Earnings Announcement Informativeness and Expected Stock Returns

Abstract

Earnings announcement returns tend to be volatile when a significant amount of information remains unknown to investors prior to the announcement. The informativeness of earnings announcements—measured by the return volatility during the announcement period relative to the pre-announcement period—reflects the extent of information not publicly disclosed ex ante. Consistent with the view that information plays a crucial role in stock pricing, we find a positive correlation between the informativeness of earnings announcements and expected stock returns in the U.S. market from 1977 to 2022. This effect is more pronounced among younger firms, firms with higher return volatility, and during periods of heightened investor risk aversion. Additionally, we find supporting evidence that reductions in information supply, driven by Regulation FD, lead to higher expected stock returns.

Keywords

earnings announcement informativeness information risk expected stock returns regulation FD information disclosure G12 G14 M41

Introduction

Beginning with the seminal work of Beaver (1968), numerous studies have examined the information content of earnings announcements, as reflected by the abnormal return variance on announcement dates. This body of literature has successfully identified a variety of significant economic determinants of earnings announcement informativeness (EAI), encompassing a wide range of market conditions and firm-level characteristics (e.g., Atiase, 1985; DeFond et al., 2007; Landsman et al., 2012; Pevzner et al., 2015). Despite this progress, a significant research gap remains regarding the economic consequences of varying levels of earnings announcement informativeness. Given that EAI has increased substantially since 2001 (Beaver et al., 2018, 2020) and plays a significant role in financial markets, it is essential to understand its economic implications. In this study, we investigate the impact of EAI on expected stock returns, which are central to firm valuation.

Informative earnings announcements often indicate that investors lacked sufficient information prior to the announcement, and a firm’s information disclosure practices may contribute to this information deficiency. As demonstrated by Beaver et al. (2020), the upward trend in EAI among U.S. stocks in the new century can be attributed to the increasing volume of concurrent disclosures (e.g., earnings guidance) accompanying earnings announcements. Assuming the total amount of information possessed by the firm remains unchanged, the bundled release of information during earnings announcements suggests that some information could have been disclosed earlier. For instance, earnings guidance could be issued more promptly to create a more equitable information environment for investors (Ertan et al., 2025; Rogers & Van Buskirk, 2013).¹

Given their association with limited pre-announcement information availability, informative earnings announcements may imply higher expected stock returns. Generally, when investors lack sufficient information, they tend to avoid stock investments to protect themselves unless prices are low enough to justify the associated risk premium (Easley & O’Hara, 2004). In specific situations where pre-announcement information disclosures are minimal—due to bundled disclosure practices—investors may have a limited understanding of the firm. Consequently, they may perceive the firm’s stock as highly risky, leading to increased return uncertainty, particularly for younger firms and among risk-averse investors (Cohen et al., 2007; Diamond & Verrecchia, 1991; Jorgensen & Kirschenheiter, 2003; Levi & Zhang, 2015; Stoumbos, 2023). As a result, investors demand a premium to compensate for low information availability (Amihud & Mendelson, 1989; Barber et al., 2013; Campbell et al., 2001; Da Silva, 2019; Pevzner et al., 2015; Savor & Wilson, 2016).

However, one could argue that informative earnings announcements do not necessarily lead to higher expected stock returns. First, an abundance of information is not always in investors’ best interest. Excessive disclosures can harm a firm if exploited by competitors or if they result in increased agency costs (Hermalin & Weisbach, 2012). Other unintended consequences of excessive disclosures include reduced liquidity for long-term investors (Monnet & Quintin, 2017), diminished private information production, compromised risk-sharing, and the promotion of destabilizing beauty-contest incentives (Goldstein & Yang, 2017). In contrast, moderate disclosures can alleviate capital market pressure on managers and mitigate their myopic behavior (Arif & De George, 2020; Fu et al., 2020). Second, informative earnings announcements imply that earnings reports have high precision, which can help reduce a firm’s beta or covariance risk (Amir & Levi, 2019; Evans, 2016; Livnat & Zarowin, 1990). Third, information risk may be negligible to investors due to market competition or portfolio diversification (Hughes et al., 2007). Given the debate outlined above, there is a need for more empirical evidence.

In this study, we measure EAI using Atiase’s (1985) revaluation index, calculated as the ratio of return volatility during earnings announcements to normal return volatility in the absence of news. A higher index indicates less pre-announcement information availability and more information revealed on the announcement date (Atiase et al., 1989; Efendi et al., 2016; El-Gazzar, 1998). This index closely aligns with the measure developed by Beaver (1968), with the additional consideration of estimation errors in risk loadings. Consistent with Beaver et al. (2020), we observe a notable increase in earnings announcement informativeness—measured by the average revaluation index—after 2001. Our analyses indicate that earnings announcement informativeness is inversely related to pre-announcement information disclosure levels, as captured by the normal information flow measure proposed by Thomas et al. (2022), the number of analyst forecasts, and the frequency of management guidance. This evidence supports our assertion that earnings announcement informativeness reflects previously undisclosed information.

In a sample of U.S. stocks from 1977 to 2022, portfolio and regression analyses both demonstrate that our measure of earnings announcement informativeness—specifically, the average revaluation index over the past 12 quarters—exhibits a positive correlation with future stock returns. The return spread between the top and bottom deciles of stocks sorted by EAI is 4.80% per year. This difference remains statistically significant even after controlling for other important risk premiums using various risk factor models. Furthermore, the finding remains robust when accounting for earnings quality as measured by discretionary accruals (Dechow et al., 1995; Sloan, 1996).

Further analyses reveal intriguing variations in the EAI premium stemming from firm-level and market-level frictions. First, the EAI premium tends to be lower for more mature firms. Mature firms—characterized by established earnings and operational records—are easier for investors to understand and value. Consequently, investor demand for pre-announcement information is relatively low, which in turn reduces the EAI premium. We measure firm maturity by firm age (Easley & O’Hara, 2004) or by whether the firm pays cash dividends (Grullon et al., 2002). Second, the EAI premium is positively related to information uncertainty, which is captured by return volatility (Gao et al., 2017; Jiang et al., 2005; Li et al., 2021; Lyle, 2025). Higher information uncertainty leads to greater investor demand for information and a higher information risk premium. Finally, the risk premium for low information availability rises when investors exhibit higher levels of risk aversion, measured using the risk aversion data from Bekaert et al. (2022).

Moreover, we find that controlling for firm-level earnings response coefficients (ERC) does not alter our conclusions, indicating that EAI is distinct from ERC. Additionally, EAI differs from post-earnings announcement drift (PEAD). While PEAD is typically attributed to investors’ behavioral biases, the return predictability of EAI stems from investors’ rational pricing of information risk. Our findings remain robust after accounting for prior-quarter earnings surprise—measured either by the 3-day cumulative abnormal returns surrounding the earnings announcement or based on the random-walk earnings model outlined by Chan et al. (1996)—underscoring the unique significance of EAI in explaining future stock returns.

A key concern is that a firm’s EAI and its cost of equity capital may be endogenously determined. To mitigate this concern, we examine Regulation Fair Disclosure (Reg FD), enacted in 2000, which significantly altered the information environment for U.S. firms. Reg FD prohibited U.S. firms from making selective information disclosures but exempted non-U.S. American Depositary Receipt (ADR) firms (Francis et al., 2006; Litvak, 2007). As noted by Easley and O’Hara (2004), the reduction in selective disclosures diminishes investor information availability and heightens information risk. By designating U.S. firms as the treatment group and ADR firms as the control group, we anticipate—and observe—increases in both EAI and expected stock returns for U.S. firms relative to ADR firms following the implementation of Reg FD. Furthermore, using pre-regulation informativeness as a proxy for treatment intensity, we find that firms with more pre-regulation information disclosures experience more pronounced and positive changes in their expected stock returns post Reg FD.

Our study contributes to the literature in three key ways. First, it expands existing research on EAI, which has primarily focused on its determinants rather than its economic consequences. Previous studies often regard informative earnings announcements as beneficial due to their positive correlation with high earnings quality and credibility (e.g., Ball & Brown, 1968; Beaver, 1968; DeFond et al., 2007; Hanlon et al., 2008). However, our study argues that these announcements can be detrimental, as they may be associated with limited pre-announcement information availability (Han & Wild, 2000; Holthausen & Verrecchia, 1988).

Second, our study contributes to the literature on the cost of equity capital by examining specific informational attributes. This line of research has primarily focused on three attributes: information quantity, information precision, and information asymmetry (Schreder, 2018). Our study extends prior literature by demonstrating that the timeliness of information disclosure—indicated by the volume of information released during earnings announcements relative to other periods—can significantly impact the cost of capital.

Finally, our study contributes to research on the economic impact of disclosure regulations (Anilowski et al., 2007). Our investigation of Reg FD confirms that it exerts a “chilling effect” on information availability (Koch et al., 2013), resulting in increased information risk for investors. This evidence has important implications for policymakers seeking to comprehensively evaluate the economic consequences of disclosure policies.

Hypothesis Development

New information about a firm may arrive continuously in small, random doses (Fama, 1970). However, a firm may choose to postpone information disclosure, resulting in a substantial amount of information being released during earnings announcements. This practice renders such announcements highly informative, with very little information disclosed beforehand. Limited firm-provided information can also hinder the ability of information intermediaries (e.g., financial analysts and news media) to generate insights (Francis et al., 2002). Consequently, the shift in information flow—specifically, disclosing pre-announcement information during the announcement period—can enhance EAI but at the cost of reduced pre-announcement information availability. The positive relationship between high EAI and low pre-announcement information availability aligns with the following argument: when investors have only vague expectations about upcoming news, they have much to learn upon the public release of earnings announcement information, leading to significant price reactions (e.g., Holthausen & Verrecchia, 1988; Kim & Verrecchia, 1991a, 1991b).

