Does Litigation Risk Affect Meeting-or-Beating Earnings Expectations? Evidence from Quasi-Natural Experiment

Abstract

This paper examines the impact of litigation risk on meeting-or-beating earnings expectations (MBE). Whether and how litigation risk affects MBE is largely inconclusive due to potential reverse causality and measurement error. To find a causal effect, this study employs differences-in-differences tests based on the unanticipated legal event that decreased litigation risk for firms headquartered in the U.S. Ninth Circuit states. The result indicates that Ninth Circuit firms are more likely to meet or beat their earnings target after the ruling, suggesting that litigation risk negatively affects the likelihood of MBE. Further analyses show that investors’ reactions are less positive to MBE premium for firms with a decrease in litigation risk. Taken together, this study contributes to the literature by documenting that litigation risk is a significant factor influencing managers’ benchmark beating behavior.

1. Introduction

Securities class action lawsuits in the U.S. offer a mechanism by which owners (i.e., shareholders) can sue firms or managers in the event of managerial misconduct. The underlying premise of litigation risk is that substantial financial and reputational penalties associated with litigation can discipline managers and deter them from future wrongdoing (e.g., Hopkins, 2018; Huang et al., 2020; Zhang et al., 2023). Prior studies also document that litigations brought by shareholders play a crucial role in resolving the conflicts of interest between managers and shareholders (e.g., C. S. A. Cheng et al., 2010). By posing a credible legal threat to managers, shareholders can efficiently discipline them. Also, shareholders require reliable and timely disclosures to monitor managers (Armstrong et al., 2010). In this paper, I utilize unanticipated legal changes that affect shareholders’ ability to file class action lawsuits and investigate the effect of a firm’s litigation risk on managerial reporting strategies regarding mandatory disclosure (i.e., reported earnings around bright-line targets). Specifically, this study examines the effect of litigation risk on meeting-or-beating earnings expectations (hereafter “MBE”).

MBE is an important reporting outcome for firms and managers, and the incentives or benefits of MBE are well documented in prior studies (e.g., Bannister et al., 2023; Graham et al., 2005). Further, prior studies document a discontinuity around earnings benchmarks (e.g., Degeorge et al., 1999; Jacob & Jorgensen, 2007) and interpret the results as evidence that managers manage earnings to meet or beat their benchmarks. In addition, according to survey evidence presented by Graham et al. (2005), approximately 70% of chief financial officers (CFOs) believe that firms should take actions (i.e., manage earnings) to report desirable earnings even though such earnings management may sacrifice some economic value, but not too large, of their firms. Prior studies also find that firms enjoy financial benefits of MBE by managing earnings expectations downward and that firms tend to warn investors on news about upcoming negative earnings (Kasznik & Lev, 1995; Richardson et al., 2004; Tang et al., 2021).

Despite the importance of MBE and shareholder lawsuits, whether and how litigation risk affects MBE is largely inconclusive due to potential reverse causality and measurement error. Specifically, while litigation risk can directly affect firms’ financial policies and outcomes, including MBE, it is also possible that MBE changes the level of litigation risk. For example, Houston et al. (2019) find that litigation risk influences firms’ voluntary disclosures, whereas O’Brien and Hodges (1991) show that missing earnings benchmarks, which are accompanied by sudden stock price declines, trigger shareholder class action filings.1 In addition, empirically constructing a firm-level litigation risk measure is challenging because such a measure is based on certain industries and firm characteristics that might be determined endogenously. Accordingly, this study exploits a quasi-natural legal event that generates a plausibly exogenous shock to corporate litigation risk to establish a causal effect of litigation risk on MBE.

On 2 July 1999, the U.S. Ninth Circuit Court of Appeals released an unexpected ruling (re: Silicon Graphics Inc. Securities Litigation). This ruling, which applied only to firms in the Ninth Circuit states, required plaintiffs to prove that defendants were “deliberately reckless” in making the alleged misstatement or omitting material information in the disclosures.2 In contrast, proving “mere recklessness” was sufficient for firms located in other states. The requirement from the 1999 ruling significantly increased hurdles for shareholders filing class action lawsuits against firms located in Ninth Circuit states and therefore decreased litigation risk of such firms (Pritchard & Sale, 2005). Since the ruling only affects the litigation risk of a subset of U.S. firms (i.e., the Ninth Circuit firms), this study can control for other concurrent factors and time trends to mitigate the omitted variable bias by using this legal event.3 There are two competing hypotheses in relating litigation risk to MBE.

On one hand, the risk of securities litigation increases managers’ motivation/incentive to MBE (motivation-based hypothesis). A sudden stock price drop can be deemed as omission of negative information (the “fraud on the market” theory), and missing earnings expectations (i.e., a failure in MBE) likely triggers securities lawsuits (e.g., Field et al., 2005; Skinner, 1994). According to the “fraud on the market theory”, a large one-time stock price decline is more likely to precipitate the filing of securities class action lawsuits because it can be argued as evidence of omission of negative information, given that the stock market is informationally efficient (O’Brien & Hodges, 1991). In this context, subsequent to the lowered litigation risk by the 1999 ruling, managers in the Ninth Circuit firms face less pressure to MBE to avoid a stock price drop. Following this logic, litigation risk may positively affect the propensity to MBE.4

On the other hand, a threat of litigation may increase the costs of MBE borne by managers (cost-based hypothesis). Litigation risk, as a governance mechanism, is likely to deter managers facing severe capital market pressure from opportunistically managing earnings because of potential securities lawsuits regarding misstatement or omitting material information from earnings management (Appel, 2019). In this setting, a decline in litigation risk for firms located in Ninth Circuit states after the 1999 ruling increases the degree of earnings management for MBE purposes (Hopkins, 2018; Huang et al., 2020). Following this story, litigation risk may negatively affect MBE under the cost-based hypothesis.

Using a sample of firm-quarters over the period 1995–2003, this study finds that Ninth Circuit firms are more likely to meet or just-beat earnings expectations relative to other firms after the 1999 ruling, supporting the cost-based story that litigation risk decreases the propensity of MBE.5 The effect of litigation risk on MBE is incrementally significant after controlling for various firm aspects. To test a robustness of the results, the study also considers intervals of MBE (e.g., MBE by 2 and 5 cents versus missing by 2 and 5 cents, respectively) to examine the effect of litigation risk on narrow beaters versus narrow missers in which opportunistic incentives for earnings management are better captured in discontinuity around earnings benchmarks (e.g., Degeorge et al., 1999).6 Further analyses show that firms facing a decrease in litigation risk are more likely to report small positive earnings surprises (just-MBE) than small negative surprises (just missing benchmark).

