Next Article in Journal
Value Creation Through Environmental, Social, and Governance (ESG) Disclosures
Previous Article in Journal
Predicting Environmental Social and Governance Scores: Applying Machine Learning Models to French Companies
Previous Article in Special Issue
What Drives Cost System Sophistication? Empirical Evidence from the Greek Hotel Industry
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Re-Examination of the “Informational” Role of Non-GAAP Earnings in the Post-Reg G Period

1
School of Business, Belhaven University, Jackson, MS 39202, USA
2
Else School of Management, Millsaps College, Jackson, MS 39210, USA
3
Department of Finance, College of Business and Economics, Boise State University, Boise, ID 83725, USA
4
School of Accountancy, College of Business, Louisiana Tech University, Ruston, LA 71272, USA
5
School of Business, Mississippi College, Clinton, MS 39058, USA
*
Author to whom correspondence should be addressed.
J. Risk Financial Manag. 2025, 18(8), 414; https://doi.org/10.3390/jrfm18080414
Submission received: 19 May 2025 / Revised: 10 June 2025 / Accepted: 22 July 2025 / Published: 26 July 2025
(This article belongs to the Special Issue Innovations and Challenges in Management Accounting)

Abstract

In this study, we utilize a unique quarterly dataset of non-GAAP earnings to re-examine the “informational” role of non-GAAP earnings from the perspective of value relevance and earnings predictability in the post-Reg G period. We find that non-GAAP earnings are more value relevant and can better predict future operating earnings of a firm compared to equivalent GAAP earnings. Additionally, we also find empirical evidence suggesting that the difference in the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings can vary across but cannot be completely mitigated by firm-level characteristics, such as the market value of equity, accruals quality, analyst coverage, and managerial ability of a firm. Moreover, our supplementary analysis reveals that the superior value relevance and predictive power of non-GAAP earnings persist even after the SEC’s release of the Compliance and Disclosure Interpretations (C&DI) in 2010. Overall, our empirical evidence suggests a superior “informational” role of non-GAAP earnings to equivalent GAAP earnings in terms of valuation and predictability on future operating performance in the post-Reg G period.

1. Introduction

Non-GAAP earnings disclosures have become an increasingly common feature in contemporary financial reporting. This approach allows management to share performance metrics that they believe can better reflect the company’s core operations. While these earnings measures can improve operational transparency and offer valuable insights to investors, the aggressive exclusion of expense items in calculating such earnings measures can pose significant informational risks, for instance, management can manipulate non-GAAP earnings to present an overly favorable view of company performance (D. E. Black & Christensen, 2009; D. E. Black et al., 2018; Bhattacharya et al., 2003; Brosnan et al., 2024; Doyle et al., 2003).
The tension between the informativeness and manipulation roles of non-GAAP earnings stood out prominently during Groupon’s 2011 IPO, when the company aggressively excluded some material and recurring expenses in calculating its non-GAAP earnings measure, known as Adjusted Consolidated Segment Operating Income (Adjusted CSOI). By omitting these core operating expenses, Groupon painted an overly favorable picture of its financial health. As a result, the Securities and Exchange Commission (SEC) intervened, compelling Groupon to amend its filings and adjust its disclosures, thereby turning the case into a notable instance of aggressive non-GAAP reporting.
This incident occurred shortly after the SEC issued the revised Compliance & Disclosure Interpretations (C&DI) in 2010 to clarify how the SEC would interpret and expect companies to apply Regulation G, the first regulation governing non-GAAP reporting enacted in 2003 in response to the Sarbanes-Oxley Act. While Regulation G imposed strict requirements for transparency and reconciliation in non-GAAP reporting, the 2010 C&DI seemed to offer interpretive leniency. These revisions clarified existing rules and, in certain cases, eased the stringent enforcement of Regulation G, as well as granting companies greater discretion in presenting non-GAAP measures. The incident thereby marked a potentially pivotal shift in the regulatory environment governing firms’ non-GAAP reporting.
Importantly, the actions of Groupon and the related reactions of the SEC (noted above), as well as the pivotal shift in the regulatory environment for non-GAAP reporting, motivate the need to assess the relative importance of the “informational” role of non-GAAP earnings. While researchers have investigated this field in the past by comparing the “informativeness” of non-GAAP versus equivalent GAAP earnings, most of the literature uses research data in the pre-Reg G period and has reported inconsistent outcomes on which earnings convey better information to the market (D. E. Black et al., 2018). Thus, we are interested in using a quantitative method to re-examine the topic in the post-Reg G period. Specifically, we emphasize examining the value relevance and earnings predictability of non-GAAP earnings since these two attributes can link earnings to investor decision-making and future firm performance, thus providing a more comprehensive measure of earnings informativeness (FASB, 2010, 2018).
To construct the sample for our empirical analysis, we collect data from a variety of databases. To begin, we obtain the raw data of quarterly non-GAAP earnings disclosures from Bentley et al. (2018), which mainly contains the information regarding the per share non-GAAP earnings, or non-GAAP EPS. Additionally, we collect firms’ financial and price information from the Thomson Reuters Worldscope database to create the GAAP earnings measures, as well as other variables, for the analysis. Moreover, we merge the two datasets by the appropriate firm-time identifier (firm -year and -quarter), to construct the final dataset for our empirical analysis. Lastly, following previous literature, we exclude firms in financial (SIC 6000-6999) and utility (SIC 4400-4499) industries, as well as any observations with missing data for required variables.
By applying the Ohlson (1995) price model and related earnings predictability models, we compare how well GAAP and non-GAAP earnings predict a firm’s future value and operating performance. We find that non-GAAP earnings are more value relevant and have better predictability towards future operating earnings of a firm than equivalent GAAP earnings. Additionally, we also find empirical evidence suggesting that the difference in the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings can be altered but cannot be completely mitigated by firm-level characteristics such as market value of equity, accruals quality, analyst coverage, and managerial ability of a firm. Specifically, the difference in the value relevance between two earnings measures is more pronounced when a firm has a higher market value of equity, lower accruals quality, higher analyst coverage, and higher managerial ability. In contrast, the difference in the earnings predictability between two earnings measures is more evident when a firm has a lower market value of equity, lower accruals quality, higher analyst coverage, and lower managerial ability. Moreover, our supplementary analysis reveals that non-GAAP earnings are more value relevant and can better predict future operating earnings of a firm than equivalent GAAP earnings, even after the release of C&DI in 2010. Overall, our empirical evidence suggests a superior “informational” role of non-GAAP earnings to equivalent GAAP earnings in terms of valuation and predictability of future operating performance.
This study makes several important contributions. Our empirical evidence confirms the superior “informational” role of non-GAAP earnings to GAAP earnings (Bhattacharya et al., 2003; L. D. Brown & Sivakumar, 2003; Doyle et al., 2003), even after controlling for certain firm-level characteristics and after the release of the Compliance and Disclosure Interpretations in 2010. Hence, our empirical evidence helps address the concerns from market participants and regulators regarding the potential decline in the quality of non-GAAP earnings following the implementation of formal guidance on non-GAAP reporting (D. E. Black et al., 2018; Brosnan et al., 2024).
While some prior research suggests that non-GAAP earnings disclosures may be self-serving, rather than incrementally informative (Bhattacharya et al., 2003; D. E. Black & Christensen, 2009; Heflin & Hsu, 2008), our empirical evidence lends support to the opposite notion that non-GAAP earnings are informative and useful to investors (Bhattacharya et al., 2003; Doyle et al., 2003; Lougee & Marquardt, 2004).
To some extent, our empirical evidence implies that regulatory efforts to govern non-GAAP reporting practices in the U.S. (e.g., Reg G and the subsequent C&DIs) have been at least partially successful. The success is evident in both the increased adoption of non-GAAP reporting by firms and the improvements in the quality of these earnings (Coleman & Usvyatsky, 2015; Arena et al., 2021). Although regulations of non-GAAP reporting have improved the quality of non-GAAP reporting, continued regulatory oversight remains crucial to enhance consistency, promote transparency, and reduce the risk of misleading adjustments of non-GAAP reporting (D. E. Black et al., 2018; Brosnan et al., 2024). The SEC’s C&DI updates in 2016 and 2022 reinforce this need by clarifying requirements on prominence, reconciliation, and the prohibition of tailored accounting principles that can distort investor perception.
Moreover, our empirical findings have practical implications for both equity investors and firms that report non-GAAP earnings. For investors, non-GAAP earnings can be used as a more reliable tool to forecast future operating performance (e.g., predict future operating earnings), particularly for firms with a lower market value of equity, lower accruals quality, higher analyst coverage, and lower managerial ability. For companies, given that investors place greater weight on self-reported non-GAAP earnings to evaluate equity value, particularly for firms with a higher market value of equity, lower accruals quality, higher analyst coverage, and higher managerial ability, it is crucial to ensure the quality and credibility of these disclosures to maximize shareholder value.
The remainder of the paper is organized as follows: Section 2 introduces relevant background information and develops the hypotheses; Section 3 presents the sample construction and the research methodology; Section 4 reports the main empirical results; and Section 5 summarizes the conclusions and limitations of our research and suggests future research opportunities.

2. Background Information and Hypothesis Development

2.1. Background Information

In the United States of America, public companies are required by the SEC to report earnings according to the Generally Accepted Accounting Principles (GAAP). GAAP, therefore, represents the “gold” standard for financial reporting. However, there have been skeptics on whether GAAP earnings can provide investors with reliable and sufficient information to access companies’ performance and to make investment decision, an example of which is the requirement for immediate recognition of research and development (R&D) expense that potentially leads to the overestimation of periodic expenses and the underestimation of “intangible” assets for a company (e.g., Holthausen & Watts, 2001).
Since the early 1990s, some public companies in United States started to voluntarily report in their Management’s Discussion &Analyzes (MD&A), earnings releases, or other communications selected earnings measures that do not conform with GAAP rules as a supplement to the “noisy” GAAP earnings, and such earnings measures are referred to as non-GAAP earnings. To develop non-GAAP earnings measures, companies normally start with their equivalent GAAP earnings measures but adjust for what they believe to be non-recurring or non-cash items, such as stock-based compensation expenses, amortization and impairment of intangible assets, restructuring charges, and realized gains or losses from sales of assets (D. E. Black et al., 2018; Brosnan et al., 2024). One of the primary purposes for the use of non-GAAP reporting is to better inform the stakeholders of “the management’s standpoint of business” or sustainable core earnings. However, with non-GAAP reporting being a voluntary and potentially self-serving reporting firm behavior, some non-GAAP earnings reported by companies may be subjective, lack of transparency and consistency, non-comparability between firms, or even misleading (Bhattacharya et al., 2003; D. E. Black & Christensen, 2009; Heflin & Hsu, 2008). A common concern regarding non-GAAP reporting is that managers may aggressively exclude recurring items in calculating non-GAAP earnings to meet or beat strategic earnings targets or expectations (Bhattacharya et al., 2003; D. E. Black & Christensen, 2009; Doyle et al., 2013; E. L. Black, 2016).
The aggressive application of non-GAAP earnings reporting, particularly in the early stages of such reporting practices (i.e., before 2003), had drawn attention from official accounting standard rule-setting bodies (Securities and Exchange Commission, 2001). The SEC adopted strict rules and amendments in 2003, most notably Regulation G and Item 10(e) of Regulation S-K, to govern the use of non-GAAP earnings reporting. These rules posted strict stipulations on the use of non-GAAP reporting, requiring a quantitative reconciliation of the difference between non-GAAP and equivalent GAAP earnings (Final Rule: Conditions for Use of Non-GAAP Financial Measures 2003). Prior studies reported an initial decrease in the frequency but an increase in the quality of non-GAAP reporting subsequent to the strict regulations (Entwistle et al., 2006; Heflin & Hsu, 2008; E. L. Black et al., 2017; Kolev et al., 2008).
Interestingly, Reg G only applied to all public communications of an issuer, but did not require any non-GAAP earnings to be included in reports filed with the SEC. Therefore, this difference in application resulted in inconsistencies in non-GAAP earnings disclosures within SEC filings and other information made available to the public. As an effort to improve the quality and usefulness of information in reports filed with the SEC, the SEC issued the revised Compliance and Disclosure Interpretations (C&DI) in 2010, 2016, and 2022. These revisions, particularly the one in 2010, moderately loosened the reporting standards on the non-GAAP reporting, and that, in turn, led to the rebound of non-GAAP reporting (Bentley et al., 2018; D. E. Black et al., 2012). Additionally, there had been an increasing concern about non-GAAP earnings being misused to artificially inflate earnings and misleading investors (Morgan et al., 2018).
In recent years, non-GAAP reporting has become a common practice for companies. The growth of non-GAAP reporting in the past decades has reflected a widespread acceptance of such non-standard reporting as a method to evaluate a company’s performance; however, it is still a debatable topic whether non-GAAP earnings can fulfill the “informational” role more effectively than GAAP earnings in painting a better operational picture of a company. Thus, while the regulatory framework surrounding non-GAAP reporting has evolved to accommodate the informational needs of stakeholders, it continues to demand careful balance to prevent misuse and maintain investor trust.