The uneven temporal distribution of information—evidenced by highly informative earnings announcements—has significant implications for investors. In principle, investors at an informational disadvantage often mitigate risk by avoiding stock investments unless prices are sufficiently low to provide adequate compensation for that risk (Easley & O’Hara, 2004). Specifically, when pre-announcement information disclosure is minimal, investors may perceive asset values as riskier. This perception results in lower stock liquidity and heightened return uncertainty, a concern that is particularly acute for less mature firms and among risk-averse investors (Cohen et al., 2007; Diamond & Verrecchia, 1991; Jorgensen & Kirschenheiter, 2003; Levi & Zhang, 2015; Stoumbos, 2023). As suggested by Easley and O’Hara (2004), less-informed investors end up holding an excessive amount of underperforming stocks and an insufficient amount of outperforming stocks compared to their well-informed counterparts. Consequently, the risks associated with low information availability compel investors to demand a premium. This required compensation for bearing information risk ultimately translates into higher expected stock returns (Amihud & Mendelson, 1989; Barber et al., 2013; Campbell et al., 2001; Da Silva, 2019; Pevzner et al., 2015; Savor & Wilson, 2016).

An alternative perspective posits that more informative earnings announcements may not inherently translate into higher expected shareholder returns. This view is supported by several countervailing arguments. For instance, heightened transparency can sometimes destroy firm value if it reveals proprietary information to competitors or intensifies agency conflicts (Hermalin & Weisbach, 2012). Additional drawbacks may include reduced market liquidity for long-term investors (Monnet & Quintin, 2017), suppressed private information production, reduced risk-sharing efficiency, and the encouragement of short-term market dynamics (Goldstein & Yang, 2017). In contrast, a measured approach to disclosure can shield managers from excessive market pressures and curb short-sighted decision-making (Arif & De George, 2020; Fu et al., 2020). Moreover, high-quality, precise earnings reports could potentially lower a firm’s systematic risk—by reducing its beta or covariance with the market—thereby decreasing rather than increasing expected returns (Amir & Levi, 2019; Evans, 2016; Livnat & Zarowin, 1990). Finally, in well-diversified portfolios, firm-specific information risk may be largely irrelevant, as it can be diversified away (Hughes et al., 2007). Given these competing theoretical channels, the net effect on expected returns remains an unresolved empirical question. Our testable hypothesis is therefore as follows:

Hypothesis

There is a positive relationship between EAI and expected stock returns.

Informativeness Measure and Sample Description

The Measure of Earnings Announcement Informativeness

The cornerstone of our measure of earnings announcement informativeness is Atiase’s (1985) revaluation index. This index is a slight variation of Beaver’s (1968) measure and assesses return volatility during earnings announcements in comparison to the normal level of return volatility in the absence of earnings news. It is defined as follows

R I_{i t} = \frac{C A R_{i t}^{2}}{E x p V a r (C A R_{i t})} = \frac{C A R_{i t}^{2}}{3 σ_{i t, ε}^{2} + V a r A d j_{i t}},

(1)

Where

C A R_{i t}

is the cumulative daily abnormal returns of stock

i

in the 3-day period [-1,+1] around the earnings announcement date

t

E x p V a r (C A R_{i t})

is the expected variance of

C A R_{i t}

in the absence of news,

σ_{i t, ε}^{2}

is the variance of daily residual returns in the estimation window based on a four-factor model, and

V a r A d j_{i t}

is the adjustment needed to account for the estimation errors of risk loadings—this enhances the accuracy of the revaluation index over Beaver’s measure (as detailed in Appendix A). The estimation window comprises the non-earnings announcement days from the sixth trading date after the previous announcement date to the sixth trading date before the current announcement date.

If there is no significant new information, $C A R_{i t}^{2}$ is close to the expected variance of $C A R_{i t}$ . Conversely, if there is significant new information, $C A R_{i t}^{2}$ exceeds the expected variance, resulting in an RI greater than one. Given that RI is positively skewed, we use its square root in all subsequent applications. Figure 1 illustrates the quarterly average $\sqrt{R I}$ for our sample stocks, revealing an upward trend consistent with earlier studies (e.g., Beaver et al., 2020).

Figure 1.

Mean $\sqrt{R I}$ over time.

Our primary measure of earnings announcement informativeness is the average of the square root of RI over the past 12 quarters²

E A I_{i t} = \frac{1}{N^{i t}} \sum_{j = - 12}^{- 1} \sqrt{R I_{i t + j}},

(2)

Where

N^{i t}

is the number of earnings announcements in the 12 quarters prior to quarter

t

, and

R I_{i t + j}

denotes the revaluation index for stock

i

in quarter

t + j

. The primary purpose of averaging

\sqrt{R I}

over specific periods is to minimize measurement errors in the informativeness of earnings announcements. Our findings remain robust when the average is calculated over a period preceding quarter

t

of a different length than 12 quarters, for example, over a 20-quarter period prior to quarter

t

.³ Additionally, we employ Beaver’s measure as a robustness check. The average of the square root of Beaver’s measure over the past 12 quarters is denoted as EAI2. It is highly correlated with EAI, with a pooled correlation of 0.998 and a time-series average of cross-sectional correlation of 0.995.

The Sample Formation Process

Our primary dataset consists of monthly U.S. stock data spanning from January 1977 to December 2022. Stock return data is sourced from the Center for Research in Security Prices (CRSP), while accounting information and earnings announcement dates are obtained from Compustat, and analyst forecasts are derived from the Institutional Brokers’ Estimate System (IBES). We selected 1977 as the starting point for our sample for two main reasons: first, NASDAQ companies have been included in CRSP since December 1972, and earnings announcement data have been available in Compustat since the fourth quarter of 1971; second, the EAI measure relies on data from the preceding 20 quarters in robustness tests.

Following the methodology established by Fama and French (1992), we integrate the CRSP monthly stock file with the Compustat annual file to construct a monthly stock dataset for cross-sectional analysis. The sample comprises U.S. firms with stock codes of 10 or 11, excluding those in the financial sector (SIC codes 6000–6999), that have at least 24 monthly returns over the preceding 5 years and positive book equity from the most recent fiscal year. For each stock-month, we calculate the EAI measure as the average root of the RI over the prior 12 quarters.

Summary Statistics of the Sample

We calculate various control variables that potentially explain expected stock returns. BETA represents the market beta; BM represents the book-to-market ratio; and ME stands for market equity, a proxy for firm size. The three variables are defined as in Fama and French (1992). RET12 signifies cumulative stock returns from month $t$ -12 to month $t$ -1, and IVOL is the monthly measure of idiosyncratic volatility as defined by Ang et al. (2006). DISC_ACCR represents the discretionary accruals estimated following Dechow et al. (1995). FDISP refers to analyst forecast dispersion, as per Diether et al. (2002). SPRD represents the bid-ask spread estimate as defined by Corwin and Schultz (2012). AGE refers to the number of years for which a stock has monthly stock returns in the CRSP file. DIVYIELD denotes dividend yield, as defined by Dhaliwal et al. (2006). Appendix B provides more detailed definitions of these variables. All continuous variables, except for stock returns (RET), are winsorized at the 1st and 99th percentiles each month to mitigate the impact of outliers in regression analyses.

Panel A of Table 1 presents the summary statistics for our main sample. The mean of EAI is 1.39, and the mean of the alternative informativeness measure EAI2 is 1.53. The latter is larger because its denominator is not adjusted upward (Atiase, 1985). The other variables are selectively described as follows: the average market beta (BETA) is 1.30; idiosyncratic volatility (IVOL) averages 10.40% per month, while the standard deviation of discretionary accruals (DISC_ACCR) is 11%; firms remain in the monthly CRSP file for an average of 21.26 years (AGE); the average dividend yield (DIVYIELD) is 1.47%, with over 50% of the observations showing no dividends; the implied bid-ask spread, denoted by SPRD, is on average 96 basis points. Panel B of Table 1 displays the time-series average of EAI and selected firm characteristics for the 10 monthly equally weighted stock portfolios (D1–D10) sorted by EAI. The D1 stocks with the lowest average EAI are characterized by small market capitalization, high book-to-market ratios, and high IVOL. However, there is no clear monotonic relationship between EAI and these variables.

Table 1.

