CLIMATE CHANGE RISK DISCLOSURE AND ACCOUNTING CHOICE: EVIDENCE FROM U.S. OIL AND GAS COMPANIES

_


INTRODUCTION
In September 2016, the United States Securities and Exchange Commission (SEC) launched an investigation into ExxonMobil and its auditor PricewaterhouseCoopers, questioning how the country's largest oil group reported the value of its assets and disclosed the impact of climate change on its business (Olson and Viswanatha, 2016).
https://ielas.org/ijbe/index.php/ijbeISSN (online) 2545-4137 90 In 2018, the investigation was terminated due to insufficient evidence (Michaels and Olson, 2018).However, the SEC's scrutiny of the oil giant ExxonMobil instigated an uproar in the U.S. oil and gas industry.Since it is difficult for investors to determine which companies are taking action against climate change, investors are urgently requesting oil and gas companies to disclose the impact of CCR on business operations and corporate reporting (Diaz-Rainey et al., 2023;Matsumura et al., 2014;Pankratz et al., 2023;van Benthem et al., 2022).
As an exploration of U.S. oil and gas firms' CCR disclosure following the SEC's investigation of ExxonMobil, this study investigates how CCR disclosure affects oil and gas companies' procedural accounting choices.We delve into one of the most critical ESG (Environmental, Social, and Governance) risks confronting the sustainability of society-climate change (e.g., Borghei, 2021;SEC, 2022;TCFD, 2018).With the aim of preventing global temperatures from rising by 2°C above the pre-industrial level, nearly 200 countries signed the "Paris Agreement" in 2015, agreeing to reduce greenhouse gas emissions and encourage the transition to a low-carbon economy.The U.S. SEC released the Climate Change Disclosure Guidance in 2010 and proposed new rules to standardize climate-related disclosures in 2022 (SEC, 2022).The CDP, the Global Reporting Initiative (GRI), the Intergovernmental Panel on Climate Change (IPCC), and various other institutions are also striving to increase corporate awareness and improve disclosure guidelines regarding CCR matters.
Meanwhile, oil and gas companies are suffering from global warming and extreme weather, and they can choose the successful-efforts (SE) method or the full-cost (FC) method to record oil and gas exploration and development (E&D) activities.This crucial accounting choice affects how companies report their net income and cash flows and delineate E&D costs and assets across exploration, development, and production categories (Begley, 1990;Bryant, 2003).Since procedural accounting methods play a pivotal role in shaping reported costs, revenues, expenses, profits, and balance sheet valuations (Bryant, 2003;Lilien and Pastena, 1982;Misund et al., 2008), it is imperative to investigate the flexibility of procedural accounting choice under the influence of CCR disclosure, helping investors and other stakeholders better understand oil and gas firms' CCR disclosure and the E&D accounting choice that affect corporate earnings and cash flows (Diaz-Rainey et al., 2023;van Benthem et al., 2022).
To test our research question, after the SEC terminated the CCR investigation of ExxonMobil in 2018, we manually collected and investigated all U.S. oil and gas companies' 10-K financial reports, CDP reports, and corporate sustainability reports to examine whether these companies mentioned CCR issues and their E&D accounting choices.CCR disclosure information was collected from various sources, including the SEC's Edgar database, CDP database, Ceres database, GRI database, and companies' sustainability websites.Other financial indicators regarding accounting choice were collected from the Compustat and CRSP databases.
Our study makes important contributions to the previous literature.First, this study contributes to the previous research regarding corporate CCR disclosure and accounting choice topics.Although scholars have begun to study carbon disclosure and its corporate financial impact (e.g., Borghei, 2021;Downar et al., 2021;Matsumura et al., 2014;Tang and Demeritt, 2018;Wang, 2023), our study fills the gap between the CCR disclosure and E&D accounting choices by oil and gas firms, following the 2018 SEC's climate disclosure investigation of ExxonMobil.We manually collected companies' CCR disclosures from multiple sources and updated CCR disclosure research data with richer dimensions, providing a new influencing factor for oil and gas firms' procedural accounting choice (Begley, 1990;Bryant, 2003;Lilien and Pastena, 1982;Misund et al., 2008).
Further, our study enriches corporate financial reporting guidance for investors in the oil and gas industry, helping eliminate opacity in the disclosure practices of oil and gas companies facing severe CCRs (e.g., Lemma et al., 2020;Luo and Wu, 2019;Velte, 2021).Our study offers investors and other stakeholders a better understanding of the association between oil and gas CCR disclosure and accounting choices, supplying the public with effectual guidance to reveal the impact of CCR disclosure on the reporting trends of oil and gas companies' earnings and cash flows.Additionally, this study provides evidence for SEC, FASB, and government regulators by digging into the association between CCR disclosure and accounting choice in the energy sector.Following the release of the SEC's Climate Change Disclosure Guidance (SEC, 2010(SEC, , 2022)), this study responds to the urgent call from the SEC and global environmental organizations to encourage more oil and gas companies to disclose CCR to facilitate the transition towards a low-carbon economy (COP 27, 2022).
The remainder of this study is organized as follows.Section 2 provides the theoretical framework and Section 3 presents background information and hypothesis development.Sample selection and research design are discussed in Section 4. Section 5 presents and discusses our empirical results.Section 6 shows additional robustness test results, and Section 7 concludes the paper and highlights future research directions.