Next, the study investigates how managers opportunistically employ earnings management tools when they face MBE incentives and/or costs associated with different levels of litigation risk. This study examines two commonly used earnings management channels, i.e., accrual-based earnings management (hereafter “AEM”) and real activity-based earnings management (hereafter “REM”). The results show that after the 1999 ruling, Ninth Circuit firms, compared to other firms, are more likely to MBE but miss the benchmark without considering discretionary accruals, suggesting that the MBE is achieved through AEM. However, the analyses do not provide similar results when the earnings management tool is REM.7

Furthermore, the study compares the stock market reaction to earnings announcements between the Ninth Circuit firms and other firms subsequent to the ruling. The results show that the three-day market reaction to MBE is less positive (i.e., smaller MBE premium) for firms located in Ninth Circuit states after the ruling. Taken together, the results imply that investors discount the reward MBE achieved when the litigation risk is unexpectedly reduced.

Finally, prior studies find that firms enjoy financial benefits of MBE by managing earnings expectations downward and warning investors about upcoming negative earnings news (Bartov et al., 2024; Richardson et al., 2004). Managers can engage not only in AEM and REM, but also in expectation management (hereafter “EXM”) to meet or beat earnings benchmarks. The additional test results show that Ninth Circuit firms reduce their engagement in EXM as a channel for MBE after the ruling, suggesting that firms are less likely to rely on EXM as AEM becomes less costly and can be sufficient to meet or beat earnings benchmarks.

This paper makes contributions to the literature. First, this study contributes to the literature on MBE by identifying firms’ litigation risk as an important determinant of firms’ MBE. While prior studies on MBE consider litigation risk as one of the factors when examining the association between their variable of interests and MBE (Barton & Simko, 2002; Q. Cheng & Warfield, 2005; Frankel et al., 2002; Matsumoto, 2002; McGregor, 2012), the reported association between MBE and litigation risk has largely been inconclusive. By using a quasi-natural experiment setting, this study extends the literature by documenting that the propensity of MBE increases as the threat of litigation dissipates. This paper has the potential to add insight into corporate governance literature because it documents that an important dimension of external corporate governance—the risk associated with shareholder class action lawsuits—influences firms’ reporting outcomes.

In addition, this study has implications for the literature on the role of litigation risk in different types of earnings management. Managers trade off the costs and benefits of AEM and REM (e.g., Barton & Simko, 2002; Cohen et al., 2008; Cohen & Zarowin, 2010; Gunny, 2010; Zang, 2012). Relatedly, Hopkins (2018) and Huang et al. (2020) show that litigation risk deters managers from AEM and REM, respectively, by using the same legal event adopted in this paper.8 While these studies are related to my earnings management analysis, this paper extends these studies by jointly focusing on both AEM and REM as a tool for MBE.9 The results of this study imply that investors and regulators need to scrutinize the possibilities of opportunistic earnings management as litigation risk changes.

The remainder of this paper proceeds as follows: Section 2 reviews the related literature and the legal background of the securities litigation regulation and develops my hypothesis. Section 3 describes data sources and outlines my research design. Section 4 and Section 5 describe empirical results and additional tests, and Section 6 concludes.

2. Background and Hypothesis Development

2.1. Legal Background of the Ninth Circuit Court Ruling

The Private Securities Litigation Reform Act of 1995 is crucial to the reform of federal securities laws governing securities class actions. When firms or managers breach their fiduciary duties, shareholder lawsuits play a role as a potential governance force (e.g., Appel, 2019; C. S. A. Cheng et al., 2010). However, individual shareholders might not properly utilize such a tool to protect the interests of all the shareholders. Before the 1995 Reform Act, shareholders could file a lawsuit against firms or managers with minimal evidence of fraud, including a large stock price drop based on the “fraud on the market” theory. Thus, some shareholders filed a lawsuit with dubious merit to target firms with “deep-pocket” (Johnson et al., 2001), imposing huge burdens on firms. Such problems caused the birth of the 1995 Reform Act introduced several hurdles to plaintiffs. The 1995 Reform Act required shareholders to provide facts that give rise to a strong inference of fraudulent intent of the defendants.

However, interpretations of the pleading standards are independent from U.S. circuit courts (Chu, 2017; Crane & Koch, 2018). Among various interpretations, the interpretation in 1999 by the Ninth Circuit Court of Appeals in the case of Silicon Graphics Inc. was the most surprising and stringent. On 2 July 1999, the Ninth Circuit Court ruling required plaintiffs to plead strong facts inferring that the defendants were “deliberately reckless” when they made the alleged misstatement or omitted material information in the statement. In other words, for firms headquartered in the Ninth Circuit, shareholders are not legally allowed to form a class without presenting evidence that managers behaved with deliberate, conscious recklessness. The 1999 ruling affected all class action lawsuits subsequently filed with the Ninth Circuit Court. In contrast, for firms located in other states, proving mere recklessness was sufficient to form a lawsuit. Such a change was important because evidence of a manager’s intention often only comes in the process of discovery, and the discovery process occurs after the class has been formed. Thus, shareholders of the Ninth Circuit firms must possess the evidence before they form a class.

Following the 1999 ruling, the number of securities class actions declined by more than 50 percent in the Ninth Circuit Court relative to the other circuit courts (Crane & Koch, 2018). Put differently, firms in the Ninth Circuit states could expect a lower probability of shareholder lawsuits after the 1999 ruling. The 1999 ruling was not anticipated, and many U.S. circuit courts faced challenges in interpreting the 1995 pleading standard (Pub. L. No. 104-67; Private Securities Litigation Reform Act in 1995). Hence, the Ninth Circuit Court ruling in 1999 is a valid legal event to examine the impact of a decrease in litigation risk on MBE.