Summary Statistics of the Sample

Panel A: Summary statistics of the variables
Variables	Mean	Std	P25	Median	P75	N
RET (%)	1.11	13.31	−5.49	0.65	7.06	1,131,821
EAI	1.39	0.73	0.88	1.18	1.69	1,132,023
EAI2	1.53	0.78	0.98	1.30	1.86	1,132,023
BETA	1.30	0.30	1.10	1.28	1.50	1,132,023
BM	0.70	0.54	0.32	0.56	0.92	1,132,023
ME	3,548	13,587	110	430	1,728	1,131,899
RET12	0.23	0.64	−0.10	0.12	0.40	1,132,022
IVOL (%)	10.40	6.58	5.85	8.71	13.06	1,132,016
DISC_ACCR	0.01	0.11	−0.04	0.00	0.04	1,104,871
FDISP	0.13	0.34	0.02	0.04	0.10	763,900
AGE	21.26	17.40	8.34	16.10	28.27	1,132,023
DIVYIELD (%)	1.47	2.22	0	0	2.33	1,131,588
SPRD (%)	0.96	0.65	0.52	0.79	1.21	1,132,016
AvgAbsEAR_12q	5.39	2.94	3.23	4.78	6.90	1,156,209
Ivol_preEA_12q	2.36	1.18	1.50	2.09	2.95	1,156,209
$\sqrt{R I}$	1.42	1.47	0.42	0.96	1.88	1,110,983
AbsEAR	5.32	5.50	1.52	3.57	7.16	1,110,983
Ivol_preEA	2.24	1.31	1.33	1.92	2.80	1,110,983
NmlInfo	8.91	5.38	5.24	7.60	11.06	1,156,124
Loss	0.19	0.39	0	0	0	1,109,891
COVEAGE	7.07	7.72	0	5	11	1,156,209
LagRpt	32.39	13.82	23	30	39	1,110,983
RETVAR (%)	11.71	6.86	7.23	10.16	14.29	1,156,199

Panel B: Average firm characteristics for monthly stock portfolios (D1–D10) sorted by EAI
	D1	D2	D3	D4	D5	D6	D7	D8	D9	D10
EAI	0.64	0.85	1.00	1.13	1.26	1.4	1.54	1.72	1.96	2.53
BETA	1.29	1.29	1.29	1.29	1.29	1.29	1.28	1.29	1.29	1.30
BM	0.81	0.78	0.76	0.74	0.72	0.71	0.70	0.70	0.68	0.72
ME	1,843	3,448	4,686	5,062	5,196	5,282	5,070	4,699	4,071	3,787
IVOL (%)	11.66	10.92	10.52	10.31	10.14	10.03	9.86	9.84	9.84	10.10

This table provides summary statistics for the sample. Panel A presents summary statistics for the monthly stock sample from January 1977 to December 2022. RET is one-month-ahead stock returns; EAI is the measure of earnings announcement informativeness. All other variables are defined in Appendix B. Panel B reports the time-series average of the cross-sectional mean of selected firm characteristics for monthly equally weighted stock portfolios sorted by EAI (D1–D10).

Empirical Analyses

Determinants of EAI

We conduct analyses to identify the factors influencing EAI. The normal information flow (NmlInfo), as defined by Thomas et al. (2022), measures the extent of information disclosures during the pre-announcement period, while discretionary accruals (DISC_ACCR) indicate the quality of earnings disclosure.⁴ We anticipate that both variables will exhibit a negative relationship with EAI. This expectation is based on the rationale that a larger pre-announcement information flow renders the upcoming earnings announcement itself less informative, and similarly, lower earnings quality diminishes the informational value of the disclosure.

According to Beaver et al. (2018), the loss dummy (Loss), the number of analysts following (COVERAGE), and the reporting lag (LagRpt)—defined as the number of days between the fiscal period end and the announcement date—capture the temporary nature of losses, the monitoring effects of financial analysts, and information leakage prior to earnings announcements, respectively. Consequently, we expect that earnings announcements will be less informative for firms reporting a loss, more informative for those with greater analyst coverage, and less informative when there is a longer lag between the fiscal period end and the announcement.

In addition, we use market beta (BETA) to capture a firm’s systematic risk, while also controlling for the book-to-market equity ratio (BM) and firm size (ME). We posit that the earnings announcements of high-beta stocks, growth stocks (with lower BM), and smaller firms are more informative, as they help resolve a greater degree of uncertainty regarding the valuation of these particular types of stocks. Finally, the dividend yield (DIVYIELD) reflects the maturity of a firm and the challenge of firm valuation for investors (Grullon et al., 2002). We anticipate this variable will be negatively related to EAI, as there is inherently less uncertainty surrounding a more mature firm in the periods leading up to its earnings announcements.

The sample used in the determinant analyses is derived from our primary monthly sample by retaining observations from the final month of each quarter. The t-statistics in square brackets are adjusted for serial correlations using Newey–West standard errors with 12 lags.

Columns (1) and (2) in Table 2 show that when EAI or

\sqrt{R I}

is the dependent variable, the estimated coefficients on NmlInfo, Loss, LagRpt, and DIVYIELD are significantly negative, while those on BETA and COVERAGE are significantly positive.⁵ The coefficient estimates for other variables show a mix of signs and statistical significance. In line with Beaver et al. (2018), the findings suggest that EAI rises with a firm’s systematic risk and analyst monitoring, and declines with accounting losses, reporting delays, and firm maturity. Importantly, NmlInfo is negatively correlated with EAI, suggesting that increased information disclosures during non-earnings announcement periods are linked to less informative earnings announcements.⁶

Table 2.

Determinant Analyses on EAI

Variables	(1)	(2)
Variables	EAI	$\sqrt{R I}$
Intercept	1.503*** [8.48]	1.633*** [8.00]
NmlInfo	−0.011*** [−3.38]	−0.018*** [−3.90]
DISC_ACCR	−0.055* [−1.73]	0.046 [1.37]
BETA	0.048*** [2.84]	0.079*** [3.50]
BM	−0.017 [−1.61]	−0.037** [−2.27]
ME	−0.001 [−0.07]	−0.020* [−1.95]
Loss	−0.080*** [−4.84]	−0.123*** [−5.32]
COVERAGE	0.038** [2.26]	0.061*** [3.08]
LagRpt	−0.003*** [−4.62]	−0.003*** [−4.73]
DIVYIELD	−0.035*** [−6.37]	−0.040*** [−6.55]
# Quarters	184	184
# OBS	360,075	360,075
Adj. R²	0.064	0.021

This table reports results from quarterly Fama–MacBeth regressions analyzing the determinants of EAI. The sample is derived from the main monthly sample by retaining the final-month observations for each quarter. EAI is the measure of earnings announcement informativeness; $\sqrt{R I}$ is the square root of the revaluation index as of the end of the most recent quarter. Detailed definitions of all other variables are provided in Appendix B. T-statistics in square brackets are adjusted for serial correlation using Newey–West standard errors with 12 lags. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively

Risk Premium of EAI

Our hypothesis is that there is a risk premium associated with the informativeness of earnings announcements. We begin by conducting stock portfolio analyses, in which we sort stocks into 10 portfolios based on their EAI and rebalance these portfolios on a monthly basis. Subsequently, we collect one-month-ahead returns for the composite stocks to calculate the equally weighted portfolio returns.

Table 3 displays the monthly average returns of stock portfolios sorted by EAI. Column (1) shows that the top decile of stocks with the lowest EAI earns 0.96% per month, while the bottom decile of stocks with the highest EAI earns 1.36% per month. The difference in returns between the top and bottom deciles is 0.40% per month, or 4.80% per year. The spread is statistically significant, as indicated by the t-value calculated using Newey–West standard errors with 12 lags.⁷ Columns (2) to (6) utilize various risk factor models, including the CAPM model, the Fama–French three-factor model, the Carhart four-factor model, the Fama–French five-factor model (Fama & French, 2015), and the mispricing factor model proposed by Stambaugh and Yuan (2017).⁸ Notably, the difference in returns between the top and bottom deciles of stocks remains statistically significant even after adjusting for various risk factors.⁹

Table 3.

Portfolio Analyses on Abnormal Returns of EAI

EAI decile	(1)	(2)	(3)	(4)	(5)	(6)
EAI decile	RAW	CAPM	ALPHA3F	ALPHA4F	ALPHA5F	SY-Alpha
1 = lowest	0.96*** [4.71]	−0.00 [−0.01]	−0.10 [−1.63]	−0.11* [−1.78]	−0.02 [−0.40]	−0.04 [−0.58]
2	1.07*** [4.87]	0.04 [0.38]	−0.07 [−1.17]	−0.06 [−0.92]	−0.05 [−0.78]	−0.00 [−0.08]
3	1.14*** [5.43]	0.10 [0.77]	−0.02 [−0.31]	0.02 [0.28]	−0.01 [−0.12]	0.04 [0.71]
4	1.16*** [5.26]	0.11 [0.87]	−0.02 [−0.34]	0.03 [0.51]	−0.02 [−0.34]	0.04 [0.59]
5	1.19*** [5.41]	0.12 [1.10]	−0.00 [−0.03]	0.05 [1.00]	−0.01 [−0.13]	0.00 [0.05]
6	1.30*** [5.94]	0.23* [1.95]	0.12** [2.27]	0.20*** [3.70]	0.10** [2.04]	0.15** [2.56]
7	1.26*** [5.71]	0.19 [1.59]	0.08 [1.49]	0.14*** [2.68]	0.06 [1.07]	0.09 [1.34]
8	1.20*** [5.56]	0.13 [1.16]	0.02 [0.34]	0.10* [1.85]	−0.03 [−0.50]	0.03 [0.42]
9	1.19*** [5.42]	0.12 [0.92]	0.01 [0.22]	0.09 [1.54]	−0.01 [−0.22]	0.01 [0.25]
10 = highest	1.36*** [5.92]	0.29** [2.11]	0.17** [2.56]	0.27*** [3.62]	0.15** [2.21]	0.22*** [2.65]
10-1	0.40*** [4.65]	0.29*** [3.41]	0.27*** [3.16]	0.38*** [4.27]	0.18** [1.99]	0.25** [2.50]

This table shows one-month-ahead returns for monthly stock portfolios sorted by EAI (earnings announcement informativeness). Portfolios are equally weighted. Raw = raw stock returns; CAPM = abnormal returns based on the market factor model; ALPHA3F = abnormal returns based on the Fama–French three-factor model; ALPHA4F = abnormal returns based on the Carhart four-factor model; ALPHA5F = abnormal returns based on the Fama–French five-factor model; SY-Alpha = abnormal returns based on the Stambaugh and Yuan (2017) mispricing factor model. T-statistics in square brackets are adjusted for serial correlation using Newey–West standard errors with 12 lags. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively

Next, we conduct monthly cross-sectional regressions to further investigate the effect of EAI on future stock returns

{R E T}_{i t + 1} = β_{0} + β_{1} {E A I}_{i t} + β_{2} {C o n t r o l s}_{i t} + ε_{i t,}

(3)

Where

R E T_{i t + 1}

is the return of stock

i

in month

t + 1

, and

E A I_{i t}

is the EAI measure of stock

i

in month

t

. In Fama–MacBeth regressions, the coefficient estimate for each variable represents the average of the coefficient estimates on that variable across all months. The t-values of the coefficient estimates are calculated using Newey–West standard errors with 12 lags.