2.2. The Informativeness of Earnings

In the accounting framework, the informativeness of earnings refers to how well earnings convey useful information to investors and other stakeholders for decision-making. Two widely studied dimensions of informativeness are value relevance and earnings predictability, both of which align with the relevance qualitative characteristic in the FASB Conceptual Framework (FASB, 2010, 2018). Value relevance captures the extent to which accounting earnings are associated with market values or stock prices, often assessed through regression models linking earnings and book value to equity prices (e.g., Ohlson, 1995; Barth et al., 2001). If earnings are value relevant, then they can provide information useful for equity valuation, suggesting that earnings reflect economically meaningful performance measures. Meanwhile, earnings predictability refers to the ability of current and past earnings to predict future earnings or cash flows, a trait valued by investors for its role in reducing uncertainty and improving forecasts (J. Francis et al., 2004).
A key strength of value relevance and earnings predictability lies in their close alignment with how users interact with financial statements, making them both theoretically grounded and empirically testable. Value relevance provides a direct link between accounting information and investor behavior by measuring how well earnings and book values explain variations in stock prices. This empirical association sheds light on the extent to which accounting standards can produce useful information for equity valuation, thereby contributing to the debates on standard-setting. Meanwhile, earnings predictability fulfills investors’ needs to forecast future performance, one of the core objectives of financial reporting. Predictable earnings facilitate the use of valuation models, reduce estimation uncertainty, and often signal operational stability and transparency in financial reporting. Together, these two concepts offer a powerful framework for evaluating the decision usefulness of earnings, combining market-based validation with forward-looking informational value.
Despite their strengths, both value relevance and earnings predictability have limitations. Value relevance relies on the assumption of market efficiency and can be affected by non-accounting factors like investor sentiment or macroeconomic noise, potentially overstating the informativeness of earnings (Holthausen & Watts, 2001). Hence, a strong association with price does not always mean that the earnings are of high quality. Similarly, predictable earnings may result from managerial smoothing rather than economic stability, masking underlying risks or volatility (Dechow & Dichev, 2002). As a result, the use of value relevance and earnings predictability to proxy for the informativeness or quality of earnings should be interpreted cautiously and may need to be complemented with other broader measures of earnings informativeness.

2.3. Hypothesis Development

The “informational” role of non-GAAP reporting is rooted in its potential to enhance transparency and improve the relevance of financial disclosures. According to the Conceptual Framework of Accounting (FASB, 2010, 2018), the primary objective of financial reporting is to provide information that is useful to existing and potential investors, lenders, and other creditors in making decisions about providing resources to the entity. Thus, the economic function of financial accounting is to reduce information asymmetry between firms and external capital providers and to facilitate more efficient resource allocation in capital markets (Ball & Brown, 1968).
Complementing this notion, signal theory (Spence, 1973) posits that firms have incentives to voluntarily disclose high-quality, value-relevant information to differentiate themselves from lower-quality firms. Under these frameworks, when non-GAAP disclosures are transparent, faithfully represented, and consistently defined over time, they can serve as credible signals of management’s private information about firm performance and future prospects.
Many investors and analysts find non-GAAP earnings useful for understanding management’s view of normalized performance and making more informed decisions about a firm’s prospects (D. E. Black et al., 2018; Brosnan et al., 2024). Additionally, academic research also documents empirical evidence suggesting that certain non-GAAP measures are incrementally informative, especially when they are clearly reconciled to GAAP results and are consistently defined over time (e.g., L. D. Brown & Sivakumar, 2003; Bradshaw & Sloan, 2002; Bhattacharya et al., 2003; Kolev et al., 2008; Jennings & Marques, 2011). For example, L. D. Brown and Sivakumar (2003) find that operating income reported by managers and analysts is more value relevant than the ones derived from Standard and Poor’s Compustat. Bradshaw and Sloan (2002) show that I/B/E/S actual EPS, as a proxy for non-GAAP earnings, elicits a stronger market reaction than GAAP earnings. Similarly, Bhattacharya et al. (2003) find that pro forma earnings are more informative and persistent, better reflecting firms’ core performance. Supporting these findings, Kolev et al. (2008) and Jennings and Marques (2011) report that, when used consistently and transparently, non-GAAP disclosures improve the informativeness of earnings announcements and help investors better assess future performance. Collectively, these studies suggest that well-constructed non-GAAP measures can enhance the decision usefulness of financial reporting.
The preceding discussion leads to the first alternative format of our main hypothesis:
H1a. 
Compared with equivalent GAAP earnings, non-GAAP earnings fulfill the “informational” role more effectively to convey information to the market.
However, skeptics of non-GAAP reporting also raise concerns about earnings manipulation and the potential to mislead stakeholders. Because these measures are not standardized, companies have significant discretion in how they define and present them. This flexibility can lead to overly aggressive exclusions that portray a more favorable financial position than reality.
From the perspective of agency theory (Jensen & Meckling, 1976), managers, as agents, may exploit the discretion allowed in non-GAAP reporting to serve their own interests at the expense of principals (shareholders). When agency conflicts are present, managers may use non-GAAP measures aggressively to obscure poor performance or meet earnings expectations, rather than to improve transparency.
Supporting these concerns, a different stream of academic research finds evidence suggesting that managers adjust non-GAAP earnings to meet strategic targets or to mislead investors, particularly when GAAP earnings fall short of expectations (e.g., D. E. Black & Christensen, 2009; Doyle et al., 2003; Lougee & Marquardt, 2004). For instance, D. E. Black and Christensen (2009) show that managers often exclude recurring expense items, in addition to the one-time expense items, to meet strategic targets. Doyle et al. (2003) find that large exclusions in non-GAAP earnings predict lower future cash flows, which investors may not fully recognize at the time of disclosure. Lougee and Marquardt (2004) report that firms missing GAAP benchmarks or with lower earnings quality are more likely to issue pro forma earnings, an alternative descriptor for non-GAAP earnings in the earlier literature. Subsequent studies (e.g., D. E. Black et al., 2012; Christensen et al., 2014; Frankel et al., 2011) further reinforce these findings, suggesting that aggressive non-GAAP adjustments can mislead investors and increase information asymmetry.
The preceding discussion leads to the second alternative format of our main hypothesis:
H1b. 
Compared with equivalent GAAP earnings, non-GAAP earnings fulfill the “informational” role less effectively to convey information to the market.

3. Sample Construction and Research Methodology

3.1. Sample Construction

To construct the sample for our empirical analysis, we collect data from a variety of databases. To begin, we obtain the raw data of quarterly non-GAAP earnings disclosures from Bentley et al. (2018). While the detailed sample procedure can be found in Appendix B of Bentley et al. (2018), we briefly discuss the key sample procedures below. Specifically, Bentley et al. (2018) starts by downloading the text of all 8-K documents from the SEC’s EDGAR database and retaining only those that are likely to be earnings announcements. When multiple documents appear as earnings announcements for a given firm-quarter, the document closest to the earnings announcement date, with the highest proportion of earnings-related content, is retained. To identify the disclosure of non-GAAP EPS, Bentley et al. (2018) employs a programmatic search of the earnings announcement text, targeting sentences or table rows that meet three criteria: (1) containing an EPS metric; (2) featuring a non-GAAP trigger word or phrase such as “adjust”, “exclude”, “non-GAAP”; and (3) specifically discussing the current quarter. If a disclosure meets these criteria, the non-GAAP EPS value is extracted either programmatically or through manual review by trained research assistants (RAs). If discrepancies arise between the manager-reported non-GAAP EPS and the I/B/E/S non-GAAP EPS, particularly if differences exceed $0.50, further manual verification is conducted. In cases where no non-GAAP sentences are identified but the I/B/E/S non-GAAP EPS is found within the announcement, Bentley et al. (2018) assume that managers reported non-GAAP EPS that matches the I/B/E/S figure. Conversely, if the I/B/E/S non-GAAP metric is absent, it is assumed that only GAAP EPS was reported. Observations are excluded from the sample when non-GAAP EPS cannot be identified. The data contains about 115,370 firm quarterly non-GAAP earnings disclosures from 2003 to 2016, which, in turn, serves as the sample period for the study.
In the next step, we collect firms’ financial and price information from the Thomson Reuters Worldscope database to create the equivalent GAAP earnings measures, as well as other required variables, for the study. Lastly, we merge the preceding two datasets by the proper unique firm-time identifier (firm-year and -quarter), to construct the final dataset for the empirical study. Following the previous literature, we exclude firms in financial (SIC 6000-6999) and utility (SIC 4400-4499) industries, as well as any observations with missing information for required variables. The final sample contains a total of 31,366 firm-quarter observations. Table 1 describes the detailed sample construction process.

3.2. Research Methodology

To test the “informational” role of non-GAAP earnings, we adopt two major tests in the study: the value relevance and the earnings predictability tests.
For the value relevance test, we adopt the Ohlson (1995) model as the baseline model.1 Additionally, incorporating with the framework of its subsequent refinements (Feltham & Ohlson, 1995, 1996; Ohlson, 1999), we also include some additional control variables into the regression model to control for other firm-level characteristics that can potentially affect the firm valuation. The regression model is shown as follows:
P r i c e i , t = β 0 + β 1 B o o k   v a l u e   p e r   s h a r e i , t + β 2 E a r n i n g s   p e r   s h a r e i , t + C o n t r o l i , t + i , t ,
where P r i c e i , t is the equity price at three months after the quarter-end, adjusted for stock splits and stock dividends; B o o k   v a l u e   p e r   s h a r e i , t is calculated by the common equity scaled by the number of shares outstanding; E a r n i n g s   p e r   s h a r e i , t is either the per share non-GAAP earnings (if non-GAAP earnings, the regression model also includes the non-GAAP exclusions) or the equivalent GAAP earnings; C o n t r o l i , t includes firm size (natural log of one plus the book value of the assets), earnings volatility (rolling standard deviation of return on assets), book-to-market equity ratio (book value of equity scaled by market value of equity), sales growth (sales growth rate in the current year), and operating loss (whether the firm has operating loss in the year); and i , t is the error term of the regression. Please refer to Appendix A for the definitions of the variables.
We employ the preceding regression model separately using the non-GAAP and equivalent GAAP earnings per share in each regression and compare the coefficients of the two earnings between the two regressions. If non-GAAP earnings are more value relevant than equivalent GAAP earnings, or stated differently, if the capital market relies more on non-GAAP earnings than the GAAP earnings to price the equity of the firms, then we expect the β1 for the non-GAAP earnings to be significantly higher than the one for the GAAP earnings. For the value relevance test, we also address the overall explanatory power of the regression model (R2) as it presents the combined explanatory power of the equity price by all variables in the regression model.
For the earnings predictability test, we follow prior research (Kolev et al., 2008; Frankel et al., 2011; N. C. Brown et al., 2012; Doyle et al., 2003) and estimate the following regression model:
F u t u r e   O p e r a t i n g   E a r n i n g s i , t + 1 = α + β 1   E a r n i n g s i , t + C o n t r o l i , t + i , t ,
where F u t u r e   O p e r a t i n g   E a r n i n g s i , t + 1 is either the future operating income or cash flow from quarter t + 1 to quarter t + 4 scaled by total assets; E a r n i n g s i , t is either the per share non-GAAP earnings (if non-GAAP earnings, the regression model also includes the non-GAAP exclusions) or equivalent GAAP earnings times the number of shares outstanding and scaled by total assets; C o n t r o l i , t includes the same variables as those used in the value relevance test.
Like the value relevance test, we conduct regression analyses separately using the non-GAAP and equivalent GAAP earnings in each regression and compare the coefficients of the two earnings between the two regressions. If non-GAAP earnings have better predictability towards the future operating earnings than equivalent GAAP earnings, then we expect the β1 for the non-GAAP earnings to be significantly higher than the one for the GAAP earnings.

4. Main Empirical Results

4.1. Summary Statistics and Correlation Matrix

Table 2 reports the summary statistics (in Panel A) and the Pearson correlation matrix between the key variables (in Panel B). According to the summary statistics in Panel A, the 31,366 firms included in the sample have an average per share book value of equity and future equity price of 11.39 and 32.09 dollars. Firms’ future operating income and future operating cash flow equal to about 2% and 19% of their total assets, respectively. Further, firms’ average per share GAAP earnings are about 0.19 below their non-GAAP earnings, suggesting a difference in total earnings of about 1% of the total assets. Interestingly, the average quarterly sales growth rate for the firms is 4%, while 29% of the firms report an operating loss in the sample.
According to the summary statistics in Panel B, the dependent variables of Future Price, Future Operating Income, and Future Operating Cash Flow are highly correlated (p < 0.01). Similarly, the key independent variables of GAAP Earnings per share, Non-GAAP Earnings per share, Non-GAAP Exclusions per share, and Book Value per share are highly correlated with the dependent variables as well. It is worth noting that the pairwise correlations between Non-GAAP Earnings per share and the dependent variables are higher than the ones between GAAP Earnings per share and the dependent variables, providing the initial evidence that non-GAAP earnings reported by firms may be more value relevant and have better predictability towards future operating earnings than equivalent GAAP earnings. The correlations between other control variables and the dependent variables have the expected signs. For example, ROA sd and Book to Market Ratio both have a significantly negative correlation (p < 0.01) with the dependent variables.