2.2. Related Literature and Hypothesis Development

The incentives or benefits of MBE are well documented in the literature. Kasznik and McNichols (2002) document that MBE firms enjoy higher abnormal returns than firms missing benchmarks and that market premium for MBE is incremental to future expectation. The MBE premium is also incremental to the overall earnings surprise (Bartov et al., 2002). Graham et al. (2005) survey 312 executives in public firms and show that MBE helps managers to “build credibility with the capital market” and “maintain or increase stock price”. Also, missing expectations by even a small amount can result in a large stock price reduction (Skinner & Sloan, 2002). Given these benefits and costs, prior empirical evidence suggests that managers manage earnings to meet or beat various benchmarks (e.g., Degeorge et al., 1999; Myers et al., 2007). These studies investigate patterns of reported earnings around various earnings targets and document that management tries to avoid disclosing losses, a decrease in earnings from the previous year, and missing the analyst earnings forecasts.

Furthermore, other studies have examined various factors that constrain earnings management with a primary focus on corporate governance (e.g., Bushee, 1998; Chen et al., 2015; Q. Cheng et al., 2016; Lohwasser & Zhou, 2024). Among them, a threat of litigation also plays a role as a governance mechanism through shareholder monitoring (e.g., Shleifer & Vishny, 1997), and therefore threat of litigation affects both firms’ MBE behavior and their tools to achieve MBE (i.e., earnings management). However, the empirical evidence is mixed in prior studies. Kasznik (1999) finds that potential litigation cost motivates managers to manage discretionary accruals to increase reported earnings to meet their forecasts. Frankel et al. (2002) document a significantly positive impact of litigation risk on income-decreasing accruals choices (but no impact on income-increasing accruals), whereas Matsumoto (2002) and Q. Cheng and Warfield (2005) document a significantly positive impact of litigation risk on MBE. Barton and Simko (2002) and McGregor (2012) indicate no significant impact of litigation risk on MBE. Overall, prior studies do not show consistent results on the association or causal relationship between litigation risk and MBE.

This study examines two competing hypotheses regarding the effect of litigation risk on MBE. The motivation-based hypothesis explains the notion that litigation risk directly increases the motivation of MBE, suggesting that litigation risk positively affects MBE. Stock price decline can be evidence of misstatement or omission of negative material information and increase the probability of future lawsuits (e.g., Field et al., 2005; Skinner, 1994; Skinner, 1997). When firms miss their benchmarks (i.e., failing MBE), stock price reductions and shareholder lawsuits are more likely to occur. Thus, as managers try to mitigate litigation threats from missing earnings expectations, litigation risk increases managers’ MBE incentives. Following this logic, for Ninth Circuit firms after the 1999 ruling, a decrease in litigation risk reduces the potential shareholder litigation from failing MBE, expecting a reduction in the motivation of MBE. Therefore, litigation risk positively affects MBE under the motivation-based hypothesis.

However, the cost-based hypothesis explains the indirect channel that litigation risk increases the cost of MBE, suggesting that litigation risk negatively affects MBE. Previous studies document that managers manage earnings upward to meet or just-beat their earnings benchmarks, especially when pre-managed earnings miss the benchmarks by a small amount (e.g., Das & Zhang, 2003; Hansen, 2010; Phillips et al., 2003; Roychowdhury, 2006). Litigation risk is one of the powerful governance mechanisms to decrease managers’ engagement in earnings management because of potential lawsuits from misstatement or omission of material information in firms’ disclosures. Hence, litigation risk decreases the use of earnings management to meet or beat earnings expectations, and the threat of lawsuits from engaging in earnings management can be a cost of MBE. Following the same reasoning, for firms in Ninth Circuit states after the ruling, a decline in litigation risk increases the degree of earnings management for MBE (i.e., increases the propensity of MBE). Therefore, litigation risk negatively affects MBE under the cost-based story.

As the motivation-based (cost-based) story predicts the positive (negative) causal relationship between litigation risk and MBE, this study’s hypothesis proceeds as follows in null form:

Hypothesis 1.

Litigation risk does not affect benchmark-beating behavior.

3. Research Design and Methods

3.1. Sample

The study starts the sample selection process by obtaining data on all public U.S. firms from the Compustat-CRSP-I/B/E/S universe with information on historical locations of firms’ headquarters over the sample period (1995–2003).10 Empirical test compares the post-ruling period, 18 quarters after the ruling (from the third quarter, 1999 to the fourth quarter, 2003), with the pre-ruling period, 18 quarters before the ruling (from the first quarter, 1995 to the second quarter, 1999). I obtain firms’ financial information from the Compustat quarterly file, stock price and return information from the CRSP daily file, and analysts’ forecasts and firms’ actual earnings information from the I/B/E/S unadjusted file, respectively.

The sample excludes firms in financial (i.e., sic 6000–6999) and utilities (i.e., sic 4900–4949) industries and firms with stock prices less than $1.11 Also, the study requires the data availability of Compustat, CRSP, and I/B/E/S, which are necessary to calculate control variables. The final sample consists of 20,525 firm-quarter observations. All continuous variables are winsorized at the 1st and 99th percentiles.12

3.2. Empirical Model

The study investigates whether a decrease in litigation risk affects managers’ opportunistic incentives for MBE. It classifies firms as a treated group (i.e., Ninth Circuit firms) if firms’ headquarters are in one of the states controlled by the Ninth Circuit Court, and the remaining firms are considered a control group. The research design employs a differences-in-differences design to compare the changes in the propensity of MBE after the 1999 ruling for Ninth Circuit firms with the corresponding changes for other firms. Specifically, the study estimates the following regression model:

MBE_i,t = δ₁ Nine × Post_i,t + δ₂ Controls_i,t + δ _year-quarter + δ _firm + ε _i,t

(1)

The indicator variable, Nine, equals one if the firm’s headquarters are in one of the Ninth Circuit states and zero otherwise. The indicator variable, Post, takes the value of one from the third quarter of 1999 to the fourth quarter of 2003 and zero from the first quarter of 1995 to the second quarter of 1999. The main variable of interest is the interaction term, Nine × Post. The dependent variable, MBE, is an indicator variable that takes the value of one if actual earnings per share are equal to or larger than the median value of analyst EPS forecasts, and zero if actual earnings are lower than the median value of analyst EPS forecasts. This study also defines just-MBE (JMBE), using narrower intervals surrounding earnings forecasts by ±2 cents and ±5 cents.13 The regression model includes firm and year-quarter fixed effects and excludes Nine and Post in the model because these two variables are absorbed in the firm and year-quarter fixed effects, respectively. Standard errors are clustered by the state of location, and Fama-French 48 industry classifications are used. The motivation-based (cost-based) story expects δ₁ to be negative (positive).

