This article investigates how climate-related financial risks are addressed in statutory audit reporting through the lens of Key Audit Matters (KAMs), with a focus on those explicitly referencing climate concerns (C-KAMs). The study explores whether and how these disclosures reflect a meaningful engagement with complex and multi-layered climate-related materiality.
The study applies a mixed-method approach combining descriptive sorting and categorization with interpretive textual analysis. The sample comprises 678 C-KAMs identified from 80,628 total KAMs disclosed in European auditor reports between 2015 and 2023. Descriptive analyses were conducted using Excel-based filtering and cross-tabulation. Finally, audit procedures and communicative tone were qualitatively coded and interpreted from the C-KAMs.
C-KAMs have begun to emerge across European audit reports, but in a markedly uneven and cautious fashion. They are concentrated in a narrow set of high-carbon and infrastructure sectors, most notably utilities, oil and gas, mining and selected manufacturing and real estate firms, while sectors such as finance, insurance and agriculture are almost absent. Where C-KAMs are reported, climate-related risks are predominantly translated into familiar accounting containers such as impairment testing, provisions and valuation estimates, expressed in generic, neutral-to-negative language and supported by limited procedural detail. This pattern suggests that current practice often fits climate uncertainty into conventional accounting containers rather than engaging deeper with its strategic and system-level dimensions.
The study is limited to publicly disclosed C-KAMs in European auditor reports and therefore cannot capture the internal deliberations, omitted climate concerns or firm-level judgments that shape whether climate issues become reportable audit matters. The interpretation of tone and procedural depth also remains dependent on textual disclosures rather than underlying audit files. These limitations nonetheless underscore an important implication: current audit reporting appears to render climate risk visible only when it can be translated into established financial statement categories. This highlights the need for further research on how audit practice may accommodate systemic, long-horizon and strategically diffuse climate-related uncertainties.
For practitioners, the findings suggest that climate-related audit reporting remains narrowly anchored in conventional accounting estimates such as impairment, provisions and valuation, which may understate broader strategic and systemic exposures. Auditors may therefore need more explicit guidance on when and how climate issues should be reflected in KAM reporting, especially beyond high-carbon sectors. For regulators and standard setters, the results indicate a need to clarify expectations around climate-related materiality in audit communication. For preparers and audit committees, the study highlights the importance of ensuring that climate-related assumptions and financial statement effects are documented in a way that supports transparent audit scrutiny.
This is among the first studies to systematically examine C-KAMs as distinct research subjects. The findings contribute to a growing understanding of how audit practices engage with climate risks and highlight emerging tensions in evidencing long-horizon, uncertain phenomena within existing financial audit frameworks. Our research offers a novel perspective on C-KAMs by examining them not merely as technical disclosures, but as sites where auditors negotiate financial materiality in the face of climate uncertainty.
1. Introduction
According to the Global Risks Report (2024) published by the World Economic Forum, extreme weather events are ranked among the most likely and severe risks for businesses, exhausting financial resources to mitigate and adapt to climate change. Heightened public awareness of climate change and its related risks has also prompted governments and regulatory bodies to take measures against their escalation and foster a gradual transition toward a decarbonized economy. Examples include the Sustainable Finance Disclosure Regulation or the EU Taxonomy, both of which aim to steer capital flows toward sustainable activities but simultaneously introduce new compliance burdens and strategic uncertainties for firms, particularly those with carbon-intensive business models (Stroebel and Wurgler, 2021). As a result, companies are increasingly exposed to both physical risks and transition risks associated with climate change, each carrying potentially material financial implications (TCFD, 2017, 2021). Physical risks may be acute, such as floods, hurricanes, drought or wildfires – extreme weather events with increased severity due to human-induced climate change, or chronic, such as sustained higher temperatures that lead to chronic heatwaves or sea level rise. Such physical risks may have detrimental effects on business activities, for instance, disrupted logistics, unstable earnings, volatile cash flows and lower asset prices (Addoum et al., 2023; Huang et al., 2017; Baldauf et al., 2020; Ortiz-Bobea et al., 2020; Karydas and Xepapadeas, 2022). In contrast, transition risks emerge from the societal, regulatory and technological shifts required to move toward a low-carbon economy, potentially affecting asset valuations, business models and access to capital (Reboredo and Ugolini, 2022; Huang et al., 2023).
With widespread recognition of the material consequences of climate-related risks for business operations, investors have intensified their demands for more comprehensive climate-related disclosures and their strong connectivity to financial disclosures (Siddique et al., 2023; Octavio and Setiawan, 2025; O'Dwyer and Unerman, 2020; Lombardi et al., 2022). At the global level, the TCFD (2017) provides the first structured guidance on integrating climate considerations into mainstream financial reporting, notably emphasizing the use of scenario analysis to assess the resilience of business strategies under different climate scenarios (O'Dwyer and Unerman, 2020; Lehner et al., 2023). This approach not only enhances transparency but also improves the capacity of capital markets to evaluate firms' exposure to climate-related financial risks, thereby facilitating the incorporation of climate-related risks into investment and lending decisions (Amel-Zadeh and Serafeim, 2018; TCFD, 2017, 2021). In 2022, TCFD-aligned disclosures became mandatory in the United Kingdom for certain large companies, making it the first major economy to require climate-related reporting in line with TCFD recommendations.
Within the EU, the TCFD has informed the evolving sustainability reporting architecture, from the early Non-Financial Reporting Directive to the recent Corporate Sustainability Reporting Directive/European Sustainability Reporting Standards. However, unlike the TCFD's sole focus on financial materiality, these EU frameworks adopt a double-materiality perspective, emphasizing both financial and impact materiality. As a result, they offer comparatively less detailed guidance on the assessment, measurement and disclosure of climate-related financial risks than the TCFD provides. In parallel, large French companies have long been required to disclose environmental information, including climate-related risks, under the Grenelle II law in 2010 and the Energy Transition Law in 2015 (Kaya, 2016). Although these French laws focus on impact materiality, climate-related impacts with financial implications ultimately enter the scope of financial materiality and thus must be considered during the statutory financial audit process (Simnett and Huggins, 2015; AMF, 2021). Across these regulatory developments, auditors increasingly operate in a setting where climate-related assumptions, scenarios and valuation effects are expected to be reflected not only in sustainability statements but also in the financial statements themselves. Under ISA 720, auditors must read and consider such disclosures as other information, remaining alert to potential implications for the financial statements (FRC, 2022).
Nevertheless, climate-related financial disclosures present challenges for both preparers and auditors (Harrer and Lehner, 2024). Such disclosures rest on forward-looking information based on probabilities, spanning short and long-term horizons, high uncertainty of assumptions, systemic interdependencies and particularly the inevitable use of external (partly even unverified) data sources and climate models (TCFD, 2017, 2021; Lehner et al., 2023). Many forward-looking judgments, for example, about stranded assets, carbon pricing and environmental liabilities, affect a range of areas subject to financial audit scrutiny, such as impairments, provisions and asset valuations (Lehner et al., 2023). What is more, the financial effects of complex climate-related risks are required to be quantified in monetary amounts (TCFD, 2017, 2021), entailing a great deal of management judgment and estimation. As a result, the risk of making material misstatements or omissions is high.