Column (1) of Table 4 shows that EAI is significantly and positively correlated with one-month-ahead stock returns, with a t-value of 4.41. The coefficient estimates for control variables are consistent with previous studies. Specifically, the book-to-market (BM) ratio is positively related to one-month-ahead stock returns, while market equity (ME) is negatively related, confirming the existence of value and size premiums (Fama & French, 1992). RET12 is positively associated with stock returns 1 month ahead, in line with the momentum effect identified by Jegadeesh and Titman (1993). In columns (2) to (5), where more control variables are specified, the coefficient estimate for EAI remains significantly positive. In column (2), the coefficient estimate on idiosyncratic volatility (IVOL) is significantly negative, confirming the findings of Ang et al. (2006). In column (3), the coefficient estimate on discretionary accruals (DISC_ACCR) is significantly negative, consistent with Sloan (1996). In column (4), the coefficient estimate for forecast dispersion (FDISP) is significantly negative, which aligns with Diether et al. (2002). In column (5), industry fixed effects are controlled for, and EAI is still positively associated with future stock returns, with a t-value of 4.53. In column (6), the coefficient estimate on the alternative measure EAI2 is significantly positive, with a t-value comparable to the previous ones. In column (7), four variables are additionally controlled for: idiosyncratic volatility (IVOL), discretionary accruals (DISC_ACCR), the bid-ask spread (SPRD), and past 1-month returns (PRET). The finding regarding EAI remains robust, suggesting that it is not attributed to these factors.

Table 4.

Monthly Regressions of Future Stock Returns on EAI

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Variables	RET	RET	RET	RET	RET	RET	RET
Intercept	1.230*** [3.77]	1.842*** [5.16]	1.290*** [3.95]	0.987*** [2.63]	1.035*** [3.26]	1.035*** [3.26]	1.934*** [4.82]
EAI	0.186*** [4.41]	0.155*** [3.61]	0.178*** [4.14]	0.141** [2.46]	0.161*** [4.53]		0.109*** [2.77]
EAI2						0.148*** [4.53]
BETA	−0.117 [−0.50]	0.052 [0.24]	−0.126 [−0.54]	−0.022 [−0.10]	−0.073 [−0.37]	−0.073 [−0.37]	0.084 [0.51]
BM	0.142* [1.90]	0.099 [1.41]	0.120 [1.60]	0.174** [2.17]	0.202*** [3.31]	0.202*** [3.31]	0.162*** [2.72]
ME	−0.044 [−1.58]	−0.101*** [−3.73]	−0.052* [−1.84]	−0.106*** [−2.97]	−0.029 [−1.09]	−0.029 [−1.10]	−0.100*** [−3.51]
RET12	0.532*** [3.10]	0.550*** [3.27]	0.536*** [3.08]	0.391** [2.00]	0.400** [2.54]	0.400** [2.54]	0.382** [2.35]
IVOL		−0.054*** [−6.98]					−0.040*** [−6.69]
DISC_ACCR			−0.430** [−2.10]				−0.308 [−1.58]
FDISP				−0.217*** [−5.17]
SPRD							−0.206 [−1.54]
PRET							−3.656*** [−8.06]
Industry FE	No	No	No	No	Yes	Yes	Yes
# Months	552	552	552	552	552	552	552
# OBS	1,131,699	1,131,694	1,104,607	734,248	1,131,699	1,131,699	1,104,592
Adj. R²	0.044	0.050	0.045	0.065	0.100	0.100	0.117

This table presents results from monthly cross-sectional regressions of future stock returns (RET) on EAI and additional control variables. EAI is the measure of earnings announcement informativeness; all other variables are defined in Appendix B. T-statistics in square brackets are adjusted for serial correlation using Newey–West standard errors with 12 lags. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively

Variations in Risk Premium

This subsection examines variations in the EAI risk premium associated with both firm-level and market-level frictions. Specifically, we inspect firm maturity and information uncertainty as two indicators of firm-level frictions, based on the premise that they increase investors’ demand for information, thereby commanding a higher risk premium. Regarding market-level frictions, we investigate whether the pricing effect of EAI is amplified during periods of heightened investor risk aversion.

We expect a negative relationship between firm maturity and the EAI premium, as greater maturity indicates lower firm-level frictions. If a firm has operated for a long period and follows a stable business model, investors exhibit a lower appetite for information prior to earnings announcements. This is because investors possess tight priors regarding the firm’s value, and the marginal effect of learning from public disclosures or non-public information observed through stock prices is limited. Conversely, for investors in a young firm, lower information availability prior to an earnings announcement poses a more significant valuation challenge. This heightened uncertainty results in investors demanding a higher information risk premium. We use two proxies for firm maturity: firm age and a dividend payment dummy. A long operating history facilitates valuation (Easley & O’Hara, 2004; Kumar et al., 2008), as does the act of paying dividends, which signals financial stability (Grullon et al., 2002).

Furthermore, we expect a positive relationship between return volatility and the EAI premium. A large body of accounting literature interprets return volatility as a manifestation of information uncertainty, stemming either from a poor information environment or from inherent operational complexity (e.g., Gao et al., 2017; Jiang et al., 2005; Li et al., 2021; Lyle, 2025). Higher information uncertainty should command a greater information risk premium. When return volatility is elevated, the potential for both gains and losses in trading increases, thereby heightening investors’ stakes in the firm’s future prospects. In such an environment, investors demand more information to resolve uncertainty and require compensation for bearing the residual risk associated with the lack of information. Return volatility is specifically measured by the standard deviation of monthly stock returns over the past 12 months (RETVAR) or the idiosyncratic volatility from the last month (IVOL), as described in Ang et al. (2006).¹⁰

Table 5 displays regressions with the interaction terms between EAI and the four proxies discussed earlier.¹¹ We acknowledge that the chosen variables are only indirect proxies for the underlying cross-sectional factors in which we are interested. Column (1) shows a significantly negative coefficient estimate for the interaction term between EAI and firm age, suggesting that the impact of EAI on future stock returns is more pronounced among younger firms. Column (2) reveals a significantly negative interaction term between EAI and the dummy variable for paying cash dividends (PAYDIV), indicating that the impact of EAI on future stock returns is significantly diminished when firms pay cash dividends. Column (3) shows that the variability of returns (RETVAR) enhances the risk premium of EAI, as inferred from its positive interaction effect with EAI. This is further confirmed in column (4), which uses IVOL as the measure of return volatility.

Table 5.

Cross-sectional Variations in Risk Premium

Variables	(1)	(2)	(3)	(4)
Variables	RET	RET	RET	RET
Intercept	1.750*** [4.35]	1.833*** [4.78]	2.152*** [5.40]	2.266*** [5.61]
EAI	0.297** [2.34]	0.178*** [2.86]	−0.029 [−0.40]	−0.136** [−2.44]
EAI × Ln(AGE)	−0.070* [−1.88]
EAI × PAYDIV		−0.151** [−2.42]
EAI $\times$ RETVAR			0.003* [1.78]
EAI $\times$ IVOL				0.029*** [4.59]
Ln(AGE)	0.069 [1.30]
PAYDIV		0.194** [2.14]
RETVAR			−0.009*** [−3.09]
BETA	0.071 [0.45]	0.077 [0.48]	0.133 [0.90]	0.074 [0.45]
BM	0.171*** [2.94]	0.161*** [2.75]	0.145** [2.50]	0.160*** [2.69]
ME	−0.093*** [−3.26]	−0.099*** [−3.42]	−0.103*** [−3.61]	−0.101*** [−3.55]
RET12	0.379** [2.33]	0.383** [2.37]	0.441*** [2.73]	0.383** [2.35]
IVOL	−0.040*** [−6.72]	−0.040*** [−6.71]	−0.035*** [−6.56]	−0.073*** [−7.58]
DISC_ACCR	−0.32 [−1.63]	−0.312 [−1.62]	−0.283 [−1.44]	−0.319 [−1.65]
SPRD	−0.203 [−1.55]	−0.197 [−1.54]	−0.108 [−0.98]	−0.193 [−1.44]
PRET	−3.671*** [−8.08]	−3.662*** [−8.09]	−3.720*** [−8.02]	−3.654*** [−8.06]
Industry FE	Yes	Yes	Yes	Yes
# Months	552	552	552	552
# OBS	1,104,592	1,104,592	1,104,592	1,104,592
Adj. R²	0.119	0.119	0.122	0.118

This table reports results on cross-sectional variations in the EAI risk premium across firm maturity and return volatility. AGE = firm age; PAYDIV = dummy variable equal to 1 if the firm pays cash dividends, 0 otherwise; RETVAR = standard deviation of monthly returns over the past 12 months; IVOL = idiosyncratic volatility measure. All other variables are defined in Appendix B. T-statistics in square brackets are adjusted for serial correlation using Newey–West standard errors with 12 lags. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively

Regarding market-level frictions related to investor risk aversion, we expect a positive relationship with the EAI premium. This is consistent with the fundamental financial principle that higher risk aversion necessitates greater compensation. Each month, a cross-sectional regression of stock returns on EAI and control variables—as shown in column (7) of Table 4—produces a coefficient estimate on EAI denoted as $β_{E A I}$ , indicating the degree to which EAI is priced during that month. To investigate the relationship between investors’ risk aversion and the risk premium of EAI, $β_{E A I}$ is regressed on the monthly estimate of investors’ risk aversion provided by Bekaert et al. (2022). Additionally, the CBOE volatility index (VIX), known as the “fear index,” serves as an alternative measure of investors’ risk aversion.

