Earnings Management and Analyst Following: A Simultaneous Equations Analysis

• Date: 01 June 2014
• DOI: http://doi.org/10.1111/fima.12043
• Author: Yongtao Hong,Fariz Huseynov,Wei Zhang
• Published date: 01 June 2014

Earnings Management and Analyst

Following: A Simultaneous Equations

Analysis

Yongtao Hong, Fariz Huseynov, and Wei Zhang∗

Weuse a simultaneous equations system to examine the relationship between earnings management

and analyst following. We find that analysts’ decisions to follow firms and managerial decisions

to manage earnings are jointly determined. Firms with lower levels of accrual-based earnings

management offera better information environment to attract analyst following.Analyst following,

in turn, has important monitoring effects on managerial behavior and results in lower levels of

both accrual-based and realearnings management. The information intermediary effect on analyst

following is much weaker for expected “suspect firms” that manage their earnings continuously.

There is a considerable body of literature addressing how corporate disclosure policy af-

fects analyst behavior and vice versa (Bhushan, 1989; Fishman and Hagerty, 1989; Lang and

Lundholm, 1996; Bushman, Ptotroski, and Smith, 2005; Yu, 2008; Charoenrook and Lewis,

2009). Information about firms, processed or acquired by f inancial analysts,is a critical deter mi-

nant of portfolio allocation in capital markets. Corporate managers, however, have considerable

discretion regarding the informativeness of corporate disclosure. One of the ways in whichcor po-

rate managers manage the informativeness of corporate disclosure is earnings management. Prior

studies confirm that corporate managers may engage in accrual-based ear nings management

(Jones, 1991; Dechow, Solan, and Sweeny, 1995; Dechow and Dichev, 2002) and real activi-

ties manipulation (Roychowdhury, 2006; Cohen, Dey, and Lys, 2008; Zang, 2012) motivated by

various incentives.

In this paper, we analyzethe relationship between earnings management and analyst following.

Specifically, we examine whether firms vary the level of disclosure by managing earnings to

attract analysts and whether the magnitude and type of earnings management (real vs. accrual

based) affects the relationship between analyst following and disclosure. We also explore how

the relationship between analyst followingand disclosure changes if firms manage their earnings

continuously as opposed to only when they run into trouble. Prior studies have not typically

analyzed the joint determination of effects between analyst coverage and earnings management

by successfully addressing the potential endogeneity between analysts’ decisions and manage-

rial behavior. In our paper, we emphasize this interactive relationship. Our primary hypothesis

stipulates that analysts’ decisions to follow firms and managerial decisions to manage earnings

are jointly determined as managers set the level of disclosure to attract analysts, while analysts

We are grateful to Raghu Rau (Editor) and an anonymous referee for their excellent comments and suggestions. We

acknowledge Thomson Financials for providing analyst coverage data through I/B/E/S as part of a broad academic

programto encourage earnings expectation research. Wealso thank Heidi Mann for excellent copy editing.

∗Yongtao Hong is an Assistant Professor of Accounting in the College of Business at North Dakota State University

in Fargo, ND. Fariz Huseynov is an Assistant Professor of Finance in the College of Business at North Dakota State

University in Fargo, ND. Wei Zhang is an Associate Professor of Finance in the College of Business at North Dakota

State University, in Fargo, ND.

Financial Management •Summer 2014 •pages 355 - 390

356 Financial Management rSummer 2014

choose firm coverage based on the costs and benefits of operating in an existing information

environment impacted by the level of disclosure.

The previous literature concerning disclosure and analyst following suggests that the equi-

librium number of analysts is determined by the interaction between the aggregate supply and

demand curves for analyst services (Bhushan, 1989; Fisherman and Hagerty, 1989; Khanna,

Slezak, and Bradley, 1994; Lang and Lundholm, 1996). Prior literature regarding analyst behav-

ior (Stickel, 1992; Stickel, 1995; Mikhail, Walther, and Willis, 1999; Hong and Kubik, 2003)

suggests that the reputation associated with forecast accuracy is of consequence to analysts and

that accuracy is rewarded. If lower levels of earnings management can be interpreted as greater

levels of firm-provided disclosure, thus lowering the cost for analysts(e.g., the costs of receiving

information and the reputation cost associated with inaccurate forecast), then the analyst supply

curve will shift to the right, resulting in a greater supply of analyst services.

Lang and Lundholm (1996) argue that the impact of firm-provided disclosure on the demand

of analyst services depends upon the role that analysts play in the capital market. If analysts

primarily serve as information intermediaries through which information is transmitted from the

firm to the market, then more earnings management (i.e., less disclosure) will result in analysts

having less valuable reports to sell resulting in less demand for analyst services. If analysts

generally serve as information producers who compete with firm-provided disclosure, then more

earnings management (i.e., less disclosure) will result in the analyst report being, in part, a

more valuable alternative to firm-provided disclosure resulting in a greater demand for analyst

services.