The tightening policy regime, together with the highly complex nature of climate risks, provides a broader institutional context within which Key Audit Matters (KAMs) under ISA 701 emerge as promising communicative devices signaling how auditors address climate-related financial risks. KAMs are designed to highlight the most significant audit risks and the corresponding auditor responses, offering stakeholders insights into areas of complex judgments and sources of estimation uncertainty (Simnett and Huggins, 2014; Pinto et al., 2020). However, prior empirical research demonstrates that KAM disclosures vary widely in specificity, readability and communicative value, often becoming generic or boilerplate in judgment-intensive areas (Pinto and Morais, 2018; Velte and Issa, 2019; Sierra-García et al., 2019). Studies further show that KAMs are shaped by contextual factors, such as industry regulation, client risk-profiles and national audit traditions (Pinto and Morais, 2018; Sierra-García et al., 2019). Moreover, KAMs can also shape auditors' behavior, sometimes prompting greater diligence but also creating risks of reduced professional skepticism (Asbahr and Ruhnke, 2019; Ratzinger-Sakel and Theis, 2019). Yet, no study to date has examined how these dynamics unfold when climate-related uncertainties become material to the financial audit, where long-term assumptions, regulatory shifts and complex estimates challenge the production of decision-useful audit reports (Harrer and Lehner, 2024). This leaves a gap in understanding how auditors integrate climate-related considerations into the statutory financial audit process and communicate them through KAMs.
Therefore, the aim of this study is to provide a systematic empirical examination of how climate-related financial risks are rendered auditable through Climate-related Key Audit Matters (C-KAMs) in European financial audit reports, and what this reveals about the possibilities and limits of current practice. We organize this inquiry around one overarching research question and three guiding sub-questions:
Overarching research question:
How do European auditors use C-KAMs to make climate-related financial risks visible, and what tensions or limitations characterize this emerging practice?
Sub-questions:
How are C-KAMs distributed across financial statement themes, sectors and regions, and what patterns of concentration or absence does this reveal?
How do auditors describe the audit procedures used to address climate-related financial risks within C-KAMs, particularly with regard to the scope and depth of those procedures?
How do auditors frame and communicate climate-related risks within C-KAMs, in terms of tone, clarity and emphasis, and how does this framing shape the portrayal of climate uncertainty in the audit report?
Drawing on a dataset of 80,628 KAMs disclosed by European publicly listed entities from 2015 to 2023, we identify 678 C-KAMs through systematic keyword filtering based on the SBTi® climate lexicon with additional keywords on physical climate risks. Using Python-based textual analysis together with human validation, we analyze their disclosure patterns and trends, procedural detail and communicative setup to shed light on how auditors currently navigate the emerging terrain of climate-related financial auditing.
The remainder of the article is structured as follows. First, we outline the relevant literature on KAMs and climate-related disclosure assurance. Second, we describe the data, sample construction and methodological approach. Third, we present the findings on the distribution of C-KAMs, the audit procedures disclosed and their communicative tone. Fourth, we discuss these results in relation to prior research and emerging assurance practices. Finally, we conclude by outlining the study's contributions, implications and avenues for future research.
2. Theoretical background
KAMs under ISA 701 generally aim to enhance audit transparency (Coram and Wang, 2021; Green and Li, 2011; Dwyer et al., 2023; Minutti-Meza, 2021) by providing more detailed insights into the financial audit process and clarifying audit limitations, thus enhancing transparency and drawing readers' attention to critical areas with significant auditor judgments (Pinto and Morais, 2018; Pinto et al., 2020; Rezaee and Homayoun, 2025). The disclosure quality of KAMs plays an important role in shaping stakeholder perceptions of KAMs' communicative values and usefulness (Pinto and Morais, 2018; Sirois et al., 2018; Mwintome et al., 2023). As Velte and Issa (2019) pointed out, if KAMs are too generic, they may not enhance transparency as intended, but instead increase stakeholder skepticism (Simnett and Huggins, 2014), especially in the context of complex climate-related risks where detailed insights and adapted procedures are crucial (Harrer and Lehner, 2024). As another potential communicative pitfall, Sirois et al. (2018) caution that while KAMs increase user focus on specific risks, they may also lead investors to overlook other important financial disclosures. Yet, the extent of transparency and readability of KAMs may also be perceived differently among professional and nonprofessional investors (Köhler et al., 2020). For C-KAMs, these dynamics imply that clarity and readability become particularly critical, as climate-related risks are often unfamiliar, forward-looking and difficult for financial statement users to interpret without accessible audit explanations (Harrer and Lehner, 2024; Fan et al., 2021).
Regarding the potential influence of KAMs on auditor behavior, Asbahr and Ruhnke (2019) illustrate that the anticipation of KAM disclosures leads auditors to exercise heightened caution, particularly in areas where complex risks require high levels of professional judgment. In the same vein, Ma et al. (2024) find that KAMs have the potential to incentivize auditors to adhere more closely to professional standards, enhance audit quality through increased diligence and provide transparent documentation of procedures undertaken in areas of significant audit risk. For C-KAMs, such heightened scrutiny should, at least in principle, translate into more substantive response disclosures, with clearer explanations of the procedures performed and the judgments applied to climate-related assumptions (Dutta and Dutta, 2021), although whether this occurs in practice remains an open empirical question.
Nevertheless, KAMs may also produce unintended consequences, such as a “judgment-action gap,” where auditors reduce their overall professional skepticism after identifying KAMs, viewing them as sufficient disclosure of risk (Asbahr and Ruhnke, 2019). Ratzinger-Sakel and Theis (2019) attribute this effect to moral licensing, wherein KAMs may justify less rigorous assessments of certain client-preferred treatments. Bepari et al. (2024) furthermore emphasize that while KAMs generally improve audit quality, they also risk masking significant issues if presented in a standardized format. This dynamic suggests that while KAMs foster transparency, auditors must carefully balance openness with sustained critical assessment, particularly given the complex nature of climate-related financial disclosures. In the context of C-KAMs, these behavioral effects may be amplified, as auditors navigate high uncertainty and limited evidence, creating risks that climate-related matters are either over-disclosed without substantive action or under-examined despite their potential materiality (Junior et al., 2013).
The regulatory environment also significantly influences how KAMs are disclosed and perceived. Pinto and Morais (2018) show that both the precision of accounting standards and the level of industry regulation shape KAM reporting, with stricter regulatory frameworks often resulting in more detailed KAMs. In highly regulated industries like banking, for instance, auditors may disclose fewer KAMs due to the perceived lower risk in these sectors or because existing regulations already address key concerns. Sierra-García et al. (2019) find that client characteristics further influence KAM reporting, with larger, riskier clients prompting more extensive KAM disclosures. This suggests that C-KAMs may show similar heterogeneity, with sector characteristics, regulatory expectations and national audit environments shaping whether and how deeply climate-related risks are highlighted in audit reports. Our study context at the early stage of climate-related assurance regulations, when the existing audit guidance is mostly generic and geographically uneven, further provides an interesting setting to see how such heterogeneity unfolds.
Taken together, these insights offer a useful starting point for analyzing C-KAMs. However, climate-related financial risks introduce additional epistemic challenges that may amplify known limitations of KAM reporting. Climate risks are inherently forward-looking, model-dependent and characterized by indeterminacy, long time horizons and reliance on external assumptions (TCFD, 2017; O'Dwyer and Unerman, 2020; Harrer and Lehner, 2024). These features make the evidence underlying climate judgments less verifiable and more assumption-sensitive than in most traditional audit areas. As a result, tensions already identified in the KAM literature, between transparency and vagueness or between heightened judgment and the potential for moral licensing, are likely to be more pronounced when auditors address climate-related matters. Accordingly, it cannot be taken for granted that C-KAMs simply replicate traditional KAM formats, thus raising questions about how effectively auditors can convey judgment under climate uncertainty within the constraints of the expanded audit report. This provides a theoretical lens for interpreting C-KAMs as disclosures produced under heightened uncertainty, where auditors must translate complex climate assumptions into concise communicative artifacts.
Against this backdrop, our study builds on prior KAM research by examining how these general communicative and behavioral dynamics manifest in the emerging context of climate-related audit risks. Given the absence of empirical work specifically investigating C-KAMs, our analysis provides initial evidence on how auditors navigate these amplified uncertainties in practice, and whether existing KAM mechanisms are capable of conveying decision-useful information when judgments depend on long-range scenarios, climate models and regulatory forecasts. By focusing on C-KAMs as distinct research objects, our study also responds to recent calls for more rigorous, transparent and decision-useful climate-related disclosures within mainstream audit practices (Pham et al., 2024).
