relationship between carbon emissions and firm profitability follows a curvilinear rather than a linear pattern. We further explore how this relationship is moderated by temperature anomalies in the airline’s home region and by the adoption of sustainable aviation fuel (SAF). To address these questions, we analyze panel data on major international airlines using multiple regression models with interaction terms and non-linear specifications. This paper proceeds as follows. Section 2 reviews the literature and introduces the hypotheses. Section 3 explains the data and analytical methods. Section 4 presents the findings, and Section 5 interprets the results and considers their broader significance.

2. Literature Review and Hypotheses

Since the 1970s, scholars have produced extensive research on the relationship between corporate environmental performance (CEP) and financial performance (FP). Despite the vast empirical evidence, no consensus has emerged. Some studies identify a negative association between emissions and FP (Konar & Cohen, 2001; Lankoski, 2008), while others report no significance. Horváthová’s (2010) meta-analysis of more than 150 studies illustrates this fragmentation: roughly 55% find a positive relationship, 30% a negative one, and the remaining 15% no statistical association. Recent empirical work also highlights methodological choices, industry heterogeneity, institutional pressures, disclosure practices, and measurement differences as the primary sources of these mixed findings (Meng et al., 2023). These insights suggest that the CEP–FP nexus is complex and likely contingent on multiple organizational and contextual factors.

2.1. Positive Linkages

A substantial line of research supports a positive association between CEP and FP (Trumpp & Guenther, 2017). The win–win hypothesis (Porter & van der Linde, 1995) suggests that environmentally responsible practices can lead to operational efficiencies, innovation, and competitive advantages. Empirical studies confirm this mechanism: environmental strategies can stimulate innovation, reduce energy use, strengthen stakeholder relations, and enhance firm reputation (Hart & Ahuja, 1996; Guenther & Hoppe, 2014; Busch & Hoffmann, 2011). Busch and Hoffmann (2011) documented a positive relationship between carbon mitigation and FP among large global firms, and Alvarez (2012) found that emissions reductions positively influenced ROA. Similar results appear in Margolis et al. (2009), Hilmi (2016), Shi & Jing (2016), and Nichita et al. (2021).

Environmental initiatives—including emissions reduction, pollution prevention, waste management, recycling, water purification, integrated environmental management systems, and cleaner production—can reduce operating costs and improve firm value (Duque-Grisales et al., 2020). From the stakeholder perspective, carbon disclosure is an important signal of credibility and long-term commitment (Roberts, 1992; Saka & Oshika, 2014; Benlemlih et al., 2018). Firms that provide voluntary disclosures reduce information asymmetry and enhance transparency, which strengthen investor confidence. According to Aydıngülü Sakalsız and Özçelik (2024), while high carbon emissions can depress ROA and ROE, firms that engage in innovation can offset these negative financial impacts.

The resource-based view (RBV) and instrumental stakeholder theory propose that environmental performance creates intangible resources—such as innovative capabilities, employee engagement, legitimacy, and relational capital—that generate long-term competitive benefits (Abban & Hasan, 2021; Dixon-Fowler et al., 2013). Studies also highlight that carbon-efficient firms tend to outperform their peers due to superior resource allocation and reputational advantages (Lewandowski, 2017; Ganda & Milondzo, 2018; Perlin et al., 2021). Castilho and Barakat (2022) find that climate adaptation enhances firm value, while Ghosh et al. (2023) report evidence of non-linear improvements. Busch & Lewandowski (2018) also conclude that firms with better carbon management enjoy stronger financial outcomes.

2.2. Negative Effects

Despite the potential benefits of environmental initiatives, a substantial stream of research highlights the financial burdens and competitive disadvantages that such strategies may impose, particularly in the short term. Early work by Freedman and Bikki (1992) and Konar and Cohen (2001) suggests that firms engaging in eco-friendly practices face increased operating costs that can erode profitability. Classical economic arguments and trade-off theory (Friedman, 1970; Lankoski, 2008) further posit that allocating resources to environmental activities may crowd out investments in core operations, thereby reducing financial performance. Several scholars also emphasize the broader risks posed by climate change and carbon emissions for firm stability (Bebbington & Larrinaga-González, 2008; McFarland, 2009; Burton, 2010).

Empirical evidence reinforces this cost-based perspective. Carbon emission reduction often requires substantial capital expenditures, advanced technologies, and wide-ranging organizational adjustments, all of which increase operational costs and reduce cash flow (Delmas et al., 2015; Lewandowski, 2017; Deng & Li, 2020). Esty and Porter (1998) similarly argue that the large upfront investments required for environmental improvements may outweigh expected financial returns, especially in industries with tight margins. Fujii et al. (2013) show that many emissions-reducing technologies remain prohibitively expensive, undermining practical adoption.

Regulatory compliance can also impose significant burdens. According to Palmer et al. (1995), regulatory-driven increases in CEP may raise compliance costs, while Chan et al. (2013) find that participation in the EU Emissions Trading System (ETS) results in compliance expenses representing 5–8% of total material costs for carbon-intensive firms. Eccles et al. (2013) also highlight that ESG and sustainability initiatives entail substantial investment, which may depress firm performance in the near term.

Industry-specific evidence further supports these concerns. In the airline sector, Kuo et al. (2021) report a decline in ROA during the early stages of ESG adoption, even though performance gradually improves over time. Damert et al. (2017) and Trumpp and Guenther (2017) note that when environmental costs exceed associated benefits, the financial gains from decarbonization efforts may be limited or delayed. Managerial opportunism theory (King & Lenox, 2001) additionally suggests that some environmental actions are symbolic rather than efficiency-enhancing, raising concerns regarding the genuine financial value of such initiatives.

Research across global markets identifies negative effects as well. Ganda and Milondzo (2018) find significant negative associations between Scope 1 and 2 emissions and ROE, ROI, and ROS among South African firms. McLaughlin (2011) argues that carbon disclosures may expose firms to reputational scrutiny and regulatory pressure. Desai et al. (2022) show that carbon emissions significantly reduce financial performance, while Trinks et al. (2020) highlight that the financial impact of carbon efficiency varies considerably depending on market dynamics and regulatory regimes.

Taken together, these studies suggest that environmental initiatives may entail substantial financial risks, particularly when initial investments, compliance costs, and technological expenditures exceed short-term benefits. Such findings underscore the importance of assessing both the timing and magnitude of environmental investments when evaluating their financial implications.

2.3. Curvilinear Perspectives

Research increasingly suggests that the CEP–FP relationship may follow a curvilinear rather than a purely linear pattern. Scholars argue that the link between CEP and FP is more complex than a simple positive or negative association (Moore, 2001; Lankoski, 2008; Orlitzky, 2013). Instead of examining whether environmental efforts improve or impair financial outcomes, recent studies explore how varying levels of environmental engagement produce non-linear effects.