The regression model considers several control variables that are expected to be associated with MBE. The model includes the firm’s size (Size) as prior research suggests that larger firms have less optimistic biases in analysts’ forecasts, which makes firms meet or beat analyst forecasts easily (e.g., Matsumoto, 2002). The model also includes leverage (Leverage) to proxy for the firm’s closeness to debt default, which might serve as an incentive or a constraint to meet or beat benchmarks (McAnally et al., 2008). Matsumoto (2002) shows that institutional ownership (IO) and implicit claims with stakeholders (R&D and Labor) motivate managers to MBE.14 Firm growth may affect investors’ reaction to earnings information (McAnally et al., 2008; Skinner & Sloan, 2002), and it is difficult for managers to manage earnings when there are large numbers of stocks because a larger amount of earnings should be managed per one cent of EPS (Barton & Simko, 2002). Therefore, the book-to-market ratio (BM), sales growth (Sales Growth), and the number of shares outstanding (#Shares) are included. The number of analysts following (Coverage), the variation in analysts’ earnings forecasts (Dispersion), and earnings volatility (Vol_ROA) are included to control for difficulties in the forecasting environment (Matsumoto, 2002). Management earnings forecasts (MF) impact analyst forecasts, which may serve as an incentive or a constraint to meet earnings expectations (Richardson et al., 2004). Also, the model includes the fourth quarter dummy as a control variable because the fourth quarter might affect MBE and/or earnings management. Moreover, the firm’s profitability (ROA) and loss indicators (Loss) are included because MBE increases with firms’ performance (Barton & Simko, 2002). Finally, the indicator variable for Regulation FD (RegFD) is included since management earnings forecasts may not capture the effect of Regulation FD.

4. Empirical Results

4.1. Descriptive Statistics and Univariate Analysis

Table 1. Summary statistics.

Variables	N	Mean	Std. Dev	Min	P25	P50	P75	Max
Nine	20,525	0.39	0.49	0.00	0.00	0.00	1.00	1.00
Post	20,525	0.50	0.50	0.00	0.00	0.00	1.00	1.00
MBE	20,525	0.74	0.44	0.00	0.00	1.00	1.00	1.00
MBE_SRW	20,525	0.55	0.50	0.00	0.00	1.00	1.00	1.00
Size	20,525	5.95	1.71	2.87	4.63	5.73	7.05	10.52
Leverage	20,525	0.16	0.18	0.00	0.00	0.08	0.27	0.79
BM	20,525	0.41	0.32	−0.09	0.19	0.33	0.53	1.78
Sales Growth	20,525	0.40	1.02	−0.68	0.00	0.15	0.41	7.10
ROA	20,525	0.00	0.06	−0.27	−0.01	0.01	0.03	0.08
Coverage	20,525	1.84	0.59	1.10	1.39	1.79	2.20	3.30
Dispersion	20,525	0.03	0.04	0.00	0.01	0.01	0.03	0.26
R&D	20,525	0.03	0.03	0.00	0.01	0.02	0.04	0.16
Labor	20,525	0.78	0.19	0.15	0.70	0.84	0.92	0.99
#Shares	20,525	3.58	1.19	1.47	2.74	3.39	4.22	7.34
Vol_ROA	20,525	0.04	0.07	0.00	0.01	0.02	0.04	0.50
IO	20,525	0.55	0.25	0.04	0.35	0.57	0.74	1.00
MF	20,525	0.16	0.37	0.00	0.00	0.00	0.00	1.00
Loss	20,525	0.32	0.47	0.00	0.00	0.00	1.00	1.00
FourthQ	20,525	0.26	0.44	0.00	0.00	0.00	1.00	1.00
RegFD	20,525	0.36	0.48	0.00	0.00	0.00	1.00	1.00

Notes: Table 1 reports descriptive statistics for variables in the analysis. The sample includes 20,525 firm-quarter observations with the required information to compute the main and control variables. The sample period is from 1995 to 2003. I exclude firms in financial and utility industries (sic 4000–4949 and sic 6000–6999), small price stocks (share price less than one dollar), and firms with missing headquarters locations. All variables in this table are defined in Appendix A.

reports the summary statistics for the main and control variables used in empirical analyses. The full sample has 20,525 observations, and 39% of the sample are firm-quarters in Ninth Circuit states. About 74% of firm-quarters in the sample meet or beat the median value of analyst earnings forecasts. The sample of MBE firms with AEM and REM has 10,492 and 12,000 observations, respectively.

Table 2. Univariate tests.

	Pre-1999 Ruling				Post-1999 Ruling				Diff-in-Diff
Variable	Ninth Circuit	Others	Difference	p-Value	Ninth Circuit	Others	Difference	p-Value	Diff-in-Diff	p-Value
MBE	0.687	0.715	−0.028 ***	0.002	0.796	0.774	0.022 **	0.011	0.050 ***	0.000
(N = 20,525)
By 2 cents
JMBE	0.774	0.791	−0.017	0.145	0.845	0.810	0.035 ***	0.001	0.051 ***	0.001
(N = 11,051)
By 5 cents
JMBE	0.758	0.769	−0.011	0.253	0.833	0.805	0.028 ***	0.003	0.039 ***	0.004
(N = 15,791)

Notes: Table 2 reports the results of univariate differences-in-differences tests. It shows the comparison between pre- and post-1999 rulings for firms located in Ninth Circuit states and those in other states. The sample period is from 1995 to 2003. I exclude firms in financial and utility industries (sic 4000–4949 and sic 6000–6999), small price stocks (share price less than one dollar), and firms with missing headquarters locations. ** and *** denote two-tailed significance level at 5%, and 1%, respectively.