4.2. Multi-Variant Regressions

4.2.1. Comparison of the Value Relevance and Earnings Predictability Between Non-GAAP and Equivalent GAAP Earnings in the Entire Sample

Table 3 reports the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings in the entire sample, in which Panels A and B report results for the value relevance and earnings predictability tests, respectively. To test the “informational” role of non-GAAP earnings, we emphasize comparing the coefficients on GAAP earnings (defined as GAAP Earnings per share in Panel A and as GAAP Earnings in Panel B) and non-GAAP earnings (defined as Non-GAAP Earnings per share in Panel A and as Non-GAAP Earnings in Panel B) in both panels, respectively.
As shown in Panel A, both GAAP and non-GAAP earnings have significantly positive associations (p < 0.01) with future equity prices (Future Price). Specifically, holding other factors constant at mean values, on average, a dollar increase in GAAP earnings and non-GAAP earnings is associated with an increase of 14.48 dollars and 40.90 dollars in the future equity price of a firm. The result of the T-test in the following column confirms that there is a significant difference (p < 0.01) between the mean value of the coefficient on GAAP earnings (GAAP Earnings per share) and the mean value of the coefficient on non-GAAP earnings (Non-GAAP Earnings per share). Stated differently, non-GAAP earnings are more value relevant compared to their equivalent GAAP earnings. Additionally, the analysis of the adjusted R2 also confirms that the overall explanatory power (0.506) of the regression in column (2) is higher compared to the corresponding value (0.423) of the regression in column (1). The empirical results from Panel A thus reject the null hypothesis and support the alternative hypothesis H1a from the perspective of the value relevance of the earnings.
For Panel B, columns (1) and (2) report results for the earnings predictability tests with the dependent variable as the future operating income (Future Operating Income), whereas columns (3) and (4) present results for the ones with the dependent variable as the future operating cash flow (Future Operating Cash Flow). Like the result in Panel A, both GAAP earnings and non-GAAP earnings have significantly positive associations (p < 0.01) with the future operating income (Future Operating Income) and the future operating cash flow (Future Operating Cash Flow). Specifically, columns (1) and (2) show that a one dollar increase in GAAP earnings and non-GAAP earnings is associated with the increase of 0.722 dollars and 2.511 dollars in the future operating income of a firm while columns (3) and (4) unveil that a one dollar increase in GAAP earnings and non-GAAP earnings is associated with the increase of 1.661 dollars and 6.069 dollars in the future operating cash flow of a firm. The results of the T-test in the following columns also confirm the significant difference (p < 0.01) between the mean value of the coefficient on GAAP earnings (GAAP Earnings) and the mean value of the coefficient on non-GAAP earnings (Non-GAAP Earnings). Overall, the empirical results from Panel B reinforce the findings from Panel A to reject the null hypothesis and support the alternative hypothesis H1a from the perspective of earnings predictability.
It is worth noting that the results (un-tabulated) remain identical when we apply the Fama and MacBeth (1973) cross-sectional regression with the Newey–West standard error. Moreover, the control variables on both Panels have the expected signs. For example, the coefficients on the book-to-market equity ratio (Book to Market Ratio) are significantly negative, indicating that earnings from value firms are less relevant to future equity prices.

4.2.2. The Value Relevance and Earnings Predictability of Non-GAAP and Equivalent GAAP Earnings: Conditioning on Firm-Level Characteristics

The analyses in the preceding section suggest that there are substantial differences between non-GAAP and equivalent GAAP earnings in terms of value relevance and earnings predictability. In this section, we investigate whether such differences can vary across various firm-level characteristics, such as market value of equity, accruals quality, analyst coverage, and managerial ability of a firm, and report the results in Table 4, Table 5, Table 6 and Table 7, respectively. Like the format in Table 3, Panels A and B of each table report results for the value relevance and the earnings predictability tests, respectively. It is worth noting that the total observations within each cross-sectional regression vary due to the availability of the data for each firm-level characteristic.
Table 4 repeats the analysis of Table 3 but splits the sample into two sub-samples based on whether a firm has a market value of equity above or below the median market value of equity in the sample each quarter. The market value of equity can either positively or negatively affect a firm’s earnings quality. On one hand, firms with a higher market value of equity tend to make income-decreasing adjustments to their earnings when facing greater regulatory scrutiny, but such income-decreasing adjustments can negatively affect the quality of earnings (Watts & Zimmerman, 1986). On the other hand, the market value of equity can also be a proxy for information uncertainty since firms with a lower market value of equity are less diversified and have less available information for the market (Zhang, 2006). It is thus plausible that firms with a lower market value of equity can more aggressively inflate their earnings due to higher internal control deficiencies or lower disclosure preparation costs (Zhang, 2006; Doyle et al., 2007).
Panel A of Table 4 reveals that the value of the coefficient on GAAP earnings (GAAP Earnings per share) increases from 2.473 in the low-market-value-of-equity sub-sample to 18.09 in the high-market-value-of-equity sub-sample. Similarly, the value of the coefficient on non-GAAP earnings (Non-GAAP Earnings per share) increases from 12.18 in the low-market-value-of-equity sub-sample to 40.90 in the high-market-value-of-equity sub-sample. The result is consistent with the findings from extant literature that firms with a high market value of equity report more value relevant non-GAAP and equivalent GAAP earnings than firms with a low market value of equity (Zhang, 2006; Doyle et al., 2007). Interestingly, the magnitude of the difference for the coefficients between non-GAAP and equivalent GAAP earnings is larger in the high-market-value-of-equity sub-sample, although the significance level of the difference remains identical (p < 0.01) in both sub-samples. Further, the analysis of the adjusted R2 confirms that the overall explanatory powers of the regression in columns (2) and (4) are higher compared to the corresponding values in columns (1) and (3), thus suggesting that the difference in the value relevance between non-GAAP and equivalent GAAP earnings is not completely mitigated by the market value of equity.
Panel B of Table 4 also confirms the findings from Panel A that the value of the coefficients on both GAAP earnings (GAAP Earnings) and non-GAAP earnings (Non-GAAP Earnings) is higher in the high-market-value-of-equity sub-sample than the value of the corresponding coefficients in the low-market-value-of-equity sub-sample. Panel B, however, documents a slightly different result to the one in Panel A that the magnitude of the difference for the coefficients between non-GAAP and equivalent GAAP earnings is larger in firms with a low market value of equity than in firms with a high market value of equity, although the difference remains highly statistically significant (p < 0.01) in both sub-samples.
Overall, the empirical evidence from Table 4 reveals that the market value of equity of a firm can alter but cannot completely mitigate the phenomenon that non-GAAP earnings are more value relevant and have higher predictive power towards future operating earnings of a firm than equivalent GAAP earnings.
Table 5 repeats the analysis in Table 3 but with separate estimates for firms with above and below median values of accruals quality in a year. When a firm has lower quality accruals, the firm is normally viewed as reporting lower quality GAAP earnings. It is reasonable to assume that if a firm reports lower quality GAAP earnings, investors may rely more heavily on non-GAAP earnings to assess the value of the firm, and the non-GAAP earnings of the firm are therefore more relevant to the equity value. Similarly, lower quality GAAP earnings may contain a lot of non-recurring “noisy” items and thus have lower predictability towards future operating income or cash flow. It is thus plausible to assume that non-GAAP earnings have better predictability towards the future operating earnings under this circumstance. However, if managers of a firm intentionally report low-quality GAAP earnings to manipulate earnings, the same group of managers could have reported low-quality non-GAAP earnings as well. Thus, it is an empirical question to examine the conditioning effect of accruals quality on the value relevance and earnings predictability of the non-GAAP and equivalent GAAP earnings.
To create the proxy for accruals quality, we adopt the method from Dechow and Dichev (2002) by considering accruals as a function of past, present, and future cash flows in a regression model. Additionally, we incorporate the thoughts of McNichols (2002) by including in the regression model the change in sales and the level of property, plant, and equipment. Specifically, we estimate the following cross-sectional regression by each industry based on the 2-digit SIC code and year.
Δ W C t = β 0 + β 1 C F O t 1 + β 2 C F O t + β 3 C F O t + 1 + β 4 Δ R e v t + β 5 P P E t + i t
where ΔWC is the change in working capital from year t − 1 to year t, CFO is the cash flow from operations, ΔREV is the change in sales from year t − 1 to year t, and PPE presents the level of property, plant, and equipment at year t. All variables are scaled by the average total assets between the year t − 1 and t. To include each industry-year into the regression model, we require the industry-year to have at least ten observations. After estimating the above regression, we generate the residuals and calculate the standard deviation of the residuals over a rolling window of three years as the measure of accruals quality. The higher value for the variable indicates the lower quality of earnings.
Panel A of Table 5 reports a similar pattern to the one in Panel A of Table 4 that the value of the coefficient on GAAP earnings (GAAP Earnings per share) increases from the low-quality sub-sample to the high-quality sub-sample, whereas the value of the coefficient on non-GAAP earnings (Non-GAAP Earnings per share) decreases from the low-quality sub-sample to the high-quality sub-sample. Therefore, the magnitude of the difference for the coefficients between non-GAAP and equivalent GAAP earnings is larger in the low-quality sub-sample but the significance level of the difference remains identical (p < 0.01) in both sub-samples. The analysis of the adjusted R2 also confirms that the overall explanatory powers of the regression in columns (2) and (4) are higher compared to the corresponding value in columns (1) and (3), thus suggesting that the difference in the value relevance between non-GAAP and equivalent GAAP earnings is not completely mitigated by the accruals quality of a firm. The empirical evidence partially corroborates the previous conjecture that investors rely more heavily on the non-GAAP earnings to estimate the equity value of a firm if the firm reports low-quality GAAP earnings.
Panel B of Table 5 reports that both values of the coefficients on GAAP earnings (GAAP Earnings) and on non-GAAP earnings (Non-GAAP Earnings) increase from the low-quality sub-sample to the high-quality sub-sample, except that the ones between columns (8) and (6) are nearly indifferent from each other. Further, the magnitude of the difference for the coefficients between non-GAAP and equivalent GAAP earnings is larger in the low-quality sub-sample while the significance level of the difference remains identical (p < 0.01) in both sub-samples.
Overall, the empirical evidence from Table 5 unveils that the difference in the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings is more evident when a firm reports lower-quality GAAP earnings, however, the phenomenon persists that non-GAAP earnings are more value relevant and have better predictability towards future operating earnings of a firm than equivalent GAAP earnings, even when accruals quality of a firm is high. It is worth noting that the result (un-tabulated) remains identical when we use alternative measures for accruals quality, such as the one estimated from the performance matched modified Jones (1991) model (Dechow et al., 1995; Kothari et al., 2005).
Table 6 repeats the analysis of Table 3 but splits the sample into two sub-samples based on whether the number of analysts following a firm is above or below the median value in the sample each year. The extant literature demonstrates that greater analyst coverage serves as an external governance mechanism to monitor a firm’s financial reporting practice, including non-GAAP reporting (Healy & Palepu, 2001; Yu, 2008; Bradshaw et al., 2018; Christensen et al., 2021). Similarly, fundamental corporate theories (Jensen & Meckling, 1976; Fama, 1980) suggest that stronger governance of a firm can better monitor CEOs’ behaviors, thus reducing their rent-seeking behaviors. It is thus reasonable to conjecture that stronger analyst coverage can improve the quality of earnings and potentially increase the value relevance and the earnings predictability of both non-GAAP and equivalent GAAP earnings. However, other factors, such as whether investors care about a firm’s quality of governance, can also affect the value relevance of earnings. Hence, it is an empirical question to examine whether analyst coverage can moderate the value relevance and earnings predictability of the non-GAAP and equivalent GAAP earnings. To further examine the topic, we obtain the analyst coverage information from the IBES dataset and apply it to the empirical analysis.
Panel A of Table 6 shows that both values of the coefficients on GAAP earnings (GAAP Earnings per share) and on non-GAAP earnings (Non-GAAP Earnings per share) increase from the low analyst coverage sub-sample to the high analyst coverage sub-sample. The empirical evidence thereby confirms that the value relevance of both non-GAAP and equivalent GAAP earnings improves when analyst coverage is higher. Interestingly, the value relevance of non-GAAP earnings increases more than the value relevance of GAAP earnings when more analysts cover a firm, thus leading to an increase in the magnitude of the difference in the coefficients between non-GAAP and equivalent GAAP earnings in the high analyst coverage sub-sample. However, in both sub-samples, the significance level of the difference remains identical (p < 0.01). The analysis of the adjusted R2 also confirms that the overall explanatory powers of the regression in columns (2) and (4) are higher compared to the corresponding value in columns (1) and (3), thus suggesting that the difference in the value relevance between non-GAAP and equivalent GAAP earnings remains statistically identical after controlling for the analyst coverage of a firm. The empirical evidence lends support to the previous conjecture that the value relevance of earnings improves when analyst coverage is higher but does not support the notion that the difference in value relevance between non-GAAP and equivalent GAAP earnings is completely mitigated by the analyst coverage.
Panel B of Table 6 reports a similar pattern to the one in Panel A of Table 6. Specifically, both values of the coefficients on GAAP earnings (GAAP Earnings) and on non-GAAP earnings (Non-GAAP Earnings) increase from the low analyst coverage sub-sample to the high analyst coverage sub-sample. Further, the magnitude of the difference in the coefficients between non-GAAP and equivalent GAAP earnings is larger in the high analyst coverage sub-sample, whereas the significance level of the difference remains identical (p < 0.01) in both sub-samples.
Overall, the empirical evidence from Table 6 reveals the difference in the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings is more pronounced when more analysts cover a firm. Admittedly, the phenomenon persists that non-GAAP earnings are more value relevant and have better predictability towards future operating earnings of a firm than equivalent GAAP earnings, even when analyst coverage of the firm is low.
Table 7 repeats the analysis of Table 3 but splits the sample into two sub-samples based on whether a firm has above and below median values of the managerial ability in a year.2 The extant literature reveals that management quality has a positive impact on corporate policies and outcomes, as well as accounting choices (Bamber et al., 2010; P. R. Demerjian et al., 2013; Ge et al., 2011; B. Francis et al., 2019; Kim, 2023; Abdel-Meguid et al., 2021). Specifically, P. R. Demerjian et al. (2013) suggest that superior managers have better knowledge about their firms and businesses, which leads to better judgment and estimates on how to conduct business and, eventually, to a better quality of earnings. Their empirical evidence suggests that the presence of more capable managers is associated with better predictability of earnings and accruals on future operating earnings. Additionally, B. Francis et al. (2019) document evidence to suggest that managerial ability is a positive intangible asset of a firm that can increase the value relevance of earnings of the firm. Reversely, Cheng (2017) reports mixed evidence regarding the impact of managerial ability on the quality of non-GAAP earnings. Based on the preceding discussion, it is an empirical question to examine how managerial ability affects the value relevance and earnings predictability of both non-GAAP and equivalent GAAP earnings. To further examine the topic, we obtain the proxy for managerial ability from P. Demerjian et al. (2012) and apply it to the empirical analysis.
Panel A of Table 7 demonstrates that both values of the coefficients on GAAP earnings (GAAP Earnings per share) and on non-GAAP earnings (Non-GAAP Earnings per share) increase from the low managerial ability sub-sample to the high managerial ability sub-sample. The empirical evidence thereby confirms the findings from extant literature that the value relevance of earnings improves when managerial ability is higher (B. Francis et al., 2019). Additionally, the magnitude of the difference for the coefficients between non-GAAP and equivalent GAAP earnings is larger in the high managerial ability sub-sample while the significance level of the difference remains identical (p < 0.01) in both sub-samples. The analysis of the adjusted R2 also confirms that the overall explanatory powers of the regression in columns (2) and (4) are higher compared to the corresponding value in columns (1) and (3), thus suggesting that the difference in the value relevance between non-GAAP and equivalent GAAP earnings remains statistically identical after controlling for the managerial ability of a firm. The empirical evidence supports the previous conjecture that the value relevance of earnings improves when managerial ability is higher, but it does not support the notion that the difference in value relevance between non-GAAP and equivalent GAAP earnings is completely mitigated by managerial ability.
Panel B of Table 7, however, reports slightly different results from the one in Panel A of Table 7. Specifically, the values of the coefficients on GAAP earnings (GAAP Earnings) increase from the low managerial ability sub-sample to the high managerial ability sub-sample, whereas the values of the coefficients on non-GAAP earnings (Non-GAAP Earnings) decrease. This situation leads to a lower magnitude of the difference for the coefficients between non-GAAP and equivalent GAAP earnings in the high managerial ability sub-sample, although the significance level of the difference remains identical (p < 0.01) in both sub-samples.
Overall, the empirical evidence from Table 7 indicates that the difference in the value relevance between non-GAAP earnings and equivalent GAAP earnings is more pronounced when firms are operated by more capable managers, whereas the difference in the earnings predictability between the non-GAAP and equivalent GAAP earnings is more evident in the sub-sample consisting of firms operated by less capable managers. Importantly, the difference in the value relevance and the earnings predictability between non-GAAP and equivalent GAAP earnings remain significantly identical after controlling for the managerial ability of the firm.
To summarize, the empirical evidence from this section reveals that the difference in the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings can be altered by firm-level characteristics, such as market value of equity, accruals quality, analyst coverage, and managerial ability of a firm, but these factors can not completely mitigate the phenomenon that non-GAAP earnings are more value relevant and have better predictability towards the future operating earnings of a firm than equivalent GAAP earnings. Specifically, the difference in the value relevance between non-GAAP and equivalent GAAP earnings is more pronounced when a firm has a higher market value of equity, lower accruals quality, higher analyst coverage, and higher managerial ability. Additionally, the difference in the earnings predictability between non-GAAP and equivalent GAAP earnings is more evident when a firm has a lower market value of equity, lower accruals quality, higher analyst coverage, and lower managerial ability.

