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

Xuan Song, Huan Qiu, Ying Chou Lin, Michael S. Luehlfing, Marcelo EduardoJournal of Risk and Financial Management 2025 · DOI

TL;DR

This paper investigates whether non-GAAP earnings provide more valuable information to investors than standard GAAP earnings, focusing on their relevance for stock valuation and predicting future performance. The study finds that non-GAAP earnings are indeed more informative, a trend that persists even after regulatory changes aimed at standardizing their use.

What Problem Does This Paper Address?

In financial reporting, companies often provide "non-GAAP" earnings figures alongside the standard Generally Accepted Accounting Principles (GAAP) earnings. These non-GAAP figures are management's adjustments to GAAP earnings, intended to better reflect core operational performance by excluding certain items. However, this flexibility raises concerns: while potentially offering valuable insights, non-GAAP earnings could also be manipulated by management to present an overly optimistic view of the company's financial health. This tension between informativeness and potential manipulation has been a subject of debate, especially after regulatory interventions like Regulation G and subsequent guidance (C&DI) aimed at ensuring transparency. Previous research has yielded mixed results, particularly regarding the period after these regulations were implemented. This paper aims to re-examine the "informational" role of non-GAAP earnings in the post-Regulation G period, specifically investigating their value relevance and earnings predictability.

Methods

The study employs two primary econometric tests to assess the "informational" role of earnings:

  1. Value Relevance Test: This test uses a modified version of the Ohlson (1995) model. It regresses the equity price of a firm on its book value per share and earnings per share. The key innovation here is running this regression twice: once using GAAP earnings per share and again using non-GAAP earnings per share. By comparing the coefficients and the overall explanatory power (R-squared) of these regressions, the researchers can determine which earnings measure is more strongly associated with stock prices. Control variables like firm size, earnings volatility, and sales growth are included to account for other factors influencing valuation.
  2. Earnings Predictability Test: This test uses a regression model to assess how well current earnings predict future operating income or cash flow. Similar to the value relevance test, this is performed separately for GAAP and non-GAAP earnings. A higher coefficient for earnings in this model indicates better predictive power.

The researchers also conduct supplementary analyses to see if the differences in value relevance and predictability between non-GAAP and GAAP earnings are influenced by firm-level characteristics such as market value of equity, accruals quality, analyst coverage, and managerial ability. They also compare results from the pre- and post-2010 periods to assess the impact of the SEC's C&DI issuance.

Datasets

The study utilizes a unique quarterly dataset of non-GAAP earnings disclosures compiled by Bentley et al. (2018). This dataset was derived from SEC 8-K filings. Financial and price information for the equivalent GAAP earnings measures and control variables were sourced from the Thomson Reuters Worldscope database. Analyst coverage data was obtained from the IBES dataset.

Key Insights and Ideas

  • Non-GAAP Earnings Shine: The core finding is that non-GAAP earnings are not just a management tool for presenting a rosy picture, but genuinely provide more useful information to the market than standard GAAP earnings.
  • Predictive Power Matters: Beyond just reflecting current value, non-GAAP earnings are better at forecasting a company's future financial performance.
  • Regulation's Impact: The study suggests that regulatory efforts (like Regulation G and the C&DI) haven't diminished the informational value of non-GAAP earnings; in fact, their superior role persists even after these regulations were put in place.
  • Firm Characteristics Play a Role: While non-GAAP earnings are generally superior, the degree of this superiority can be influenced by factors like a firm's size (market value of equity), the quality of its accounting (accruals quality), how much attention it gets from analysts, and the competence of its management. However, these factors don't completely erase the advantage of non-GAAP earnings.
  • Countering Skepticism: The findings challenge the view that non-GAAP disclosures are primarily self-serving or manipulative, lending support to the idea that they can be informative and useful for investors.

Key Results

  • Value Relevance: Non-GAAP earnings are significantly more value relevant than GAAP earnings. For every dollar increase in non-GAAP earnings, the firm's future equity price increases by approximately $40.90, compared to $14.48 for GAAP earnings. The adjusted R-squared for the non-GAAP regression is also higher, indicating better overall explanatory power.
  • Earnings Predictability: Non-GAAP earnings also demonstrate superior predictive power for future operating income and cash flow. A one-dollar increase in non-GAAP earnings is associated with a $2.51 increase in future operating income and a $6.07 increase in future operating cash flow, compared to $0.72 and $1.66 for GAAP earnings, respectively.
  • Firm-Level Effects: The difference in value relevance and predictability between non-GAAP and GAAP earnings is more pronounced in specific scenarios:
  • Value Relevance: Stronger for firms with higher market value of equity, lower accruals quality, higher analyst coverage, and higher managerial ability.
  • Earnings Predictability: More evident for firms with lower market value of equity, lower accruals quality, higher analyst coverage, and lower managerial ability.
  • Post-Regulation Impact: The superior informational role of non-GAAP earnings persists in the post-2010 period, even after the SEC's C&DI issuance.

Limitations and Future Work

  • Data Accuracy: The study acknowledges that its findings are contingent on the accuracy of the non-GAAP data provided by Bentley et al. (2018) and the financial data from Thomson Reuters Worldscope. There's a possibility that some non-GAAP reporting firms were missed.
  • Sample Period: The study's sample period ends in 2016. Future research could extend this analysis to more recent years to capture evolving trends.
  • Scope of Roles: The paper primarily focuses on the valuation and prediction roles of non-GAAP earnings. Future work could explore other roles, such as income smoothing, earnings volatility, and earnings conservatism.
  • Valuation Model: The study uses a price model for value relevance. Future research could employ return models or event studies for a more comprehensive analysis.
  • Alternative Non-GAAP Measures: The study focuses on non-GAAP EPS. Future research could investigate other non-GAAP measures like adjusted EBITDA.

Summary updated

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.
BibTeX
@article{song2025re,
  author = {Xuan Song and Huan Qiu and Ying Chou Lin and Michael S. Luehlfing and Marcelo Eduardo},
  title = {{A Re-Examination of the “Informational” Role of Non-GAAP Earnings in the Post-Reg G Period}},
  journal = {Journal of Risk and Financial Management},
  year = {2025},
  doi = {10.3390/jrfm18080414},
}
Figure