presents the univariate results of differences-in-differences tests and provides preliminary insights into the changes in MBE for Ninth Circuit firms after the 1999 ruling. The table reports the propensity of MBE for firms located in Ninth Circuit states and for those in other states, as well as the changes in MBE separately for pre- and post-1999 rulings. Subsamples based on narrower intervals of JMBE surrounding earnings forecasts by ±2 cents and ±5 cents have 11,051 and 15,791 observations, respectively. The decreases in subsamples’ observations are due to the narrow intervals of MBE (i.e., ±2 cents (±5 cents). The sample excludes firm-quarter observations of firms that beat or miss the target by more than 2 cents (5 cents). The results show that Ninth Circuit firms exhibit increases in MBE subsequent to the 1999 ruling. Such increases are statistically significant (p < 0.01). The table also reports the results of the main differences-in-differences tests, which compare the changes in MBE for Ninth Circuit firms relative to those for non-Ninth-Circuit firms between pre- and post-1999 ruling. The propensity to MBE significantly increases (p < 0.01) with all intervals. Taken together, the univariate tests show preliminary evidence consistent with a decrease in litigation risk causing a significant increase in MBE, supporting the cost-based hypothesis.

4.2. Multivariate Analysis

4.2.1. Meeting-or-Beating Earnings Expectations (MBE)

Table 3. The changes in meeting-or-beating earnings expectation (MBE).

	(1)	(2)	(3)
		By 2 Cents	By 5 Cents
Variable	MBE	JMBE	JMBE
Nine × Post	0.398 ***	0.435 ***	0.321 ***
	(4.92)	(4.06)	(2.59)
Size	−0.104	−0.409 ***	−0.184 *
	(−1.01)	(−3.55)	(−1.74)
Leverage	−0.082	−0.501	−0.429
	(−0.30)	(−1.10)	(−1.38)
BM	−0.429 **	−0.406	−0.636 ***
	(−2.47)	(−1.51)	(−3.22)
Sales Growth	0.094 **	0.097	0.089
	(2.01)	(1.25)	(1.33)
ROA	6.129 ***	3.109 **	4.570 ***
	(6.70)	(2.34)	(5.13)
Coverage	−0.018	0.173 *	0.081
	(−0.24)	(1.65)	(0.93)
Dispersion	−11.973 ***	−13.173 ***	−13.999 ***
	(−12.89)	(−7.13)	(−8.28)
R&D	4.579 ***	1.617	2.853 **
	(3.98)	(0.92)	(2.05)
Labor	2.084 ***	2.551 ***	2.463 ***
	(3.51)	(3.77)	(4.32)
#Shares	−0.338 ***	−0.007	−0.326 **
	(−3.06)	(−0.06)	(−2.36)
Vol_ROA	0.271	−1.245	−0.324
	(0.42)	(−0.96)	(−0.42)
IO	−0.314 *	0.362	0.351 *
	(−1.80)	(1.05)	(1.71)
MF	0.035	0.043	0.092
	(0.56)	(0.52)	(1.52)
Loss	−0.645 ***	−0.409 ***	−0.470 ***
	(−7.62)	(−2.79)	(−4.39)
FourthQ	0.105	0.070	0.096
	(1.58)	(0.61)	(1.20)
RegFD	0.221	−0.278	0.203
	(0.96)	(−0.88)	(0.82)
Year-Quarter FE	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes
N	20,525	11,051	15,791
Pseudo R-squared	0.175	0.140	0.154

Notes: Table 3 reports the results of logistic regression that examines the changes in opportunistic MBE following the 1999 Ninth Circuit Court ruling. Z-statistics are presented beneath the coefficients within parentheses. *, ** and *** denote two-tailed significance level at 10%, 5%, and 1%, respectively. Standard errors are clustered at the state of location level. A total of 18,087, 8269, and 12,985 observations are used for the regression (i.e., 2438, 2782, and 2806 observations are omitted) because of no within-group variance in columns (1), (2), and (3), respectively, which include firm and year-quarter fixed effects.

provides the logistic regression results of the changes in just-MBE for Ninth Circuit firms after the ruling (i.e., estimating Equation (1)). All coefficients on the interaction term, Nine × Post, are positive and significant for all three specifications (p < 0.01). The results indicate that the odds of MBE versus missing expectations for Ninth Circuit firms after the ruling are 49% higher than the odds for other firms based on the coefficient on MBE in column (1). The results suggest statistically significant increases in just-MBE and MBE, in general, following the 1999 ruling for firms located in Ninth Circuit states relative to those in other states, supporting the cost-based hypothesis that litigation risk reduces the propensity of MBE by increasing the cost of MBE.15 In other words, firms facing a decrease in litigation risk are more likely to report small positive earnings surprises than small negative surprises, suggesting that a decrease in litigation risk gives managers an edge in MBE at the margin.16

4.2.2. Alternative Earnings Benchmarks

The study examines the role of litigation risk in MBE by using analyst earnings forecasts as a proxy for earnings expectations. Even though analysts provide benchmarks widely used to evaluate earnings, prior studies also suggest that other earnings benchmarks, such as earnings for the same quarter of the prior year and zero earnings, may be important to firms (Brown & Caylor, 2005; Degeorge et al., 1999). Therefore, as a sensitivity check, this study investigates whether managers of firms facing a decrease in litigation risk are more likely to meet or beat alternative earnings benchmarks. I reestimate Equation (1) but redefine the MBE indicator based on an alternative earnings benchmark. MBE_SRW takes the value of one if EPS is equal to or larger than EPS for the same quarter of the prior year, and zero if EPS is smaller than EPS for the same quarter of the prior year.

Table 4. The changes in meeting-or-beating earnings expectation (MBE)—alternative benchmark.