3. Methodology
3.1 Sample selection and justification
The empirical analysis in this study leverages a comprehensive dataset of 80,628 KAMs, extracted from statutory financial audit reports of publicly listed companies across Europe between 2015 and 2023. The data is accessed via the Ideagen® Audit Analytics database. From this, 803 KAMs with explicit reference to climate-risks (C-KAMs) are identified using systematic keyword filtering. The keywords are extracted from the Science Based Targets Initiative (SBTi) Glossary (SBTi, 2024) and comprise terms such as “climate,” “carbon,” “CO2,” “emission,” “net-zero,” “ETS” (EU Emission Trading Scheme), “transition risk,” “physical risk” or “global warming.” In addition to the keyword list derived from the SBTi glossary, we implemented an iterative, snowball-style refinement of the filter to mitigate the risk of overlooking C-KAMs that emphasize physical rather than transition risks. Starting from the initial matches, we manually reviewed C-KAMs in sectors with elevated exposure to physical climate risks and recorded further recurrent terms such as “flood,” “drought,” “storm,” “heatwave” (primed by our TCFD experience), which were then added to the search dictionary and rerun across the full dataset. This procedure reduced, but cannot fully eliminate, the possibility of false negatives, as climate-related physical risks may still be described in more implicit language that does not contain any of the predefined or subsequently snowballed terms; we therefore treat our sample as comprehensive for explicit C-KAM references to climate-related risks, while acknowledging that some more obliquely phrased cases may remain outside our scope.
Then, second, we apply manual filtering by reading and interpreting the filtered KAMs to eliminate those that contain the key terms but do not actually deal with climate-related audit risks, for example dealing with “economic climate” or “geopolitical climate.” The final subset contains 678 C-KAMs selected for further analysis.
The selection of European listed entities has been strategically chosen to represent a region characterized by both an overall strong regulatory drive toward climate-related disclosures and yet considerable cross-national variance in specific regulatory conditions (Pinto and Morais, 2018; Mihret et al., 2021). This provides an analytically rich context to explore auditor responses to climate-related financial risks. The timeframe captures critical regulatory developments, including the implementation of mandatory TCFD reporting in the United Kingdom from 2022, thus allowing for comparative analysis.
3.2 Descriptive analysis approach
The initial descriptive analysis has been conducted using Microsoft Excel, leveraging data filtering and sorting functionalities to identify general distribution patterns, including temporal trends, sectoral frequency and regional variance of C-KAM disclosures.
This stage involves calculating relative frequencies, identifying dominant themes and conducting cross-tabulations between sectors, regions and disclosure years to ascertain the distributional characteristics and allow insights into the evolution of C-KAM practices. Cross-tabulation analysis is used to assess sectoral and regional concentrations and temporal changes pre- and post-implementation of key regulatory measures such as mandatory TCFD-aligned reporting.
3.3 Interpretive analysis procedure
For the next, interpretive phase, a structured coding framework has been applied using qualitative content analysis assisted by Python-based textual analysis. We employed Python 3.11 within a Jupyter® 6 Notebook environment to conduct textual analysis and categorization of C-KAMs. The interactive coding process involved iteratively validating Python-generated classifications against our manually coded subsets, thus ensuring the robustness and interpretive accuracy of our automated text analysis. This analysis focuses on two main dimensions: (1) audit procedures and their depth of detail, and (2) communicative tone within C-KAM disclosures (please see details in Appendix A).
The inductive coding of audit procedures categorizes audit methodologies and assumptions referenced in the text, such as sensitivity analyses, scenario testing or engagement with external experts. Python tools, particularly the SpaCy and NLTK libraries, facilitated this textual classification by identifying recurring procedural phrases and contextually categorizing them into high, medium or low depth based on explicitness and specificity of disclosed procedures. The second dimension, communicative tone, utilized Python-assisted sentiment analysis techniques tailored specifically to financial and sustainability texts. This included established lexicon-based analysis (Loughran and McDonald, 2011), adapted for capturing positivity, negativity and neutrality, reflecting how auditors linguistically framed climate risks and opportunities. The Loughran-McDonald Financial Sentiment Dictionary was explicitly developed by finance scholars to reflect the semantic characteristics of accounting, finance and regulatory discourse.
In our study, this lexicon has been implemented using Python's NLTK library, through custom dictionaries and token-matching routines. Each C-KAM is tokenized into individual words, and then those tokens are matched against the Loughran and Mcdonald (2011) word list to score the prevalence of different sentiment categories. This has been particularly effective in distinguishing audit texts that discussed climate risks in terms of impairments and litigation from those that framed decarbonization investments or sustainability transitions in more forward-looking or positive ways.
The precision of the Loughran-McDonald method further lies in its alignment with financial discourse, helping avoid false positives that would arise from applying generic sentiment models. For example, if a C-KAM contains the word “uncertainty,” a general sentiment model may overemphasize negativity, while the Loughran–McDonald model correctly classifies it under a distinct risk probability category, allowing for more refined tonal diagnostics.
Both dimensions have been iteratively refined, where initial automated classifications have been manually validated by researchers, ensuring interpretative accuracy and theoretical coherence. This combined automated-human interpretive approach is adopted to maintain analytical rigor while accommodating the complexity and variability inherent in textual disclosures (Channuntapipat et al., 2020; Harrer and Lehner, 2024).
3.4 Validation and intercoder reliability
To assure the robustness of qualitative interpretations, an intercoder reliability assessment has been conducted. Following established methodological guidelines (Krippendorff, 2018), two independent coders analyze a randomly selected subset (approximately 10%) of the identified C-KAMs. Cohen's kappa is calculated to statistically assess intercoder agreement levels across both procedural and communicative tone categorizations, resulting in high intercoder agreement scores (κ > 0.80), indicating strong reliability. Discrepancies identified through this validation process have been reviewed jointly by coders and resolved through consensus, ensuring analytical consistency and reducing potential coder biases (Neuendorf, 2017). Appendix B shows a flowchart (Figure A1) of our research design.
4. Findings
In what follows, we present a neutral, empirically grounded account of the patterns that emerge from the 678 C-KAMs identified in our sample. The focus in this section is on describing how climate-related financial risks are currently distributed across financial statement themes, sectors and regions, and how auditors report related procedures and communicative choices, based on the methods outlined above. We first map the six most frequent financial themes in which climate-related risks are articulated, before turning to sectoral and country-level variation and then to the reported audit procedures and linguistic tone. At this stage, we deliberately refrain from offering an extended interpretation; this, however, will be taken up in the discussion.
Our analysis of 678 C-KAMs reveals six recurring themes in which climate-related risks significantly impact the auditor's assessment. These themes jointly account for 490 C-KAMs, or approximately 72% of the filtered sample, indicating that a large share of climate-related audit disclosures concentrate on a limited set of risk-sensitive financial areas (see Table 1).
Frequency of main themes in the climate-related Key Audit Matters
| No. | Theme | C-KAMs |
|---|---|---|
| 1 | Impairment and valuation of goodwill | 176 |
| 2 | Impairment and valuation of PPE | 173 |
| 3 | Estimation of hydrocarbon reserves | 118 |
| 4 | Provisions for environmental obligations and litigation | 56 |
| 5 | Valuation of investment property and other investments | 52 |
| 6 | Going concern | 34 |
The most prevalent themes relate to asset impairments, particularly in the valuation of goodwill (No.1) and property, plant and equipment (PPE) (No.2). Auditors frequently cite climate risks, such as extreme weather, regulatory transitions or market shifts, as factors affecting projected cashflows, discount rate assumptions and the useful lives of assets. These risks directly influence impairment tests and the recoverability of both tangible and intangible assets, especially in carbon-intensive sectors. To anchor our findings in the data, we provide an exemplary C-KAM in Table 2.