However, whether serving as information intermediaries and/or producers, it is also plau-

sible that financial analysts do not just passively accept the impact of an information envi-

ronment. Analysts often communicate with firm managers and convey their opinions about

the current and future state of covered firms through forecasts or public releases. Prior stud-

ies suggest that analysts may influence the information environment in two ways. According

to the pressure hypothesis, financial analysts can impose excessive pressure on managers by

forcing managers to engage in earnings management to meet or beat analyst forecasts. Due

to the severe reaction by the market to missing an earnings target, firms are sometimes will-

ing to sacrifice economic value in order to meet a short run earnings target (Graham, Harvey,

and Rajgopal, 2005). Earnings manipulation takes this one step further. Degeorge, Patel, and

Zeckhauser (1999) find that managers have strong incentives to meet analyst forecasts by man-

aging earnings. Overall, this suggests that financial analysts may cause an increase in earnings

management.

Alternatively, financial analysts may see through earnings management and reduce fir ms’

opportunities to manipulate earnings. Firms may choose to reveal more inside information through

disclosures and management communication, seeking an increase in the number of analysts

following the firm (Graham et al., 2005; Francis, Nanda, and Olsson, 2008) and in analyst forecast

accuracy (Lang and Lundholm, 1996). While higher transparency constrains managers’ ability to

take advantage of their insider positions, higher analyst coverage,and g reater transparencyresult

in important benefits to the company, including lower target debt ratios (Chang, Dasgupta, and

Hilary, 2006), improved liquidity (Balakrishnan et al., 2013), and higher market value (Chung

and Jo, 1996). Analysts also play a monitoring role by facilitating the detection of corporate

fraud (Dyck, Morse, and Zingales, 2010). Yu (2008) finds that firms followed by more analysts

manager their earnings.

In summary, the impact of less earnings management may either increase or decrease the

demand for analysts’ service, depending upon whether analysts serve as information intermedi-

aries or information producers. The presence of analyst scrutiny may, in turn, result in more or

Hong, Huseynov, & Zhang rEarnings Management and Analyst Following 357

less earnings management. In the end, the relationship between analyst following and earnings

management depends upon the results of these competing elements and is an empirical question

addressed in the paper.

In this paper, we use a two-stage least-squares (2SLS) method to investigate the joint de-

termination of analyst following and earnings management by simultaneously analyzing four

possible effects between analyst coverage and earnings management: 1) the presence of earnings

management reduces firm-provided disclosure and reduces the level of analyst coverage (infor-

mation intermediary hypothesis), 2) the presence of earnings management reduces firm-provided

disclosure and increases the level of analyst coverage (information producer hypothesis), 3) the

presence of financial analysts reduces earnings management by providing additional monitoring

(monitoring hypothesis), and 4) the presence of financial analysts puts more pressure on managers

and makes them more likely to manage earnings (pressure hypothesis). Following prior studies

on earnings management (Yu, 2008; Zang, 2012), we focus our analysis on “suspect firms” that

report small positive earnings, small positive changes in earnings, or marginally meet or beat the

earnings forecast (Burgstahler and Dichev, 1997; Degeorge et al., 1999). These firms are more

likely to use accruals and real activities to manage earnings (Francis et al., 2005; Cohen et al.,

2008). In this setting, we are more likely to observe the effect of earnings management on analyst

following. Our major finding is that analysts’ decisions to follow firms and managerial decisions

to manage earnings are jointly determined. We find strong evidence that firms with a lower level

of accrual-based earnings management offer a better information environment and attract analyst

following. This implies that firms choose the level of disclosure in order to attract analysts and

analysts respond accordingly. There is also compelling evidence that the presence of analyst

following reduces the level of both accrual-based and real earnings management suggesting that

analysts play the role of external monitors. Further, our results from the 2SLS method provideno

evidence that real earnings management plays any role in attracting analyst following,suggesting

that accrual-based earnings management is the preferred method used by fir ms to calibrate their

level of disclosure.

To develop further insight into the relationship, we also examine the interaction of analyst

following and earnings management when firms run into trouble. Our results for fir ms that meet

or beat analyst forecasts just once in five years are similar to those in the entire sample in that

analyst following drops as the level of accrual-based management increases. The association

between analyst following and earnings management variables in firms that meet or beat analyst

forecasts repeatedly in a five-year period is statistically much weaker, suggesting that analysts

may be less attracted to following these firms due to their poorer information environment. Our

results indicate that the number of analysts following a firm decreases when a firm manipulates

earnings. This downward trend in analyst followingis more pronounced and lasts longer for f irms

that manage earnings continuously.

We also analyze the impact of the Sarbanes-Oxley Act (SOX) and the Securities and Ex-

change Commission (SEC) regulations (e.g., the Global Settlement [GS]) passed in 2002 on

the relation between analyst following and earnings management. Our results in the pre- and

post-SOX subperiods suggest that while analysts play the role of external monitors through-

out the entire period from 1990 to 2010, only after 2001 is there evidence that accrual-based

earnings management influences analyst coverage. Prior to 2001, it is plausible that conflicts of

interest, real or perceived, diminished the information impact on analyst following as analysts

would cover a firm simply to maintain relationships and drive investment banking activities.

Furthermore, analysts might have had access to a sufficient amount of information from var-

ious other sources such that earnings management (either accrual-based or real) would have

little or no impact on their choice of coverage. After 2001, the decision of analysts to cover