From a resource-based perspective, firms may initially experience decreases in FP when increasing CEP due to substantial early-stage investments. Over time, however, accumulated environmental capabilities, improved resource efficiency, and stronger stakeholder relationships may enhance financial outcomes (Barnett & Salomon, 2006). One major theoretical mechanism is the “too-much-of-a-good-thing” (TMGT) effect, which proposes an inverted U-shaped relationship between CEP and FP: performance increases up to an optimal turning point (TP), after which excessive environmental investments yield diminishing returns or resource misallocation (Deng & Li, 2020; Trumpp & Guenther, 2017).

To explain the opposite side of the curve, Trumpp and Guenther (2017) also introduce the “too-little-of-a-good-thing” (TLGT) effect. Under TLGT, insufficient environmental engagement may expose firms to regulatory penalties, reputational harm, stakeholder distrust, and greater long-term vulnerability. As a result, the CEP–FP relationship may initially appear negative but eventually turns positive as engagement increases. This dual mechanism reflects the reality that both extremely low emissions (due to high abatement costs) and excessively high emissions (due to societal and regulatory pressures) can adversely impact FP.

Taken together, these perspectives emphasize the importance of identifying an optimal level of environmental engagement, where firms achieve positive financial outcomes without incurring the costs associated with under- or over-investment. These theoretical insights suggest that CEP and FP may be linked through a curvilinear pattern, rather than a strictly monotonic one.

Empirical work supports this view. Jin and Xu (2020) document a U-shaped relationship between environmental activities and FP, driven by improvements in organizational culture, pollution-prevention technologies, and stakeholder management. Khatib et al. (2023) show that managerial environmental training enhances carbon disclosure quality and, in turn, improves FP. These findings reinforce the notion that firms benefit from environmental engagement only up to an optimal point—after which the trade-off hypothesis and managerial opportunism perspectives explain why excessive investment may reduce financial returns.

To provide a clearer overview of prior findings,

Table 1. CEP-FP Literature Summary.

Category	Authors (Year)	Main Findings
Positive	Porter and van der Linde (1995); Guenther and Hoppe (2014); Trumpp and Guenther (2017)	CEP enhances FP via resource efficiency, innovation, stakeholder relations, and reputational benefits
	Busch and Hoffmann (2011)	Positive correlation between carbon mitigation and financial performance
	Alvarez (2012)	Positive effect on ROA; no effect on ROE
	Nichita et al. (2021)	Positive correlation between mitigation and FP
	Busch and Lewandowski (2018)	Better carbon management linked to stronger financial outcomes
	Castilho and Barakat (2022)	Adaptation improves financial performance
Negative	Freedman and Bikki (1992)	Eco-friendly practices may raise costs, reducing profitability
	Delmas et al. (2015); Lewandowski (2017)	Investments increase costs and reduce cash flow
	Esty and Porter (1998)	High initial investments may outweigh returns
	Chan et al. (2013)	Compliance costs = 5–8% of total material costs for carbon-intensive firms
	Kuo et al. (2021)	Short-term ROA decline in initial ESG phase, recovery later
	Ganda and Milondzo (2018)	Significant negative link with ROE, ROI, and ROS
Curvilinear	Trumpp and Guenther (2017); Deng and Li (2020)	Optimal CEP level maximizes FP; over- or under-investment reduces returns (TMGT/TLGT)
	Jin and Xu (2020)	U-shaped link; cultural & technological advances improve FP over time
	Khatib et al. (2023)	Training improves disclosure quality, enhancing FP
	Ghosh et al. (2023)	U-shaped relationship; both under- and over-investment suboptimal
	Orlitzky (2013)	CEP–FP link more complex than a simple positive or negative pattern (conceptual)
	Barnett and Salomon (2006)	Resource-based theory explanation for optimal CEP–FP balance (theoretical)

summarizes key studies across positive, negative, and curvilinear CEP–FP relationships.

2.4. Industry and Regional Differences

The financial implications of carbon emissions differ substantially across industries and regions, reflecting variation in regulatory environments, operational structures, and stakeholder expectations. Prior research emphasizes that industry- and region-specific characteristics play a critical role in shaping the environmental–financial performance nexus (Lopes de Senna & de Araujo Moxotó, 2025). In many carbon-intensive sectors, compliance obligations and reputational pressures elevate credit risk, prompting credit rating agencies to incorporate environmental indicators when assessing firm stability (Weber et al., 2008; Goss & Roberts, 2011; Safiullah et al., 2021).

Industries with inherently high emissions—such as energy, manufacturing, and aviation—exhibit particularly distinct patterns. Iwata and Okada (2011) show that greenhouse gas mitigation yields stronger financial returns in environmentally sensitive industries than in high-polluting ones, suggesting that industry context moderates the profitability of emission-reduction strategies. Miah et al. (2021) further find that carbon emissions are negatively associated with ROE, Tobin’s Q, credit ratings, and financial stability metrics, especially within non-financial industries. Regional factors shape these dynamics as well. Evidence from China indicates that carbon emission policies have had positive effects on energy-intensive firms by fostering cost reductions and facilitating participation in carbon trading schemes (Liu et al., 2023). In the United States, Aswani et al. (2024) report that high-emitting firms with substantial emission reductions achieved stronger stock market performance, challenging the conventional assumption that lower emissions always correspond to higher profitability. Similarly, Włodarczyk et al. (2024) show that high-emission firms experienced sharper declines in profitability during the COVID-19 period, reflecting heightened vulnerability under crisis conditions.

Cross-sectional evidence reinforces this pattern. Oestreich and Tsiakas (2024) find that among S&P 500 companies, carbon emission intensity is strongly negatively associated with profitability, indicating that high-emission firms face greater financial pressure across a wide range of sectors. Overall, industry- and regional-level heterogeneity plays a significant role in determining how carbon emissions translate into financial outcomes. These insights highlight the importance of sector-specific analysis—particularly for carbon-intensive industries such as aviation—where the emissions–performance relationship may exhibit unique and complex dynamics.

2.5. Measurement Challenges and Disclosure Quality

Corporate environmental performance (CEP) is a multifaceted concept, and different measurement approaches can lead to divergent empirical results (Guenther & Hoppe, 2014). Depending on whether reporting is mandatory or voluntary, firms use varying indicators that reflect different emission scopes. ESG and carbon disclosures provide stakeholders with greater visibility into operations and help reduce information asymmetry (van Duuren et al., 2016). Voluntary carbon disclosure also enhances transparency and mitigates agency problems (Healy & Palepu, 2001; Huang & Watson, 2015), serving both as a proxy for corporate openness (Freeman & Evan, 1990) and as a forward-looking signal of long-term commitment. High-quality disclosure may further strengthen reputational capital (Ludwig & Sassen, 2022).

Despite these benefits, significant challenges remain. The lack of standardized carbon reporting practices undermines comparability and limits the usefulness of emissions data (Andrew & Cortese, 2011). Although mandatory reporting provides more structured information (Sullivan & Gouldson, 2012), it is not always more reliable than voluntary disclosures (Perrault & Clark, 2010). Technical issues related to boundary selection, indicator choice, and verification further complicate carbon accounting (Schaltegger & Csutora, 2012).