Exemplary C-KAM for impairment and valuation of PPE
| Item | Information |
|---|---|
| Entity: EnBW energie baden-wuerttemberg AG | Year: 2021 |
| Industry: Electric power distribution | Title: Valuation of the individual EnBW offshore wind farms |
| Description: “We classified the valuation of the EnBW offshore wind farms as a key audit matter because the determination of the recoverable amount is highly dependent on the assessment of future cash flows by management and in our view poses an increased risk of incorrect accounting with respect, in particular, to the regular adjustment to assumptions for the short, medium and long-term planning forecasts. In order to create these planning forecasts, it is necessary to derive scenarios that describe potential future developments as realistically as possible. The scenarios derived by management differ primarily regarding the degree of climate protection and the sustainable economic growth that is achievable in the long term …” | |
On the other end of the spectrum, a smaller but still notable group of C-KAMs discusses the going concern assumption. Auditors acknowledge that climate-related risks, ranging from operational disruptions to rising compliance costs, can impair an entity's ability to maintain solvency over a reasonable period. In these cases, C-KAMs highlight how climate considerations are factored into liquidity analysis and strategic forward planning. A further anchoring exemplary C-KAM is presented in Table 3.
Exemplary C-KAM on-going concern
| Item | Information |
|---|---|
| Entity: De La rue PLC | Year: 2023 |
| Industry: Commercial Printing | Title: Conclusions relating to going concern |
| Description: “… We considered the extent to which emerging climate-related risks may affect the Group's assessment and the assumptions around the costs anticipated in meeting the Group's target to become carbon neutral for its own operations by 2030. This includes the capital expenditure required to enable the Group to reduce its carbon footprint, energy usage, waste, and reliance on plastics … The audit procedures performed in evaluating the director's assessment were performed by the Group audit team, however we also considered the financial and non-financial information communicated to us from our component teams of overseas locations as sources of potential contrary indicators which may cast doubt over the going concern assessment. We determined going concern to be a key audit matter.” | |
Together, these six themes demonstrate that climate-related audit matters increasingly intersect with all core financial judgment areas. Rather than forming a separate category, climate-related risks are being integrated into existing audit risk domains, yet, with sector- and region-specific variation in how deeply and transparently they are addressed, as explored in the subsequent sections.
4.1 Comparative analysis: sectoral and regional perspectives on climate-related Key Audit Matters
4.1.1 Sectoral variability
The dataset reveals significant variability in the prevalence of C-KAMs across different industries. Table 4 highlights that energy-intensive sectors such as Mining, Quarrying, and Oil and Gas Extraction (3.52%) and Utilities (4.21%) demonstrate the highest concentration of C-KAMs.
Sectoral distribution of C-KAMs and KAMs, all years, sorted by absolute C-KAM frequency. Multiple C-KAMs were possible per company, and multiple years are taken into account for each company
| Sector | C-KAMs | Total KAMs | Relative |
|---|---|---|---|
| Mining, Quarrying, and Oil and Gas Extraction | 210 | 5,970 | 3.52% |
| Manufacturing | 181 | 19,666 | 0.92% |
| Utilities | 97 | 2,303 | 4.21% |
| Finance and Insurance | 45 | 17,415 | 0.26% |
| Transportation and Warehousing | 32 | 2,629 | 1.22% |
| Real Estate and Rental and Leasing | 25 | 4,011 | 0.62% |
| Construction | 20 | 2,708 | 0.74% |
| Retail Trade | 15 | 4,074 | 0.37% |
| Accommodation and Food Services | 13 | 1,193 | 1.09% |
| Management of Companies and Enterprises | 13 | 1,369 | 0.95% |
| Wholesale Trade | 11 | 1960 | 0.56% |
| Administrative and Support and Waste Management and Remediation Services | 7 | 2039 | 0.34% |
| Information | 4 | 5,884 | 0.07% |
| Professional, Scientific, and Technical Services | 3 | 6,334 | 0.05% |
| Arts, Entertainment, and Recreation | 1 | 1,060 | 0.09% |
| Agriculture, Forestry, Fishing and Hunting | 0 | 782 | 0.00% |
| Educational Services | 0 | 170 | 0.00% |
| Health Care and Social Assistance | 0 | 535 | 0.00% |
| Other Services | 0 | 266 | 0.00% |
| Public Administration | 0 | 128 | 0.00% |
Surprisingly, sectors like Finance and Insurance (0.26%) and Transportation and Warehousing (1.22%) – and even more so with Agriculture and Forestry (0% each), which are also systematically exposed to climate-related risks yet show an insignificant representation of C-KAMs.
4.1.2 Regional concentration and evolution
From a geographic perspective, the United Kingdom leads in the proportion of companies disclosing C-KAMs – especially post-2022 as we will examine more in the next section. As illustrated in Table 5, sectors such as Real Estate and Rental and Leasing (15%) and Mining, Quarrying, and Oil and Gas Extraction (13.92%) are particularly prominent in UK disclosures. Other countries, such as France and Germany, show moderate engagement with C-KAMs in sectors like Manufacturing, with relative frequencies ranging from 6.2% to 9.59%.
Top six countries/sector with C-KAM per unique companies for the whole database scope from 2015 to 2023, sorted by relative frequency
| Country | Sector | Companies with at least one C-KAM | Companies with KAMs | Relative |
|---|---|---|---|---|
| United Kingdom | Real Estate and Rental and Leasing | 12 | 80 | 15.00% |
| United Kingdom | Mining, Quarrying, and Oil and Gas Extraction | 27 | 194 | 13.92% |
| United Kingdom | Manufacturing | 22 | 320 | 6.88% |
| Germany | Manufacturing | 10 | 150 | 6.67% |
| France | Manufacturing | 7 | 108 | 6.48% |
| Italy | Manufacturing | 2 | 65 | 3.08% |
4.1.3 Temporal variations
The findings further indicate a notable temporal variation: the majority of C-KAMs were disclosed from FY2022 onward, despite the significantly larger volume of pre-2022 KAMs overall in the dataset (64,812 compared to 15,816 post-2022). For example, 8 out of 12 C-KAMs in the Real Estate sector in the United Kingdom were written only from FY2022, and all of the 7 French C-KAMs in the Manufacturing sector stem from this recent period. This potentially reflects growing regulatory pressures and stakeholder demands for transparency, emphasizing the transformative impact of recent mandates like the TCFD from 2022 in the United Kingdom and anticipated changes under ESRS in the EU (Baboukardos et al., 2023).
4.2 Related audit procedures
The audit procedures observed across various C-KAMs overall reflect varying approaches designed to address both immediate and long-term risks associated with climate impact, indicating differing levels of auditor proactiveness in integrating climate considerations into the audit process. Common audit procedures include impairment testing under different carbon pricing or climate scenarios, sensitivity analyses and assessments of asset useful lives based on evolving carbon regulations, indicating the importance of accounting for regulatory and market uncertainties. For instance, in sectors with high exposure to carbon costs, such as energy or heavy industry, auditors often conduct in-depth sensitivity analyses to examine how stricter future regulations or carbon taxes might affect asset valuations.
Additionally, auditors frequently review management's assumptions about future climate regulations and market demand shifts, particularly for long-lived assets like PP&E or goodwill. This involves scrutinizing assumptions underlying cash flow projections and evaluating whether provisions are adequate to cover environmental obligations and future decommissioning costs in high-emission sectors. However, these procedures show considerable reliance on management's projections and the incorporation of complex assumptions, challenging their effectiveness in achieving complete transparency and reducing information gap between management and stakeholders.