4.2.3. Comparison of the Value Relevance and Earnings Predictability Between Non-GAAP and Equivalent GAAP Earnings for Both Pre- and Post-2010 Periods

As an effort to improve the quality and usefulness of information in reports filed to the SEC, the SEC issued revised Compliance and Disclosure Interpretations (C&DI) in 2010, 2016, and 2022. To some extent, these actions, especially the issuance of C&DI 2010, had loosened the reporting standards on non-GAAP reporting, and in turn, led to the rebound of non-GAAP reporting (Bentley et al., 2018; D. E. Black et al., 2012), and the increasing concern of non-GAAP earnings being misused to artificially inflate earnings and being misleading to investors (Morgan et al., 2018). Thus, to better understand whether the quality of non-GAAP earnings decreased after the issuance of C&DI in 2010, we decide to conduct a supplementary analysis in this section to compare the value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings for both pre- and post-2010 periods. Specifically, we create an indicator variable that equals one if it is post-2010 and zero if it is pre-2010. We use the indicator variable to split the sample into two sub-samples, conduct the regression analysis in each sub-sample, and compare the coefficients on non-GAAP and equivalent GAAP earnings. Table 8 reports the results.
Panel A of Table 8 shows that the coefficients on both GAAP (GAAP Earnings per share) and non-GAAP earnings (Non-GAAP Earnings per share) increase in the post-2010 period, thus suggesting that the value relevance of both non-GAAP and equivalent GAAP earnings increase in the post-2010 period. Interestingly, the value relevance of non-GAAP earnings increases more than the value relevance of GAAP earnings in the post-2010 period, thus leading to an increase in the magnitude of the difference in the coefficients between non-GAAP and equivalent GAAP earnings in the post-2010 period. But in both pre- and post-2010 periods, the significance level of the difference remains identical (p < 0.01), which indicates non-GAAP earnings are more value relevant than GAAP earnings in both pre- and post-2010 periods. Moreover, the analysis of the adjusted R2 confirms that the overall explanatory powers of the regression in columns (2) and (4) are higher compared to the corresponding value in columns (1) and (3), thus corroborating the hypothesis that non-GAAP earnings are more value relevant than GAAP earnings in both pre- and post-C&DI periods.
Panel B of Table 8 reports that the coefficients on GAAP earnings (GAAP Earnings) increase from column (1) to (3) and from column (5) to (7), indicating that the earnings predictability of GAAP earnings increase in the post-2010 period. The coefficients on non-GAAP earnings (Non-GAAP Earnings), however, decrease from column (2) to (4) and from column (6) to (8), although the decrease is insignificant from a statistical perspective in an untabulated test. As a result, the magnitude of the difference in the earnings predictability between non-GAAP and equivalent GAAP earnings decreases in the post-2010 period, but the significance level of the difference remains identical (p < 0.01) in both pre- and post-2010 periods. The results indicate that non-GAAP earnings have better predictability for future operating earnings than GAAP earnings in both pre- and post-2010 periods.
Importantly, the preceding analyses do not lend support to the notion that the quality of non-GAAP earnings decreases significantly after the release of the C&DI in 2010. The non-GAAP earnings are more value relevant and have better predictability towards the future operating earnings of a firm than GAAP earnings in both pre- and post-2010 periods. In summary, the superior “informational” role of non-GAAP earnings to GAAP earnings persists in the post-2010 period.

5. Conclusions, Limitations, and Future Research

In this study, we utilize a unique quarterly dataset of non-GAAP earnings to re-examine the “informational” role of non-GAAP earnings from the perspective of their value relevance and abilities to predict future operating performance. We find empirical evidence suggesting that non-GAAP earnings are more value relevant and can better predict future operating earnings of a firm compared to GAAP earnings. Additionally, we also find that the difference in the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings can be altered but cannot be completely mitigated by firm-level characteristics, such as market value of equity, accruals quality, analyst coverage, and managerial ability of a firm. Specifically, the difference in the value relevance between two earnings measures is more pronounced for firms with a higher market value of equity, lower accruals quality, greater analyst coverage, and higher managerial ability. In contrast, the difference in the earnings predictability between two earnings measures is more evident among firms with a lower market value of equity, lower accruals quality, greater analyst coverage, and lower managerial ability. Moreover, our supplementary analysis reveals that the superior value relevance and predictive power of non-GAAP earnings persist even after the SEC’s release of the Compliance and Disclosure Interpretations (C&DI) in 2010.
Our empirical evidence has important implications from both academic and practical perspectives. Academically, our empirical evidence confirms the superior “informational” role of non-GAAP earnings to GAAP earnings (Bhattacharya et al., 2003; L. D. Brown & Sivakumar, 2003; Doyle et al., 2003). Thus, our evidence counters the notion that non-GAAP disclosures are merely self-serving (Bhattacharya et al., 2003; D. E. Black & Christensen, 2009; Heflin & Hsu, 2008). Instead, our evidence suggests that non-GAAP earnings convey useful information to investors. We also find evidence that regulatory efforts have improved the quality and use of non-GAAP reporting. Practically, our empirical evidence suggests that investors can benefit from using non-GAAP earnings for valuation purposes, while firms should focus on the credibility of their non-GAAP disclosures to support informed investment decisions and enhance shareholder value.
Our study has several limitations. First, the accuracy of the results depends highly on the accuracy of the coverage of the non-GAAP data provided by Bentley et al. (2018). There is a possibility that some firms report non-GAAP earnings in their SEC reporting but were not captured by Bentley et al. (2018). The inclusion of these firms into the sample can potentially change the results of the empirical analyses. Similarly, the accuracy of the results also depends highly on the accuracy of the financial data provided by the Thomson Reuters World Scope database. Second, since our sample period concludes at the end of 2016, future research could focus on years subsequent to 2016 in order to gain additional insights regarding the information content of non-GAAP earnings. Third, due to scope limitations, we focus on evaluating the valuation role of the non-GAAP earnings but ignore some other potential roles of non-GAAP earnings, such as the stewardship role (Ribeiro et al., 2019). Thus, it may be beneficial to conduct future research to compare non-GAAP and equivalent GAAP earnings from the perspective of income smoothing, earnings volatility, and earnings conservatism, to comprehensively understand the role of non-GAAP earnings. Fourth, we only use the price model to test the value relevance of earnings. Future research can be extended to using the return model to examine the association between the change in equity price, that is, the investment return, and the earnings (e.g., Kothari & Zimmerman, 1995). Conducting an event study is another potential method to evaluate the “informational” role of the non-GAAP earnings within a short-term event window. Fifth, due to data limitations, we only emphasize examining the “informational” role of non-GAAP earnings using the non-GAAP EPS in this study. Future research can be extended to study the “informational” role of some alternative non-GAAP earnings measures, such as the adjusted EBITDA.

Author Contributions

Conceptualization, X.S. and H.Q.; methodology, X.S., H.Q., and Y.L.; software, X.S. and H.Q.; validation, X.S., H.Q., and Y.L.; formal analysis, X.S. and H.Q.; investigation, H.Q., Y.L., and M.S.L.; resources, H.Q. and M.S.L.; data curation, X.S. and H.Q.; writing—original draft preparation, X.S. and H.Q.; writing—review and editing, H.Q., M.S.L., and M.E.; visualization, H.Q., M.S.L., and M.E.; supervision, H.Q. and Y.L.; project administration, H.Q.; funding acquisition, M.S.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data presented in this study is available on request from the corresponding author.