Column (1) of Table 6 shows a positive relationship between Bekaert et al.’s (2022) risk aversion estimate and

β_{E A I}

. Column (2) reports a positive relationship between the VIX and

β_{E A I}

. Together, these findings suggest that when investors exhibit high risk aversion, the pricing effect of EAI, as measured by

β_{E A I}

, is amplified.¹²

Table 6.

Investors’ Risk Aversion and Risk Premium

Variable	(1)	(2)
Variable	$β_{E A I}$	$β_{E A I}$
Intercept	−0.217 [−1.30]	−0.109 [−0.95]
Risk Aversion	0.114** [2.04]
VIX		0.013** [2.08]
# OBS	438	395
Adj. R²	0.007	0.012

This table reports results on the role of investor risk aversion in determining the EAI risk premium. $β_{E A I}$ denotes the EAI risk premium, defined as the monthly coefficient estimate on EAI from the monthly regressions of future stock returns on EAI and control variables (Column 7, Table 4). Risk Aversion = monthly investor risk aversion estimate (Bekaert et al., 2022), starting June 1986; VIX = CBOE Volatility Index (the “fear index”), starting January 1990. T-values in square brackets are calculated using robust standard errors. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively

Control for ERC or PEAD

We examine whether the observed return predictability of EAI can be attributed to ERC or PEAD, both of which may be correlated with EAI and stock returns following earnings announcements (e.g., Bernard & Thomas, 1990; Chan et al., 1996). To ensure that our results are not influenced by ERC, we construct a firm-level ERC measure. Specifically, for each firm-month, we analyze earnings announcements from the previous 12 quarters, aligning with the timeframe of EAI. We compute the announcement returns (denoted as CAR) and a measure of earnings surprise (denoted as SUR) for each announcement. CAR represents the cumulative abnormal returns over a 3-day window surrounding the announcement date, while SUR is calculated as the earnings per share (EPS) currently announced minus the EPS of the same quarter in the previous year, divided by the standard deviation of EPS over the past eight quarters (Chan et al., 1996). We regress CAR on SUR to obtain an estimate of ERC for each stock-month, which we then use as an additional control variable. We require a minimum of eight unique CAR and SUR pairs to estimate the ERC. As shown in column (1) of Table 7, the findings regarding EAI remain consistent with the inclusion of the firm-level ERC as a control variable. Furthermore, ERC is not significantly priced.

Table 7.

Robustness Checks on ERC and PEAD

Variables	(1)	(2)	(3)
Variables	RET	RET	RET
Intercept	1.699*** [4.18]	1.921*** [4.78]	1.780*** [4.60]
EAI	0.111*** [2.73]	0.101*** [2.64]	0.109*** [2.85]
BETA	−0.003 [−0.02]	0.084 [0.52]	0.076 [0.47]
BM	0.156** [2.55]	0.163*** [2.72]	0.164*** [2.76]
ME	−0.066** [−2.41]	−0.097*** [−3.42]	−0.098*** [−3.46]
RET12	0.281 [1.58]	0.332** [2.12]	0.369** [2.34]
IVOL	−0.033*** [−4.83]	−0.179*** [−6.66]	−0.179*** [−6.43]
DISC_ACCR	−0.171 [−0.80]	−0.294 [−1.50]	−0.350* [−1.82]
SPRD	−0.147 [−1.05]	−0.201 [−1.49]	−0.199 [−1.46]
PRET	−3.877*** [−8.78]	−3.773*** [−8.17]	−3.719*** [−8.02]
ERC	0.000 [0.88]
CAR		0.023*** [6.48]
SUR			0.040*** [5.93]
Industry FE	Yes	Yes	Yes
# Months	531	552	552
# OBS	883,584	1,089,735	1,079,728
Adj. R²	0.129	0.119	0.119

This table shows regression results after controlling for earnings response coefficients (ERC) or post-earnings announcement drift (PEAD) (Chan et al., 1996). RET = one-month-ahead stock returns; EAI = earnings announcement informativeness measure. Firm-level ERC is computed using unique combinations of announcement returns (CAR) and earnings surprise (SUR) over the last 12 quarters. CAR = 3-day announcement returns from the previous quarter; SUR = earnings surprise, calculated as the change in EPS relative to the same quarter of the prior year divided by the standard deviation of EPS over the past eight quarters. All other variables are defined in Appendix B. T-statistics in square brackets are adjusted for serial correlation using Newey–West standard errors with 12 lags. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively

To check the robustness of our results with respect to PEAD, we control for CAR or SUR from the preceding quarter. As shown in columns (2) and (3) of Table 7, the coefficient associated with EAI remains statistically significant. It is important to note that PEAD is generally attributed to behavioral biases among investors (Fink, 2021). In contrast, we believe that the return predictability of EAI is a product of investors’ rational response to information risk. Furthermore, while PEAD exhibits an inverse relationship with firm size, our unreported analysis indicates that the return predictability of EAI does not show a significant relationship with firm size.¹³

Evidence Around Reg FD

The relationship between EAI and expected stock returns could be endogenous. To strengthen causal inferences, we rely on Reg FD, which became effective on October 23, 2000. This regulation prohibits selective disclosures of material information to investors—including financial analysts and institutional investors—by U.S. firms, while allowing such disclosures for foreign ADR firms (Francis et al., 2006; Litvak, 2007). As a result, investors in U.S. firms have to rely more heavily on public disclosures (Kross & Suk, 2012), making upcoming earnings announcements more informative. This, in turn, elevates information risk and further increases expected stock returns (Easley & O’Hara, 2004).

Following Francis et al. (2006) and Litvak (2007), we conduct a difference-in-difference (DID) test. In particular, we consider U.S. firms as the treatment group and non-U.S. ADR firms as the control group. We analyze the difference between the two groups of firms around Reg FD to determine the relationship between EAI and expected stock returns. The regression model is as follows

R E T_{i t + 1} = β_{0} + β_{1} U S_{i} \times F D_{t} + β_{2} U S_{i} + β_{3} F D_{t} + β_{4} C o n t r o l s_{i t} + ε_{i t}

(4)

Where

R E T_{i t + 1}

is stock returns in month

t + 1

for firm

i

U S_{i}

is the dummy variable indicating that the firm is based in the U.S., and

F D_{t}

is the dummy variable indicating the post-regulation period.

The sample comprises monthly data from U.S. and ADR firms (including Canadian companies) spanning the years 1996 to 2005. In accordance with Fama and French (1992), we stipulate that the lagged book equity must be positive and that the lagged stock price exceeds US $5. Following the methodology of Francis et al. (2006), we exclude the last two quarters of 2000 from our analyses, as they are associated with regulatory discussions and implementation. The analyses are confined to a 16-quarter period centered around the enactment of the regulation. This narrow time frame minimizes the influence of unobserved confounding factors (Brogaard et al., 2017).

Figure 2 illustrates the quarterly average $\sqrt{R I}$ for U.S. and non-U.S. firms surrounding Reg FD. We can see that the average $\sqrt{R I}$ of U.S. firms is close to that of non-U.S. firms pre-Reg FD, but it surpasses the latter afterward, in line with existing findings (Beaver et al., 2020; Kross & Suk, 2012). Hence, the parallel-trend assumption holds for the DID test.

Figure 2.

Mean $\sqrt{R I}$ of U.S. and ADR firms around Reg FD.

Panel A of Table 8 shows that the coefficient estimate for the interaction term (

U S \times F D

) is positive and statistically significant, regardless of whether industry or firm fixed effects are controlled for. Hence, the expected returns of U.S. firms increased significantly after Reg FD compared to non-U.S. firms.¹⁴

Table 8.

Evidence around Reg FD

Panel A: Difference-in-differences of stock returns around the regulation
Variables	(1)	(2)
Variables	RET	RET
Intercept	−0.102 [−0.37]	0.866[1.26]
US $\times$ FD	1.089*** [4.28]	2.355*** [3.49]
BETA	−0.000 [−0.00]	−0.793* [−1.88]
BM	0.034*** [3.30]	0.211*** [6.34]
ME	−0.000*** [−6.28]	−0.000*** [−12.09]
RET12	0.623*** [6.68]	−0.349*** [−2.72]
IVOL	6.132 [1.30]	29.088*** [5.34]
DISC_ACCR	−0.691 [−1.44]	−3.156*** [−5.05]
SPRD	−0.094 [−0.76]	−0.320 [−1.09]
PRET	−2.046*** [−4.64]	−4.777*** [−11.56]
Cross-section FE	Industry	Firm
Time FE	Month	Month
# OBS	121,306	121,117
Adj. R²	0.141	0.154

Panel B: Difference-in-differences analyses using the sample of U.S. firms only
Variables	(1)	(2)
Variables	RET	RET
Intercept	25.262*** [32.92]	25.302*** [32.97]
Treatment $\times$ Post	0.386* [1.66]
Treatment $\times$ y ⁺¹		−0.151 [−0.49]
Treatment $\times$ y ⁺²		0.456 [1.41]
Treatment $\times$ y ⁺³		0.254 [0.94]
Treatment $\times$ y ⁺⁴		0.699** [2.33]
Treatment $\times$ y ⁺⁵		0.806*** [2.61]
Beta	−1.893*** [−8.84]	−1.906*** [−8.89]
BM	0.094 [0.85]	0.093 [0.83]
ME	−3.382*** [−32.26]	−3.386*** [−32.30]
RET12	−0.194** [−2.39]	−0.193** [−2.37]
IVOL	0.022** [2.51]	0.022** [2.50]
DISC_ACCR	0.238 [0.61]	0.238 [0.61]
SPRD	0.025 [0.21]	0.023 [0.19]
PRET	−4.610*** [−13.34]	−4.612*** [−13.35]
Cross-section FE	Firm	Firm
Time FE	Month	Month
# OBS	215,902	215,902
Adj. R²	0.142	0.142