Firms use a range of quantitative indicators—absolute emissions, relative emissions, annual emissions, and actual reductions—to evaluate CEP. Financial markets tend to rely on intensity metrics such as emissions relative to sales, which offer more consistent comparisons across firms. Delmas et al. (2015) show that firms with lower carbon intensity achieve higher Tobin’s Q, highlighting the financial relevance of efficiency measures. Ultimately, stakeholder perceptions remain central: corporate reputation strongly influences financial outcomes (Donaldson & Preston, 1995; Jones, 1995), and firms with stronger reputational standing often experience higher sales (Galama & Scholtens, 2021).

2.6. The Role of Sustainable Aviation Fuel (SAF)

Sustainable aviation fuels (SAF) are widely regarded as a key mechanism for substantially reducing lifecycle greenhouse gas emissions in the aviation sector. Produced from renewable sources such as used cooking oil and agricultural waste, SAF lowers net CO₂ output by relying on carbon from renewable cycles rather than extracting new carbon from fossil reserves. By substituting traditional jet fuel with SAF, airlines can significantly reduce their overall emissions. According to the International Air Transport Association (IATA), SAF is expected to account for approximately 65% of the emission reductions required for the aviation industry to achieve net-zero CO₂ emissions by 2050. Achieving this target will require rapid expansion of SAF-related technologies, which may become feasible within the next decade if strong international policy support is implemented (Earth.Org, 2024).

Despite its environmental advantages, broad adoption of SAF remains challenging due to high production costs and associated impacts on airline profitability. Jaume et al. (2024) show that using 100% SAF becomes economically unsustainable when SAF prices exceed 1.5 times those of conventional fuel, as airlines cannot fully pass these costs to passengers due to demand elasticity. Although SAF can reduce emissions by at least 27%, most bio-jet fuel production technologies currently cost more than 120% above fossil-based jet fuel (Watson et al., 2024). Efforts to raise fares or reduce CASK (unit cost per available seat-kilometer) are constrained by competitive market pressures, making it difficult for airlines to transfer higher fuel costs to consumers.

2.7. Climate Stress and Moderating Effects of Temperature Anomalies

Temperature anomalies have become an important contextual factor shaping how environmental performance translates into financial outcomes. Prior studies show that climate-related temperature shocks impose significant operational and financial pressures on firms. Liu et al. (2025) find that extreme temperature shocks increase the likelihood of financial distress, particularly among energy-intensive industries. These vulnerabilities are exacerbated by investor sensitivity to climate risk, as institutional investors increasingly consider climate exposures when making asset allocation decisions (Krueger et al., 2020). Additional evidence suggests that corporate social responsibility (CSR) initiatives can help mitigate the adverse financial impacts of climate risk, as firms face stronger stakeholder expectations under environmentally challenging conditions (Ozkan et al., 2023). Temperature fluctuations and broader climate uncertainty also heighten financing constraints, further limiting firms’ ability to adapt (Wu et al., 2025).

Taken together, these findings indicate that temperature anomalies may moderate the relationship between carbon emissions and financial performance. Depending on firms’ adaptive capacity and stakeholder expectations, climatic stress can either amplify or dampen the financial consequences of emissions.

Given the carbon-intensive nature of airline operations, the aviation sector is particularly exposed to climate-related risks and sustainability pressures. Recent research highlights the challenges airlines face in addressing environmental concerns (Chen et al., 2022). These characteristics make the industry a compelling setting in which to examine how emissions, financial performance, and climate stress interact.

Based on the aforementioned factors, the following hypotheses are formulated:

H1. 

A curvilinear relationship exists between carbon emissions and firm financial performance in the airline industry.

H2. 

Temperature anomalies moderate the non-linear association between carbon emissions and financial performance, with the nature and strength of this moderation potentially varying across climatic conditions.

H3. 

The adoption of sustainable aviation fuel (SAF) is expected to mitigate the impact of CO₂ emissions on the financial performance of airlines, potentially influencing the profitability impact of emissions.

3. Research Design

3.1. Sample Selection

This study investigates the environmental performance of global airline companies using CO₂ emissions as a proxy, and examines its impact on financial outcomes over the period from 2011 to 2024. The sample consists of 29 major carriers headquartered across seven regions—Asia, Europe, North America, South America, Africa, Oceania, and the Middle East. While most of the selected firms operate as full-service carriers, a small number of low-cost carriers are also included. For an additional robustness check, we include a dummy variable (LCC_Dummy) coded as 1 for low-cost carriers (LCCs) and 0 otherwise. Among the 29 airlines in our sample, only five are classified as LCCs—four Korean carriers and one European airline. Although these firms represent a minority, they are retained to maximize sample coverage and preserve variation in business strategies. The inclusion of Korean LCCs was facilitated by the availability of standardized emissions data through the National GHGs Management System (NGMS), which ensures consistent and comparable information across companies, while the European LCC was included due to its transparent public disclosure of environmental data. We acknowledge that the LCC dummy may partially capture country-level effects due to the geographic concentration of LCCs in the sample. For analytical purposes, carriers based in the Middle East were classified under Asia, particularly to align with the regional temperature anomaly variable. The regional imbalance—especially the concentration of airlines in Asia—partly reflects this classification choice as well as the relatively high availability of consistent CO₂ disclosure data among Asian carriers. In particular, all publicly listed airlines in Korea were included in the sample due to data accessibility. CO₂ emissions data were manually extracted from sustainability or annual reports, as disclosure practices vary and some firms do not report such data annually. The final dataset comprises 312 firm-year observations.

In addition to emissions, this study incorporates two moderating variables: temperature anomalies and the adoption of Sustainable Aviation Fuel (SAF). Temperature anomalies are measured annually at the continental level using data from the National Oceanic and Atmospheric Administration (NOAA), reflecting deviations from the 1901–2000 historical average. These time-varying values capture regional climate pressure specific to each year and serve as a proxy for the environmental intensity faced by airlines in their home regions. The adoption of Sustainable Aviation Fuel (SAF) is captured through a firm-level, time-varying dummy variable coded as 1 in the year following the airline’s first publicly disclosed SAF use, or pilot testing, and 0 otherwise. These disclosures are identified through sustainability reports, annual filings, or other public sources. As SAF remains in the early stages of implementation, this variable reflects symbolic and strategic engagement with low-carbon transition rather than full operational substitution. To mitigate the influence of outliers, selected continuous variables, including emissions and financial performance metrics, were winsorized at the top and bottom 1%. The geographical distribution of the airline sample across seven global regions are reported in

Table 2. Airline Distribution by Continent (2011–2024).

Geographic Area	Number of Firms	%
Asia	15	52%
Europe	4	14%
North America	4	14%
South America	1	3%
Africa	3	10%
Oceania	2	7%
Total	29	100%

.

3.2. Regression Model and Measurement of Variables

This study compiles firm-level CO₂ emissions data following the approaches of Oestreich and Tsiakas (2024) and Bolton and Kacperczyk (2021, 2023). Carbon emissions are classified according to the International Local Governments Greenhouse Gas Emissions Analysis Protocol (IEAP) into Scope 1 (direct emissions from owned or controlled sources), Scope 2 (indirect emissions from purchased electricity, heat, or steam), and Scope 3 (other indirect emissions occurring across the value chain) (GHG Protocol, 2015).