THEORETICAL FRAMEWORK
Agency theory states that when an economic entity (principal) has an agency relationship, another entity (agent) is authorized to make decisions on behalf of the principal (Subramaniam, 2006).Generally, agents have an information advantage over principals because agents are more directly involved in day-to-day business operations.Considering the information asymmetry between shareholders (principals) and managers (agents) in a shareholder-manager relationship, oil and gas companies' CCR disclosure can increase information transparency and shareholders' oversight, thereby reducing potential costs of conflicts (Benston, 1982;Broadstock et al., 2018 , 2021;Prado-Lorenzo and Garcia-Sanchez, 2010).Thus, shareholders can reduce investment risks by understanding oil and gas firms' CCR disclosure and accounting choice strategies (Eleftheriadis and Anagnostopoulou, 2015;Griffin et al., 2017;Krishnamurti and Velayutham, 2018).
Debt covenant theory indicates that firms negotiate financial contracts with bondholders (Leftwich, 1981;Smith and Warner, 1979).These contracts have certain restrictions that affect the firms' financial investment, production policies, bond dividends, asset disposal and maintenance, and merger activities (Begley, 1990;Dyreng et al., 2022;Smith and Warner, 1979).Debt covenants help control the conflicts of interest between bondholders and shareholders.Once the firm comes close to breaching the contract, the debt contract will limit the firm's ability to pay dividends while adding additional agency costs and non-compliance costs (Dichev and Skinner, 2002).Managers' different accounting choices have different impacts on shareholders' interests and corporate wealth distribution.Managers have economic incentives to choose accounting methods more conducive to the firm to reduce the non-compliance costs from potential contract breach risks (Begley, 1990;Pittman and Zhao, 2020;Robin et al., 2017).
According to the debt covenant theory, since intense CCR increases the difficulty of oil and gas firms' mining and exploration activities, huge exploration costs and dry well costs make these high-risk companies more likely to face contractual default risks.As a result, oil and gas company managers have incentives to reduce default risks and mitigate institutional constraints through accounting choices (Collins et al., 1981;Gilje et al., 2020;Franz et al., 2014).Choosing the FC method to capitalize oil and gas exploration expenses can help maximize corporate income and smooth financial volatility, ensuring that oil and gas companies are not exposed to default risks and restrictions imposed by bondholders in their financial contracts (Beatty et al., 2002;Begley, 1990;Daley and Vigeland, 1983;Dyreng et al., 2022).

