Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Earnings Management and Performance Incentives — The Monitoring Effect of Analysts’ Coverage

Earnings Management and Performance Incentives — The Monitoring Effect of Analysts’ Coverage EARNINGS MANAGEMENT AND PERFORMANCE INCENTIVES — THE MONITORING EFFECT OF ANALYSTS’ COVERAGE By ERASMUS ATTAH-GYAMFI Electronic copy available at: https://ssrn.com/abstract=3724612 © 2020 Erasmus Attah-Gyamfi Electronic copy available at: https://ssrn.com/abstract=3724612 ABSTRACT: I examine the effect of analysts’ coverage on managerial behavior to manage earnings to meet or beat analysts’ estimates. I manipulate analysts’ coverage (high versus low) and performance incentives (present versus absent) to investigate their effect on managers’ choices in a revenue recognition scenario. The findings show that when performance incentives are present, managers are likely to manage earnings to meet or beat targets regardless of the level of analysts’ coverage. However, I find that incentivized managers are sensitive to the level of analysts’ coverage, which improves the quality of the information environment and disciplines managerial behavior. This suggests that analysts’ coverage can serve as a useful monitor of managers’ reporting choices. The findings extend accounting literature by using a controlled experiment to examine earnings management, allowing for more reliable causal connections to be inferred between it and the manipulations. The results are potentially useful to regulators and standard setters as they evaluate alternative devices to curtail managerial opportunism in external reporting. Keywords: earnings management, revenue recognition, analysts’ coverage, equity incentives, monitoring. Data availability: Contact the Author. Electronic copy available at: https://ssrn.com/abstract=3724612 I. INTRODUCTION Firms' desire to align managerial performance and shareholder interests, has led to a substantial increase in their use of performance incentives, notably stock-based and option-based executive compensation (Bergstresser and Philippon 2006). Moreover, evidence suggests that “firm performance is positively related to the share of equity held by managers and the share of manager compensation that is equity-based” (see Bergstresser and Philippon 2006; Mehran 1995). Nevertheless, regulators and other stakeholders have raised concerns that performance incentives lead to earnings manipulations, thereby reducing the quality of financial reports with its associated consequences (see Cheng and Warfield 2005; Levitt 1998; Bergstresser and Philippon 2006). Recent research has examined the potential role of governance, regulatory, and external practices to reduce the incidence of earnings management (Hales, Koka, and Venkataraman 2018). This paper extends this literature by examining whether the extent of analysts’ coverage can curb performance-incentivized managers’ tendency to manage earnings. Potentially providing an alternate tool to mitigate opportunistic managerial behavior. Security analysts are legitimate market players whose actions influence the stock price of companies they follow (Graham, Harvey, and Rajpoal 2005). Meeting or missing analysts’ forecasts is value relevant (Lee 2007; Bartov, Givoly, and Hayn 2002). Thus, managers compensations are affected by whether they meet or fail to meet analysts’ expectations (Matsunaga and Park, 2001). Analysts base their forecasts on publicly available information as well as formal presentations by company executives (Payne and Robb 2000; Lees 1981). They are, therefore, seen as knowledgeable professionals with industry expertise. I posit that analysts’ Electronic copy available at: https://ssrn.com/abstract=3724612 coverage improves the information environment of a company, which reduces the attractiveness of earnings management. Moreover, the increased likelihood that analysts will detect and report earnings management to the public serves as a deterrent, thereby providing a monitoring effect. There is reason to believe that analysts’ coverage improves firms’ information environment. First, they are typically the first to detect corporate fraud and thus serve as whistleblowers (Dyck, Morse, and Zingales 2010). Second, archival research shows that analysts' coverage results in fewer earnings management (Degeorge, Ding, Jeanjean and Stolowy 2005; Irani and Oesch 2013). This study uses an experiment to attempt to replicate this finding. Third, analysts serve as monitors, alongside traditional mechanisms of corporate governance, to deter misreporting and to discipline managerial misbehavior (Yu 2008). Fourth, analysts create a stronger and more transparent operating environment, which enhances the efficiency of a firm’s processes, including their investment decisions (Firth, Xie, and Zhang 2016). I use an experiment to test my prediction that analysts’ coverage curbs the negative tendencies of management to manipulate earnings. Thus, the higher the analysts’ coverage, the fewer managers are inclined to manipulate earnings even if they are incentivized with equity- based compensation. I also test the predictions that high analysts’ coverage is associated with an improvement in the quality of the information environment and managers’ perception of being monitored. In the experiment, I manipulate, between participants, performance incentives (present or absent), and analysts’ coverage (high or low). Across all experimental conditions, participants assume the role of a CFO of a hypothetical firm who is asked to recommend an amount of revenue to be recognized for a percentage of the completed contract, juxtaposed against traditional practices. Participants in the performance incentives condition are told that they are Electronic copy available at: https://ssrn.com/abstract=3724612 entitled to equity-based bonuses each year that are very competitive in the industry; this is in addition to an annual salary. The bonuses