	(1)	(2)	(3)
		By 2 Cents	By 5 Cents
Variable	MBE_SRW	JMBE_SRW	JMBE_SRW
Nine × Post	0.541 ***	−0.232	0.322 **
	(4.36)	(−0.55)	(2.13)
Size	−0.412 ***	−0.153	0.019
	(−2.72)	(−0.18)	(0.09)
Leverage	0.883 **	−0.662	−0.327
	(2.27)	(−0.44)	(−0.47)
BM	−0.528 ***	0.329	−0.445
	(−4.66)	(0.44)	(−1.27)
Sales Growth	0.396 ***	1.268 ***	0.513 ***
	(9.45)	(4.44)	(4.18)
ROA	30.910 ***	15.240	14.251 ***
	(9.85)	(1.47)	(2.66)
Coverage	−0.492 ***	−0.324	−0.305 *
	(−4.39)	(−0.65)	(−1.73)
Dispersion	−5.609 ***	0.272	−10.876 ***
	(−7.50)	(0.07)	(−3.81)
R&D	−5.628 ***	−35.987 **	−15.220 ***
	(−3.33)	(−2.10)	(−2.68)
Labor	−0.542	−0.745	−0.255
	(−1.22)	(−0.46)	(−0.24)
#Shares	−0.701 ***	−0.413	−0.559 *
	(−4.37)	(−0.53)	(−1.88)
Vol_ROA	4.445 ***	−2.586	−2.447
	(7.17)	(−0.59)	(−1.35)
IO	−0.032	0.187	−0.097
	(−0.20)	(0.25)	(−0.19)
MF	−0.328 ***	−0.079	−0.166
	(−6.37)	(−0.23)	(−1.14)
Loss	−1.066 ***	−0.882	−0.820 ***
	(−8.29)	(−1.31)	(−3.98)
FourthQ	0.019	0.294	0.352 *
	(0.74)	(0.82)	(1.85)
RegFD	1.663 ***	−0.112	0.122
	(5.25)	(−0.11)	(0.30)
Year-Quarter FE	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes
N	20,525	2,115	4,770
Pseudo R-squared	0.266	0.145	0.141

Notes: Table 4 reports the results of a logistic regression that examines the changes in opportunistic MBE following the 1999 Ninth Circuit Court ruling. MBE_SRW and JMBE_SRW are based on the earnings for the same quarter of the prior year. Z-statistics are presented beneath the coefficients within parentheses. *, ** and *** denote two-tailed significance level at 10%, 5%, and 1%, respectively. Standard errors are clustered at the state of location level. A total of 19,076, 1125, and 3408 observations are used for the regression (i.e., 1449, 990, and 1362 observations are omitted) because of no within-group variance in columns (1), (2), and (3), respectively, which include firm and year-quarter fixed effects.

reports the results. The coefficients on the interaction term, Nine × Post, are positive and significant (except for column (2)), suggesting that firms in Ninth Circuit states are more likely to meet or beat earnings expectations based on earnings for the same quarter of the previous year after the ruling. The results indicate that the odds of MBE versus missing expectations for Ninth Circuit firms after the ruling are 72% higher than the odds for other firms based on the coefficient on MBE_SRW in column (1). The result provides further evidence that the main results for the propensity of MBE are not primarily driven by the choices of earnings benchmark or analyst and manager interaction. Firms facing a decrease in litigation risk may focus their attention on meeting-or-beating not only analyst expectations but also alternative benchmarks (i.e., earnings for the same quarter of the prior year).17

5. Additional Analysis

5.1. Opportunistic Earning Management for Meeting-or-Beating Earnings Expectation (MBE)

To further investigate the method to achieve MBE, this study turns to the impact of litigation risk on two earnings management methods, AEM and REM. While the motivation hypothesis does not directly explain the relationship between litigation risk and earnings management, the cost-based hypothesis predicts that a threat of litigation decreases earnings management. Therefore, the study expects that Ninth Circuit firms are more likely to achieve MBE through managing earnings (AEM and REM) after the ruling relative to other firms, which is consistent with prior studies (Hopkins, 2018; Huang et al., 2020).

To test the impact of litigation risk on opportunistic AEM and REM as a method for MBE, the study estimates the following logistic regressions:

OppMBE_AEM_i,t = ϕ₁ Nine × Post_i,t + ϕ₂ Controls_i,t + ϕ _year-quarter + ϕ _firm + ε _i,t

(2)

OppMBE_REM_i,t = τ₁ Nine × Post_i,t + τ₂ Controls_i,t + τ _year-quarter + τ _firm + ε _i,t

(3)

The dependent variable, OppMBE_AEM, takes the value of one if the firm meets or beats analyst forecasts, but would have missed expectations if abnormal discretionary accruals were removed from reported earnings, and zero otherwise. This study follows the modified Jones model (Dechow et al., 1995) to estimate abnormal discretionary accruals. OppMBE_AEM is a proxy for suspicious firms that opportunistically engage in AEM to meet or beat earnings expectations. Similarly, OppMBE_REM is a proxy for suspicious firms that opportunistically engage in REM to meet or beat earnings expectations. Specifically, OppMBE_REM takes the value of one if the firm meets or beats analyst forecasts, but would have missed expectations if abnormal cuts of discretionary expenses were removed from reported earnings, and zero otherwise. The study closely follows Roychowdhury (2006) to estimate discretionary expenses.18 All control variables in Equation (1) are included as well.

Table 5. The changes in opportunistic accrual and real earnings management.

	(1)	(2)
Variable	OppMBE_AEM	OppMBE_REM
Nine × Post	0.295 **	−0.105
	(2.52)	(−0.24)
Size	−0.490 ***	−0.585 *
	(−4.70)	(−1.79)
Leverage	1.946 ***	−1.139
	(6.44)	(−1.13)
BM	0.125	−0.066
	(0.83)	(−0.17)
Sales Growth	0.072	−0.429 *
	(1.14)	(−1.70)
ROA	5.107 ***	−6.621 ***
	(3.06)	(−2.72)
Coverage	−0.013	0.547 ***
	(−0.17)	(3.25)
Dispersion	−1.853	−8.496 **
	(−1.25)	(−2.00)
R&D	−7.018 ***	−93.974 ***
	(−3.08)	(−4.39)
Labor	−3.214 ***	0.537
	(−5.68)	(0.42)
#Shares	0.012	1.145 ***
	(0.07)	(3.16)
Vol_ROA	1.811 **	−5.061 ***
	(2.45)	(−3.07)
IO	0.513 ***	−0.455
	(3.03)	(−0.88)
MF	0.095	−0.104
	(1.13)	(−0.59)
Loss	−0.214	−0.644 ***
	(−1.52)	(−3.33)
FourthQ	0.382 ***	−0.064
	(4.31)	(−0.28)
RegFD	0.915 ***	−1.338 *
	(3.39)	(−1.95)
Year-Quarter FE	Yes	Yes
Firm FE	Yes	Yes
N	10,492	12,000
Pseudo R-squared	0.143	0.289