BACKGROUND AND HYPOTHESIS DEVELOPMENT
Accounting choices for oil and gas companies have been controversial since the 1970s (Bryant, 2003).Under U.S. GAAP, oil and gas companies can choose either the SE or the FC accounting method to describe E&D activities.The main difference between the two accounting methods, SE and FC, is how they calculate exploration costs and oil and gas asset impairments.Under the SE method, only exploration costs associated with successful wells are recorded as assets.All exploration costs for unsuccessful or dry wells are included in expenses.Companies using the SE method apply the guidelines from FASB ASC Topic 932 (Extractive Activities -Oil and Gas) and Topic 360 (Property, Plant, and Equipment) to explain the impairment of their oil and gas assets.
In contrast, the FC approach allows exploration companies to capitalize all costs associated with exploration activities for oil and gas reserves, regardless of the success of efforts.While the FC approach allows unsuccessful exploration costs to be capitalized and incorporated into oil and gas assets, the SEC imposed an upper limit on the total costs that FC companies can capitalize, known as the "ceiling test."Oil and gas firms using the FC method should apply the guidelines in SEC Regulation S-X Rules 4-10, SAB (Staff Accounting Bulletin) Subject 12.D, and FRC (Financial Reporting Codification) Section 406.01.c.
Managers have different incentives to choose particular accounting methods (Fields et al., 2001;Libby et al., 2015;Wells, 2020;Zhang and Zhang, 2023).Lilien and Pastena (1982) mention that management's accounting choices can result indifferent cumulative earnings based on firms' specific circumstances and the SEC's procedural standards.Compared with the FC method, the SE method leads to more conservative accounting results for total assets and income on average.Selecting the SE method accelerates the recording of oil and gas costs, and the rapid spending on these costs reduces reported income (Bryant, 2003;Misund et al., 2008).Hence, companies adopting the SE method are classified as income minimizers, while companies choosing the FC method are classified as income maximizers (Foster, 1980;Lilien and Pastena, 1982).Lilien and Pastena (1982) point out that oil and gas extraction risks affect firms' choices of oil and gas methods.Companies facing higher oil extraction and detection risks are more inclined to choose the FC accounting method to smooth operating income.According to the IPCC Assessment Reports, global warming increases the frequency and magnitude of extreme weather events.Severe weather risks challenge oil and gas exploration and production facilities (IPCC, 2023).The IPCC states that CCR brings multiple physical hazards to traditional oil and gas exploration equipment and pipeline transportation.Severe waves and storm surges also challenge deepwater floating systems and the production facilities on offshore oil platforms.Due to the adverse weather impact and the challenges brought by sea ice, subsea facilities can no longer enter some deep sea areas to exploit resources, and the drilling operations on offshore platforms have been forced to stop (IPCC, 2014(IPCC, , 2023;;Masnadi and Brandt, 2017).At the same time, onshore oil extraction facilities are suffering from permafrost instability and accelerated flooding from severe climate risks.Due to the extreme climate crisis on land and sea, actively exploring oil and gas companies tend to detect a higher proportion of dry wells, increasing early exploration failures and higher mining costs.To overcome these physical challenges, offshore and onshore oil drilling facilities and oil production investments can be exceptionally expensive (Erickson et al., 2018;Kang et al., 2020;Maxwell, 1997).
As exploratory risks and drilling costs increase, oil and gas companies choosing the SE method accelerate the recording of dry well costs and exploration expenses, which will damage corporate income and account balances (Lilien and Pastena, 1982;Misund et al., 2008).According to the debt covenant theory, oil and gas companies under global warming and extreme CCR will likely face contract default risks due to increased oil exploration costs.Contract default risks will limit corporate financial capability and production activities (Bhaskar et al., 2017;Pittman and Zhao, 2020;Robin et al., 2017).High exploration costs will pressure oil and gas companies as they may face higher debts and institutional investment constraints.Oil and gas companies with higher levels of CCR disclosure are more likely to realize the exploration risks posed by CCR and thus have an incentive to mitigate the debt covenant restrictions imposed by bondholders through internal accounting choices.Managers are willing to adopt looser accounting policies, such as the FC method that maximizes earnings, to help the firm reduce default risk costs and ease corporate financial constraints (Begley, 1990;Dyreng et al., 2022;Guttman and Marinovic, 2018).Furthermore, corporations consider CCR disclosure and income maximization as positive signs to reduce agency costs and help investors evaluate firms' ability to cope with CCR, thereby enhancing the potential market value of companies (Borghei, 2021;Brooks and Oikonomou, 2018;Eleftheriadis and Anagnostopoulou, 2015;Griffin et al., 2017;Matsumura et al., 2014).Therefore, based on the above discussion, our hypothesis is illustrated as follows: H1: Climate change risk disclosure is positively associated with the full-cost (FC) accounting method that maximizes income.