are tied to the stock price appreciating, and there are no mandatory stock redemptions. Those in the no performance incentives condition are told that they are entitled to only salary and perks commensurate with being a CFO. Participants in the high analysts’ coverage conditions are told that they are heavily followed by 26 high-quality analysts who develop consensus estimates. This level of coverage is significantly above the industry average; Yu (2008) posit that the average analysts’ coverage is 9.44. On the other hand, participants in the low analysts’ coverage conditions are told that the company is followed by only three analysts, whose coverage is often characterized by high forecast dispersion that rarely results in consensus estimates. All other information is consistent across all conditions. I find that when performance incentives are present, managers are likely to manage earnings to meet or beat targets regardless of the level of analysts’ coverage. I also find that when managers are incentivized, high analysts’ coverage improves the quality of the information environment than when coverage is low. In addition, I find that high analysts’ coverage impacts the decisions of incentivized managers than when coverage is low. Further, I find that managers factor analysts’ consensus estimates in their decision making when analysts’ coverage is high, and when performance incentives are present than when they are low and absent, respectively. Similarly, I find that when analysts’ coverage is high, managers are sensitive to their presence, which disciplines their behavior than when analysts’ coverage is low; this is true when performance incentives are present compared to when absent. Taken together, these results suggest that high analysts’ coverage serves as effective monitors of managerial reporting Electronic copy available at: https://ssrn.com/abstract=3724612 activities than when analysts’ coverage is low because managers are sensitive to the presence of analysts, potentially curbing their opportunistic behaviors. This paper contributes to the accounting literature by using a controlled experiment to examine whether analysts’ coverage serves as a transparency tool that mitigates managers’ opportunistic behavior. The use of an experimental approach extends the earnings management literature, given that most results in the arena have used archival data, where controls are difficult to achieve. The results are useful to regulators, and standard setters whose mandated interest is in the propriety of financial information shared with the investment community (e.g., Pozen 2008; SEC 2003; Koonce 2016). The remainder of this paper is organized as follows: In Section II, I review the literature and develop the research hypothesis. Section III discusses the method followed by the results in section IV. The last section discusses the findings and conclusions. II. THEORY AND HYPOTHESIS DEVELOPMENT Earnings Management Prior research shows that managers have a tendency to manage earnings (see Cheng and Warfield 2005; Richardson, Tuna, and Wu 2002). They do so to maximize their interests from explicit contractual arrangements including bonus plans and debt covenants (e.g., Watts and Zimmerman 1986; Richardson et al. 2002; Dechow and Skinner 2000, Healy and Whalen 1998), antitrust/government regulation, and implicit management compensation contracts, including equity incentives tied to stock price (e.g., Healy and Whalen 1998; Dechow and Skinner 2000). In addition to these incentives based rational for earnings management, concentrated ownership, and smaller boards are associated with more earnings management (Yu 2005). In effect, prior Electronic copy available at: https://ssrn.com/abstract=3724612 research consistently shows that performance incentives, which can align managers’ and investors' interests, also have the effect of leading to opportunistic managerial reporting. Analysts’ Coverage Research shows that analysts’ coverage, among other benefits, serves as a robust monitoring mechanism that results in fewer earnings management. Degeorge et al. (2005) posit that analysts’ coverage serves as transparency tools that check earnings management. The effect was amplified when board monitoring was sufficient, and disclosure requirements were high. I expect that when analysts’ coverage is high for a firm, it will deter managers from managing earnings. Based on this, I formally hypothesize that: H1: When equity incentives are present, analysts’ coverage will reduce the opportunistic behavior of managers to manage earnings to meet or beat analysts’ estimates. Expert external financial analysis, a product of analysts’ coverage, enhances the efficiency of a firm's processes by encouraging a stronger and more transparent operating environment. Firth et al. (2016) argue that analysts are responsible for high-quality information gathering and monitoring, which can improve firms’ investment efficiency hence fewer earnings management. For example, Firth et al. (2016) posit that premium-level analysis correlates strongly with optimal investment decisions, ensuring an equilibrium that prevents terrible decisions such as over or under-investment. Irani and Oesch (2013) also suggest that analysts' coverage reduces information asymmetry by making managers afraid that their inflated earnings would be discovered and revealed to the public. Accordingly, I expect that analysts’ coverage improves the information environment. This expectation is formally hypothesized as: Electronic copy available at: https://ssrn.com/abstract=3724612 H2: When analysts’ coverage is high, the quality of the information environment will be higher than when the coverage is low. Yu (2008) suggests that analysts’ coverage serves as a monitoring tool, thus the more the availability of analysts' forecasts, the less the management of earnings. Analysts’ activities play a significant role in serving as whistleblowers by being the first to detect