The procedural depth also varies significantly. While some auditors engage critically with and challenge management's assumptions, others limit their procedures to reviewing management's calculations for mathematical accuracy. For example, at EnBW Energie Baden-Wuerttemberg AG, auditors assessed pricing assumptions using multiple climate scenarios and collaborated with energy market specialists to validate key parameters impacting offshore wind farm valuations. Meanwhile, at Air Liquide SA, the procedures described in the C-KAMs included analyzing the Group's climate strategy impacts on useful lives and recoverable values, ensuring disclosures aligned with financial statements.
Furthermore, although the use of scenario analysis and sensitivity testing, particularly within sectors highly exposed to climate-related risks, allows auditors to evaluate potential impacts under various regulatory or market-driven scenarios, the extent to which scenario analysis is applied varies, with some sectors implementing rigorous, multi-scenario assessments while others apply only basic sensitivity tests:
At Low & Bonar PLC (Plastics Material and Resin Manufacturing), for example, auditors conducted rigorous multi-scenario assessments. Their procedures included applying sensitivities to medium and long-term growth rates, discount rates, and budgeted profits. Specific actions involved leveraging external data, such as OECD country GDP forecasts, and engaging valuation specialists to validate discount rates. Additionally, they calculated a range of sensitivities reflecting risks of underperformance and aligned discounted cash flows with market capitalization for reasonableness. The auditors critically analyzed budget assumptions and historical trends, reflecting a thorough approach. In contrast, at Antofagasta PLC (Copper and Nickel Ore Mining), the approach was less complex, focusing mostly on basic sensitivity analyses and impairment indicator assessments. While procedures included verifying capital and operating expenditure and applying sensitivity tests to key market assumptions, the scope and depth were comparatively limited.
These findings further suggest that auditors face inherent challenges in applying scenario analyses and assessing management's assumptions, although these complexities are not explicitly detailed in the C-KAMs themselves. At Hammerson PLC, for example, procedures included engaging external valuation experts to assess external appraisals and ensure market assumptions incorporated climate change and other sustainability considerations in investment property valuations.
Finally, auditors, perhaps based on tradition and training, typically frame C-KAMs in terms of “liabilities,” “constraints” and “stress testing,” focusing heavily on the uncertainties and financial risks tied to carbon costs and regulatory shifts. Even when C-KAMs acknowledge potential opportunities, their treatment remains largely descriptive rather than quantitative. In addition, the findings here point to a pronounced regional divergence. For example, C-KAMs in the United Kingdom, aligned with the TCFD's recommendations, tend to adopt more balanced approaches that consider both climate-related risks and opportunities. In contrast, disclosures in countries such as France place greater emphasis on climate-related risks, particularly those related to regulatory compliance.
4.3 Communicative setup
C-KAMs often provide a high-level view of audit procedures without conveying the subjective decisions or interpretative flexibility likely shaping the audit process. C-KAMs frequently summarize procedures without revealing detailed assumptions or sector-specific considerations behind them. While C-KAMs refer to the application of future carbon pricing scenarios, they often lack transparency about the criteria used to select those scenarios or auditors' judgment of management's assumptions regarding long-term regulatory impacts:
At Mayr-Melnhof Karton AG, for example, auditors performed sensitivity analyses to evaluate goodwill recoverability under varying discount rates, validating these assumptions against selected market data without specifying selection criteria. Consequently, C-KAMs alone provide an incomplete picture, highlighting the need for additional disclosures or context to fully understand the complexities of scenario analysis and assumption testing in climate-related audits.
To further enhance the linguistic analysis of C-KAMs, we also examined tone variations to assess how auditors communicate the risks and opportunities associated with climate disclosures. By utilizing sentiment analysis techniques, including lexicon-based methods tailored for financial text, we evaluated whether the tone of C-KAMs leans toward negative (risk-focused), positive (opportunity-focused) or neutral framing. This approach provides insights into how auditors frame climate-related challenges in a manner that influences stakeholder perceptions. The findings reveal that C-KAMs predominantly adopt a neutral to negative tone, with significant variations across industries and regions. For instance, C-KAMs addressing asset impairments or regulatory compliance risks often exhibit a more negative tone, emphasizing the financial and operational vulnerabilities linked to climate risks. In contrast, industries like renewable energy and technology, which focus on climate opportunities, such as innovation and market expansion (albeit rarely represented in C-KAMs), tend to present C-KAMs with a more balanced or positive tone.
The findings also reveal a factual tone overall, avoiding subjective language, a strategy critical in sustainability reporting to counteract perceptions of greenwashing. Neutral language, as seen in phrases such as “provisions for liabilities,” ensures that disclosures focus on quantifiable data and verifiable assessments, promoting trust in the disclosures' objectivity. However, neutral standardized phrases such as the above, could also simply mask nonengagement with more complex climate-related risks. Future-oriented language within C-KAMs was also found, including references to “future regulations” and “transition risks,” which underscores a proactive focus some auditors already place on regulatory shifts and their long-term impacts on company viability. This approach resonates with the TCFD-aligned reporting framework emphasizing scenario-based disclosures. For instance, audit procedures like “impairment testing under carbon scenarios” serve as tools for evaluating potential future risks, offering stakeholders insights into how climate challenges could alter company valuations under various scenarios.
5. Discussion
5.1 Current C-KAM practice
Although C-KAMs in theory present a promising avenue to integrate climate considerations into financial audits (Simnett and Huggins, 2014; Pinto and Morais, 2018), our findings suggest evident limitations that must be addressed to realize their full potential.
The most striking feature of our findings is the thematic concentration of C-KAMs in traditional accounting containers such as impairments, valuation of goodwill and PPE, provisions, and reserves. This pattern does more than signal where climate-related effects currently crystallize in the financial statements; it also indicates how climate uncertainty is rendered auditable in practice. Climate risks are primarily made visible at the point where they can be expressed as adjustments to recoverable amounts, discount rates or decommissioning obligations, rather than as questions about business model resilience, transition strategy or portfolio reallocation. The audit work described in C-KAMs mirrors this framing. Procedures focus on testing discounted cash-flow models, challenging selected parameters, or checking compliance with IAS 36 and IAS 37, while forward-looking climate transition plans or scenario-based strategic choices almost never appear as audit objects in their own right. In this sense, the “auditability” of climate risk (Power, 1997) is achieved by folding it into existing impairment and provisioning routines, which offers auditors familiar evidential anchors but simultaneously narrows the horizon of what can be examined and communicated about climate as a strategic, system-level risk.
Moreover, this predominant framing in a risk-oriented manner, focusing on downside exposure such as asset impairments or regulatory liabilities, aligns with earlier critiques by Hodge et al. (2009) and O'Dwyer and Unerman (2020). They argue that climate-related disclosures often neglect to reflect potential opportunities related to innovation or transition strategies, which may be of equal interest to investors and stem, for example, from TCFD's obligation to report both risks and opportunities.
This discursive asymmetry may partly reflect a broader conservatism bias embedded in the audit profession, where emphasis is placed on downside financial risk and prudence (Power, 1997; Gibbins et al., 2001; Khalifa et al., 2024) as guided by most national GAAPs, whereas IFRS would advocate a true and fair view with a strong orientation toward decision-usefulness for investors. While such caution aligns with the principle of professional skepticism, it may obscure important forward-looking aspects of corporate strategy that are material to investors, such as the capacity of firms to benefit from green subsidies or low-carbon innovation. The current framing in C-KAMs thus also raises questions about how auditors interpret materiality in the climate context (O'Dwyer and Unerman, 2020; Harrer and Lehner, 2024) and to what extent the new domain of climate change is simply seen as an extension of existing domains in financial auditing.