Notes: Table 5 reports the results of a logistic regression that examines the changes in accrual earnings management (AEM) and real earnings management (REM) following the 1999 Ninth Circuit Court ruling. Z-statistics are presented beneath the coefficients within parentheses. *, ** and *** denote two-tailed significance level at 10%, 5%, and 1%, respectively. Standard errors are clustered at the state of the location level. In total, 9372 and 3260 observations are used for the regression (i.e., 1120 and 8740 observations are omitted) because of no within-group variance in columns (1) and (2), respectively.

shows the logistic regression results of the changes in opportunistic AEM and REM for firms located in Ninth Circuit states after the ruling relative to firms in other states. The table presents the measures of opportunistic accrual and real earnings management, OppMBE_AEM and OppMBE_REM, in columns (1) and (2), respectively. The coefficient on the variable of interest, Nine × Post, is positive and significant in column (1) (p < 0.05), suggesting that the probability of meeting-or-beating analyst forecasts but would have missed the forecasts without accrual earnings management, increases as litigation risk reduces. In other words, firms are more likely to manage accruals as a tool for MBE after litigation risk decreases. The coefficient on the interaction term Nine × Post is not significant in column (2), suggesting that, as litigation risk decreases, firms are not likely to engage more in REM as a tool for MBE.19

5.2. Market Reaction to Meeting-or-Beating Earnings Expectation (MBE)

A large body of literature documents that the incentive to meet or beat expectations is associated with equity market benefits (e.g., Kasznik & McNichols, 2002; Skinner & Sloan, 2002). MBE increases contemporaneous stock returns, and a string of positive earnings surprises also increases the valuation premium for such firms (Bartov et al., 2002; Kasznik & McNichols, 2002). Prior research has also examined the equity market reaction to firms’ use of earnings management. When earnings may have been increased through earnings management, investors discount premiums paid for MBE accordingly (e.g., Baber et al., 2006). Under a motivation-based (cost-based) story, as litigation risk decreases, firms are less (more) likely to manage earnings to achieve MBE. However, managing accruals or cutting discretionary expenses as a tool for MBE induces transitory components to reported earnings, causing an increase in the likelihood that future earnings and/or performance will reverse. Therefore, if equity investors rationally expect the lower (higher) likelihood of earnings management, at least partially, for firms with a decrease in litigation risk, then they will increase (discount) the MBE premium for such firms.

To investigate the changes in MBE premium (i.e., investors’ reaction) for Ninth Circuit firms after the ruling, the study performs a market reaction test surrounding the earnings announcement date. Specifically, the empirical analysis estimates the following OLS regression:

CAR(−1, +1)_i,t = η₁ Nine × Post × JMBE_i,t + η₂ Nine × Post_i,t + η₃ MBE_i,t +
η₄EarnSURP_i,t + η₅ Controls_i,t + η_year-quarter + η_firm + ε_i,t

(4)

CAR(−1, +1) is the three-day market-adjusted cumulative abnormal return surrounding the earnings announcement date. EarnSURP is the earnings surprise (actual earnings minus the median value of I/B/E/S one-quarter-ahead street earnings forecasts). Controls include the same control variables in previous tests as well as PreRet and Beta. PreRet is the cumulative returns for the prior six months before the earnings announcement dates, and Beta is estimated for the prior 36 months. Consistent with the prior studies, I expect η3 and η4 to be significantly positive because the MBE premium after controlling for earnings surprises exists (e.g., Bartov et al., 2002). The variable of interest is the interaction term, Nine × Post × MBE. If investors rationally expect lower litigation risk and a higher degree of opportunistic earnings management for Ninth Circuit firms after the ruling than other firms, the market reaction to MBE premium is less positive for Ninth Circuit firms after the ruling.

Table 6. The changes in market reaction to meeting-or-beating earnings expectations (MBE).

	(1)	(2)	(3)
		By 2 Cents	By 5 Cents
Variable	CAR (−1,+1)	CAR (−1,+1)	CAR (−1,+1)
Nine × Post × MBE	−0.006 **
	(−2.31)
MBE	0.028 ***
	(10.82)
Nine × Post × JMBE		−0.011 ***	−0.010 ***
		(−3.10)	(−2.83)
JMBE		0.011 **	0.018 ***
		(2.13)	(4.94)
EarnSURP	2.355 ***	11.941 ***	7.395 ***
	(8.65)	(4.55)	(6.60)
Nine × Post	0.009 **	0.016 ***	0.017 ***
	(2.43)	(2.87)	(3.98)
Size	−0.012 ***	−0.009 **	−0.010 ***
	(−4.54)	(−2.68)	(−3.28)
Leverage	0.015 **	−0.006	0.010
	(2.38)	(−0.43)	(0.99)
BM	0.038 ***	0.041 ***	0.035 ***
	(8.95)	(6.08)	(6.90)
ROA	0.035	−0.013	−0.031
	(0.77)	(−0.24)	(−0.55)
Coverage	−0.003	−0.005	−0.004
	(−0.99)	(−1.26)	(−1.47)
Dispersion	0.053 ***	0.048	0.014
	(2.80)	(0.89)	(0.47)
R&D	−0.049	−0.098	−0.157 **
	(−0.79)	(−1.25)	(−2.25)
Labor	0.000	0.004	−0.012
	(0.02)	(0.15)	(−0.56)
#Shares	−0.014 ***	−0.008 **	−0.008 **
	(−3.64)	(−2.25)	(−2.27)
Vol_ROA	−0.008	0.017	−0.001
	(−0.16)	(0.42)	(−0.04)
IO	−0.005	−0.006	−0.006
	(−0.84)	(−0.45)	(−0.70)
MF	−0.002	−0.002	−0.000
	(−0.78)	(−0.57)	(−0.10)
Loss	−0.006	−0.004	−0.004
	(−1.37)	(−0.44)	(−0.70)
FourthQ	0.004	0.004	0.003
	(1.64)	(1.35)	(1.14)
PreRet	0.001	−0.014 ***	−0.003
	(0.46)	(−4.06)	(−1.13)
Beta	−0.004 **	−0.007 ***	−0.005 *
	(−2.38)	(−3.13)	(−1.97)
Year-Quarter FE	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes
N	14,131	7,561	10,851
Adj. R-squared	0.081	0.069	0.071