corporate fraud (Dyck et al. 2010). Based on these findings, I predict that when analysts’ coverage is high, managers will be sensitive to their presence, which will discipline their behavior. Based on this prediction, I formally hypothesize that: H3: When analysts’ coverage is high, the monitoring of the reporting activities of incentivized managers will be better than when coverage is low. III. METHODS Participants The participants are 162 managers recruited by using Qualtrics. The sample consists of 74 females (45.6 percent) and 88 males (54.3 percent); 59 (36.4 percent) participants have at least a Master's degree, and 67 (41.3 percent) possess a Bachelor's degree. All participants are mid-upper level managers with a mean age of 40.98 years (Standard Deviation = 10.05, Minimum = 25.00, Maximum = 75.00, Skewness = 0.71, Kurtosis = 0.02). Participants have 10.48 mean years of experience in finance and accounting (Standard Deviation = 6.00, Skewness = 0.36, Kurtosis = -1.01). Table 1 summarizes these demographic statistics. The skewness and kurtosis suggest that the presence of outliers is unlikely. While incentives are likely to induce When the skewness is greater than 2 in absolute value, the variable is considered to be asymmetrical about its mean. When the kurtosis is greater than or equal to 3, then the variable's distribution is markedly different from a normal distribution in its tendency to produce outliers (Westfall & Henning, 2013). Electronic copy available at: https://ssrn.com/abstract=3724612 participants to exert additional effort on the experimental task (Camerer and Hogarth 1999). I, nevertheless, incentivized each individual with a $30 visa gift card. Institutional Review Board (IRB) approval was obtained for the experiment. All study participants provided informed consent for their participation, and they were assured of confidentiality. This is consistent with the IRB requirements. Task and Materials Participants are asked to assume the role of a CFO of a hypothetical firm — Gates, Inc. Their hypothetical firm is structured exactly as a brick and mortar firm would be, including standard operation procedures, history of gains and losses, and examples of employer/employee interaction. The first part of the case materials provides the participants with information about the company, including comparative financial statements, its operations, the board of directors, and internal control mechanisms. The company’s stock is traded on NYSE. Internal controls are described as typical for a company of its size and standing. The financial information suggests that Gates Inc. has performed well over the past five years, consistently meeting or beating analysts’ earnings per share (EPS) estimates. They are advised that the firm faces competitive pressures from a resurgent market player; as a result, profits of the firm have been affected. They are also informed that they have won a contract which will be 75 percent complete by year-end. The case materials include the current FASB guidance on Revenue Recognition for Contracts (ASU 2014-09; Revenue from Contracts with Customers (Topic 606)). Participants are told that, in accordance with GAAP, while managers can recognize contract revenue when each performance of a deliverable is complete, historically, the managers Electronic copy available at: https://ssrn.com/abstract=3724612 at Gates, Inc. have been very conservative and recognize revenues related to each contract when it is 100 percent complete (all performance deliverables are complete). Table 1 Demographic Descriptive Statistics Panel A: Frequencies Variable N % Sex Male 88 54.32 Female 74 45.68 Missing 0 0 Education High school graduate (high school diploma or equivalent including GED) 6 3.70 Some college but no degree 9 5.56 Associate degree in college (2-year) 11 6.79 Bachelor's degree in college (4-year) 67 41.36 Master's degree 59 36.42 Doctoral degree 5 3.09 Professional degree (JD, MD) 5 3.09 Missing 0 0 Professional Credential If Yes, state credential 52 32.10 No 110 67.90 Missing 0 0 Panel B: Summary Statistics Table for Age and Years of Experience Variable Mean Std. Deviation N Min Max Skewness Kurtosis Age 40.98 10.05 162 25.00 75.00 0.71 0.02 Years of Experience 10.48 6.00 162 0.00 20.00 0.36 -1.01 Participants are then informed that they are in the fourth quarter, for the year-end financial results, and that the results of Gates, Inc., are expected to fall short of analysts' Participants are new in their roles even though they have prior experience. Electronic copy available at: https://ssrn.com/abstract=3724612 consensus EPS estimate for the year-end (by 10 cents). Participants are informed that the analysts’ EPS consensus estimate is $3.15, whereas the internal year-end analysis yields an estimate of $3.05. An inability to meet the $3.15 will probably cause the value of the stock to drop by 7%. Historically, values have dropped by 7% for such forecast misses. The task for participants will be deciding how to deal with the shortfall by recognizing an amount of revenue to meet or beat the analysts' estimate. Participants then respond to additional questions that relate to the information environment, the monitoring effect of analysts’ coverage, and manipulation checks. Appendices 1-3 shows the respective scales used for these tests. Participants are also asked to provide demographic information. Experimental Design I use a 2x2 between-subjects research design to test my predictions. The two independent variables, performance incentives, and analysts’ coverage are manipulated between subjects. I manipulate performance incentives at two levels: present versus absent. Participants in the “present” performance incentives condition are told that they are entitled to equity-based bonuses each year in addition to an annual salary. The bonuses are tied to Gates Inc.’s equity (stock) price change of +/-5%. Participants are advised that the company has no policy of mandatory stock