Acknowledgments

We thank the editors and anonymous reviewers of the Journal of Risk and Financial Management (JRFM) for their valuable and constructive feedback on our manuscript. We are especially grateful to Xuan Song’s DBA dissertation committee members, Brett Andrews and. Mark Shannon of Belhaven University, for their insightful comments during the dissertation defense. For this project, Michael S. Luehlfing received funding through his George E. Breazeal Family Endowed Professorship which is made available through the State of Louisiana Board of Regents Support Funds.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A. Variables Descriptions

The Appendix documents the definitions of variables used in this study (all continuous variables are winsorized at 1% and 99% level).
VariablesDescriptions
Key Dependent and Independent Variables:
Future PriceFuture equity price in quarter t + 1
Future Operating Income Sum of future operating income from quarter t + 1 to quarter t + 4 scaled by average total assets in quarter t
Future Operating Cash FlowSum of future operating cash flow from quarter t + 1 to quarter t + 4 scaled by average total assets in quarter t
GAAP Earnings per shareGAAP earnings per share in quarter t
Non-GAAP Earnings per shareNon-GAAP earnings per share in quarter t
Non-GAAP Exclusions per shareThe difference between the non-GAAP earnings per share and the equivalent GAAP earnings per share in quarter t
GAAP EarningsGAAP earnings per share multiplied by shares outstanding and scaled by average total assets in quarter t
Non-GAAP Earningsnon-GAAP earnings per share multiplied by shares outstanding and scaled by average total assets in quarter t
Non-GAAP ExclusionsThe difference between non-GAAP earnings per share and GAAP earnings per share multiplied by the number of shares outstanding and scaled by average total assets in quarter t
Control Variables:
SizeNatural log of one plus book value of assets in quarter t
ROA sdRolling standard deviation of ROA for past eight quarters including current quarter
Book-to-Market RatioBook value of equity to market value of equity in quarter t
Sales GrowthSales growth of a firm from quarter t−1 to quarter t
Op LossIndicator variable equals to 1 if a firm incurs operational loss in quarter t, 0 otherwise
Other Variables:
Market Value of EquityMarket value of equity in quarter t
Accruals QualityStandard deviation of the residuals over a three-year rolling window from cross-sectional regressions (by industry and year) of the change in working capital on the past, current, and future cash flow, the change in sales and the level of property, plant, and equipment (PP&E), as suggested by Dechow and Dichev (2002) and incorporated the thoughts of McNichols (2002)
Analyst CoverageNumber of analysts cover a firm in year t
Managerial AbilityManagerial ability for a firm in year t as calculated by P. R. Demerjian et al. (2013)

Notes

1
The Ohlson (1995) model is more suitable for empirical analysis because it provides a linear valuation equation based on readily available accounting data: book value of equity and contemporary earnings. This model is derived under the clean surplus relation and is rooted in the discounted cash flow (DCF) framework. The DCF model, while theoretically foundational, is difficult to implement empirically due to its reliance on unobservable inputs like future cash flows and firm-specific discount rates. In contrast, the Feltham and Ohlson (1995) model, while extending the framework to account for accounting conservatism and operating activities, is more theoretical and less empirically implementable due to its reliance on unobservable variables such as dirty surplus items and economic rents.
2
The concept of managerial ability, as developed by P. Demerjian et al. (2012), refers to the efficiency with which managers utilize a firm’s resources to generate revenues. To quantify this construct, the authors develop a two-step approach grounded in data envelopment analysis (DEA). First, they estimate firm-level efficiency scores by comparing each firm’s output (i.e., sales) to a set of inputs (e.g., assets, employees, operating expenses), controlling for industry-specific production possibilities. In the second step, they isolate the portion of efficiency attributable to managerial ability by controlling for firm-specific characteristics such as firm size, market share, and business complexity. The resulting managerial ability score reflects a manager’s talent in converting inputs into outputs, beyond what is explained by observable firm traits. This measure has been widely adopted in empirical research to study how managerial talent influences various corporate outcomes, including financial reporting quality, investment efficiency, and strategic decision-making.