For Korean airlines, CO₂ emissions were obtained from the National GHGs Management System (NGMS), which provides standardized and government-verified disclosures. For non-Korean airlines, emissions data were manually collected from annual and sustainability reports, as well as other publicly available disclosures, and standardized to metric tons of CO₂ equivalent (tCO₂e). When emissions were reported in alternative units (e.g., per unit of operational output), the data were recalculated using the corresponding operational metrics to ensure consistency across all firm-year observations. This standardization process enhances comparability and minimizes inconsistencies arising from heterogeneous disclosure practices. Nevertheless, disclosure practices varied across firms, resulting in differences in the scope and detail of emissions data, which may introduce some measurement heterogeneity.

To investigate Hypothesis 1, an ordinary least squares (OLS) regression model is constructed, with sustainable profitability serving as the dependent variable. The specification is designed to test the potential curvilinear relationship between carbon emissions and financial performance:

SP_i,t+1 = α + β₁CO_2i.t + β₂CO₂²_i.t + ∑α_jX_j + ∑α_kCONTI_k + ∑α_lYEAR_l + ε_i,t

(1)

Here, SP_i,t+1 denotes represents the firm’s sustainable profitability in the following year, which captures the persistence of performance over time. This variable is derived based on return on assets (ROA), with persistence measured through the following auxiliary regression:

ROA_i,t+1 = α + β₁ROA_i.t + ε_i,t

(2)

In our empirical analysis, financial performance (FP) is operationalized using return on assets (ROA). The calculation of return on assets involves dividing net income by total assets.

Carbon intensity (CO₂) is defined as total CO₂ emissions divided by revenue, and its squared term is included to capture potential non-linear effects. While emissions per RPK is a commonly used aviation-specific intensity measure, our analysis follows Oestreich and Tsiakas (2024) and Bolton and Kacperczyk (2021, 2023) and operationalizes CO₂ intensity as total emissions divided by revenue. Firm-specific controls include leverage (total liabilities relative to total assets), firm size (log-transformed total assets), annual asset growth, and operating cash flow (cash from operations scaled by total assets). To account for time and regional effects, the model incorporates year and continent fixed effects through dummy variables.

To examine Hypothesis 2, the baseline model is extended by introducing interaction terms that capture the moderating effect of climate conditions—specifically, regional temperature anomalies—on the relationship between carbon emissions and financial performance. The following model is estimated:

SP_i,t+1 = α + β₁CO_2i.t + β₂CO₂²_i.t + β₃HighTempCO_2i.t + β₄HighTempCO₂²_i.t + ∑α_jX_j + ∑α_kCONTID_k + ∑α_lYEAR_l + ε_i,t

(3)

In this specification, the variable HighTemp is a binary indicator equal to 1 when the annual temperature anomaly for a firm’s home continent is at or above the cross-continental median for that year, and 0 otherwise. Interaction terms between this climate dummy and both CO₂ intensity and its squared term are included to assess whether the emissions–performance relationship varies depending on the level of regional temperature anomaly. This approach assumes that firms in regions experiencing relatively higher temperature deviations may face greater climate-related pressures, which could influence the financial effects of their emissions. As in the previous model, control variables include leverage, size, sales growth, and operating cash flow. Year and continent fixed effects are incorporated to account for unobserved temporal and regional heterogeneity.

To test Hypothesis 3, the analysis further expands the prior model by introducing an additional moderating variable that captures whether a firm has adopted sustainable aviation fuel (SAF). Specifically, the following regression model is estimated:

SP_i,t+1 = α + β₁CO_2i.t + β₂CO₂²_2i.t + β₃HighTempCO_2i.t + β₄HighTempCO₂²_i.t + β₅HighTempSAFCO_2i.t + β₆HighTempSAFCO₂²_i.t + ∑α_jX_j + ∑α_kCONTI_k + ∑α_lYEAR_l + ε_i,t

(4)

Here, SAF is a time-varying dummy variable coded as 1 for firm-year observations occurring after a firm has adopted sustainable aviation fuel, and 0 otherwise. In our models, SAF adoption enters as a time-varying binary variable indicating whether an airline has publicly initiated SAF-related activities. As described in the sample selection section, this variable reflects early-stage engagement with low-carbon transition efforts. While it does not capture actual usage volume, it serves as a proxy for environmental signaling and strategic commitment, consistent with prior studies on early green technology adoption. The triple interaction terms test whether the financial implications of carbon emissions differ for firms that not only face heightened climate exposure but also take proactive steps toward decarbonization through SAF. As in previous models, control variables include leverage, firm size, asset growth, and operating cash flows. Year and continent fixed effects are also included to account for temporal and regional heterogeneity. To account for unobserved regional heterogeneity, we include continent-level dummy variables representing the location of each airline’s headquarters. These dummies control for broad differences in policy environments, climate risks, and SAF infrastructure across regions.

4. Empirical Results

4.1. Descriptive Statistics and Correlations

Table 3. Descriptive Statistics.

Variables	Mean	StdDev	Median	Q1	Q3
SP	0.1557	0.2356	0.0892	0.0721	0.1170
CO2	0.5107	2.4516	0.0017	0.0001	0.0066
HighTemp	0.5370	0.4993	1	0	1
HighTempCO2	0.2326	1.6155	0.0001	0	0.0013
SAF	0.2365	0.2365	0	0	0
HighTempSAFCO2	0.0250	0.2847	0	0	0
LEV	0.8271	0.2535	0.7930	0.7096	0.9048
SIZE	13.0166	4.4235	11.7173	10.2153	14.2473
GROW	0.2247	1.5044	0.0584	−0.0370	0.2020
OCF	0.2257	0.8626	0.0948	0.0556	0.1381

Notes. SP: Sustainable profitability, measured as the persistence of return on assets (ROA). CO₂: Carbon intensity, calculated as total Scope 1–3 CO₂ emissions divided by total revenue. CO₂ intensity standardized to tCO₂e; conversions applied where original units varied. HighTemp: A dummy variable equal to 1 if the regional temperature anomaly for a given year is above the sample median, and 0 otherwise. HighTempCO₂: Interaction terms between the HighTemp dummy and carbon intensity. SAF: A dummy variable equal to 1 for firm-years after the adoption of Sustainable Aviation Fuel (SAF), 0 otherwise. HighTempSAFCO₂: Triple interaction terms capturing how SAF adoption under higher temperature anomalies affects the emissions–performance relationship. Size: Log-transformed total assets. Leverage: Total liabilities relative to total assets. Growth: Annual asset growth, defined as (Asset_t − Asset_t−1)/Asset_t−1. OCF: Operating cash flows scaled by total assets.

displays the summary statistics for the key variables used in the investigation. The mean (median) value of sustainable profitability (SP) is –0.1557 (0.0892), while the corresponding value for carbon emissions (CO₂) is 0.5107 (0.0017). The mean (median) of the HighTemp dummy variable is 0.5370 (1), indicating that more than half of the observations correspond to firms operating under elevated temperature anomalies. The interaction term HighTempCO₂ shows a mean (median) of 0.2326 (0). The mean (median) of the SAF dummy variable is 0.2365 (0), suggesting that approximately 24% of the sample adopted Sustainable Aviation Fuel (SAF) during the observation period. The three-way interaction term HighTempSAFCO₂ has a mean (median) of 0.0250 (0).