redemptions; as such meeting, the set target is the way to cashing in on the bonuses. The equity-based bonus serves as the proxy for performance incentives. According to Cornet et al. (2008), “equity-based compensation has been used as a proxy for incentives to manage earnings in several papers” (see, for example, Bergstresser and Philippon 2006; Cheng and Warfield 2005; Cohen, Dey, and Lys 2005). Participants in the “absent” performance incentives condition are told that they are only entitled to industry-standard competitive annual Electronic copy available at: https://ssrn.com/abstract=3724612 salary and perks commensurate with being a CFO. There is no provision for bonuses or additional rewards tied to equity performance. I manipulate analysts’ coverage at two levels: high versus low. Participants in the “high” analysts’ coverage condition are told that they are heavily followed by 26 analysts who develop consensus estimates. This level of coverage is significantly above the industry average; Yu (2008) posit that the average analysts’ coverage is 9.44. On the other hand, participants in the “low” analysts’ coverage condition are told that the company is followed by only three analysts, whose coverage is often characterized by high forecast dispersion that rarely results in consensus estimates. I validate my manipulations by conducting a pilot experiment where Amazon Mechanical Turk (MTurk) workers rate the extent of analysts’ coverage and the extent of CFO compensation package influenced by performance incentives on 10-point scales (see, Asay, Elliot and Rennekamp 2017; Henry and Peytcheva 2020, for a similar out of sample manipulation check validations). Appendix 1 shows the scales used in obtaining the manipulation checks. Non- tabulated results from the pilot experiment reveal that both the analysts’ coverage (F (1,158) =15.178, p< 0.001), and the performance incentives (F (1,158) =10.39, p= 0.002) significantly impacts the responses to the manipulation checks. A comparison of the means shows that participants in the high analysts’ coverage condition rated the extent of analysts’ coverage at a mean of 7.60 (2.14) compared to those in low coverage conditions who rated it at a mean of 5.56 (2.60) (t= 3.28, p=.002). Similarly, participants in performance incentives present condition rated the extent of CFOs compensation influenced by performance incentives at a mean of 7.40 (1.80) compared to a mean of 5.50 (2.54) by those in absent performance incentives condition (t= 3.311, p= 0.004). These results suggest that the manipulations were successful Electronic copy available at: https://ssrn.com/abstract=3724612 Dependent Variables There are three dependent variables in this study. First is the amount of revenue recognized. I ask participants to record how much of the $12 million potential revenue they would recognize for the percentage of contract completed by year-end. I assume that no earnings management is consistent with sticking with the company’s conservative posture and historical practices of not recognizing revenue until the contract is completed. Conversely, the degree of earnings management is in increasing the amount of revenue recognized. The second dependent variable is managers’ rating of the quality of the information environment. Participants are asked to rate on a 10-point scale, how much the level of analysts’ coverage enhanced the quality of information available to potential investors to support their investment decisions. High (low) ratings are indicative of managers’ belief that high (low) analysts’ coverage improves (diminishes) the quality of information available to investors. The last dependent variable is managers’ perception of the effect of analysts’ coverage on their reporting activities. I ask participants to rate on a 10-point scale, three questions related to the impact of analysts’ coverage on the revenue recognition decision they make, the impact of analyst consensus estimates on their choices, and how much the presence of analysts enhanced discipline in their reporting activities. I infer that high (low) ratings imply a strong (weak) effect of analysts’ coverage on managerial reporting activities. IV. RESULTS Manipulations Checks After making revenue recognition decisions and answering questions about the monitoring effect of analysts’ coverage and the information environment, I ask participants to Electronic copy available at: https://ssrn.com/abstract=3724612 characterize the extent to which analysts cover the firm and to rate the extent to which the CFO compensation is influenced by performance incentives; on 10-point scales (see appendix 1). I test the effectiveness of the manipulations with a 2 x 2 multiple analysis of variance (MANOVA) that employs analysts’ coverage and performance incentives as independent variables. The dependent variables are the responses to the manipulation checks. Table 2 reveals that both the analysts’ coverage (F (2, 158) = 151.93, p < .001) and the performance incentives (F (2, 158) = 111.73, p < .001) significantly impact the responses to the manipulation checks. Univariate comparison of the means reveals that participants in the high analysts’ coverage condition rate the level of analysts’ coverage at a mean of 7.95 (1.50) compared to a mean of 3.49 (1.66) by those in low coverage condition. Participants in the present performance incentives condition rate the level of CFOs compensation influenced by performance incentives at 7.67 (1.54) compared to a mean of 3.33 (2.02) by those in the no performance incentives condition. Table 3 shows descriptive statistics. Overall, these results provide evidence of the success of the manipulations. Table 2. 