The implications for investor behavior are noteworthy. A narrowly risk-focused narrative may overstate firm vulnerability and under-communicate resilience or transition preparedness, potentially distorting market perceptions. In turn, this could affect capital allocation decisions, especially as climate-aware investment grows (Krueger et al., 2020). For climate-related risk disclosures to be decision-useful, auditors may need to move beyond conservative defaults and adopt a more balanced framing aligned with financial materiality principles. The emphasis on risks rather than potential business opportunities for innovation or adaptation, therefore, suggests an area where disclosures could evolve to provide a more balanced view. Integrating both risks and opportunities would not only align with stakeholder expectations for transparent reporting but would also address concerns about the completeness and objectivity of climate-related disclosures for shareholders.
Another challenge is the limited transparency surrounding how scenarios or forecasts are selected and evaluated. Several C-KAMs in our sample referred to scenario analysis or market assumptions without specifying which data sources or modeling frameworks were used. This echoes concerns raised by Krasodomska et al. (2021) and Harrer and Lehner (2024), who argue that sustainability assurance, including climate-related disclosures, often suffers from a lack of methodological clarity. Looking ahead, emerging standards such as ISSA 5000 (finalized in 2025) present an opportunity to address these limitations. Regulators and standard-setters should encourage the development of sector-specific guidance for C-KAMs and promote clearer expectations about the inclusion of forward-looking audit procedures. This would align with calls from Cuadrado-Ballesteros et al. (2017) and Hertl and Maniora (2024) for more robust and comparable assurance practices in sustainability contexts.
To reflect on these limitations, realizing the full potential of C-KAMs will require not only clearer regulatory guidance but also a willingness within the profession to revisit how materiality is assessed, communicated and evidenced when the underlying risks are long-horizon, model-dependent and inherently uncertain.
5.2 Variations across sectors, countries and audit practices
Context significantly influences the prevalence and quality of C-KAMs (Sierra-García et al., 2019; Pinto and Morais, 2018; Pinto et al., 2020; Mihret et al., 2021; Minutti-Meza, 2021). The disproportionately large share of C-KAMs in high-carbon sectors, as our findings show, reflects their greater exposure to climate transition and physical risks and perhaps more public awareness. This aligns with earlier research that shows auditors respond to industry risk profiles when determining KAMs (Pinto and Morais, 2018). In contrast, the low representation of C-KAMs in service or consumer sectors like finance, insurance or even agriculture, despite their systemic exposure to climate-related financial risks, raises concerns about underreporting or insufficient audit attention. After all, the use of C-KAMs needs to be based upon auditors’ specific judgment regarding potential areas of material misstatements and not upon public awareness or industry trends (Octavio and Setiawan, 2025).
As financial institutions face growing pressure to disclose transition risks linked to lending and investment portfolios, the puzzling lack of corresponding C-KAMs in the finance sector may further signal a misalignment between audit reporting and stakeholder expectations (including regulators) and eventually hint toward audit errors. This asymmetry supports Pinto and Morais (2018) observation that the relevance of KAM disclosures can vary significantly, depending on the regulatory stringency and the perceived risk of nondisclosures, and may not always align with users' informational needs. One fitting structural explanation is that banking supervisors, notably the European Central Bank and the European Banking Authority, currently channel climate risk assessment primarily through prudential instruments such as climate stress tests and Pillar 3 disclosure requirements, which operate outside the financial statements and therefore outside the scope of matters that auditors would typically elevate to KAM status. What is more, and related to that, is that climate risks in financial institutions express themselves mostly in loan portfolios with limited company-specific data. This also points to a capability gap. Translating portfolio-level into entity-specific materiality and corresponding audit matters requires modeling skills that typically lie beyond traditional audit competencies.
Geographical differences were also notable. In line with the findings of Pinto et al. (2020) and Coram and Wang (2021), countries with more advanced climate reporting regulations, especially the United Kingdom, where TCFD-aligned disclosure is mandatory, exhibited higher-quality and more frequent C-KAMs. Although auditors in the United Kingdom were only liable for reviewing the consistency between TCFD-aligned disclosures and the financial statements (not assuring the climate-related disclosures themselves), it seems that such mandatory climate-related disclosures act as strong signals and source of “other information” (ISA 720) for auditors to watch out for financial impacts and risks in the financial statements, and consequently they identify more areas for closer audit scrutiny. Conversely, relatively few C-KAMs were disclosed in Nordic countries, despite their reputation for sustainability leadership (Strand et al., 2015). This may reflect institutional factors such as audit culture or perceived stakeholder trust, as discussed by Minutti-Meza (2021), suggesting that the effectiveness of C-KAMs may also be shaped by national norms and assurance expectations.
Finally, related audit procedures described in C-KAMs varied considerably despite the same underlying risks of material misstatements. While some auditors detailed their review of future carbon regulation impacts or impairment scenarios, others gave only generic descriptions such as “assessment of assumptions” or “evaluation of forecasts.” This procedural inconsistency may arise from the unintended “judgment-action gap” as shown by Asbahr and Ruhnke (2019), who argue that KAM disclosures can serve as a moral license, meaning auditors may still maintain skeptical judgments about an accounting estimate's reasonableness but are less likely to take corresponding skeptical actions. These disparities may also signal uneven assurance coverage, where climate-related audit attention is not consistently applied across the corporate landscape. A lot is at stake here, and ISSA 5000 will certainly provide better guidance.
From a professional judgment standpoint, this unevenness may also indicate a lack of consensus within the profession about when and how climate-related risks warrant audit emphasis (which is indeed the case as one author, a CPA director with one of the Big4 experiences in his daily practice). Inconsistencies in materiality judgments can erode the comparability and credibility of disclosures, leaving users uncertain about whether climate-related risks are immaterial or merely under-examined. It also points to potential regulatory arbitrage, where firms in less stringent jurisdictions may avoid scrutiny despite having material exposure.
The inconsistency further complicates the integration of climate-related risks into investor decision-making. If climate-related audit signals are present in some reports but absent in others with similar risk exposure, the reliability of audit reports as comparative tools is diminished. Standard-setters may need to issue more explicit criteria or guidelines for climate risk inclusion in KAMs to address this emerging assurance gap (Simnett and Huggins, 2015; Maroun, 2020).
5.3 On the potential of C-KAMs
By disclosing how auditors test assumptions linked to carbon pricing, impairment models or environmental obligations, well-developed C-KAMs can serve as decision-useful disclosures for investors, regulators and other stakeholders (Pinto and Morais, 2018; Krasodomska et al., 2021). This is consistent with studies suggesting that KAMs can improve the transparency and perceived credibility of audit reports, especially when the underlying audit procedures are clearly explained (Sirois et al., 2018; Köhler et al., 2020). In our sample, C-KAMs that detailed specific procedures, such as scenario analyses or external expert consultations, were notably more informative and aligned with user expectations.
Nevertheless, in practice, low specificity and depth of C-KAMs remain a concern, rendering some C-KAMs unfit for the task. Some disclosures defaulted to high-level, generic language, a trend previously criticized by Velte and Issa (2019) for undermining the communicative intent of KAMs. This observation holds especially for firms operating in less carbon-intensive industries or in regions where climate issues are not yet mainstreamed into financial risk management. The procedural vagueness, compared to the high complexity settings of climate projections and quantitative simulations or management judgments, potentially reflects auditors' uncertainty or limited expertise in evaluating forward-looking climate-related assumptions, especially those involving scenario analysis, regulatory risk or technological disruption.
From an audit quality perspective, this raises concerns about the extent to which C-KAMs provide users with insight into how auditors engage with climate-related risks of material misstatements (Minutti-Meza, 2021). ISA 701 emphasizes that KAMs should not only identify significant matters but also describe how those matters were addressed in the audit. When C-KAMs omit clear explanations of procedures or risk assessments, they fall short of enhancing transparency, potentially undermining the informational value of audit reports (Christensen et al., 2014; Velte and Issa, 2019). Moreover, low-specificity disclosures may erode user confidence and contribute to the audit expectation gap, particularly in an area where investor scrutiny is intensifying (Sirois et al., 2018).