Among the control variables, the mean (median) values of leverage (LEV), firm size (SIZE), growth (GROW), and operating cash flows (OCF) are 0.8271 (0.7930), 13.0166 (11.7173), 0.2247 (0.0584), and 0.2257 (0.0948), respectively.

Table 4. Correlations.

Variable	SP	CO2	HighTempCO2	HightempSAFCO2	LEV	SIZE	Growth	OCF
SP	1
CO2	0.5612 *	1
HighTempCO2	0.4943 *	0.6426 *	1
HightempSAFCO2	0.2540 *	0.0987	0.1649 *	1
LEV	−0.0420	0.0512	0.0245	0.0379	1
SIZE	−0.3052 *	−0.3178 *	−0.2281 *	−0.1257 *	−0.0868	1
Growth	0.0899	−0.0207	−0.0301	0.0233	−0.0398	−0.0055	1
OCF	0.6460 *	0.8011 *	0.7866 *	0.2110 *	−0.0775	−0.2325 *	−0.0141	1

Note. Variable definitions are provided in Table 2. * p < 0.05.

displays the pairwise correlation coefficients among the key variables. Sustainable profitability exhibits statistically significant positive associations with CO₂ emissions as well as the interaction terms involving temperature anomalies and SAF adoption. These initial patterns lend indicative support to the proposed hypotheses. Furthermore, multicollinearity is not a concern. Variance inflation factors (VIFs) for all predictors—including interaction terms, squared CO₂ terms, and the moderating variables—remain well below conventional thresholds. In all specifications, the maximum VIF is below 5, indicating a very low degree of collinearity and ensuring that the estimated coefficients are not distorted by redundant explanatory power among the regressors.

4.2. Regression Results and Discussion

Table 5. Regression Results.

Panel A. OLS Regression Results
Variables	Dependent Variable: SP
	Model 1.	Model 2.	Model 3.
Constant	0.1586 *** (3.93)	0.1422 *** (3.62)	0.1477 *** (4.00)
CO2	0.2419 *** (12.49)	0.0925 *** (2.50)	0.0862 *** (2.47)
CO22	−0.0163 *** (−12.57)	−0.0058 ** (−2.24)	−0.0054 ** (−2.20)
HighTempCO2	-	0.1691 *** (4.46)	0.2414 *** (6.21)
HighTempCO22	-	−0.0134 *** (−4.68)	−0.0188 *** (−6.36)
HighTempSAFCO2	-	-	−0.4536 *** (−5.85)
HighTempSAFCO2	-	-	0.0873 *** (4.78)
LEV	−0.0692 ** (−1.92)	−0.0700 *** (−2.01)	−0.0751 *** (−2.30)
SIZE	0.0006 * (0.31)	0.0006 (0.33)	0.0002 (0.11)
GROW	−0.0057 (−1.46)	0.0017 (0.41)	0.0028 (0.72)
OCF	0.1027 *** (8.72)	0.1339 *** (9.20)	0.1324 *** (9.47)
Continent dummies	Included	Included	Included
Year dummies	Included	Included	Included
F value	58.88 ***	59.22 ***	63.85 ***
Adjusted R2	0.8171	0.8296	0.8498
n	312	312	312
Panel B. Fixed Effect Regression Results
Variables	Dependent Variable: SP
	Model 1.	Model 2.	Model 3.
Constant	0.4323 (1.52)	0.3040 (1.12)	0.3980 (1.52)
CO2	0.2798 *** (11.05)	0.1175 *** (3.07)	0.0942 *** (2.50)
CO2 *	−0.0182 *** (−11.78)	−0.0064 ** (−2.42)	−0.0056 ** (−2.19)
HighTempCO2	-	0.2077 *** (5.24)	0.2439 *** (6.02)
HighTempCO22	-	−0.0163 *** (−5.44)	−0.0189 *** (−6.16)
HighTempSAFCO2	-	-	−0.4144 *** (−4.54)
HighTempSAFCO2	-	-	0.0804 *** (3.95)
LEV	−0.1011 ** (−1.77)	−0.1152 *** (−2.11)	−0.1042 *** (−1.98)
SIZE	−0.0155 (−0.74)	−0.0065 (−0.33)	−0.0135 (−0.71)
GROW	−0.0068 * (−1.67)	0.0021 (0.50)	0.0030 (0.73)
OCF	0.1138 *** (8.77)	0.1546 *** (9.80)	0.1367 *** (8.63)
Continent dummies	Included	Included	Included
Year dummies	Included	Included	Included
F value	18.70 ***	20.19 ***	20.79 ***
R2	0.5907	0.5943	0.6143
n	312	312	312
Panel C. Dynamic Panel-Data Estimation (GMM—Generalized Method of Moments) Results
Variables	Dependent Variable: SP
	Model 1.	Model 2.	Model 3.
Constant	0.4996 *** (3.25)	0.2945 * (2.05)	0.2214 (1.64)
SPt−1	−0.2519 *** (−6.10)	0.7213 *** (11.37)	0.6585 *** (10.58)
SPt−2	0.5522 *** (9.03)	−0.2953 *** (−7.06)	−0.4001 *** (−9.19)
CO2	0.3339 *** (14.01)	0.1497 *** (4.05)	0.1642 *** (5.02)
CO22	−0.0221 *** (−14.64)	−0.0095 ** (−3.83)	−0.0107 *** (−4.90)
HighTempCO2	-	0.1465 *** (4.10)	0.1655 *** (4.81)
HighTempCO22	-	−0.0134 *** (−5.01)	−0.0147 *** (−5.69)
HighTempSAFCO2	-	-	−0.5382 *** (−7.41)
HighTempSAFCO22	-	-	0.1069 *** (7.25)
LEV	−0.1126 ** (−1.95)	−0.0458 (−0.85)	−0.0366 (−0.77)
SIZE	−0.0136 (−1.45)	−0.0051 (−0.57)	0.0042 (0.53)
GROW	−0.0140 *** (−3.57)	−0.0042 (−1.03)	−0.0031 (−0.87)
OCF	−0.0083 (−0.52)	0.0443 ** (2.55)	0.0291 * (1.73)
Continent dummies	Included	Included	Included
Year dummies	Included	Included	Included
AR(2)	p = 0.397	p = 0.736	p = 0.935
Sargan	χ2(5) = 7.36 (p = 0.195)	χ2(3) = 3.93 (p = 0.269)	χ2(2) = 3.45 (p = 0.179)
Hansen	χ2(5) = 3.68 (p = 0.596)	χ2(3) = 1.87 (p = 0.600)	χ2(2) = 0.14 (p = 0.933)
Instruments	30	30	31
n	280	280	280