2x2 MANOVA for Manipulation Check Questions by Experimental Conditions Pillai’s Residual p- Variable F df Trace df value Analysts’ Coverage 0.66 151.93 2 158 < .001 0.66 Performance Incentives 0.59 111.73 2 158 < .001 0.59 Analysts’ Coverage x Performance 0.004 .304 2 158 .738 0.004 Incentives Table 3. Mean (Standard Deviation) for Manipulation Checks Analysts' Coverage Manipulation Checks Low High Total Performance Incentives Absent 3.03 (1.96) n=32 3.65 (2.06) n=31 3.33 (2.02) n=63 Present 4.00 (1.51) n=31 8.15(1.49) n=68 7.67 (1.54) n=99 Total 3.49 (1.66) n=63 7.95 (1.50) n=99 Electronic copy available at: https://ssrn.com/abstract=3724612 Tests of Hypotheses Hypothesis 1 H1 states that when equity incentives are present, analysts’ coverage will reduce the opportunistic behavior of managers to manage earnings to meet or beat analysts' estimates. I test the hypothesis with an analysis of variance (ANOVA) that employs analysts’ coverage and performance incentives as two levels between subjects' factors. The dependent variable is the amount of revenue recognized. Table 4 presents the ANOVA model, which shows that there is a significant main effect for performance incentives (F (1, 158) = 6.77, p = .010). The reported mean (standard deviation) = 7.44 (3.13) of the present condition is significantly larger than the absent condition 5.80(3.28). Contrary to expectations, the interaction effect is not significant (F (1, 158) = 2.50, p = .116). The main effect for analysts’ coverage is also not significant (F (1, 158) = 1.33, p = .251). Thus, hypothesis 1 is not supported. Table 5 presents the descriptive statistics which show that managers in high incentive conditions are likely to report a higher revenue than those in the low incentive conditions, regardless of the level of analysts’ coverage. Table 4. Analysis of Variance for Revenue Recognized Term SS df F p-value η 2 Analysts’ Coverage 13.28 1 1.33 0.251 0 Performance Incentives 67.66 1 6.77 0.01 0 Analysts’ Coverage x Performance Incentives 25 1 2.50 0.116 0 Residuals 1579.72 158 Table 5. Mean (Standard Deviation) for Revenue Recognized Analysts' Coverage Revenue Recognized Low High Total Performance Incentives Absent 5.92 (3.23) n=32 5.69 (3.39) n=31 5.80 (3.28) n=63 Present 6.45 (3.08) n=31 7.89 (3.06) n=68 7.44 (3.13) n=99 Total 6.18 (3.14) n=63 7.20 (3.31) n=99 Electronic copy available at: https://ssrn.com/abstract=3724612 Hypothesis 2 H2 proposes that when analysts’ coverage is high, the quality of the information environment will be higher than when the coverage is low. I test the hypothesis with an analysis of variance (ANOVA) that employs analysts’ coverage and performance incentives as two levels between subjects' factors. The dependent variable is the managers' ratings of the quality of the information environment. Table 6 presents the results of the ANOVA model. The main effect for analysts’ coverage is significant (F (1, 158) = 25.39, p < .001). The main effect for performance incentives is not significant (F (1, 158) = 0.29, p = .589). However, because of the significant interaction described below, the main effects cannot be given a normal interpretation. The interaction effect between analysts’ coverage and performance incentives is significant (F (1, 158) = 9.96, p = .002). To further understand the nature of the significant interaction, I conduct a simple effects analysis. Table 6. Analysis of Variance for Information Environment Quality Term SS df F p-value η 2 Analysts’ Coverage 87.42 1 25.39 < .001 0.1 Performance Incentives 1.01 1 0.29 0.589 0 Analysts’ Coverage x Performance Incentives 34.29 1 9.96 0.002 0.1 Residuals 544.08 158 Table 7 presents results from the tests of the simple effects, which reveal that for incentivized managers, high analysts’ coverage led to a stronger information environment rating than low analysts’ coverage (F (1, 158) = 8.04, p = .005). For non-incentivized managers, analysts’ coverage had no effect (F (1, 158) = 2.97, p = .087). A comparison of the means indicates that participants in high analysts’ coverage and performance incentives present conditions reported a mean (standard deviation) of 7.72 (1.68) compared to low analysts’ coverage and performance Electronic copy available at: https://ssrn.com/abstract=3724612 incentives present conditions 5.19 (2.04). Participants in high analysts’ coverage and performance incentives absent conditions reported 6.58 (1.78) compared to those in low analysts’ coverage and performance incentives absent conditions 6.00 (2.08). These results suggest that when analysts’ coverage is high, the quality of the information environment is more robust than when coverage is low. The robustness is likely amplified when performance incentives are present compared to when absent. However, when analysts’ coverage is low, performance incentives likely reduces the quality of the information environment compared to when absent. Figure1 illustrates this observation. Esti mated Margi nal Means of Informati on Envi ronment Pe rformance Ince ntive s Condition 8.0 Absent Present 7.5 7.0 6.5 6.0 5.5 High Low Analys ts ' Cove rage Condition Figure 1. Information Environment Quality Means by factors levels of Analysts' Coverage and Performance incentives Electronic copy available at: https://ssrn.com/abstract=3724612 Estimated Marginal Means Table 8 presents the descriptive statistics, which shows that when analysts’ coverage is high, managers are more likely influenced to improve the information environment than when coverage is low. This result provides support for hypothesis 2. Table 7. Simple Effect Test for Information Environment Quality Analysts' Coverage SS df F p-value η 2 High Contrast 27.67 1 8.04 0.005 0.048 Error 544.08 158 Low Contrast 10.24 1 2.97 0.087 0.018 Error 544.08 158 Table 8. Mean (Standard Deviation) for Information Environment Quality Analyst Coverage Information Environment Low High Total Performance Incentives Absent 6.00 (2.08) n=32 6.58 (1.78) n=31 6.29 (1.95) n=63 Present 5.19 (2.04) n=31 7.72(1.68) n=68 6.93 (2.14) n=99 Total 5.60 (2.08) n=63 7.36 (1.79) n=99 Hypothesis 3 H3 predicts that when analysts’ coverage is high, the reporting activities of incentivized managers are effectively monitored compared to when coverage