Sample Collection
Our research investigates U.S. oil and gas companies' CCR disclosure and assesses the impact of CCR disclosure on corporate E&D accounting choice.In 2016, the SEC investigated ExxonMobil for misleading investors about the CCR issue and its implications (Olson and Viswanatha, 2016).After the SEC dropped this investigation in 2018, we manually collected the CCR disclosure of all U.S. oil and gas companies for fiscal years 2014 through 2017 and observed the impact of these disclosures on firms' accounting choices.
To test our research questions, we collected all U.S. oil and gas companies for the selected years.Table 1 presents the sample selection criteria.After searching the oil and gas industry SIC code (2-digit SIC code 13) from the Compustat North America database, we collected 1034 annual observations of U.S. oil and gas companies from 315 unique companies.We first ruled out bankrupt or start-up companies and their sample observations.This brought our total sample to 804 complete annual company observations, representing 201 unique U.S. oil and gas companies for the year range.
Other observation samples we excludedwere either due to the lack of 10-K annual reports in the Edgar database or the unavailability of financial variable data needed for the empirical model in the Compustat database.Next, as shown in Table 1, the complete sample for testing Hypothesis 1 includes 479 firm-year observations from 129 unique oil and gas companies.Regarding the measurement of CCR disclosure, we manually collected all U.S. oil and gas companies' CCR-related disclosures from 10-K financial reports, CDP reports, corporate websites, and corporate sustainability reports.We searched for CCR-related content through multiple channels and compared these voluntary disclosures to understand the major risks and subsequentimpactsbrought by climate change.We first collected the financial reports of sample companies from the SEC's Edgar database and then collected the companies' CDP climate disclosure reports from the CDP database.Additionally, we collected corporate sustainability reports through the company websitesand the GRI's Sustainability Disclosure Database (discontinued in 2020).All other financial indicators of the sample companies were collected from Compustat and CRSP databases.

Research Design
Our hypothesis investigates the association between U.S. oil and gas firms' CCR disclosure and the accounting choice of adopting the SE or FC method, which either minimizes or maximizes income.The CCR disclosure proxy has been added to the accounting choice model according to Deakin (1979) and Lilien and Pastena (1982) as follows: where AC t is coded as 0 if the firm chooses the SE method and 1 if the firm chooses the FC method in year t.We examine the impact of CCR disclosure on accounting choice and use the CCRD t variable to measure the firm's CCR disclosure in 10-K annual reports, CDP reports, and corporate sustainability reports.For the remaining variables, https://ielas.org/ijbe/index.php/ijbeISSN (online) 2545-4137 96 REV t is the firm's revenue in year t to control firm size, according to Lilien and Pastena (1982).AGE t measures the firm's age as the consistency variable.EXPR t is calculated as oil and gas exploration expenditures divided by oil and gas operating income.We use EXPR t to test the exploratory risk of U.S. oil and gas firms.LEV t is the leverage variable, calculated as the firm's debt divided by shareholders' equity in year t, as shown below in Lilien and Pastena (1982).Lilien and Pastena (1982) predict whether firms choose accounting methods that maximize or minimize income, and the general model is as follows: where Z i = 1 if the company minimized cumulative income, 0 if the company maximized cumulative income, and X 1 = Revenue (political variable), X 2 =Age (consistency variable), X 3 = Exploratory risk, X 4 = Debt/Shareholders' equity (leverage).

Descriptive Statistics
Table 2 presents descriptive statistics.All continuous variables in the tests were winsorized at the 1% level at two tails of the distribution to remove the noise from outliers.Panel A of Table 2 shows summary statistics for the full sample of U.S. oil and gas companies from 2014 to 2017.Panel A indicates that most sample companies have chosen at least one CCR disclosure channel from the firm's 10-K annual disclosure, CDP report, or corporate sustainability report.In addition, the average value of REV t is much larger than the median, designating that in the sample of U.S. oil and gas firms, some large companies with high revenue increase the mean value of the REV t variable.Panel B of Table 2 divides sample observations into oil and gas firms with CCR disclosure (CCRD group) and oil and gas firms without CCR disclosure (non-CCRD group) in the four-year range.According to this panel, we find that the average REV t of the CCRD group is significantly higher than that of the non-CCRD group, denoting that high-revenue oil firms are more inclined to disclose climate risks compared with lowrevenue oil firms.Table 3 provides the correlation coefficients for all tested variables in the main hypothesis model.Because the variables in the empirical model include continuous and indicator variables, Pearson's and Spearman's rank correlations are shown below and above the diagonal in Table 3.  1) below (above) the diagonal.Correlations in bold are statistically significant at the 0.10 level.Coefficients in plain text are not significant (p > 0.10).All variables are defined in Appendix A.