Note. SP_t−1: Lagged sustainable profitability (t − 1), proxied by return on assets in the previous period, included to capture performance persistence and dynamic adjustment. SP_t−2: Second lag of sustainable profitability (t − 2), included to capture deeper persistence and dynamic adjustment. Other variable definitions are provided in Table 3. t-values are shown in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.

summarizes the regression results. Panel A presents the main findings related to Hypotheses 1–3. Hypothesis 1 is supported. The regression results show a significant positive coefficient for CO₂ intensity (β₁) and a significant negative coefficient for its squared term (β₂), confirming a curvilinear (inverted U-shaped) relationship between emissions and financial performance. This pattern suggests that moderate emissions may reflect economies of scale or operational efficiency, while excessive emissions lead to declining profitability. This non-linear pattern suggests that at moderate levels, higher emissions may be associated with improved profitability—possibly due to increased operational scale or efficiency. However, beyond a certain threshold, further increases in emissions are linked to declining performance. These results imply that the financial implications of carbon emissions depend on contextual factors and are not uniformly harmful or beneficial. They reinforce theoretical perspectives suggesting that environmental efforts can generate both benefits and costs depending on their extent (Barnett & Salomon, 2006; Trumpp & Guenther, 2017; Jin & Xu, 2020). This supports a “balancing strategy” view, in which firms must carefully manage their environmental footprint to avoid adverse financial consequences. The findings also align with the “Too-Little-of-a-Good-Thing” (TLGT) and “Too-Much-of-a-Good-Thing” (TMGT) effects proposed by Trumpp and Guenther (2017), indicating that both underinvestment and over-reliance on carbon-intensive operations can be detrimental. In practice, these insights suggest that airlines may benefit from identifying an optimal emissions level—one that balances short-term operational gains with long-term sustainability concerns. The strategic implication is not to maximize or minimize emissions, but rather to optimize them in light of operational realities, stakeholder expectations, and regulatory pressures.

Among the control variables, leverage (LEV) is negatively and significantly associated with performance, while operating cash flow (OCF) shows a significant positive effect. This suggests that firms with stronger liquidity are more resilient in managing environmental costs.

This result also sets the stage for Hypotheses 2 and 3, which examine how the emissions–performance relationship may vary under different conditions. For example, firms located in regions with more pronounced temperature anomalies may face heightened regulatory scrutiny and stakeholder pressure, potentially shifting the cost–benefit balance of emissions. Hypothesis 2, which introduces regional temperature anomaly data, is well-positioned to test this heterogeneity. Furthermore, these results may carry implications for risk assessment models used by institutional investors, especially those integrating environmental risk factors into valuation frameworks. Identifying a firm’s emission-efficiency inflection point could serve as a key indicator of long-term operational sustainability and financial resilience.

Second, the results for Hypothesis 2 indicate that temperature anomalies moderate the relationship between carbon emissions and financial performance. The main terms—CO₂ intensity and its squared term—remain positive and negative, respectively, consistent with the curvilinear pattern observed in Hypothesis 1. However, the interaction terms reveal that this curvature becomes more pronounced under high-temperature anomaly conditions. Specifically, the coefficient on the interaction between CO₂ and the HighTemp dummy is positive and larger than the main effect, suggesting that in regions experiencing more extreme temperature anomalies, the initial positive impact of emissions on financial performance is stronger—potentially reflecting temporary gains such as increased seasonal demand. Moreover, the coefficient on the squared interaction term remains negative and exceeds the magnitude of the original squared term, suggesting an even steeper drop in performance once emissions surpass a certain threshold. These findings imply that temperature anomalies exert a moderating effect, intensifying both the initial benefits and the subsequent costs of emissions. In contexts with higher climate volatility, the tipping point at which emissions become financially detrimental appears to arrive sooner and with greater severity. This underscores the importance for firms to adopt more proactive and balanced carbon management strategies in light of growing global climate risks—not just in specific regions but as part of a broader response to evolving environmental and stakeholder expectations. These results may also reflect the broader socio-political climate in high-anomaly regions, where public awareness and activism regarding climate change are stronger, creating external pressures for firms to reduce their carbon footprint. These findings underscore the importance of regional climate volatility in shaping firms’ emission-performance dynamics and highlight the role of regulatory and stakeholder pressures in such regions. This may not solely reflect physical climate risks but also institutional responses. Regions with more frequent or severe climate anomalies are likely to experience stronger public awareness, tighter environmental regulations, and greater investor scrutiny. Consequently, firms headquartered in such areas may face heightened expectations for emission control and sustainability disclosure. Thus, the observed moderating effect likely captures not only physical temperature variability but also the broader socio-political context shaping corporate behavior.

These findings also support the growing emphasis on place-based climate risk assessments in corporate strategy and disclosure standards. In particular, investors and regulators may increasingly expect firms headquartered in more vulnerable areas to lead in sustainable innovation and disclosure transparency. Moreover, firms operating in climate-sensitive markets may face differentiated insurance premiums, capital access constraints, or reputational risks depending on the perceived adequacy of their climate response. By understanding how environmental performance interacts with region-specific climate volatility, companies can tailor risk management strategies to protect long-term value creation in diverse regulatory and ecological settings. As with Hypothesis 1, the control variables yield consistent results—leverage remains negatively associated with performance, while operating cash flow shows a positive and significant relationship.

The extended model for Hypothesis 3 incorporates a three-way interaction term between carbon emissions, temperature anomalies, and the adoption of Sustainable Aviation Fuel (SAF). The base effects of CO₂ and its square remain consistent with previous models, showing a positive and then negative relationship with financial performance. This pattern suggests that emissions initially contribute to profitability—possibly through economies of scale or higher operational intensity—but become detrimental beyond a certain threshold due to increasing regulatory burdens, cost inefficiencies, or reputational risks. These effects are further amplified under high temperature anomalies, consistent with Hypothesis 2. However, the three-way interaction terms reveal a distinct moderating effect of SAF adoption. Specifically, the coefficient on the HighTemp × SAF × CO₂ term is negative, while that on the HighTemp × SAF × CO₂² term is positive—both statistically significant and opposite in sign to the two-way interaction terms. This indicates that under climate stress, SAF adoption dampens both the short-term benefits and long-term costs of emissions. In effect, SAF flattens the curvature of the emissions–performance relationship, reducing financial volatility in both extremes. The findings suggest that adopting SAF can mitigate the adverse financial effects of carbon emissions under heightened climate-related risk. Although SAF adoption entails substantial costs—such as fuel price premiums, infrastructure upgrades, and procurement complexity—it appears to buffer profitability risks in environmentally vulnerable regions. This supports the emerging view that decarbonization efforts are not merely compliance tools but essential components of long-term financial strategy. These results align with prior research showing a positive link between financial performance and proactive environmental strategies, including emissions reduction and innovation investments (Nichita et al., 2021; Aydıngülü Sakalsız & Özçelik, 2024). More broadly, actions such as improving energy efficiency, waste management, and carbon disclosure can help firms lower operational costs and enhance long-term value creation (Duque-Grisales et al., 2020). This evidence reinforces the strategic importance of embedding sustainability in core business models. As prior literature suggests (Porter & van der Linde, 1995), green technologies like SAF can act as buffers against climate-related policy shocks and stakeholder scrutiny. Within the aviation sector, recent studies have shown that while SAF may not guarantee immediate financial returns, it contributes to organizational resilience by hedging against evolving regulatory pressures and investor expectations (Kuo et al., 2021; Oestreich & Tsiakas, 2024).