is low. To test this hypothesis, I ask participants to rate (1) the impact of analysts’ coverage on the revenue recognition decision they had made, (2) the impact of analysts’ consensus estimates on their choices, and (3) how much the presence of analysts enhanced discipline in their reporting activities. The dependent variables are the managers’ ratings to these questions. To obtain an overall score for the managers’ ratings, I calculate the aggregate mean score for all three questions. This score represents the overall monitoring effect of analysts’ coverage on managerial, financial reporting Each F tests the simple effects of Performance Incentives Condition within each level combination of the other effects shown. These tests are based on the linearly independent pairwise comparisons among the estimated marginal means. Electronic copy available at: https://ssrn.com/abstract=3724612 decisions. I test the hypothesis by first conducting a multiple analysis of variance (MANOVA) that employs analysts’ coverage and performance incentives as independent variables. Table 9 presents the MANOVA results which reveal that the main effects for analysts’ coverage (F (3, 156) = 8.51, p < .001) and performance incentives (F (3, 156) = 9.45, p < .001) are significant . Similarly, the interaction effect between analysts’ coverage and performance incentives (F (3, 156) = 4.25, p = .006) is significant. Table 9. 2x2 MANOVA for Overall Monitoring Effect Pillai’s Residual Variable F df p η 2 Trace df Analysts’ Coverage 0.14 8.51 3 156 < .001 0.14 Performance Incentives 0.15 9.45 3 156 < .001 0.15 Analysts’ Coverage x Performance 0.08 4.25 3 156 0.006 0.08 Incentives Given that the main and interaction effects from the MANOVA are all significant, I proceed to conduct individual ANOVAs on the three questions to determine the nature of these effects. Analysts’ Impact on Managerial Decision-making: The first question for the monitoring effect of analysts relates to whether analysts’ coverage impacts managerial decision making. An ANOVA that employs analysts’ coverage and performance incentives as two levels between subjects' factors is conducted. The dependent variable is managers’ rating of the impact of analysts' coverage on their managerial decisions. Table 10 presents the ANOVA model. Univariate comparison of means shows that participants in high analysts’ coverage conditions reported a mean (standard deviation of 6.39 (1.99) compared to low conditions 5.08 (1.91) p=0.005. Participants in performance incentives present conditions reported 6.97(1.93) compared to absent conditions 5.02 (1.95) p=0.005. Electronic copy available at: https://ssrn.com/abstract=3724612 The main effect for analysts’ coverage is significant (F (1, 158) = 22.84, p < .001). The main effect for performance incentives is also significant (F (1, 158) = 18.41, p < .001). However, because of the significant interaction described below, the main effects cannot be given a normal interpretation. The interaction effect between analysts’ coverage and performance incentives is significant (F (1, 158) = 8.25, p = .005). To further understand the nature of the significant interaction effect, I conduct a simple effects analysis. Table 10. Analysis of Variance for Analysts’ Impact Term SS df F p-value η 2 Analysts’ Coverage 90.36 1 22.84 < .001 0.1 Performance Incentives 72.85 1 18.41 < .001 0.1 Analysts’ Coverage x Performance Incentives 32.66 1 8.25 0.005 0.1 Residuals 625.1 158 Table 11 presents results from the tests of the simple effects which reveal that for incentivized managers, high analysts’ coverage led to higher monitoring scores than low analysts’ coverage (F (1, 158) = 30.18, p < .001). For non-incentivized managers, analysts’ coverage had no effect (F (1, 158) = 0.87, p = .351). A comparison of the means indicates that participants in high analysts’ coverage and present performance incentives conditions reported a mean (standard deviation) of 7.53 (1.87) compared to low analysts’ coverage and performance incentives present conditions 5.00 (2.14). Participants in high analysts’ coverage and performance incentives absent conditions reported a mean (standard deviation) of 5.16 (2.18) compared to low analysts’ coverage and performance incentives absent conditions 4.53 (1.88). These results suggest that the impact of analysts’ coverage on managerial decisions is stronger when analysts’ coverage is high, and performance incentives are present. Electronic copy available at: https://ssrn.com/abstract=3724612 Table 12 presents descriptive statistics which show that incentivized managers are sensitive to analysts' coverage, which likely impacts their financial reporting decisions. This result supports hypothesis 3. Table 11. Simple Effect Test for Analysts' Impact Analysts' Coverage SS df F p-value η 2 High Contrast 119.41 1 30.18 0.000 0.2 Error 625.10 158 Low Contrast 3.46 1 0.87 0.351 0.0 Error 625.10 158 Table 12. Mean (Standard Deviation) for Analysts’ Impact Analyst Coverage Analysts’ Impact Low High Total Performance Incentives Absent 4.53 (1.88) n=32 5.16 (2.18) n=31 4.84 (2.04) n=63 Present 5.00 (2.14) n=31 7.53(1.87) n=68 6.74(2.28) n=99 Total 4.76 (2.01) n=63 6.79 (2.25) n=99 Impact of Analysts’ Consensus Estimates: The second question for the monitoring effect of analysts is whether managers factored analysts’ consensus estimates in their decision making. An ANOVA that employs analysts’ coverage and performance incentives as two levels between subjects' factors is conducted. The dependent variable is managers' rating of the effect of analysts' consensus estimates on their managerial decisions. Table 13 presents the ANOVA model which shows that the main effect for analysts’ coverage is significant (F (1, 158) = 14.86, p < .001), indicating that the analysts’ consensus estimate effect is higher when analysts’ coverage is high (mean (standard deviation) = 6.93(2.40) compared to when analysts’ coverage is low 5.11(2.20)). The main effect for