Test Results
The empirical results of the hypothesis are shown in Table 4. Since some oil and gas companies did not disclose which oil and gas method was selected in their financial report footnotes, the complete sample has been reduced to 479 firm-year observations.According to the Lilien and Pastena (1982) model, our model uses probit regressions incorporating CCR disclosures and four influencing factors (REV t , AGE t , EXPR t , and LEV t ) to examine the effect on oil and gas accounting choices.From the results in Table 4, the coefficient of CCRD t is positive and significant (0.533, p< 0.05), which indicates that oil and gas firms with more CCR disclosure are more likely to choose the FC method to record oil and gas exploration.According to the assumptions in Section 3, the difficulty and costs of oil and gas exploration activities have increased dramatically due to global warming and natural risks.The direct impact of CCR on oil and gas drilling and exploration activities encourages companies to adopt loose accounting policies and the FC method to minimize financial volatility and smooth earnings.Stakeholder pressure based on the debt covenant theory also motivates oil and gas companies to choose the FC method to ease institutional constraints and maximize cumulative income simultaneously.For the remaining four variables, the coefficient of REV t is negative (-0.120), meaning that companies with higher revenue are more inclined to choose the SE method to record oil and gas exploration activities.Large oil and gas companies would avoid using income maximization methods because these companies may become political targets alongside industry giants (Bowen et al., 1981;Watts and Zimmerman, 1978).However; this result is not significant at the conventional level.In addition, the AGE t coefficient is negative and significant (-0.043, p< 0.05), denoting that older companies are more willing to choose the SE method.In fact, the SE method was the only option for old companies in the oil and gas industry until the late 1950s.Since transferring from the SE to the FC method involves varying levels of bookkeeping and other costs, some old oil and gas companies continued to use the SE method.
In contrast, the coefficient for exploration risk EXPR t is significantly positive (0.012, p< 0.10), illustrating that firms with higher exploration densities prefer the FC approach.Lilien and Pastena (1982) mention that the more aggressively oil and gas companies explore, the higher the probability of finding dry wells.These failure costs will increase significantly in the early stages of drilling activities.By adopting the FC approach, oil and gas firms experiencing high exploration risks can minimize corporate financial volatility.The coefficient significance of exploration risk is lower than the previous variables (CCRD t and AGE t ) at the 0.10 level.The firm's LEV t coefficient is positive (0.102), designating that oil and gas firms with higher leverage would face higher contract default risks.Therefore, managers have incentives to choose the FC method to maximize profits and ease financial burden.

ROBUSTNESS TEST
In the main model above, we applied the accounting choices of oil and gas companies as our dependent variable.Following Lilien and Pastena (1982), we collect oil and gas firms' dual accounting choices as the new dependent variable in the robustness test, including oil and gas accounting choices and adjustments to retained earnings.If the two accounting choices affect corporate income in the same direction, these samples are defined as dual choice maximization or dual choice minimization.The dual-choice maximization samples (labeled as 1) refer to firms that use the FC method and make negative adjustments to retained earnings.Conversely, the dual-choice minimization samples (labeled as 0) are firms that use the SE method and record zero or positive adjustments to retained earnings.Table 5 demonstrates the robustness test results of the oil and gas accounting choice model for Hypothesis 1.The coefficients are broadly consistent with our original predictions; only the leverage variable is in the opposite direction to previous findings.The coefficient of CCRD t in this dual-choice model remains positive and significant (1.798, p< 0.01), showing that the more CCR disclosures oil and gas firms prepare, the more likely firm managers are to choose income-maximizing accounting strategies.Since the dual accounting model considers both selection methods, the Pseudo R 2 in this model has been improved, which better explains the relationship between oil and gas firms' CCR disclosure and accounting choice.