This trend also highlights a potential capability gap in the profession. As climate-related risks require multidisciplinary insights, from environmental modeling to regulatory forecasting, traditional audit methodologies may prove insufficient. Professional accounting bodies and standard-setters (e.g. IAASB, EFRAG) may need to consider targeted guidance or training on how to assess and report on climate-related risks through C-KAMs to ensure audit quality is not compromised in the face of rising complexity. Given the evolving landscape of sustainability assurance regulation, C-KAMs use in the financial audits are uniquely positioned to act as a bridge between traditional financial audits and the assurance of new sustainability disclosure regimes (Minutti-Meza, 2021; Harrer and Lehner, 2024). As audit frameworks such as the ISSA 5000 become more widely applied, there is an opportunity to improve the standardization, depth, and relevance of C-KAM disclosures to better support stakeholders in evaluating climate-related risks and assumptions.
Looking ahead, the regulatory environment in which C-KAMs are produced will be reshaped by the combined effects of the ISSB standards and the EU's CSRD/ESRS architecture. IFRS S1 and S2 will require entities using IFRS to articulate how climate-related risks and opportunities affect business models, strategy, governance and, crucially, the measurement of assets and liabilities in the financial statements. CSRD and the ESRS, in turn, formalize double materiality and specify sector-agnostic and sector-specific disclosure requirements that cut across strategy, governance, risk management and metrics. Together, these developments will increase the volume and structure of climate-related information that auditors must consider when forming their opinion and deciding which matters warrant disclosure as C-KAMs. If climate assumptions and scenario analyses become systematically embedded in impairment tests, provisions and fair value estimates, more of the tensions we document here are likely to surface in audit reports, rather than remaining implicit in working papers.
At the same time, these regimes create new assurance arenas that may reallocate where climate-related audit signals appear. Dedicated sustainability statements prepared under ESRS and subject to limited, and later reasonable, assurance may come to host much of the explicit discussion of transition plans, decarbonization pathways and physical risk exposure. This raises an open question about whether climate-related issues will gradually migrate out of the statutory financial audit report into separate sustainability assurance reports, or whether C-KAMs will retain a distinctive role by focusing on how climate uncertainty is translated into critical accounting estimates. In a CSRD/ESRS setting, C-KAMs could become a key interface between financial-material climate effects and broader impact-oriented disclosures, highlighting where climate scenarios have actually altered recoverable amounts, useful lives or provisions, and where they have not. Similarly, under the ISSB framework, C-KAMs may serve as a testing ground for how far auditors are prepared to engage with the forward-looking, model-based aspects of IFRS S2 rather than confining their work to traditional conservatism.
6. Contributions, future research and conclusion
Taken together, our analysis shows that C-KAMs have begun to take root in statutory audit reporting, but in a markedly uneven and cautious manner. Across 80,628 KAMs, we identify 678 C-KAMs, concentrated in a narrow set of high-carbon and infrastructure sectors, most prominently utilities, oil and gas, mining and selected manufacturing and real estate firms, while large parts of the economy, including finance, insurance and agriculture, remain virtually untouched. Within this limited footprint, climate-related risks are predominantly translated into familiar audit containers such as impairment testing, provisions and valuation estimates, typically supported by limited procedural depth and a largely neutral-to-negative tone.
Our central conclusion is that while the emergence of C-KAMs signals a growing recognition that climate-related financial risks warrant audit attention, current practice continues to struggle with the fundamental challenges of auditing climate uncertainty. Rather than reworking audit approaches to accommodate long-horizon, scenario-based judgments, most disclosures adapt climate risks to conventional frameworks designed for backward-looking, point-in-time estimates. Only a smaller subset of C-KAMs engages more substantively through elaborated procedures or strategic framing, indicating that the profession remains in an early, transitional phase of articulating what it would mean to render climate risk auditable on its own terms.
Overall, the evidence points to a persistent gap between the promise of C-KAMs as mechanisms for making judgment under uncertainty visible and their current implementation in practice. This gap reflects a profession responding to regulatory pressure while still grappling with how to translate climate complexity, indeterminacy and model dependence into audit communication that is both credible and decision-useful. At the same time, it is important to distinguish between limitations that stem from prevailing audit practice and those that are structurally embedded in ISA 701 and the KAM reporting framework. Some of the observed constraints reflect how auditors currently interpret and communicate climate-related risks within C-KAMs, including conservative framing and reliance on familiar audit heuristics. Other limitations, however, arise from the design of KAMs themselves, which are anchored in backward-looking assessments of significant audit matters and constrained narrative forms. Recognizing this distinction helps situate the findings as an analysis of practice operating within an infrastructure that both enables and bounds what can be rendered auditable.
Our article makes three key contributions to the literature on KAMs and climate-related assurance:
First, we extend prior research on KAMs as mechanisms for enhanced transparency in financial audits (Simnett and Huggins, 2015; Sirois et al., 2018; Elliott et al., 2020; Noureldeen et al., 2024) by conceptualizing C-KAMs as a distinct subset and by providing an empirical account of their emergence across European audit reports. Our analysis shows that C-KAMs are shaped by the interplay of climate-related and financial materiality, and that they mediate between auditors, stakeholders and users through varying degrees of clarity, granularity and procedural depth. More fundamentally, we show that C-KAMs operate as locations where financial materiality is actively negotiated under conditions of long-horizon uncertainty. In these disclosures, auditors participate in the practical alignment of future-oriented climate risks with present-period financial reporting thresholds. C-KAMs thus constitute sites where uncertainty is provisionally stabilized, assumptions are rendered justifiable and materiality becomes auditable despite the absence of definitive evidence. In doing so, they signal a shift in the epistemic foundations of audit discourse itself (Minutti-Meza, 2021).
Second, we contribute to the growing body of literature on climate assurance by examining how statutory financial auditors interpret and incorporate climate-related risks into audit planning and disclosure practices (Asbahr and Ruhnke, 2019; Pham et al., 2024). Rather than treating climate assurance as a discrete service or assurance type, we analyze how climate concerns are absorbed into existing financial audit routines, especially in areas such as asset impairment, legal provisions and valuation estimation. We show that the communication of corresponding audit judgments remains uneven, ranging from detailed procedural descriptions to vague and abstract formulations. This finding highlights the tension between the auditable and the anticipatory: climate risks are acknowledged but rarely interrogated in terms of their strategic or forward-looking implications.
Third, our study offers timely insights for regulators and standard setters seeking to enhance the coherence and credibility of climate-related assurance within financial audit regimes. We show that while the introduction of climate risk into KAMs is increasing, particularly in jurisdictions with mandatory climate disclosure regimes, there remain structural gaps in how climate-related audit work is represented. In particular, C-KAMs often fail to communicate the scope, limitations or epistemic status of the auditor's engagement with climate uncertainty. This limits their decision-usefulness for financial statement users and exposes inconsistencies in current practice. By empirically identifying these gaps, we contribute evidence to ongoing standard-setting processes such as the development of ISSA 5000. In doing so, we respond to recent calls for assurance frameworks that go beyond procedural box-ticking and foster meaningful professional judgment and communicative transparency (Harrer and Lehner, 2024).
The contributions of this study also delineate a forward-looking research agenda that spans user reception, audit practice and the evolving epistemic conditions of climate assurance:
Our first suggestion concerns the reception and downstream effects of C-KAMs. By conceptualizing C-KAMs as sites where financial materiality is negotiated under long-horizon uncertainty, our study raises questions about how such negotiations are interpreted by users of financial statements. Future research could examine whether and how C-KAMs shape investor decision-making, analyst forecasts or credit assessments, and whether differences in clarity, granularity, and procedural depth translate into differential market responses. Prior work suggests that KAMs can influence user judgments (Coram and Wang, 2021), yet the implications of climate-related audit disclosures remain underexplored within both the KAM and sustainability assurance literatures (Köhler et al., 2020). Investigating these feedback effects would help assess whether C-KAMs function merely as transparency devices or as consequential inputs into capital market reasoning under climate uncertainty.