References

  1. Abdel-Meguid, A., Jennings, J. N., Olsen, K. J., & Soliman, M. T. (2021). The impact of the CEO’s personal narcissism on non-GAAP earnings. The Accounting Review, 96(3), 1–25. [Google Scholar] [CrossRef]
  2. Arena, C., Catuogno, S., & Moscariello, N. (2021). The unusual debate on non-GAAP reporting in the current standard practice. The lens of corporate governance. Journal of Management & Governance, 25(3), 655–684. [Google Scholar]
  3. Ball, R., & Brown, P. (1968). An empirical evaluation of accounting income numbers. Journal of Accounting Research, 6(2), 159–178. [Google Scholar] [CrossRef]
  4. Bamber, L. S., Jiang, J., & Wang, I. Y. (2010). What’s my style? The influence of top managers on voluntary corporate financial disclosure. The Accounting Review, 85(4), 1131–1162. [Google Scholar] [CrossRef]
  5. Barth, M. E., Beaver, W. H., & Landsman, W. R. (2001). The relevance of the value relevance literature for financial accounting standard setting: Another view. Journal of Accounting and Economics, 31(1–3), 77–104. [Google Scholar] [CrossRef]
  6. Bentley, J. W., Christensen, T. E., Gee, K. H., & Whipple, B. C. (2018). Disentangling managers’ and analysts’ non-GAAP reporting. Journal of Accounting Research, 56(4), 1039–1081. [Google Scholar] [CrossRef]
  7. Bhattacharya, N., Black, E. L., Christensen, T. E., & Larson, C. R. (2003). Assessing the relative informativeness and permanence of pro forma earnings and GAAP operating earnings. Journal of Accounting & Economics, 36(1–3), 285–319. [Google Scholar]
  8. Black, D. E., Black, E. L., Christensen, T. E., & Heninger, W. G. (2012). Has the regulation of pro forma reporting in the U.S. changed investors’ perceptions of pro forma earnings disclosures? Journal of Business Finance & Accounting, 39(7–8), 876–904. [Google Scholar]
  9. Black, D. E., & Christensen, T. E. (2009). US managers’ use of ‘pro forma’ adjustments to meet strategic earnings targets. Journal of Business Finance & Accounting, 36(3–4), 297–326. [Google Scholar] [PubMed]
  10. Black, D. E., Christensen, T. E., Ciesielski, J. T., & Whipple, B. C. (2018). Non-GAAP reporting: Evidence from academia and current practice. Journal of Business Finance & Accounting, 45(3–4), 259–294. [Google Scholar]
  11. Black, E. L. (2016). The ethical reporting of non-GAAP performance measures. Revista Contabilidade & Finanças, 27, 7–11. [Google Scholar] [CrossRef]
  12. Black, E. L., Christensen, T. E., Kiosse, P. V., & Steffen, T. D. (2017). Has the regulation of non-GAAP disclosures influenced managers’ use of aggressive earnings exclusions? Journal of Accounting, Auditing & Finance, 32(2), 209–240. [Google Scholar] [CrossRef]
  13. Bradshaw, M. T., Christensen, T. E., Gee, K. H., & Whipple, B. C. (2018). Analysts’ GAAP earnings forecasts and their implications for accounting research. Journal of Accounting and Economics, 66(1), 46–66. [Google Scholar] [CrossRef]
  14. Bradshaw, M. T., & Sloan, R. G. (2002). GAAP versus the street: An empirical assessment of two alternative definitions of earnings. Journal of Accounting Research, 40(1), 41–66. [Google Scholar] [CrossRef]
  15. Brosnan, M., Duncan, K., Hasso, T., & Hollindale, J. (2024). Happy 20-year anniversary non-GAAP earnings: A systematic review of the literature. Journal of Accounting Literature, 46(1), 82–104. [Google Scholar] [CrossRef]
  16. Brown, L. D., & Sivakumar, K. (2003). Comparing the value relevance of two operating income measures. Review of Accounting Studies, 8(4), 561–572. [Google Scholar] [CrossRef]
  17. Brown, N. C., Christensen, T. E., Elliott, W. B., & Mergenthaler, R. D. (2012). Investor sentiment and pro forma earnings disclosures. Journal of Accounting Research, 50(1), 1–40. [Google Scholar] [CrossRef]
  18. Cheng, Y. (2017). Managerial reputation and the quality of non-GAAP earnings disclosures. Journal of Accounting and Finance, 17(2), 117–134. [Google Scholar]
  19. Christensen, T. E., Drake, M. S., & Thornock, J. R. (2014). Optimistic reporting and pessimistic investing: Do pro forma earnings disclosures attract short sellers? Contemporary Accounting Research, 31(1), 67–102. [Google Scholar] [CrossRef]
  20. Christensen, T. E., Gomez, E., Ma, M., & Pan, J. (2021). Analysts’ role in shaping non-GAAP reporting: Evidence from a natural experiment. Review of Accounting Studies, 26, 172–217. [Google Scholar] [CrossRef]
  21. Coleman, D., & Usvyatsky, O. (2015). Trends in non-GAAP disclosures. Sutton, MA, USA. Available online: https://blog.auditanalytics.com/trends-in-non-gaap-disclosures/ (accessed on 18 May 2025).
  22. Dechow, P. M., & Dichev, I. D. (2002). The quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77(s-1), 35–59. [Google Scholar] [CrossRef]
  23. Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1995). Detecting earnings management. The Accounting Review, 70(2), 193–225. [Google Scholar]
  24. Demerjian, P., Lev, B., & McVay, S. (2012). Quantifying managerial ability: A new measure and validity tests. Management Science, 58(7), 1229–1248. [Google Scholar] [CrossRef]
  25. Demerjian, P. R., Lev, B., Lewis, M. F., & McVay, S. E. (2013). Managerial ability and earnings quality. The Accounting Review, 88(2), 463–498. [Google Scholar] [CrossRef]
  26. Doyle, J. T., Ge, W., & McVay, S. (2007). Accruals quality and internal control over financial reporting. The Accounting Review, 82(5), 1141–1170. [Google Scholar] [CrossRef]
  27. Doyle, J. T., Jennings, J. N., & Soliman, M. T. (2013). Do managers define non-GAAP earnings to meet or beat analyst forecasts? Journal of Accounting and Economics, 56(1), 40–56. [Google Scholar] [CrossRef]
  28. Doyle, J. T., Lundholm, R. J., & Soliman, M. T. (2003). The predictive value of expenses excluded from pro forma earnings. Review of Accounting Studies, 8(2/3), 145–174. [Google Scholar] [CrossRef]
  29. Entwistle, G. M., Feltham, G. D., & Mbagwu, C. (2006). Financial reporting regulation and the reporting of pro forma earnings. Accounting Horizons, 20(1), 39–55. [Google Scholar] [CrossRef]
  30. Fama, E. F. (1980). Agency problems and the theory of the firm. Journal of Political Economy, 88(2), 288–307. [Google Scholar] [CrossRef]
  31. Fama, E. F., & French, K. R. (1997). Industry costs of equity. Journal of Financial Economics, 43(2), 153–193. [Google Scholar] [CrossRef]
  32. Fama, E. F., & MacBeth, J. D. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81(3), 607–636. [Google Scholar] [CrossRef]
  33. Feltham, G. A., & Ohlson, J. A. (1995). Valuation and clean surplus accounting for operating and financial activities. Contemporary Accounting Research, 11(2), 689–731. [Google Scholar] [CrossRef]
  34. Feltham, G. A., & Ohlson, J. A. (1996). Uncertainty resolution and the theory of depreciation measurement. Journal of Accounting Research, 34(2), 209–234. [Google Scholar] [CrossRef]
  35. Financial Accounting Standards Board (FASB). (2010). Statement of financial accounting concepts No. 8: Conceptual framework for financial reporting. Financial Accounting Standards Board (FASB). [Google Scholar]
  36. Financial Accounting Standards Board (FASB). (2018). Statement of financial accounting concepts No. 8 (as amended). Available online: https://fasb.org/Page/ShowPdf?path=Concepts_Statement_No._8_(As_Amended).pdf (accessed on 18 May 2025).
  37. Francis, B., Hasan, I., Siraj, I., & Wu, Q. (2019). Managerial ability and value relevance of earnings. China Accounting and Finance Review, 21(4), 147–191. [Google Scholar]
  38. Francis, J., LaFond, R., Olsson, P. M., & Schipper, K. (2004). Costs of equity and earnings attributes. The Accounting Review, 79(4), 967–1010. [Google Scholar] [CrossRef]
  39. Frankel, R., McVay, S., & Soliman, M. (2011). Non-GAAP earnings and board independence. Review of Accounting Studies, 16(4), 719–744. [Google Scholar] [CrossRef]
  40. Ge, W., Matsumoto, D., & Zhang, J. L. (2011). Do CFOs have style? An empirical investigation of the effect of individual CFOs on accounting practices. Contemporary Accounting Research, 28(4), 1141–1179. [Google Scholar] [CrossRef]
  41. Healy, P. M., & Palepu, K. G. (2001). Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of Accounting and Economics, 31(1–3), 405–440. [Google Scholar] [CrossRef]
  42. Heflin, F., & Hsu, C. (2008). The impact of the SEC’s regulation of non-GAAP disclosures. Journal of Accounting and Economics, 46(2–3), 349–365. [Google Scholar] [CrossRef]
  43. Holthausen, R. W., & Watts, R. L. (2001). The relevance of the value-relevance literature for financial accounting standard setting. Journal of Accounting and Economics, 31(1–3), 3–75. [Google Scholar] [CrossRef]
  44. Jennings, M. M., & Marques, L. A. (2011). The joint effects of corporate governance and regulation on the disclosure of manager-adjusted non-GAAP earnings in the U.S. Journal of Business Finance & Accounting, 38(3–4), 364–394. [Google Scholar]
  45. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360. [Google Scholar] [CrossRef]
  46. Jones, J. J. (1991). Earnings management during import relief investigations. Journal of Accounting Research, 29(2), 193–228. [Google Scholar] [CrossRef]
  47. Kim, R. (2023). Do more able managers provide better non-GAAP earnings? Accounting & Finance, 63(2), 1983–2012. [Google Scholar]
  48. Kolev, K., Marquardt, C. A., & McVay, S. E. (2008). SEC scrutiny and the evolution of non-GAAP reporting. The Accounting Review, 83(1), 157–184. [Google Scholar] [CrossRef]
  49. Kothari, S. P., Leone, A. J., & Wasley, C. E. (2005). Performance matched discretionary accrual measures. Journal of Accounting and Economics, 39(1), 163–197. [Google Scholar] [CrossRef]
  50. Kothari, S. P., & Zimmerman, J. L. (1995). Price and return models. Journal of Accounting and Economics, 20(2), 155–192. [Google Scholar] [CrossRef]
  51. Lougee, B. A., & Marquardt, C. A. (2004). Earnings informativeness and strategic disclosure: An empirical examination of “pro forma” earnings. The Accounting Review, 79(3), 769–795. [Google Scholar] [CrossRef]
  52. McNichols, M. F. (2002). Discussion of the quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77(s-1), 61–69. [Google Scholar] [CrossRef]
  53. Morgan, M. I., Poole, V. B., Qiu, H., & Owen, C. A. (2018). Introduce non-GAAP metrics to business students. Business Education Innovation Journal, 10(2), 116–127. [Google Scholar]
  54. Ohlson, J. A. (1995). Earnings, book values, and dividends in equity valuation. Contemporary Accounting Research, 11(2), 661–687. [Google Scholar] [CrossRef]
  55. Ohlson, J. A. (1999). Earnings, book values, and dividends in a stewardship setting with moral hazard. Contemporary Accounting Research, 16(3), 525–540. [Google Scholar] [CrossRef]
  56. Ribeiro, A., Shan, Y., & Taylor, S. (2019). Non-GAAP earnings and the earnings quality trade-off. Abacus, 55(1), 6–41. [Google Scholar] [CrossRef]
  57. Securities and Exchange Commission. (2001). SEC cautions companies, alerts investors to potential dangers of “pro forma” financials. Press Release 2001–144. Securities and Exchange Commission. [Google Scholar]
  58. Spence, M. (1973). Job market signaling. The Quarterly Journal of Economics, 87(3), 355–374. [Google Scholar] [CrossRef]
  59. Watts, R. L., & Zimmerman, J. L. (1986). Positive accounting theory. Prentice-Hall. [Google Scholar]
  60. Yu, F. F. (2008). Analyst coverage and earnings management. Journal of Financial Economics, 88(2), 245–271. [Google Scholar] [CrossRef]
  61. Zhang, X. F. (2006). Information uncertainty and analyst forecast behavior. Contemporary Accounting Research, 23(2), 565–590. [Google Scholar] [CrossRef]
Table 1. Sample Selections.
Table 1. Sample Selections.
Firm-Quarter
Initial sample from Bentley et al. (2018) 115,370
Less: Observations with missing non-GAAP earnings information(65,453)
Less: Observations with missing information in required variables for empirical analysis(10,506)
Less: Observations in Financial and Utility Industries(8045)
Final Sample 31,366
Table 2. Summary statistics and correlation matrix.
Table 2. Summary statistics and correlation matrix.
Panel A. Summary Statistics
VariablesNMeanSDQ1Q2Q3
Key Dependent and Independent Variables
Future Price31,36632.0943.7710.1922.2641.40
Future Operating Income31,3660.020.130.000.040.08
Future Operating Cash Flow31,3660.190.270.100.200.32
GAAP Earnings per share31,3660.210.75−0.040.170.50
Non-GAAP Earnings per share31,3660.400.540.080.290.62
Non-GAAP Exclusions per share31,366−0.190.55−0.20−0.08−0.02
Book Value per share31,36611.3910.624.188.6415.43
GAAP Earnings31,3660.000.050.000.010.02
Non-GAAP Earnings31,3660.010.030.010.010.02
Non-GAAP Exclusions31,366−0.010.04−0.01−0.010.00
Control Variables
Size31,3667.171.725.937.118.30
ROA sd31,3660.030.050.010.020.04
Book to Market Ratio31,3660.500.540.240.410.67
Sales Growth31,3660.040.27−0.050.020.09
Op Loss31,3660.290.460.000.001.00
Other Variables
Market Value of Equity31,3667268.4022,495.12423.671312.244240.35
Accruals Quality30,5220.090.080.040.070.11
Analyst Coverage28,84490.76182.1521.2641.4678.08
Managerial Ability31,0970.010.15−0.08−0.020.06
Panel B. Correlation Metrics
(1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)
(1)Future Price1
(2)Future Operating Income0.24471
0.00
(3)Future Operating Cash Flow0.20640.68981
0.000.00