These insights may also guide future regulatory and policy frameworks that incentivize SAF infrastructure and adoption, particularly in high-risk regions. For firms, the results highlight that climate risk management must be integrated into strategic planning—not just as a matter of environmental compliance or corporate image, but as a determinant of financial sustainability. Early adopters of SAF and other low-carbon innovations may benefit from improved ESG ratings, stronger stakeholder trust, and greater access to climate-linked capital. Furthermore, in markets increasingly exposed to climate anomalies, proactive decarbonization investments can shield firms from abrupt profitability shocks, serving as a financial stabilizer over time. Consistent with previous models, the control variables yield similar results: leverage continues to show a significant negative association with sustainable performance, while operating cash flow remains positively associated. These findings underscore that firms with greater financial flexibility are better positioned to pursue sustainability transitions while preserving economic viability.

As a robustness check, firm and year fixed-effects regressions were estimated for the three models corresponding to Hypotheses 1 through 3. Standard errors were clustered at the airline level to account for within-firm serial correlation and heteroskedasticity. The results, reported in Table 5 Panel B, remain qualitatively consistent with the baseline OLS estimates, with the key explanatory variables preserving their signs and statistical significance. These findings reinforce the robustness of the results by controlling for unobserved firm-specific and year-specific heterogeneity.

Panel C of Table 5 presents the dynamic GMM estimations for all three hypotheses (H1–H3). The lagged dependent variables are statistically significant, confirming persistence and dynamic adjustment in sustainable profitability. In Model 1, carbon emissions exhibit a nonlinear pattern, with the linear term positive and the squared term negative, confirming an inverted-U relationship. This supports Hypothesis 1, namely that moderate emissions enhance performance, whereas excessive emissions reduce it. Among the controls, leverage is negatively related to performance, while growth is positively associated with performance. Year and continent dummies are included to capture time-specific and regional heterogeneity. The diagnostic tests further support the specification. The Arellano–Bond test shows no evidence of second-order serial correlation (AR(2), p = 0.397). Both the Sargan test (p = 0.195) and the Hansen test (p = 0.596) confirm the validity of the instrument set, alleviating concerns about instrument proliferation.

Model 2 extends the baseline model by incorporating the moderating role of temperature anomalies (Hypothesis 2). The inverted-U shape of the emissions–performance nexus persists, but the interaction terms reveal that climate stress amplifies this relationship. Specifically, HighTemp × CO₂ is positive and significant, whereas HighTemp × CO₂² is negative and significant, suggesting that in periods or regions with more severe temperature anomalies, the short-term performance benefits of emissions are stronger, but the turning point is reached earlier and the subsequent decline is steeper. Among the controls, operating cash flow is positively associated with performance. Year and continent dummies are again included. The diagnostic tests confirm robustness: the Arellano–Bond test shows no evidence of second-order autocorrelation (AR(2), p = 0.736), and both the Sargan (p = 0.269) and Hansen (p = 0.600) tests fall within the acceptable range, supporting the validity of the instruments. Taken together, these results provide strong support for Hypothesis 2, showing that the nonlinear emissions–performance relationship becomes more pronounced under climate stress conditions.

Model 3 further incorporates Sustainable Aviation Fuel (SAF) adoption as an additional moderator alongside temperature anomalies (Hypothesis 3). The interaction terms show that SAF modifies the curvature: HighTemp × SAF × CO₂ is negative and HighTemp × SAF × CO₂² is positive, implying that SAF adoption dampens the initial benefits but reduces the severity of the decline, effectively flattening the inverted-U relationship. Among the controls, operating cash flow remains positive, with year and continent dummies included. The specification passes all diagnostic checks: the Arellano–Bond test shows no evidence of second-order autocorrelation (AR(2), p = 0.935), and both the Sargan (p = 0.179) and Hansen (p = 0.933) tests indicate instrument validity. These results support Hypothesis 3, indicating that SAF adoption sharpens and stabilizes the emissions–performance nexus under climate stress.

To account for potential heterogeneity across airline business models, we included a dummy variable indicating low-cost carrier (LCC) status in the regression model (

Table 6. OLS Regression Results with LCC Dummy Variable as a Control for Business Model Heterogeneity.

Variables	Dependent Variable: SP
	Model 1.	Model 2.	Model 3.
Constant	0.1577 *** (3.75)	0.1388 *** (3.39)	0.1428 *** (3.71)
CO2	0.2419 *** (12.47)	0.0921 *** (2.48)	0.0856 *** (2.44)
CO22	−0.0163 *** (−12.54)	−0.0057 ** (−2.22)	−0.0053 ** (−2.17)
HighTempCO2	-	0.1695 *** (4.46)	0.2421 *** (6.21)
HighTempCO22	-	−0.0134 *** (−4.68)	−0.0188 *** (−6.37)
HighTempSAFCO2	-	-	−0.4542 *** (−5.85)
HighTempSAFCO2	-	-	0.0874 *** (4.78)
LEV	−0.0693 ** (−1.92)	−0.0702 *** (−2.01)	−0.0754 *** (−2.30)
SIZE	0.0006 * (0.31)	0.0008 (0.43)	0.0006 (0.32)
GROW	−0.0057 (−1.45)	0.0018 (0.43)	0.0029 (0.76)
OCF	0.1027 *** (8.71)	0.1340 *** (9.19)	0.1326 *** (9.46)
LCC	−0.0015 (−1.47)	−0.0057 (−0.30)	−0.0082 (−0.46)
Continent dummies	Included	Included	Included
Year dummies	Included	Included	Included
F value	56.33 ***	56.85 ***	61.48 ***
Adjusted R2	0.8164	0.8290	0.8494
n	312	312	312

Note. LCC = 1 for low-cost carriers and 0 otherwise. Only five airlines in the sample are classified as LCCs—four Korean carriers and one European carrier. The inclusion of Korean LCCs was facilitated by the availability of standardized emissions data through the National GHGs Management System (NGMS), while the European LCC was included due to its transparent public disclosure of environmental data. Other variable definitions are provided in Table 3. t-values are shown in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.

). The coefficients remain largely unchanged, reinforcing the robustness of our findings.