This table presents results on changes in earnings announcement informativeness and expected stock returns surrounding the implementation of Reg FD. Panel A reports difference-in-differences estimates of future stock returns between U.S. and non-U.S. firms around the regulation. US = dummy variable equal to 1 for U.S. firms, 0 otherwise; FD = dummy variable equal to 1 for the post-regulation period, 0 otherwise. Panel B reports difference-in-differences analyses for U.S. firms only. The sample period is 1996–2005, with 2000 (Reg FD enactment) as the event year (year 0). Treatment = pre-event EAI multiplied by −1 (a larger value indicates a lower pre-event EAI and a greater Reg FD impact for firms with more pre-announcement disclosure); Post = dummy variable equal to 1 for the post-event period, 0 otherwise; y^+k = the k-th year after the event. All other variables are defined in Appendix B. Standard errors are clustered at the firm level. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

The Reg FD effects described above are likely to vary across U.S. firms. To test this conjecture, we conduct DID analyses using the sample of U.S. firms only. Specifically, we use pre-event EAI (multiplied by −1) as a proxy for treatment intensity (denoted as Treatment) from Reg FD. The rationale is that firms with lower pre-event EAI previously disclosed more information prior to their earnings announcements. These firms were subject to a greater impact from Reg FD, as the regulation restricted their ability to engage in such selective disclosure.

Panel B of Table 8 presents the results of regressions controlling for firm and year fixed effects, based on a subsample of monthly U.S. stocks from the period of 1996 to 2005. Column (1) reveals a positive coefficient for the interaction term $T r e a t m e n t \times P o s t$ , indicating that firms with higher treatment intensity experience larger increases in expected stock returns. Next, we conduct a detailed analysis of the post-regulation period. $y^{+ k}$ is the $k$ -th year after the event. The interaction term $T r e a t m e n t \times y^{+ k}$ demonstrates the incremental impact on firms with higher treatment intensity in the $k$ -th year after Reg FD relative to other firms. Column (2) provides the results, indicating that significant causal effects begin to emerge in the fourth year. Therefore, our findings on Reg FD offer consistent evidence that a reduced supply of information enhances the informativeness of earnings announcements and subsequently increases the cost of capital.

Robustness Tests

This section presents several robustness tests. Kim (2006) identifies a common risk factor associated with earnings volatility that explains stock returns in January. To distinguish our findings from Kim’s, we must determine whether our results are specific to January. Additionally, it is important to ascertain whether our findings are confined to a particular historical period. Return predictability resulting from mispricing may diminish over time as investors become more informed.¹⁵ However, return predictability stemming from a risk premium can persist across all periods if the associated risks do not dissipate.

Columns (1) and (2) of Panel A of Table 9 show that the coefficient estimate on EAI is not significant in January but is significant in non-January months. Therefore, our finding is not related to the January effect (Kim, 2006). In columns (3) to (5), the sample is divided into two periods: 1977–2001 and 2002–2022. The coefficient estimates for EAI are statistically significant in both periods. To facilitate the comparison of risk premiums across periods, we use the monthly rank of EAI normalized between zero and one, denoted as EAI Rank. Columns (6) and (7) indicate that the coefficient on EAI Rank is 55% (=0.225/0.145-1) larger in the post-FD period, suggesting that the risk premium increases by this magnitude. This is consistent with the idea that a decrease in information supply after Reg FD leads to a higher risk premium.

Table 9.

Robustness Tests

Panel A: Subsample analyses
Variables	Jan.	Non-Jan.	1977-2021	2002-2022	1977-2001	2002-2022
	(1)	(2)	(3)	(4)	(5)	(6)
	RET	RET	RET	RET	RET	RET
Intercept	1.503 [1.55]	1.974*** [4.63]	2.131*** [4.37]	1.700*** [2.61]	2.198*** [4.53]	1.736*** [2.70]
EAI	0.242 [1.46]	0.097** [2.32]	0.125** [2.01]	0.090** [2.10]
EAI Rank					0.145* [1.94]	0.225* [1.91]
BETA	0.414 [1.46]	0.053 [0.29]	0.101 [0.56]	0.063 [0.22]	0.101 [0.56]	0.061 [0.21]
BM	0.692** [2.37]	0.114* [1.91]	0.273*** [3.36]	0.031 [0.40]	0.273*** [3.37]	0.031 [0.40]
ME	−0.286 [−1.66]	−0.083*** [−3.01]	−0.107*** [−2.66]	−0.091** [−2.40]	−0.107*** [−2.66]	−0.092** [−2.42]
RET12	−0.595** [−2.57]	0.471*** [2.70]	0.774*** [6.39]	−0.085 [−0.29]	0.774*** [6.39]	−0.083 [−0.29]
IVOL	−0.072 [−0.84]	−0.194*** [−7.10]	−0.261*** [−8.19]	−0.092** [−2.46]	−0.261*** [−8.18]	−0.091** [−2.43]
DISC_ACCR	−0.789 [−1.14]	−0.264 [−1.25]	−0.476** [−2.35]	−0.108 [−0.31]	−0.473** [−2.33]	−0.108 [−0.31]
SPRD	0.905* [1.82]	−0.307** [−2.30]	0.031 [0.20]	−0.487** [−2.24]	0.03 [0.20]	−0.486** [−2.23]
PRET	−11.675*** [−9.33]	−2.927*** [−6.40]	−5.135*** [−8.50]	−1.895*** [−4.54]	−5.135*** [−8.51]	−1.895*** [−4.53]
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes
# Months	46	506	300	252	300	252
# OBS	89,630	1,014,962	596075	508517	596,075	508,517
Adj. R²	0.146	0.114	0.120	0.113	0.12	0.114

Panel B: EAI based on alternative event windows or factor models
Variables	CAR [0]	CAR [−1, 0]	Market model	FF3 model
	(1)	(2)	(3)	(4)
	RET	RET	RET	RET
Intercept	1.931*** [4.80]	1.933*** [4.81]	1.955*** [4.86]	1.937*** [4.81]
EAI	0.063*** [2.78]	0.088*** [2.78]	0.095** [2.48]	0.107*** [2.75]
BETA	0.083 [0.51]	0.084 [0.51]	0.083 [0.50]	0.083 [0.51]
BM	0.162*** [2.72]	0.162*** [2.72]	0.163*** [2.73]	0.163*** [2.73]
ME	−0.100*** [−3.51]	−0.100*** [−3.51]	−0.100*** [−3.52]	−0.100*** [−3.51]
RET12	0.382** [2.35]	0.382** [2.35]	0.382** [2.35]	0.382** [2.35]
IVOL	−0.184*** [−6.68]	−0.184*** [−6.69]	−0.184*** [−6.69]	−0.184*** [−6.69]
DISC_ACCR	−0.308 [−1.58]	−0.307 [−1.58]	−0.305 [−1.57]	−0.306 [−1.57]
SPRD	−0.206 [−1.54]	−0.206 [−1.54]	−0.209 [−1.57]	−0.206 [−1.54]
PRET	−3.656*** [−8.06]	−3.655*** [−8.06]	−3.657*** [−8.07]	−3.656*** [−8.06]
Industry FE	Yes	Yes	Yes	Yes
# Months	552	552	552	552
# OBS	1,104,592	1,104,592	1,104,592	1,104,592
Adj. R²	0.117	0.117	0.117	0.117

Panel C: EAI based on revaluation indexes at different horizons
EAI window	Qtr. [−4: −1]	Qtr. [−8: −1]	Qtr. [−16: −1]	Qtr. [−20: −1]
Variables	(1)	(2)	(3)	(4)
Variables	RET	RET	RET	RET
Intercept	2.039*** [5.24]	1.980*** [4.98]	1.888*** [4.65]	1.892*** [4.62]
EAI	0.030 [1.52]	0.069** [2.32]	0.147*** [3.36]	0.147*** [3.13]
BETA	0.086 [0.52]	0.084 [0.51]	0.082 [0.50]	0.082 [0.50]
BM	0.163*** [2.74]	0.163*** [2.74]	0.162*** [2.72]	0.162*** [2.72]
ME	−0.099*** [−3.47]	−0.098*** [−3.47]	−0.100*** [−3.54]	−0.101*** [−3.58]
RET12	0.383** [2.34]	0.382** [2.34]	0.383** [2.35]	0.383** [2.35]
IVOL	−0.185*** [−6.73]	−0.184*** [−6.70]	−0.184*** [−6.68]	−0.184*** [−6.67]
DISC_ACCR	−0.306 [−1.58]	−0.308 [−1.58]	−0.305 [−1.57]	−0.306 [−1.57]
SPRD	−0.214 [−1.62]	−0.209 [−1.58]	−0.200 [−1.49]	−0.201 [−1.49]
PRET	−3.651*** [−8.06]	−3.657*** [−8.07]	−3.660*** [−8.08]	−3.663*** [−8.08]
Industry FE	Yes	Yes	Yes	Yes
# Months	552	552	552	552
# OBS	1,104,580	1,104,592	1,104,592	1,104,592
Adj. R²	0.117	0.117	0.117	0.117

This table presents robustness tests for the EAI risk premium. Panel A reports results for January vs. non-January months, and for two subsample periods (1977–2001 and 2002–2022). Panel B shows results when EAI is calculated using earnings announcement returns from alternative event windows or different risk factor models. Panel C shows results when EAI is calculated as the average of revaluation indexes over alternative past time horizons. EAI Rank = monthly EAI rank normalized between 0 and 1. All other variables are defined in Appendix B. T-statistics in square brackets are adjusted for serial correlation using Newey–West standard errors with 12 lags. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

Panel B demonstrates the robustness of our results across various event windows and risk factor models used to calculate earnings announcement returns. Panel C shows that the results remain robust across different time horizons when measuring the informativeness of earnings announcements. The coefficient estimates on EAI are all statistically significant, based on the average root of revaluation indexes taken over the past 4, 8, 16, and 20 quarters. Thus, EAI seems to have a long-lasting impact on expected stock returns, a characteristic of risk premiums.