(4)GAAP Earnings per share0.42210.34030.23561
0.000.000.00
(5)Non−GAAP Earnings per share0.65080.38350.31350.67861
0.000.000.000.00
(6)Non−GAAP Exclusions per share−0.06860.08390.01080.6892−0.06461
0.000.000.060.000.00
(7)Book Value per share0.50710.16220.10180.34570.5771−0.09961
0.000.000.000.000.000.00
(8)Size0.34710.27480.24150.27420.4759−0.0970.47481
0.000.000.000.000.000.000.00
(9)ROA sd−0.1542−0.257−0.1632−0.1716−0.2041−0.0319−0.2371−0.28171
0.000.000.000.000.000.000.000.00
(10)Book to Market Ratio−0.186−0.1632−0.1218−0.1563−0.1503−0.06410.2231−0.046−0.00131
0.000.000.000.000.000.000.000.000.82
(11)Sales Growth0.01270.0099−0.01960.10280.1070.0342−0.0291−0.03290.0445−0.05491
0.020.080.000.000.000.000.000.000.000.00
(12)Op Loss−0.2294−0.4534−0.3624−0.5988−0.4625−0.3572−0.23−0.27990.20220.1547−0.07021
0.000.000.000.000.000.000.000.000.000.000.00
Table 3. Comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings in the entire sample. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings for the entire sample, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Table 3. Comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings in the entire sample. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings for the entire sample, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Panel A. Value Relevance Test
Dependent VariableFuture Price
(1)(2)Non-GAAP v.s. GAAP
Book Value per share1.936 ***1.179 ***
(24.07)(21.82)
GAAP Earnings per share14.48 ***
(19.01)
Non-GAAP Earnings per share 40.90 ***26.24 ***
(24.08)
Non-GAAP Exclusions per share 1.426 ***
(3.61)
Size1.966 ***−0.0557
(15.74)(−0.34)
ROA sd1.236−3.504
(0.32)(−0.96)
Book to Market Ratio−18.50 ***−13.04 ***
(−20.97)(−20.20)
Sales Growth−0.794−5.863 ***
(−1.19)(−7.16)
Op Loss7.553 ***9.202 ***
(9.72)(11.88)
_cons−11.14 ***4.888 *
(−4.06)(1.65)
N31,36631,366
adj. R-sq0.4230.506
Industry Fixed Effect IncludedYesYes
Year Fixed Effect IncludedYesYes
Quarter Fixed Effect IncludedYesYes
t statistics in parentheses
* p < 0.10, *** p < 0.01
Panel B. Earnings Predictability Test
Dependent VariableFuture Operating Income Future Operating Cash Flow
(1)(2)Non-GAAP v.s. GAAP(3)(4)Non-GAAP v.s. GAAP
GAAP Earnings0.722 *** 1.661 ***
(12.05) (14.22)
Non-GAAP Earnings 2.511 ***1.789 *** 6.069 ***4.408 ***
(9.68) (13.56)
Non-GAAP Exclusions 0.197 ** 0.441 **
(1.97) (2.48)
Size0.0115 ***0.00912 *** 0.0190 ***0.0131 ***
(22.93)(14.31) (17.64)(10.85)
ROA sd−0.299 ***−0.306 *** −0.262 ***−0.272 ***
(−7.80)(−7.32) (−3.37)(−3.39)
Book to Market Ratio−0.0254 ***−0.0151 *** −0.0424 ***−0.0175 ***
(−8.70)(−5.21) (−11.10)(−5.01)
Sales Growth−0.0138 **−0.0346 *** −0.0584 ***−0.110 ***
(−2.48)(−5.25) (−4.98)(−8.34)
Op Loss−0.0616 ***−0.0298 *** −0.0963 ***−0.0154
(−18.43)(−5.75) (−15.25)(−1.62)
_cons0.0136−0.00193 0.104 ***0.0654 ***
(1.43)(−0.21) (4.03)(2.75)
N31,36631,366 31,36631,366
adj. R-sq0.3480.492 0.2850.494
Industry Fixed Effect IncludedYesYes YesYes
Year Fixed Effect IncludedYesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes
t statistics in parentheses
** p < 0.05, *** p < 0.01
Table 4. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings conditioning on market value of equity. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings conditioning on market value of equity, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for small size firms while columns (3) and (4) of Panel A report results for large size firms. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for small size firms while columns (3), (4), (7), and (8) of Panel B report results for large size firms. Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Table 4. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings conditioning on market value of equity. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings conditioning on market value of equity, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for small size firms while columns (3) and (4) of Panel A report results for large size firms. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for small size firms while columns (3), (4), (7), and (8) of Panel B report results for large size firms. Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Panel A. Value Relevance Test
Dependent VariableFuture Price
(1)(2)Non-GAAP v.s. GAAP(3)(4)Non-GAAP v.s. GAAP
Small SizeLarge Size
Book Value per share1.230 ***1.036 *** 2.661 ***1.827 ***
(61.06)(51.90) (21.29)(18.94)
GAAP Earnings per share2.473 *** 18.09 ***
(12.59) (15.64)
Non-GAAP Earnings per share 12.18 ***9.707 *** 40.90 ***22.81 ***
(24.43) (20.70)
Non-GAAP Exclusions per share 0.431 ** 2.296 ***
(2.50) (3.27)
Size−0.00771−0.283 *** −0.349−2.237 ***
(−0.09)(−3.42) (−1.26)(−8.99)
ROA sd−12.07 ***−12.70 *** 54.83 ***46.42 ***
(−6.67)(−8.25) (3.83)(3.50)
Book to Market Ratio−6.976 ***−6.035 *** −66.56 ***−51.88 ***
(−18.90)(−18.14) (−19.88)(−19.85)
Sales Growth0.158−1.194 *** −2.664 *−8.550 ***
(0.69)(−4.33) (−1.94)(−5.15)
Op Loss−2.861 ***−1.138 *** 17.85 ***16.09 ***
(−16.29)(−6.22) (12.83)(12.81)
_cons7.365 ***8.305 *** 19.66 ***34.75 ***
(4.36)(4.19) (4.91)(8.95)
N15,69315,693 15,67315,673
adj. R-sq0.5930.630 0.4450.505
Industry Fixed Effect IncludedYesYes YesYes
Year Fixed Effect IncludedYesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes
t statistics in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
Panel B. Earnings Predictability Test
Dependent VariableFuture Operating IncomeFuture Operating Cash Flow
(1)(2)Non-GAAP v.s. GAAP(3)(4)Non-GAAP v.s. GAAP(5)(6)Non-GAAP v.s. GAAP(7)(8)Non-GAAP v.s. GAAP
Small SizeLarge SizeSmall SizeLarge Size
GAAP Earnings0.634 *** 0.823 *** 1.502 *** 1.748 ***
(9.35) (7.92) (11.38) (8.31)
Non-GAAP Earnings 2.406 ***1.772 *** 2.508 ***1.685 *** 5.840 ***4.338 *** 5.876 ***4.128 ***
(6.78) (29.29) (9.73) (31.66)
Non-GAAP Exclusions 0.167 0.246 *** 0.450 ** 0.233 **
(1.48) (2.83) (2.24) (1.99)
Size0.0226 ***0.0174 *** 0.0007310.00272 *** 0.0453 ***0.0324 *** −0.00783 ***−0.00289 ***
(13.72)(9.69) (1.34)(6.19) (13.62)(9.98) (−6.28)(−2.89)
ROA sd−0.279 ***−0.278 *** −0.185 ***−0.266 *** −0.218 **−0.210 ** −0.103 **−0.298 ***
(−5.84)(−5.47) (−5.00)(−7.46) (−2.21)(−2.13) (−2.02)(−7.55)
Book to Market Ratio−0.0203 ***−0.0136 *** −0.0445 ***−0.0248 *** −0.0245 ***−0.00839 ** −0.114 ***−0.0649 ***
(−5.81)(−4.10) (−13.35)(−9.81) (−5.74)(−2.39) (−15.91)(−12.40)
Sales Growth−0.0188 **−0.0380 *** −0.0101 ***−0.0304 *** −0.0711 ***−0.118 *** −0.0451 ***−0.0953 ***
(−2.07)(−3.84) (−3.13)(−5.77) (−3.89)(−6.32) (−4.16)(−6.59)
Op Loss−0.0704 ***−0.0341 *** −0.0400 ***−0.0223 *** −0.111 ***−0.0186 −0.0574 ***−0.0154 ***
(−16.96)(−4.30) (−8.48)(−7.13) (−14.20)(−1.33) (−6.00)(−2.95)
_cons−0.0533 ***−0.0453 ** 0.109 ***0.0531 *** −0.128 **−0.109 ** 0.387 ***0.250 ***
(−2.68)(−2.51) (12.31)(5.87) (−2.49)(−2.38) (16.00)(10.53)
N15,69315,693 15,67315,673 15,69315,693 15,67315,673
adj. R-sq0.3090.444 0.3330.504 0.2700.473 0.2890.487
Industry Fixed Effect IncludedYesYes YesYes YesYes YesYes
Year Fixed Effect IncludedYesYes YesYes YesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes YesYes YesYes
t statistics in parentheses
** p < 0.05, *** p < 0.01
Table 5. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings conditioning on earnings quality. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings conditioning on earnings quality, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for firms with high earnings quality while columns (3) and (4) of Panel A report results for firms with low earnings quality. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for firms with high earnings quality while columns (3), (4), (7), and (8) of Panel B report results for firms with low earnings quality. Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Table 5. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings conditioning on earnings quality. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings conditioning on earnings quality, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for firms with high earnings quality while columns (3) and (4) of Panel A report results for firms with low earnings quality. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for firms with high earnings quality while columns (3), (4), (7), and (8) of Panel B report results for firms with low earnings quality. Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Panel A. Value Relevance Test
Dependent VariableFuture Price
(1)(2)Non-GAAP v.s. GAAP(3)(4)Non-GAAP v.s. GAAP
High Accruals QualityLow Accruals Quality
Book Value per share1.702 ***1.012 *** 2.404 ***1.573 ***
(22.26)(18.53) (13.28)(12.45)
GAAP Earnings per share16.21 *** 13.32 ***
(14.21) (12.56)
Non-GAAP Earnings per share 38.15 ***21.94 *** 43.81 ***30.49 ***
(19.65) (15.36)
Non-GAAP Exclusions per share 0.773 1.902 ***
(1.14) (3.50)
Size1.796 ***−0.0759 1.877 ***−0.217
(10.24)(−0.41) (9.15)(−0.76)
ROA sd−13.32−16.50 * 4.726−1.243
(−1.27)(−1.75) (0.90)(−0.24)
Book to Market Ratio−21.61 ***−15.48 *** −17.58 ***−12.56 ***
(−14.37)(−14.03) (−14.26)(−14.10)
Sales Growth−2.029 **−9.467 *** −0.422−4.558 ***
(−2.25)(−8.12) (−0.44)(−3.83)
Op Loss8.728 ***7.042 *** 7.758 ***12.05 ***
(8.41)(8.05) (6.49)(8.92)
_cons−11.01 ***4.774 −19.45 ***−2.990
(−3.05)(0.94) (−3.30)(−0.49)
N15,27915,279 15,24315,243
adj. R-sq0.5030.583 0.3730.456
Industry Fixed Effect IncludedYesYes YesYes
Year Fixed Effect IncludedYesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes
t statistics in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
Panel B. Earnings Predictability Test
Dependent VariableFuture Operating IncomeFuture Operating Cash Flow
(1)(2)Non−GAAP v.s. GAAP(3)(4)Non−GAAP v.s. GAAP(5)(6)Non−GAAP v.s. GAAP(7)(8)Non-GAAP v.s. GAAP
High Accruals QualityLow Accruals QualityHigh Accruals QualityLow Accruals Quality
GAAP Earnings1.196 *** 0.584 *** 2.747 *** 1.424 ***
(13.85) (9.11) (15.43) (10.88)
Non-GAAP Earnings 2.454 ***1.258 *** 2.430 ***1.846 *** 5.930 ***3.183 *** 5.982 ***4.558 ***
(13.94) (7.15) (15.29) (10.28)
Non-GAAP Exclusions 0.321 *** 0.132 0.633 *** 0.391 **
(3.87) (1.21) (4.19) (1.96)
Size0.00553 ***0.00465 *** 0.0166 ***0.0124 *** 0.00733 ***0.00509 *** 0.0294 ***0.0190 ***
(14.43)(12.09) (17.93)(9.97) (7.47)(5.24) (15.13)(8.44)
ROA sd−0.231 ***−0.321 *** −0.323 ***−0.336 *** −0.413 ***−0.630 *** −0.274 **−0.299 ***
(−5.10)(−7.94) (−6.48)(−7.11) (−4.71)(−8.85) (−2.56)(−3.15)
Book to Market Ratio−0.0247 ***−0.0165 *** −0.0264 ***−0.0162 *** −0.0457 ***−0.0253 *** −0.0401 ***−0.0153 ***
(−9.24)(−6.37) (−5.99)(−3.80) (−9.62)(−5.36) (−7.64)(−3.36)
Sales Growth−0.0159 ***−0.0339 *** −0.0150 *−0.0365 *** −0.0666 ***−0.112 *** −0.0589 ***−0.112 ***
(−4.93)(−9.05) (−1.74)(−3.71) (−6.52)(−8.63) (−3.26)(−5.81)
Op Loss−0.0242 ***−0.0184 *** −0.0789 ***−0.0368 *** −0.0252 ***−0.00815 −0.125 ***−0.0186
(−7.52)(−4.50) (−18.04)(−4.50) (−3.85)(−0.97) (−14.45)(−1.25)
_cons0.0303 ***0.0217 * −0.00311−0.00883 0.157 ***0.134 *** 0.01870.00366
(2.91)(1.80) (−0.19)(−0.58) (5.15)(4.31) (0.40)(0.08)
N15,27915,279 15,24315,243 15,27915,279 15,24315,243
adj. R-sq0.4200.536 0.3290.476 0.3190.449 0.2820.503
Industry Fixed Effect IncludedYesYes YesYes YesYes YesYes
Year Fixed Effect IncludedYesYes YesYes YesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes YesYes YesYes
t statistics in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
Table 6. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings conditioning on analyst coverage. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings conditioning on analyst coverage, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for firms with low analyst coverage while columns (3) and (4) of Panel A report results for firms with high analyst coverage. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for firms with low analyst coverage while columns (3), (4), (7), and (8) of Panel B report results for firms with high analyst coverage Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Table 6. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings conditioning on analyst coverage. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings conditioning on analyst coverage, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for firms with low analyst coverage while columns (3) and (4) of Panel A report results for firms with high analyst coverage. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for firms with low analyst coverage while columns (3), (4), (7), and (8) of Panel B report results for firms with high analyst coverage Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Panel A. Value Relevance Test
Dependent VariableFuture Price
(1)(2)Non-GAAP v.s. GAAP(3)(4)Non-GAAP v.s. GAAP
Low Analyst CoverageHigh Analyst Coverage
Book Value per share1.484 ***0.980 *** 2.374 ***1.493 ***
(39.58)(27.57) (18.16)(15.92)
GAAP Earnings per share9.367 *** 17.99 ***
(12.49) (14.98)
Non-GAAP Earnings per share 29.93 ***20.563 *** 45.67 ***27.68 ***
(15.69) (19.64)
Non-GAAP Exclusions per share 1.592 *** 2.011 ***
(4.42) (2.96)
Size1.308 ***0.0438 0.703 ***−1.619 ***
(7.76)(0.22) (3.13)(−6.10)
ROA sd−18.88 ***−19.93 *** 5.161−2.662
(−5.66)(−7.09) (0.70)(−0.42)
Book to Market Ratio−15.29 ***−11.25 *** −26.00 ***−18.67 ***
(−24.91)(−20.17) (−13.10)(−12.98)
Sales Growth−1.208 **−5.426 *** −0.523−6.283 ***
(−2.37)(−7.84) (−0.38)(−3.81)