performance incentives is also significant (F (1, 158) = 24.43, p < .001), indicating that the analysts’ consensus estimate effect is higher when performance incentives are present (mean (standard deviation) = 7.05(2.09) compared to when performance incentives are absent 4.92(2.50)). The interaction effect between analysts’ coverage and performance incentives is not significant (F (1, 158) = 2.66, p = .105). Electronic copy available at: https://ssrn.com/abstract=3724612 Table 14 presents the descriptive statistics which provide evidence that managers in high analysts’ coverage conditions, as well as those in performance incentives present conditions, are more likely to factor analysts' consensus estimates in their decisions. This result supports hypothesis 3 Table 13. Analysis of Variance for Analysts’ Consensus Estimate Term SS df F p-value η 2 Analysts’ Coverage 68.13 1 14.86 < .001 0.1 Performance Incentives 112.01 1 24.43 < .001 0.1 Analysts’ Coverage x Performance Incentives 12.2 1 2.66 0.105 0 Residuals 724.35 158 Table 14. Mean (Standard Deviation) for Analysts' Consensus Estimates Analyst Coverage Analysts' Consensus Low High Total Estimates Performance Incentives Absent 4.53 (2.27) n=32 5.32 (2.69) n=31 4.92(2.50) n=63 Present 5.71 (1.99) n=31 7.66(1.85) n=68 7.05(2.09) n=99 Total 5.11 (2.20) n=63 6.93(2.40) n=99 Disciplining managers: The last question relating to the monitoring effect is whether analysts’ coverage instilled discipline in managers. This is tested using an ANOVA that employs analysts’ coverage and performance incentives as two levels between subjects' factors. The dependent variable is the managers' rating of the level of discipline analysts’ coverage instilled in managers. Table 15 shows the ANOVA model. The main effect for analysts’ coverage is significant (F (1, 158) = 15.82, p < .001) indicating that discipline among managers is higher when analysts’ coverage is high (mean (standard deviation) = 7.08 (2.20) compared to when analysts’ coverage is low 5.37 (2.37)). The main effect for performance incentives is also significant (F (1, 158) = 18.82, p < .001) Electronic copy available at: https://ssrn.com/abstract=3724612 indicating that discipline among managers is also higher when performance incentives are present (mean (standard deviation) = 7.12 (2.11) compared to when performance incentives are absent 5.30 (2.45)). The interaction effect between analysts’ coverage and performance incentives is not significant (F (1, 158) = 0.04, p = .833). Table 16 presents descriptive statistics that provide evidence that when analysts’ coverage is high, and when performance incentives are present, managers are more likely to be disciplined than when analysts’ coverage is low, and performance incentives are absent. This result provides support for hypothesis 3. Table 15. Analysis of Variance for Managerial Discipline Term SS df F p-value η 2 Analysts’ Coverage 73.66 1 15.82 < .001 0.1 Performance Incentives 87.59 1 18.82 < .001 0.1 Analysts’ Coverage x Performance Incentives 0.21 1 0.04 0.833 0 Residuals 735.45 158 Table 16. Mean (Standard Deviation) for Managerial Discipline Analyst Coverage Managerial Discipline Low High Total Performance Incentives Absent 4.56 (2.45) n=32 6.06 (2.25) n=31 5.30(2.45) n=63 Present 6.19 (2.01) n=31 7.54(2.03) n=68 7.12(2.11) n=99 Total 5.37 (2.37) n=63 7.08(2.20) n=99 V. DISCUSSION AND CONCLUSION The experimental results show that when performance incentives are present, managers are likely to manage earnings to meet or beat targets. This is in spite of the level of monitoring of analysts, contrary to my prediction that analysts’ coverage will curb such opportunistic behavior. Electronic copy available at: https://ssrn.com/abstract=3724612 Nevertheless, managers believe analysts’ coverage significantly impacts their financial decisions, even though that impact does not seem to translate into affecting their opportunistic behavior. This phenomenon could be attributable to the effect of performance incentives being more potent than the impact of analysts’ monitoring when it comes to earnings management. The findings reveal that when analysts’ coverage is high, the presence of performance incentives improves the quality of the information environment. Conversely, when analysts’ coverage is low, the presence of performance incentives potentially reduces the quality of the information environment. This supports the assertion that analysts’ coverage provides a robust and more transparent information environment that supports investment decisions (Firth et al. 2016) and reduce information asymmetry (Irani and Oesch 2013). This study also shows that managers are sensitive to the presence of analysts, even when performance incentives are present, suggesting that analysts potentially serve as useful monitors of the reporting activities of managers. Specifically, managers are found to rate the impact of analysts’ coverage highly as well as factor analysts’ consensus estimates highly in their decision making; this is confirmed when both analysts’ coverage is high, and performance incentives are present. Further, I find that when analysts’ coverage is high, performance incentives produced stronger discipline among managers. Taken together, these suggest that high analysts’ coverage serves as effective monitors of managerial reporting activities, potentially mitigating their opportunistic behaviors- this result buttresses prior research findings that analysts' coverage disciplines managerial misbehavior (Yu, 2008). These findings enhance the understanding of managerial behavior in the face of incentives and the role of analysts in curbing earnings management. The results will be useful to regulators and standard setters as well as parties interested in corporate governance and the Electronic copy available at: https://ssrn.com/abstract=3724612 quality of financial reporting (e.g., external auditors, audit committee members, regulators). Specifically, for