5. Conclusions

Climate change has emerged as a defining global crisis, extending beyond environmental concerns to pose profound threats to human well-being, institutional resilience, and economic stability. Among the primary drivers is the increasing concentration of carbon dioxide (CO₂) emissions, particularly since the industrial revolution. In response, the global community has embraced ambitious targets—such as carbon neutrality—requiring widespread adoption of renewable energy, systemic industrial transformation, and significant changes in corporate behavior. However, the pathway toward these goals is neither linear nor costless, as it demands complex technological adaptation and imposes near-term economic trade-offs. Although multiple industries contribute to carbon emissions, the airline sector has come under particular scrutiny. While it accounts for approximately 2.5% of global CO₂ emissions, aviation is estimated to contribute nearly 4% of net climate effects when non-CO₂ impacts are considered (Transport & Environment, 2025). The sector’s carbon intensity, global visibility, and operational rigidity make decarbonization especially difficult. With limited technological substitutes, long fleet replacement cycles, and volatile fuel markets, airlines face mounting regulatory and stakeholder pressure to reduce emissions—even as their ability to do so remains constrained. This tension has inspired growing scholarly interest in understanding how carbon emissions influence firm-level financial performance, particularly within the aviation industry. The prior literature reflects three main perspectives. The first highlights a positive relationship, rooted in the win–win hypothesis, resource-based view, and stakeholder theory, which suggest that environmental practices can boost efficiency, drive innovation, and enhance reputation, thereby contributing to long-term competitiveness (Porter & van der Linde, 1995; Hart & Ahuja, 1996; Busch & Hoffmann, 2011). The second emphasizes a negative association, citing trade-offs, compliance costs, and managerial opportunism as reasons why environmental initiatives may hinder profitability in the short term (King & Lenox, 2001; Palmer et al., 1995; Chan et al., 2013). The third stream proposes a curvilinear relationship, where both underinvestment and overinvestment in environmental strategies reduce financial returns—implying that optimal engagement lies somewhere in between (Barnett & Salomon, 2006; Trumpp & Guenther, 2017; Jin & Xu, 2020). These theoretical perspectives are especially salient in high-emission, heavily regulated sectors such as aviation, energy, and manufacturing (Iwata & Okada, 2011; Ganda & Milondzo, 2018; Oestreich & Tsiakas, 2024).

Building on these theoretical foundations, this study empirically examined how carbon emissions relate to financial performance in the global airline industry. Several critical findings emerge. First, the results support the existence of a curvilinear relationship between CO₂ emissions and financial performance. While the linear term for emissions shows a positive association—likely reflecting economies of scale and activity-based revenue generation—the squared term is significantly negative. This finding is consistent with the “Too-Little-of-a-Good-Thing” (TLGT) and “Too-Much-of-a-Good-Thing” (TMGT) frameworks (Trumpp & Guenther, 2017), which suggest that both insufficient and excessive environmental engagement can adversely affect firm outcomes. The implication is that airlines must strategically manage their carbon footprint—not simply to minimize emissions, but to optimize environmental investment relative to their operational context. Second, the study finds that temperature anomalies moderate the emissions–performance relationship. In regions experiencing more severe climate deviations, the financial effects of carbon emissions are intensified. This suggests that climate volatility amplifies both the short-term benefits and long-term risks associated with emissions, likely due to localized policy responses and heightened stakeholder awareness. These results underscore the importance of incorporating regional climate dynamics into strategic carbon management. Third, the adoption of Sustainable Aviation Fuel (SAF) emerges as a meaningful moderating factor. Although SAF is currently more expensive and less widely available than conventional jet fuel, its use mitigates the negative financial effects of emissions under heightened climate stress. This aligns with previous findings that link green innovation to improved financial resilience (Nichita et al., 2021; Aydıngülü Sakalsız & Özçelik, 2024). The results suggest that firms investing in SAF and related technologies may reduce exposure to climate-driven regulatory or reputational shocks. Moreover, early adopters of SAF may enhance their competitive positioning by signaling long-term strategic orientation to investors, regulators, and customers alike (Kuo et al., 2021; Oestreich & Tsiakas, 2024).

Together, these findings point to a dynamic and contingent relationship between environmental and financial performance—one that is shaped by emissions levels, external climate variability, and organizational adaptation strategies. For firms in the airline industry, these insights support a shift from generic carbon reduction targets to more tailored, context-sensitive sustainability planning. Nonetheless, the study is not without limitations. A primary constraint lies in the availability and consistency of CO₂ emissions data. In the absence of a centralized global database, emissions figures were obtained from the National GHGs Management System for Korean airlines and manually compiled from corporate sustainability and annual reports for non-Korean airlines. While this approach ensured broad coverage, differences in the scope and detail of reported data introduced some measurement heterogeneity. To address this, all figures were standardized to metric tons of CO₂ equivalent (tCO₂e) and recalculated where necessary using reported operational metrics, in order to maximize consistency across firm-year observations. As carbon disclosure standards become more harmonized and comprehensive, and as the adoption of Sustainable Aviation Fuel (SAF) becomes more widespread and consistently reported, future research will be able to build on these findings using more standardized datasets. In particular, further studies could investigate the financial implications of various emission reduction strategies—such as carbon offsetting, route optimization, fleet electrification, and SAF adoption—to deepen understanding of how environmental initiatives translate into financial outcomes.

Although this study provides meaningful empirical insights, it is important to acknowledge several data-related limitations. A primary challenge is the absence of a unified global database for airline-level CO₂ emissions. Accordingly, this study relied on government-verified NGMS data for Korean airlines and manually collected disclosures for non-Korean airlines. Because firms differ in reporting frequency, scope coverage, and methodological detail, some heterogeneity is unavoidable. To mitigate this issue, all emissions figures were standardized to metric tons of CO₂ equivalent (tCO₂e), and recalculations were performed when airlines reported emissions in intensity-based units such as per RPK.

Regarding sample selection, the final sample of 29 airlines reflects the need for consistent and sufficiently long emissions and financial disclosures. Airlines were included if they reported firm-level CO₂ emissions for multiple consecutive years and provided the financial information required for model estimation. Airlines were excluded when emissions were disclosed only intermittently, reported only at the national or industry-association level, or when financial variables were incomplete. This approach ensured analytical rigor but may limit generalizability to carriers with more established reporting systems. As disclosure frameworks become more standardized and global datasets (e.g., Trucost, airline sustainability benchmarking platforms) expand in coverage, future studies will be able to validate and extend these findings with broader samples and more granular data.

Beyond these limitations, the findings also carry important managerial and policy implications. Airlines can operationalize the curvilinear relationship by pursuing efficiency improvements—such as fuel optimization, fleet modernization, and adaptive scheduling under climate stress—while avoiding both underinvestment and overinvestment in carbon initiatives. Managers should view SAF not as a cost burden alone but as a strategic hedge against future regulatory and reputational risks. From a policy perspective, our results suggest that governments should design SAF incentives that lower adoption costs and reduce financial risk exposure. Subsidies, tax credits, and infrastructure investment can encourage airlines to scale SAF usage in a financially sustainable way, ensuring that environmental objectives and profitability are not mutually exclusive. In particular, future studies could further explore the financial implications of alternative strategies—such as carbon offsetting, route optimization, and emerging low-carbon technologies—to deepen understanding of how airlines and policymakers can jointly balance sustainability with profitability. We hope these insights not only contribute to academic debates but also provide actionable guidance for managers and policymakers navigating the twin challenges of profitability and decarbonization in aviation. Ultimately, this study underscores that environmental and financial goals, often seen as competing, can be reconciled through context-sensitive strategies and supportive policy frameworks.