Op Loss1.363 ***3.342 *** 13.24 ***14.03 ***
(2.73)(6.29) (9.50)(10.55)
_cons0.7028.732 *** 6.31526.15 ***
(0.33)(3.52) (1.01)(2.67)
N14,43414,434 14,41014,410
adj. R-sq0.5520.638 0.4130.489
Industry Fixed Effect IncludedYesYes YesYes
Year Fixed Effect IncludedYesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes
t statistics in parentheses
** p < 0.05, *** p < 0.01
Panel B. Earnings Predictability Test
Dependent VariableFuture Operating IncomeFuture Operating Cash Flow
(1)(2)Non−GAAP v.s. GAAP(3)(4)Non−GAAP v.s. GAAP(5)(6)Non−GAAP v.s. GAAP(7)(8)Non-GAAP v.s. GAAP
Low Analyst CoverageHigh Analyst CoverageLow Analyst CoverageHigh Analyst Coverage
GAAP Earnings0.589 *** 0.876 *** 1.338 *** 1.803 ***
(9.17) (12.74) (11.02) (12.21)
Non-GAAP Earnings 1.926 ***1.337 *** 2.817 ***1.941 *** 4.715 ***3.377 *** 6.787 ***4.984 ***
(4.26) (21.89) (6.57) (22.75)
Non-GAAP Exclusions 0.175 0.347 *** 0.311 0.363 ***
(1.40) (6.23) (1.61) (3.79)
Size0.0122 ***0.0114 *** 0.0158 ***0.0121 *** 0.0111 ***0.00923 *** 0.0230 ***0.0134 ***
(15.07)(12.55) (18.77)(16.32) (5.86)(4.81) (12.61)(8.53)
ROA sd−0.212 ***−0.219 *** −0.271 ***−0.258 *** −0.206 **−0.219 ** −0.258 **−0.219 *
(−6.02)(−5.27) (−5.13)(−3.86) (−2.03)(−2.09) (−2.35)(−1.65)
Book to Market Ratio−0.0246 ***−0.0175 *** −0.0285 ***−0.0173 *** −0.0411 ***−0.0231 *** −0.0555 ***−0.0268 ***
(−11.69)(−5.55) (−5.25)(−3.46) (−9.21)(−4.10) (−9.15)(−5.98)
Sales Growth−0.0180 ***−0.0367 *** −0.0101 **−0.0336 *** −0.0658 ***−0.113 *** −0.0447 ***−0.106 ***
(−2.73)(−3.77) (−2.15)(−5.46) (−4.42)(−5.75) (−3.79)(−8.05)
Op Loss−0.0632 ***−0.0398 *** −0.0531 ***−0.0199 *** −0.0999 ***−0.0399 *** −0.0964 ***−0.0113
(−18.89)(−4.86) (−12.84)(−4.79) (−15.72)(−3.00) (−11.45)(−1.32)
_cons0.0152−0.00336 −0.0227−0.0408 * 0.162 ***0.115 *** 0.07740.0298
(1.26)(−0.31) (−1.28)(−1.91) (4.74)(4.00) (1.54)(0.50)
N14,43414,434 14,41014,410 14,43414,434 14,41014,410
adj. R-sq0.3500.464 0.4230.573 0.2520.410 0.3350.575
Industry Fixed Effect IncludedYesYes YesYes YesYes YesYes
Year Fixed Effect IncludedYesYes YesYes YesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes YesYes YesYes
t statistics in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
Table 7. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings conditioning on managerial ability. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings conditioning on managerial ability, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for firms with low managerial ability while columns (3) and (4) of Panel A report results for firms with high managerial ability. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for firms with low managerial ability while columns (3), (4), (7), and (8) of Panel B report results for firms with high managerial ability Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Table 7. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings conditioning on managerial ability. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings conditioning on managerial ability, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for firms with low managerial ability while columns (3) and (4) of Panel A report results for firms with high managerial ability. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for firms with low managerial ability while columns (3), (4), (7), and (8) of Panel B report results for firms with high managerial ability Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Panel A. Value Relevance Test
Dependent VariableFuture Price
(1)(2)Non-GAAP v.s. GAAP(3)(4)Non-GAAP v.s. GAAP
Low Managerial AbilityHigh Managerial Ability
Book value per share1.335 ***0.924 *** 2.799 ***1.756 ***
(45.95)(32.26) (16.56)(13.73)
GAAP Earnings per share7.441 *** 20.66 ***
(15.48) (15.04)
Non-GAAP Earnings per share 25.22 ***17.779 *** 48.64 ***27.98 ***
(24.27) (19.88)
Non-GAAP Exclusions per share 1.381 *** 1.315 *
(3.30) (1.66)
Size2.435 ***1.219 *** 0.455 *−1.582 ***
(20.56)(10.28) (1.83)(−5.38)
ROA sd−23.23 ***−24.04 *** 14.16 **11.53 *
(−7.71)(−8.06) (2.01)(1.82)
Book to Market Ratio−12.17 ***−9.319 *** −28.76 ***−20.95 ***
(−16.69)(−15.60) (−13.69)(−12.82)
Sales Growth−1.073 **−4.622 *** 0.553−5.234 ***
(−2.25)(−8.40) (0.41)(−3.23)
Op Loss0.5812.784 *** 13.05 ***12.78 ***
(1.41)(6.78) (10.59)(11.18)
_cons−0.8217.734 *** −23.88 ***−26.75
(−0.32)(3.01) (−6.19)(−1.02)
N15,55815,558 15,53915,539
adj. R-sq0.5890.659 0.4380.510
Industry Fixed Effect IncludedYesYes YesYes
Year Fixed Effect IncludedYesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes
t statistics in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
Panel B. Earnings Predictability Test
Dependent VariableFuture Operating IncomeFuture Operating Cash Flow
(1)(2)Non−GAAP v.s. GAAP(3)(4)Non−GAAP v.s. GAAP(5)(6)Non−GAAP v.s. GAAP(7)(8)Non−GAAP v.s. GAAP
Low Managerial AbilityHigh Managerial AbilityLow Managerial AbilityHigh Managerial Ability
GAAP Earnings0.560 *** 0.762 *** 1.045 *** 1.753 ***
(8.56) (12.28) (9.86) (12.55)
Non-GAAP Earnings 2.744 ***2.184 *** 1.896 ***1.134 *** 6.215 ***5.17 *** 4.731 ***2.978 ***
(15.99) (5.24) (17.09) (8.17)
Non-GAAP Exclusions 0.200 *** 0.317 ** 0.204 * 0.573 **
(2.87) (2.25) (1.66) (2.48)
Size0.0127 ***0.00726 *** 0.00957 ***0.00931 *** 0.0239 ***0.0112 *** 0.0137 ***0.0130 ***
(15.83)(9.45) (17.19)(15.19) (14.47)(6.95) (11.57)(10.71)
ROA sd−0.488 ***−0.454 *** −0.162 ***−0.184 *** −0.440 ***−0.344 *** −0.151 **−0.203 **
(−7.36)(−8.02) (−4.73)(−4.15) (−3.51)(−3.86) (−2.50)(−2.39)
Book to Market Ratio−0.0161 ***−0.00963 *** −0.0405 ***−0.0302 *** −0.0155 ***−0.000980 −0.0801 ***−0.0531 ***
(−4.09)(−2.73) (−10.85)(−6.48) (−3.91)(−0.34) (−12.03)(−7.11)
Sales Growth−0.00956 **−0.0284 *** −0.0141 ***−0.0346 *** −0.0526 ***−0.0978 *** −0.0545 ***−0.108 ***
(−2.05)(−5.17) (−2.61)(−3.48) (−4.41)(−8.82) (−4.26)(−5.59)
Op Loss−0.0550 ***−0.0193 *** −0.0667 ***−0.0452 *** −0.0972 ***−0.0112 −0.108 ***−0.0511 ***
(−15.03)(−4.85) (−18.68)(−6.29) (−16.15)(−1.51) (−13.66)(−4.17)
_cons−0.00348−0.00567 0.05590.0278 0.0624 *0.0559 * 0.1760.101
(−0.26)(−0.49) (0.87)(0.33) (1.82)(1.94) (1.63)(0.64)
N15,55815,558 15,53915,539 15,55815,558 15,53915,539
adj. R-sq0.3720.519 0.3770.459 0.2710.505 0.3200.446
Industry Fixed Effect IncludedYesYes YesYes YesYes YesYes
Year Fixed Effect IncludedYesYes YesYes YesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes YesYes YesYes
t statistics in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
Table 8. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings for pre- and post-2010 periods. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings for pre- and post-2010 periods, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for firms in the pre-2010 period while columns (3) and (4) of Panel A report results for firms in the post-2010 period. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for firms in the pre-2010 period while columns (3), (4), (7), and (8) of Panel B report results for firms in the post-2010 period. Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Table 8. The value relevance and earnings predictability of non-GAAP and equivalent GAAP earnings for pre- and post-2010 periods. This table presents the empirical results for the comparison of the value relevance and earnings predictability between non-GAAP and equivalent GAAP earnings for pre- and post-2010 periods, in which Panels A and B present results for the value relevance and earnings predictability tests, respectively. For Panel A, the dependent variable is Future Price, a continuous variable measured by the market price of the common equity at the end of three-months after fiscal year-end. Book value per share is calculated by common equity scaled by numbers of shares outstanding, and Earnings per share represents either the non-GAAP earnings per share (Non-GAAP Earnings per share) or the GAAP earnings per share (GAAP Earnings per share). Columns (1) and (2) of Panel A report results for firms in the pre-2010 period while columns (3) and (4) of Panel A report results for firms in the post-2010 period. For Panel B, the dependent variable is the sum of either future operating earnings (Future Operating Income) or future operating cash flow (Future Operating Cash Flow) from quarter t + 1 to quarter t + 4 scaled by total assets, and Earnings represents either the non-GAAP earnings per share (Non-GAAP Earnings) or the GAAP earnings per share (GAAP Earnings) times number of shares outstanding scaled by total assets. Columns (1), (2), (5), and (6) of Panel B report results for firms in the pre-2010 period while columns (3), (4), (7), and (8) of Panel B report results for firms in the post-2010 period. Both Panels include control variables as a vector of firm-level characteristics such as Size (natural log of 1 plus book value of assets), ROA sd (rolling standard deviation of return on assets), Book to Market Ratio (book value of equity scaled over market value of equity), Sales Growth (sales growth rate in current year), and Op Loss (whether the firm has operating loss in the year). All regressions include time and industry (as classified by Fama & French, 1997) fixed effects, and the errors are robust to firm heteroscedasticity. T-values are reported in parentheses. Statistical significance of the coefficients is designated as ***, **, and * at 1%, 5%, and 10% levels, respectively.
Panel A. Value Relevance Test
Dependent VariableFuture Price
(1)(2)Non-GAAP v.s. GAAP(3)(4)Non-GAAP v.s. GAAP
Pre- 2010Post- 2010
Book Value per share1.629 ***1.115 *** 2.049 ***1.228 ***
(16.12)(17.85) (19.30)(16.64)
GAAP Earnings per share7.825 *** 18.92 ***
(11.30) (15.66)
Non-GAAP Earnings per share 27.27 ***19.445 *** 46.54 ***27.62 ***
(13.00) (20.22)
Non-GAAP Exclusions per share 1.494 *** 1.626 **
(4.21) (2.55)
Size1.866 ***0.937 *** 2.086 ***−0.713 ***
(19.09)(7.05) (10.85)(−2.84)
ROA sd−3.542−5.025 ** 5.8340.299
(−1.34)(−2.17) (0.62)(0.03)
Book to Market Ratio−12.86 ***−10.19 *** −22.97 ***−15.49 ***
(−14.71)(−14.83) (−15.11)(−14.34)
Sales Growth0.222−2.780 *** −1.832−8.102 ***
(0.45)(−4.63) (−1.54)(−5.57)
Op Loss2.447 ***4.488 *** 10.68 ***11.58 ***
(3.05)(5.10) (9.15)(10.37)
_cons−2.6231.608 −15.89 ***3.404
(−0.70)(0.37) (−4.32)(0.95)
N13,67013,670 17,69617,696
adj. R-sq0.4740.544 0.4140.497
Industry Fixed Effect IncludedYesYes YesYes
Year Fixed Effect IncludedYesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes
t statistics in parentheses
** p < 0.05, *** p < 0.01
Panel B. Earnings Predictability Test
Dependent VariableFuture Operating IncomeFuture Operating Cash Flow
(1)(2)Non−GAAP v.s. GAAP(3)(4)Non−GAAP v.s. GAAP(5)(6)Non−GAAP v.s. GAAP(7)(8)Non-GAAP v.s. GAAP
Pre− 2010Post− 2010Pre− 2010Post− 2010
GAAP Earnings0.470 *** 1.081 *** 1.345 *** 2.117 ***
(6.83) (10.17) (8.17) (11.28)
Non-GAAP Earnings 2.547 ***2.077 *** 2.497 ***1.416 *** 6.438 ***5.093 *** 5.802 ***3.685 ***
(10.67) (6.39) (14.11) (8.68)
Non-GAAP Exclusions −0.00213 0.540 *** 0.310 0.662 **
(−0.02) (2.97) (1.44) (2.26)
Size0.0124 ***0.00985 *** 0.0101 ***0.00820 *** 0.0221 ***0.0159 *** 0.0159 ***0.0110 ***
(15.18)(14.64) (15.21)(8.79) (12.49)(10.55) (11.52)(6.46)
ROA sd−0.206 ***−0.214 *** −0.469 ***−0.469 *** −0.254 ***−0.271 *** −0.278 **−0.257 *
(−4.50)(−4.20) (−6.76)(−6.38) (−2.78)(−2.71) (−1.97)(−1.85)
Book to Market Ratio−0.0259 ***−0.0143 *** −0.0266 ***−0.0162 *** −0.0413 ***−0.0130 ** −0.0460 ***−0.0207 ***
(−11.30)(−5.76) (−5.13)(−3.37) (−7.91)(−2.48) (−7.49)(−4.30)
Sales Growth−0.0161−0.0368 *** −0.0131 **−0.0322 *** −0.0664 ***−0.117 *** −0.0543 ***−0.103 ***
(−1.61)(−3.59) (−2.43)(−3.86) (−3.19)(−5.39) (−4.62)(−6.76)
Op Loss−0.0621 ***−0.0227 *** −0.0520 ***−0.0295 *** −0.105 ***−0.00514 −0.0812 ***−0.0218 *
(−15.27)(−3.89) (−9.54)(−4.45) (−10.84)(−0.41) (−8.50)(−1.90)
_cons−0.0103−0.0410 *** 0.0282 ***0.0109 0.0671 **−0.00861 0.148 ***0.104 ***
(−0.79)(−2.84) (2.85)(1.13) (2.06)(−0.22) (5.53)(4.20)
N13,67013,670 17,69617,696 13,67013,670 17,69617,696
adj. R-sq0.2880.485 0.4230.518 0.2640.499 0.3160.495
Industry Fixed Effect IncludedYesYes YesYes YesYes YesYes
Year Fixed Effect IncludedYesYes YesYes YesYes YesYes
Quarter Fixed Effect IncludedYesYes YesYes YesYes YesYes
t statistics in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Song, X.; Qiu, H.; Lin, Y.; Luehlfing, M.S.; Eduardo, M. A Re-Examination of the “Informational” Role of Non-GAAP Earnings in the Post-Reg G Period. J. Risk Financial Manag. 2025, 18, 414. https://doi.org/10.3390/jrfm18080414

AMA Style

Song X, Qiu H, Lin Y, Luehlfing MS, Eduardo M. A Re-Examination of the “Informational” Role of Non-GAAP Earnings in the Post-Reg G Period. Journal of Risk and Financial Management. 2025; 18(8):414. https://doi.org/10.3390/jrfm18080414

Chicago/Turabian Style

Song, Xuan, Huan Qiu, Ying Lin, Michael S. Luehlfing, and Marcelo Eduardo. 2025. "A Re-Examination of the “Informational” Role of Non-GAAP Earnings in the Post-Reg G Period" Journal of Risk and Financial Management 18, no. 8: 414. https://doi.org/10.3390/jrfm18080414

APA Style

Song, X., Qiu, H., Lin, Y., Luehlfing, M. S., & Eduardo, M. (2025). A Re-Examination of the “Informational” Role of Non-GAAP Earnings in the Post-Reg G Period. Journal of Risk and Financial Management, 18(8), 414. https://doi.org/10.3390/jrfm18080414

Article Metrics

Back to TopTop