Conclusion

Volatile earnings announcements suggest that investors lacked sufficient information about the firm during the pre-announcement period. Limited information availability prior to earnings announcements exposes investors to information risk, as outlined in the theory proposed by Easley and O’Hara (2004). Their theory posits that having more information is preferable to having less at any given moment. If the informativeness of earnings announcements is a function of information that is not disclosed to all investors in advance, rational investors should anticipate a risk premium associated with this lack of information.

Consistent with the aforementioned prediction, the informativeness of earnings announcements is positively correlated with future stock returns in a sample of U.S. stocks spanning from 1977 to 2022. Furthermore, the return premium associated with earnings announcement informativeness is greater for younger firms, those that do not distribute cash dividends, and stocks exhibiting high return volatility. This effect is particularly pronounced during periods of heightened investor risk aversion. Notably, there is robust evidence indicating that, following the implementation of Reg FD, a reduction in information supply led to a higher cost of capital. Our study contributes to the existing literature on the informativeness of earnings announcements by elucidating its economic implications. It underscores the timeliness of information disclosure as a critical factor for investors and provides new insights into the impact of Reg FD on public firms, thereby enriching policy discussions.

Footnotes

Acknowledgments

We thank Xiao-Jun Zhang, Bohui Zhang, Stephen Dimmock, Chuan-Yang Hwang, Simba Xin Chang, Chenkai Ni, and the participants of the 2018 China Finance Review International Conference for their helpful comments.

ORCID iDs

Yu Cai

Renhui Fu

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We acknowledge financial support from the National Natural Science Foundation of China [72221001, 72132006, 72372097, 72521002], National Social Science Fund of China [22AZD034], the Shuguang Program [23SG31], the Program for Innovative Research Team of SUFE, and the MOE Project of Key Research Institute of Humanities and Social Science in University [23JJD790010] and the 111 Project [B18033]. All errors are our own responsibility.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Notes

Author Biographies

Yu CAI is affiliated with the Accounting Department at the School of Economics and Management, Tongji University. He holds a PhD, CFA, and CPA, having received his doctorate from Nanyang Technological University in 2012. His research spans financial management, financial analysts, asset pricing, and public policies. He has published in reputable journals such as Regional Studies, the Journal of Banking and Finance, and the Journal of Accounting, Auditing & Finance.

Renhui FU is from Antai College of Economics and Management, Shanghai Jiao Tong University. His primary research interest lies in the field of capital market accounting. He received his PhD in accounting from Nanyang Technological University in 2010. His work has appeared in esteemed journals including the Journal of Accounting and Economics, the Journal of International Business Studies, the Journal of Law and Economics, the Journal of Banking and Finance, and the Journal of Corporate Finance.

Sie Ting LAU is from Nanyang Business School, Nanyang Technological University, Singapore. He earned his PhD in Finance from the University of Texas at Arlington in May 1992. His research interests include international stock exchange listings, market microstructure of securities markets, international mutual funds and home bias, and international financial markets. He has published extensively in top-tier journals such as the Journal of Finance, the Journal of Financial Economics, and Management Science.

Huifang YIN is affiliated with the School of Accountancy at Shanghai University of Finance and Economics. She received her PhD in accounting from Nanyang Technological University in 2011. Her research covers a wide range of financial and accounting topics. She has co-authored multiple papers in renowned journals such as the Journal of Accounting and Economics, the Journal of Accounting Research, Review of Accounting Studies, and the Journal of Accounting, Auditing & Finance.

Appendix

References

Amihud

Mendelson

(1989). The effects of beta, bid-ask spread, residual risk, and size on stock returns. Journal of Finance, 44(2), 479–486. https://doi.org/10.2307/2328600

Amir

Levi

(2019). Estimating the precision of information on earnings and nonearnings announcement days, and its relation with the cost of equity. The European Accounting Review, 28(2), 223–248. https://doi.org/10.1080/09638180.2018.1461119

Ang

Hodrick

R. J.

Xing

Zhang

(2006). The cross-section of volatility and expected returns. The Journal of Finance, 61(1), 259–299. https://doi.org/10.1111/j.1540-6261.2006.00836.x

Anilowski

Feng

Skinner

D. J.

(2007). Does earnings guidance affect market returns? The nature and information content of aggregate earnings guidance. Journal of Accounting and Economics, 44(1-2), 36–63. https://doi.org/10.1016/j.jacceco.2006.09.002

Arif

De George

E. T.

(2020). The dark side of low financial reporting frequency: Investors’ reliance on alternative sources of earnings news and excessive information spillovers. The Accounting Review, 95(6), 23–49. https://doi.org/10.2308/tar-2017-0018

Atiase

R. K.

(1985). Predisclosure information, firm capitalization, and security price behavior around earnings announcements. Journal of Accounting Research, 23(1), 21–36. https://doi.org/10.2307/2490905

Atiase

R. K.

Bamber

L. S.

Tse

(1989). Timeliness of financial reporting, the firm size effect, and stock price reactions to annual earnings announcements. Contemporary Accounting Research, 5(2), 526–552. https://doi.org/10.1111/j.1911-3846.1989.tb00722.x

Ball

Brown

(1968). An empirical evaluation of accounting income numbers. Journal of Accounting Research, 6(2), 159–178. https://doi.org/10.2307/2490232

Barber

B. M.

De George

E. T.

Lehavy

Trueman

(2013). The earnings announcement premium around the globe. Journal of Financial Economics, 108(1), 118–138. https://doi.org/10.1016/j.jfineco.2012.10.006

10.

Beaver

W. H.

(1968). The information content of annual earnings annoucements. Journal of Accounting Research, 6(3), 67–92. https://doi.org/10.2307/2490070

11.

Beaver

W. H.

McNichols

M. F.

Wang

Z. Z.

(2018). The information content of earnings announcements: New insights from intertemporal and cross-sectional behavior. Review of Accounting Studies, 23(1), 95–135. https://doi.org/10.1007/s11142-017-9417-z

12.

Beaver

W. H.

McNichols

M. F.

Wang

Z. Z.

(2020). Increased market response to earnings announcements in the 21st century: An empirical investigation. Journal of Accounting and Economics, 69(1), 101244. https://doi.org/10.1016/j.jacceco.2019.101244

13.

Bekaert

Engstrom

E. C.

N. R.

(2022). The time variation in risk appetite and uncertainty. Management Science, 68(6), 3975–4004. https://doi.org/10.1287/mnsc.2021.4068

14.

Bernard

V. L.

Thomas

J. K.

(1990). Evidence that stock prices do not fully reflect the implications of current earnings for future earnings. Journal of Accounting and Economics, 13(4), 305–340. https://doi.org/10.1016/0165-4101(90)90008-r

15.

Brogaard

Xia

(2017). Stock liquidity and default risk. Journal of Financial Economics, 124(3), 486–502. https://doi.org/10.1016/j.jfineco.2017.03.003

16.

Campbell

J. Y.

Lettau

Malkiel

B. G.

(2001). Have individual stocks become more volatile? An empirical exploration of idiosyncratic risk. Journal of Finance, 56(1), 1–44. https://doi.org/10.1111/0022-1082.00318

17.

Chan

L. K.

Jegadeesh

Lakonishok

(1996). Momentum strategies. The Journal of Finance, 51(5), 1681–1713. https://doi.org/10.1111/j.1540-6261.1996.tb05222.x

18.

Cohen

Dey

Lys

Sunder

(2007). Earnings announcement premia and the limits to arbitrage. Journal of Accounting and Economics, 43(2-3), 153–180. https://doi.org/10.1016/j.jacceco.2007.01.008

19.

Corwin

S. A.

Schultz

(2012). A simple way to estimate bid-ask spreads from daily high and low prices. The Journal of Finance, 67(2), 719–760. https://doi.org/10.1111/j.1540-6261.2012.01729.x

20.

Da Silva

P. P.

(2019). Corporate governance, earnings quality and idiosyncratic crash risk during the 2007–2008 financial crisis. Journal of Multinational Financial Management, 51, 61–79. https://doi.org/10.1016/j.mulfin.2019.07.002

21.

Dechow

P. M.

Sloan

R. G.

Sweeney

A. P.

(1995). Detecting earnings management. The Accounting Review, 70(2), 193–225. https://doi.org/10.2308/tar-9505096112

22.

DeFond

Hung

Trezevant

(2007). Investor protection and the information content of annual earnings announcements: International evidence. Journal of Accounting and Economics, 43(1), 37–67. https://doi.org/10.1016/j.jacceco.2006.09.001

23.

Dhaliwal

Heitzman

O. Z.

(2006). Taxes, leverage, and the cost of equity capital. Journal of Accounting Research, 44(4), 691–723. https://doi.org/10.1111/j.1475-679x.2006.00214.x

24.

Diamond

Verrecchia

(1991). Disclosure, liquidity, and the cost of capital. The Journal of Finance, 46(4), 1325–1359. https://doi.org/10.1111/j.1540-6261.1991.tb04620.x

25.