A second one follows from our analysis of how climate risks are absorbed into existing financial audit routines. As climate assurance continues to be performed primarily within statutory financial audits rather than as a separate assurance service (Asbahr and Ruhnke, 2019; Pham et al., 2024), there is scope for more fine-grained qualitative research into the organizational and professional dynamics that underpin the formulation of C-KAMs. Building on Maroun (2020), future studies could examine how audit teams develop internal guidance, manage liability concerns and coordinate judgment when evidentiary demands rooted in financial auditing intersect with forward-looking and scenario-based climate risks. Such work would deepen understanding of how anticipatory risk is rendered auditable in practice, and why the communication of related audit judgments remains uneven across firms and jurisdictions.
A final suggestion for future research emerges at the intersection of regulation, standard setting and epistemic limits. As climate reporting becomes increasingly formalized under regimes such as the CSRD and IFRS S2, longitudinal research could trace how C-KAMs evolve in response to heightened regulatory expectations and expanded disclosure mandates. Rather than assuming convergence through compliance, future studies might explore how audit firms' institutional logics mediate the translation of these requirements into disclosure artifacts, thereby shaping what aspects of climate-related audit work become visible, standardized or obscured. At the same time, the structural gaps identified in this study point to broader questions about how assurance frameworks articulate the scope, limitations, and epistemic status of climate-related audit work. Following Harrer and Lehner (2024), this suggests a critical research agenda on how auditors confront a persistent “regime of indeterminacy,” and on whether and how professional assurance can credibly engage with risks whose financial consequences remain fundamentally uncertain.
The aforementioned directions position C-KAMs as a productive empirical and conceptual entry point for examining how audit discourse, professional judgment and regulatory ambitions are being reconfigured in response to climate risk. By extending analysis across users, organizations and standard-setting arenas, future research can further clarify whether C-KAMs will remain largely formulaic compliance texts or develop into a durable mechanism for rendering climate-related uncertainty visible, contestable and decision-useful within financial audit regimes (Minutti-Meza, 2021; Harrer and Lehner, 2024).
Appendix 1
To operationalize the notion of “depth” in the description of audit procedures, we combined manual coding with a rule-based Python routine. In a first step, we inductively developed three depth categories from close reading of the C-KAMs. Disclosures were coded as low depth when they restricted themselves to generic, largely mechanical actions, for example formulations such as “we checked the calculations prepared by management” or “we verified the mathematical accuracy of the impairment model,” where the auditor essentially reports recalculation or simple verification without engaging with underlying assumptions. Medium depth was assigned where auditors described some substantive evaluation of management's judgments, for instance “we evaluated and analyzed key assumptions, including growth rates and discount rates, and compared forecasts with historical performance and selected external benchmarks,” that is, where the wording signals analytical work on inputs but without detailing structured scenario work or specialist involvement. High depth was reserved for C-KAMs that articulated multi-layered procedures under explicit climate-related uncertainty, typically combining scenario or sensitivity analysis with external expertise, such as “we performed scenario analysis under alternative carbon pricing pathways, carried out sensitivity analyses on discount rates and long-term growth assumptions, and engaged valuation specialists to assess the discounted cash flow models used to test recoverable amounts.” In a second step, these qualitative distinctions were translated into keyword dictionaries and implemented in a Jupyter notebook. In parallel, “communicative tone” was classified as positive, negative or neutral by matching lemmatized C-KAM text against pre-specified Loughran–McDonald sentiment dictionaries and assigning the label according to the predominance of opportunity-oriented or risk-oriented terms.
Appendix 2
The flow diagram presents a vertically structured process that transitions into two parallel analytical branches. At the top, a box labeled “Universe of K A M s in E U or E E A statutory audit reports (all sectors, all years)” connects downward to “Structured data extract (K A M text plus firm, sector, country, year, auditor)”. A downward arrow leads to “Restrict to European listed entities, F Y 2015 to 23”, followed by another downward arrow to “Keyword filter v1 based on S B T i or climate transition and mitigation terminology”. Next, a downward arrow leads to a larger box labeled “Iterative snowball checks: manual review of K A M s in high physical-risk sectors (e.g. agriculture, real estate, utilities) to identify terms such as flood, drought, storm, heatwave, resilience, business interruption”. From here, a downward arrow leads to “Keyword filter v2 applied to full K A M dataset using combined transition plus physical-risk dictionary”. Another downward arrow leads to “Manual screening of all hits: remove false positives, ensure explicit link to climate-related financial risks”. A downward arrow leads to “Final sample of C-K A M s (explicit climate-related Key Audit Matters)”. From this central box, the process splits into two parallel branches. On the left branch, a downward arrow leads to “Descriptive profiling (Excel-based analysis)”, followed by “Frequency tables and cross-tabs by sector, country, year, auditor (e.g. Tables 4 to 6)”, and then “Thematic classification of C-K A M content into main climate-related risk areas (e.g. Table 1)”. On the right branch, a downward arrow leads to “Textual or interpretive analysis (Python or Jupyter)”, followed by “Python pre-processing: tokenization, lemmatization of C-K A M text”. A downward arrow leads to “Coding of audit procedures and procedural depth using rule-based keyword sets: low, medium, high depth (Appendix A)”. Next, a downward arrow leads to “Tone classification using Loughran–McDonald-based sentiment wordlists: positive, negative, neutral (Appendix A)”. Finally, a downward arrow leads to “Reliability checks: manual validation of a subsample, intercoder comparisons, refinement of dictionaries and rules”.The research workflow, source: authors
The flow diagram presents a vertically structured process that transitions into two parallel analytical branches. At the top, a box labeled “Universe of K A M s in E U or E E A statutory audit reports (all sectors, all years)” connects downward to “Structured data extract (K A M text plus firm, sector, country, year, auditor)”. A downward arrow leads to “Restrict to European listed entities, F Y 2015 to 23”, followed by another downward arrow to “Keyword filter v1 based on S B T i or climate transition and mitigation terminology”. Next, a downward arrow leads to a larger box labeled “Iterative snowball checks: manual review of K A M s in high physical-risk sectors (e.g. agriculture, real estate, utilities) to identify terms such as flood, drought, storm, heatwave, resilience, business interruption”. From here, a downward arrow leads to “Keyword filter v2 applied to full K A M dataset using combined transition plus physical-risk dictionary”. Another downward arrow leads to “Manual screening of all hits: remove false positives, ensure explicit link to climate-related financial risks”. A downward arrow leads to “Final sample of C-K A M s (explicit climate-related Key Audit Matters)”. From this central box, the process splits into two parallel branches. On the left branch, a downward arrow leads to “Descriptive profiling (Excel-based analysis)”, followed by “Frequency tables and cross-tabs by sector, country, year, auditor (e.g. Tables 4 to 6)”, and then “Thematic classification of C-K A M content into main climate-related risk areas (e.g. Table 1)”. On the right branch, a downward arrow leads to “Textual or interpretive analysis (Python or Jupyter)”, followed by “Python pre-processing: tokenization, lemmatization of C-K A M text”. A downward arrow leads to “Coding of audit procedures and procedural depth using rule-based keyword sets: low, medium, high depth (Appendix A)”. Next, a downward arrow leads to “Tone classification using Loughran–McDonald-based sentiment wordlists: positive, negative, neutral (Appendix A)”. Finally, a downward arrow leads to “Reliability checks: manual validation of a subsample, intercoder comparisons, refinement of dictionaries and rules”.The research workflow, source: authors