Notes: Table 6 reports the results of an OLS regression that examines the market reaction to the MBE premium following the 1999 Ninth Circuit Court ruling. T-statistics are presented beneath the coefficients within parentheses. *, ** and *** denote two-tailed significance level at 10%, 5% and 1%, respectively. Standard errors are clustered at the state of location level.

presents the results. In line with prior literature, the coefficients on EarnSURP and MBE are all positive and significant. More importantly, all coefficients on the interaction terms, Nine × Post × MBE, are negative and significant at the 1% level (except for 5% in Column (1)), suggesting that the MBE premium decreases for Ninth Circuit firms after the ruling relative to other firms. The results indicate that investors react to MBE premium (after controlling for forecast errors) less positively for firms with a relatively lower level of litigation risk than other firms with higher litigation risk. This study interprets the results as investors being suspicious about the quality of disclosed earnings for benchmark beaters who experience a decrease in litigation risk. Overall, the results indicate that a decrease in litigation risk results leads to a decline in MBE premium.

5.3. Expectation Management

Managers can engage not only in earnings management, AEM, and REM, but also in EXM to meet or beat earnings expectations. Kasznik and Lev (1995) and Skinner (1997) show that firms are more likely to warn investors about forthcoming negative earnings news. Richardson et al. (2004) find that firms enjoy financial benefits by managing earnings expectations downward to a beatable level. Such behavior is consistent with the notion that EXM is another tool for MBE.20

To examine the impact of litigation risk on another tool for MBE (i.e., EXM), the study first estimates the EXM proxies (EXM and OppMBE_EXM). Following Bartov et al. (2002), Brown and Pinello (2007), and Das et al. (2011), this study measures the degree of EXM (EXM) as the first analyst earnings forecast minus the last analyst earnings forecast and requires that the first analyst forecast be issued at least one day after the previous quarter’s earnings announcement. Also, this analysis requires analyst forecasts to be issued at least two times between the previous and current quarter’s earnings announcement to measure EXP proxy. For the measure of opportunistic expectation management, I identify suspect firms that manage expectations to meet or beat analyst forecasts. OppMBE_EXM takes the value of one if the firm meets or beats the latest analyst forecasts, but misses the first forecasts, and zero otherwise. Then, the study reestimates Equation (2) with OppMBE_EXM as a dependent variable.

Table 7. The changes in expectation management.

	OppMBE_EXM
Variable	Coefficients	z-Statistics
Nine × Post	−0.426 **	(−2.40)
Size	−0.682 ***	(−4.15)
Leverage	0.458	(0.95)
BM	1.630 ***	(6.50)
Sales Growth	−0.270 ***	(−7.36)
ROA	−8.417 ***	(−4.41)
Coverage	1.017 ***	(7.61)
Dispersion	11.224 ***	(10.85)
R&D	−4.368	(−1.24)
Labor	−0.834	(−1.06)
#Shares	1.056 ***	(4.03)
Vol_ROA	−1.902 *	(−1.88)
IO	0.069	(0.32)
MF	0.873 ***	(8.51)
Loss	0.163	(1.39)
FourthQ	0.091	(1.11)
RegFD	−1.352 ***	(−3.04)
Year-Quarter FE	Yes
Firm FE	Yes
N	7484
Pseudo R-squared	0.209

Notes: Table 7 reports the results of logistic regression that examines the changes in expectation management following the 1999 Ninth Circuit Court ruling. Z-statistics are presented on the right side of the coefficients within parentheses. *, ** and *** denote two-tailed significance level at 10%, 5% and 1%, respectively. Standard errors are clustered at the state of location level. A total of 6699 observations are used for the regression (i.e., 785 observations are omitted) because of no within-group variance.

reports the results. The coefficient on the interaction term, Nine × Post, is negative and significant (p < 0.05), indicating that Ninth Circuit firms are less likely to manage expectations to meet or beat analyst forecasts after the ruling. It implies that, as litigation risk decreases, managers reduce their engagement in EXM to meet or beat earnings expectations. One possible interpretation is that AEM (not REM) is sufficient to meet or beat earnings benchmark after litigation risk decreases, so that managers are less likely to rely on EXM. Such an explanation is consistent with the results of previous AEM tests in this study; managers’ engagements in AEM as a channel for MBE increase after the ruling because the cost of AEM decreases. Therefore, the tests on earnings and expectation management show that, to meet or beat analyst earnings forecasts, managers are more (less) likely to use AEM (EXP) as litigation risk declines, but do not significantly change their engagement in REM.

6. Conclusions

Shareholders can use various courses of action to deter managers or directors from breaching their fiduciary duty and engaging in actions that transfer wealth away from the owners. Shareholders can form a class and file a lawsuit against firms or managers, and such litigation brings serious reputational penalties and negative career implications to managers. This mechanism plays a role as corporate governance for firms and management, and prior studies have examined the disciplining impact of litigation risk on managers’ voluntary disclosures and earnings management (Hopkins, 2018; Houston et al., 2019; Huang et al., 2020; Skinner, 1994). Extending the literature, this paper investigates the effect of litigation risk on the propensity of firms to meet or just-beat earnings expectations by using mandatory earnings disclosures.

To establish a causal relation, this study conducts differences-in-differences tests based on the unexpected legal event that decreased litigation risk for firms headquartered in the Ninth Circuit. The study has shown that firms experiencing a decrease in litigation risk are more likely to meet or just beat their earnings expectations, suggesting that a threat of litigation negatively affects just-MBE. In other words, firms facing a decrease in litigation risk tend to report small positive earnings surprises rather than small negative surprises, indicating that a decrease in litigation risk gives managers an edge in MBE at the margin. Further, the results show that managers are more (less) likely to manage discretionary accruals (analyst expectations), but not discretionary expenses, to meet or beat earnings benchmarks when the probability of potential litigation reduces. Finally, this study finds that investors’ reactions are less positive to the MBE premium for firms with a decline in litigation risk. Overall, the study contributes to the literature by documenting that a threat of litigation is a significant factor affecting managers’ opportunistic benchmark beating behavior through earnings management.