Diether

K. B.

Malloy

C. J.

Scherbina

(2002). Differences of opinion and the cross section of stock returns. The Journal of Finance, 57(5), 2113–2141. https://doi.org/10.1111/0022-1082.00490

26.

Easley

O’Hara

(2004). Information and the cost of capital. The Journal of Finance, 59(4), 1553–1583. https://doi.org/10.1111/j.1540-6261.2004.00672.x

27.

Efendi

Park

J. D.

Subramaniam

(2016). Does the XBRL reporting format provide incremental information value? A study using XBRL disclosures during the voluntary filing program. Abacus, 52(2), 259–285. https://doi.org/10.1111/abac.12079

28.

El-Gazzar

S. M.

(1998). Predisclosure information and institutional ownership: A cross-sectional examination of market revaluations during earnings announcement periods. Accounting Review, 73(1), 119–129. https://doi.org/10.2308/tar-274319

29.

Ertan

Lee

Wittenberg-Moerman

(2025). Unexpected defaults: The role of information opacity. Review of Accounting Studies, 30(1), 899–949. https://doi.org/10.1007/s11142-024-09842-8

30.

Evans

M. E.

(2016). Commitment and cost of equity capital: An examination of timely balance sheet disclosure in earnings announcements. Contemporary Accounting Research, 33(3), 1136–1171. https://doi.org/10.1111/1911-3846.12172

31.

Fama

E. F.

(1970). Efficient capital markets. The Journal of Finance, 25(2), 383–417. https://doi.org/10.2307/2325486

32.

Fama

E. F.

French

K. R.

(1992). The cross-section of expected stock returns. The Journal of Finance, 47(2), 427–465. https://doi.org/10.2307/2329112

33.

Fama

E. F.

French

K. R.

(2015). A five-factor asset pricing model. Journal of Financial Economics, 116(1), 1–22. https://doi.org/10.1016/j.jfineco.2014.10.010

34.

Fink

(2021). A review of the post-earnings-announcement drift. Journal of Behavioral and Experimental Finance, 29, Article 100446. https://doi.org/10.1016/j.jbef.2020.100446

35.

Francis

Nanda

Wang

(2006). Re-examining the effects of regulation fair disclosure using foreign listed firms to control for concurrent shocks. Journal of Accounting and Economics, 41(3), 271–292. https://doi.org/10.1016/j.jacceco.2006.03.002

36.

Francis

Schipper

Vincent

(2002). Earnings announcements and competing information. Journal of Accounting and Economics, 33(3), 313–342. https://doi.org/10.1016/s0165-4101(02)00058-7

37.

Kraft

Tian

Zhang

Zuo

(2020). Financial reporting frequency and corporate innovation. Journal of Law and Economics, 63(3), 501–529. https://doi.org/10.1086/708706

38.

Gao

Khan

Tan

(2017). Further evidence on consequences of debt covenant violations. Contemporary Accounting Research, 34(3), 1489–1521. https://doi.org/10.1111/1911-3846.12303

39.

Goldstein

Yang

(2017). Information disclosure in financial markets. Annual Review of Financial Economics, 9(1), 101–125. https://doi.org/10.1146/annurev-financial-110716-032355

40.

Grullon

Michaely

Swaminathan

(2002). Are dividend changes a sign of firm maturity? The Journal of Business, 75(3), 387–424. https://doi.org/10.1086/339889

41.

Han

J. C.

Wild

J. J.

(2000). Predisclosure information, firm capitalization, and earnings information transfers. Journal of Business Research, 49(3), 273–288. https://doi.org/10.1016/s0148-2963(99)00015-6

42.

Hanlon

Maydew

Shevlin

(2008). An unintended consequence of book-tax conformity: A loss of earnings informativeness. Journal of Accounting and Economics, 46(2-3), 294–311. https://doi.org/10.1016/j.jacceco.2008.09.003

43.

Hermalin

B. E.

Weisbach

M. S.

(2012). Information disclosure and corporate governance. The Journal of Finance, 67(1), 195–233. https://doi.org/10.1111/j.1540-6261.2011.01710.x

44.

Holthausen

R. W.

Verrecchia

R. E.

(1988). The effect of sequential information releases on the variance of price changes in an intertemporal multi-asset market. Journal of Accounting Research, 26(1), 82–106. https://doi.org/10.2307/2491114

45.

Hughes

J. S.

Liu

(2007). Information asymmetry, diversification, and cost of capital. The Accounting Review, 82(3), 705–729. https://doi.org/10.2308/accr.2007.82.3.705

46.

Jegadeesh

Titman

(1993). Returns to buying winners and selling losers: Implications for stock market efficiency. The Journal of Finance, 48(1), 65–91. https://doi.org/10.1111/j.1540-6261.1993.tb04702.x

47.

Jiang

Lee

C. M.

Zhang

(2005). Information uncertainty and expected returns. Review of Accounting Studies, 10(2), 185–221. https://doi.org/10.1007/s11142-005-1528-2

48.

Jorgensen

B. N.

Kirschenheiter

M. T.

(2003). Discretionary risk disclosures. The Accounting Review, 78(2), 449–469. https://doi.org/10.2308/accr.2003.78.2.449

49.

Kim

(2006). On the information uncertainty risk and the January effect. The Journal of Business, 79(4), 2127–2162. https://doi.org/10.1086/503659

50.

Kim

Verrecchia

R. E.

(1991a). Market reaction to anticipated announcements. Journal of Financial Economics, 30(2), 273–309. https://doi.org/10.1016/0304-405x(91)90033-g

51.

Kim

Verrecchia

R. E.

(1991b). Trading volume and price reactions to public announcements. Journal of Accounting Research, 29(2), 302–321. https://doi.org/10.2307/2491051

52.

Koch

A. S.

Lefanowicz

C. E.

Robinson

J. R.

(2013). Regulation FD: A review and synthesis of the academic literature. Accounting Horizons, 27(3), 619–646. https://doi.org/10.2308/acch-50500

53.

Kross

W. J.

Suk

(2012). Does regulation FD work? Evidence from analysts’ reliance on public disclosure. Journal of Accounting and Economics, 53(1-2), 225–248. https://doi.org/10.1016/j.jacceco.2011.11.004

54.

Kumar

Sorescu

S. M.

Boehme

R. D.

Danielsen

B. R.

(2008). Estimation risk, information, and the conditional CAPM: Theory and evidence. The Review of Financial Studies, 21(3), 1037–1075. https://doi.org/10.1093/rfs/hhn016

55.

Landsman

W. R.

Maydew

E. L.

Thornock

J. R.

(2012). The information content of annual earnings announcements and mandatory adoption of IFRS. Journal of Accounting and Economics, 53(1-2), 34–54. https://doi.org/10.1016/j.jacceco.2011.04.002

56.

Levi

Zhang

(2015). Asymmetric decrease in liquidity trading before earnings announcements and the announcement return premium. Journal of Financial Economics, 118(2), 383–398. https://doi.org/10.1016/j.jfineco.2015.08.003

57.

O. Z.

Liu

(2021). Dividend taxes, investor horizon, and idiosyncratic volatility. The Accounting Review, 96(3), 403–430. https://doi.org/10.2308/tar-2018-0642

58.

Litvak

(2007). The effect of the Sarbanes-Oxley Act on non-US companies cross-listed in the US. Journal of Corporate Finance, 13(2-3), 0–228. https://doi.org/10.1016/j.jcorpfin.2007.03.002

59.

Livnat

Zarowin

(1990). The incremental information content of cash-flow components. Journal of Accounting and Economics, 13(1), 25–46. https://doi.org/10.1016/0165-4101(90)90066-d

60.

Lyle

M. R.

(2025). Valuation and returns on stock return volatility. The Accounting Review, 100(2), 299–328. https://doi.org/10.2308/tar-2020-0151

61.

Monnet

Quintin

(2017). Rational opacity. The Review of Financial Studies, 30(12), 4317–4348. https://doi.org/10.1093/rfs/hhx034

62.

Pevzner

Xie

Xin

(2015). When firms talk, do investors listen? The role of trust in stock market reactions to corporate earnings announcements. Journal of Financial Economics, 117(1), 190–223. https://doi.org/10.1016/j.jfineco.2013.08.004

63.

Rogers

J. L.

Van Buskirk

(2013). Bundled forecasts in empirical accounting research. Journal of Accounting and Economics, 55(1), 43–65. https://doi.org/10.1016/j.jacceco.2012.06.001

64.

Savor

Wilson

(2016). Earnings announcements and systematic risk. Journal of Finance, 71(1), 83–138. https://doi.org/10.1111/jofi.12361

65.

Schreder

(2018). Idiosyncratic information and the cost of equity capital: A meta-analytic review of the literature. Journal of Accounting Literature, 41(1), 142–172. https://doi.org/10.1016/j.acclit.2018.07.001

66.

Sloan

R. G.

(1996). Do stock prices fully reflect information in accruals and cash flows about future earnings? The Accounting Review, 71(3), 289–315. https://doi.org/10.2308/tar-9608042309

67.

Stambaugh

R. F.

Yuan

(2017). Mispricing factors. Review of Financial Studies, 30(4), 1270–1315. https://doi.org/10.1093/rfs/hhw107

68.

Stoumbos

(2023). The growth of information asymmetry between earnings announcements and its implications for reporting frequency. Management Science, 69(3), 1901–1928. https://doi.org/10.1287/mnsc.2022.4408

69.

Thomas

J. K.

Zhang

Zhu

(2022). Measuring the information content of disclosures: The role of return noise. The Accounting Review, 97(6), 417–443. https://doi.org/10.2308/tar-2021-0075