CONCLUSION
With the increasing frequency and intensity of extreme weather and natural disasters, climate change poses a continuing and significant threat to the energy industry represented by oil and gas companies (Ceres, 2023;COP 27, 2022;TCFD, 2022).Numerous stakeholders have called on oil and gas companies to disclose CCRs that affect operations and credible steps to address them (Broadstock et al., 2018;Clarkson et al., 2008;Griffin et al., 2017;Liesen et al., 2015).Taking the SEC's investigation of ExxonMobil as a starting point, our study expands the research scope to include all U.S. oil and gas companies, delving into CCR disclosure and its impact on corporate accounting choices in the oil and gas industry.
Our findings show a significant positive correlation between CCR disclosure and the FC accounting choice of U.S. oil and gas companies, suggesting that oil and gas firms with more CCR disclosure are more inclined to choose the FC method to record oil and gas exploration activities.By preparing CCR disclosure, oil and gas firms demonstrate their awareness of CCRs and the attendant rise in exploratory and drilling risks.Such disclosures highlight firms' actions and contributions to mitigate CCR, such as managing carbon emissions and developing renewable clean energy.Our results denote that these corporations tend to adopt the FC method that maximizes earnings to show investors their ability to cope with increasing CCR and consequent costs in response to public concerns.This accounting practice also helps ease the financial constraints of business and reduces potential default risks.With the development and implementation of CCR disclosure guidelines and regulations, we look forward to the broader use of CCR reports to assess the potential disclosure impact on firm-level and industry-level operations.
Our study contributes at both academic and practical levels.From an academic perspective, our study points out a new direction for research on corporate CCR disclosure and its financial reporting impact (e.g., Downar et al., 2021;Lemma et al., 2020;Matsumura et al., 2014;Tang and Demeritt, 2018;Wang, 2023) by investigating the relationship between CCR disclosure and accounting choices under increasing global warming risks.From a practical perspective, our research provides corporate valuation assistance for oil and gas industry investors facing severe CCRs and helps raise public awareness of CCR disclosure and emission reductions.A growing number of independent global organizations, encompassing GRI, IPCC, and TCFD, have begun drafting guidelines to aid companies and organizations in preparing and disclosing CCR reports.Responding to the domestic SEC Climate Disclosure Guidelines (SEC, 2010(SEC, , 2022) ) and the international organizations' call for CCR disclosure, our study encourages U.S. oil and gas companies to disclose CCR information and corporate valuation strategies to gain broader support from investors and society.
There are fruitful potential research directions related to CCR disclosure and corporate reporting strategies that can be expanded in the future.For now, our study is mainly focused on the oil and gas-based energy industry, and future research samples can be extended to other industries closely related to CCR issues, such as the five critical sectors most affected by CCRs, namely electricity, oil and gas, transportation, insurance, and agriculture (Ceres, 2023).Besides the U.S. oil and gas companies analyzed in this study, we hope to explore carbon disclosure practices in other countries and track global climate change reporting trends in the future.Additionally, potential research may scrutinize the relevance and impact of CCR disclosure on various financial aspects and unearth more exciting opportunities for accounting research on climate change and ESG issues.

Table 2 Descriptive Statistics Panel A: Full Sample a
Table 2 provides descriptive statistics for sample observations.Panel A of Table 2 shows summary statistics for the full sample of U.S. oil and gas companies from fiscal years 2014 to 2017.Panel B of Table 2 partitions the sample observations into oil and gas firms with CCR disclosure (CCRD group) and without CCR disclosure (non-CCRD group) in the four-year range. a

Table 3 Correlation Coefficients Panel: Pearson / Spearman Correlation Matrix of the Main Model a
Table 3 presents the Pearson (Spearman) correlation matrix of Equation ( a

Table 5 Probit Regression Analysis: Effects of CCR Disclosure on Dual Accounting Choice with Adjustment to Retain Earnings
Lilien and Pastena (1982)bustness test results of the oil and gas accounting choice model.FollowingLilien and Pastena (1982), we collect U.S. oil and gas firms' dual accounting choices as the new dependent variable, including oil and gas accounting methods and adjustments to retained earnings.The dual-choice maximization indicator (equal to 1) refers to firms that use the FC method and make negative adjustments to retained earnings.Conversely, the dual-choice minimization indicator (equal to 0) refers to firms that use the SE method and record zero or positive adjustments to retained earnings.
***, **, * represents statistical significance at the 1%, 5% and 10% levels, respectively.Coefficients on year dummies are not reported for parsimony.All variable definitions are provided in Appendix A.