regulators, and standard setters whose mandated interest is in the propriety of financial information shared with the investment community (e.g., Pozen 2008; SEC 2003; Koonce 2016); the findings explicitly imply that analysts are critical to the quality of information environment and serve as useful monitoring tools; hence regulators should encourage a considerable level of analysts’ coverage for firms. This study is among the first papers, if not the first, to test the interaction of analysts’ coverage, performance incentives, and earnings management using an experimental approach. This paper contributes to the accounting literature by using a controlled experiment to examine whether analysts' coverage serves as a transparency tool that mitigates management's opportunistic financial reporting. The use of an experimental approach is a critical addition to the earnings management literature, given that most results in the arena have used archival techniques, where control is difficult to achieve. The findings show that motivated managers are sensitive to analysts’ coverage. Nevertheless, they are likely to engage in opportunistic behavior even in the face of heavy analysts’ monitoring. This suggests that analysts’ coverage alone might not be enough to curb earnings management. Therefore, together with analysts’ coverage, firms should consider other corporate governance mechanisms like board monitoring and strong audit committees to minimize the incidence of earnings management. This is an avenue for further research; indeed, the focus on improving corporate governance in our current dispensation provides numerous avenues for future research. Electronic copy available at: https://ssrn.com/abstract=3724612 REFERENCES Asay, H. S., W. B. Elliott, and K. M. Rennekamp. 2017. Disclosure readability and the sensitivity of investors’ valuation judgments to outside information. The Accounting Review 92 (4): 1–25. Bartov, E., Givoly, D., Hayn, C., 2002. The Rewards to Meeting or Beating Earnings Expectations. Journal of Accounting and Economics 33, 173-204. Bergstresser, D., and Philippon, T. (2006). CEO incentives and earnings management. Journal of Financial Economics, 80(3), 511–529. Camerer, C.F., and Hogarth, R.M. (1999). The Effects of Financial Incentives in Experiments: A Review and Capital-Labor-Production Framework. Journal of Risk and Uncertainty 19, 7–42. Cheng, Q. and T. Warfield. 2005. Equity incentives and earnings management. The Accounting Review 80: 441-476. Clor-Proell, S., and L. Maines. 2014. The impact of recognition versus disclosure on financial information: A preparer's perspective. Journal of Accounting Research 52: 671-701 Cohen, D.A., Dey, A., Lys, T.Z., 2005. Trends in earnings management and informativeness of earnings announcements in the pre- and post-Sarbanes Oxley periods. Available at SSRN Cornett, M., Marcus, A., and Tehranian, H. 2008. Corporate governance and pay-for- performance: The impact of earnings management. Journal of Financial Economics, 87, pp.357– Dechow, P. M., & Skinner, D. J. 2000. Earnings Management: Reconciling the Views of Accounting Academics, Practitioners, and Regulators. SSRN Electronic Journal. Degeorge, F., Ding, Y., Jeanjean, T., and Stolowy, H. 2005. "Does Analyst Following Curb Earnings Management?," HEC Research Papers Series 810, HEC Paris. Dyck, A., Morse, A., Zingales, L. 2010. Who Blows the Whistle on Corporate Fraud? Journal of Finance 65, 2213-2253. Firth, M., Xie, L., and Zhang, Y. (2016). How do analysts' forecast characteristics relate to investment efficiency? Paper. Lingnan University Graham, J. R., C. R. Harvey, and S. Rajgopal. 2005. The economic implications of corporate financial reporting. Journal of Accounting and Economics 40: 3-73. Hales, J., Koka, B., & Venkataraman, S. 2018. Curbing Earnings Management: Experimental Evidence on How Clawback Provisions and Board Monitoring Affect Managers Use of Discretion. SSRN Electronic Journal. Electronic copy available at: https://ssrn.com/abstract=3724612 Healy, P.M and Wahlen, J.M. 1998. A Review of the Earnings Management Literature and Its Implications for Standard Setting. Accounting Horizons: December 1999, Vol. 13, No. 4, pp. 365-383 Henry. E., and Peytcheva, M. 2020. Joint Effects of Boilerplate and Text Markup on the Judgments of Novice and Experienced Users of Financial Information. Behavioral Research in Accounting: Spring 2020, Vol. 32, No. 1, pp. 1-20. https://www.fasb.org/jsp/FASB/Page/ImageBridgePage&cid=1176169257359#section_2 Irani, R., and Oesch, D. 2013. Analyst Coverage and Earnings Management: Quasi-Experimental Evidence. 2013 Accounting Conference. Lee.2007. Earnings Management to Just Meet Analyst Forecast. Thesis. Northwestern University. Lees, F. A. 1981. "Public Disclosure of Corporate Earnings Forecasts." New York: Conference Board. Levitt, A. 1998. The numbers game. The CPA Journal 68 (12): 14-18. Koonce, L., Seybert N, and Smith, J. 2016. Management Speaks, Investors Listen: Are Investors Too Focused on Managerial Disclosures? Journal of Behavioral Finance, 17:1, 31-44 Matsunaga, S. R., Park, C. W. 2001. The Effect of Missing a Quarterly Earnings Benchmark on the CEO's Annual Bonus. The Accounting Review 76, 313-332. Mehran, H., 1995. Executive compensation structure, ownership, and firm performance. Journal of Financial Economics 38, 163–184. Payne, J. L., & Robb, S. W. G. 2000. Earnings Management: The Effect of Ex Ante Earnings Expectations. Journal of Accounting, Auditing & Finance, 15(4), 371–392. Pozen, Robert C. “Final Report of the Advisory Committee on Improvements to Financial Reporting to the United States Securities and Exchange Commission, August 1.” Securities and Exchange Commission Advisory Committee, Washington, DC (2008). Richardson, S., Tuna, A., and Wu, M. 2002. Predicting Earnings Management: The Case of Earnings Restatements. SSRN Electronic Journal. Rose, J., A.M. Rose, C.s. Norman and C.R. Mazza. 2014. Will Disclosure of Friendship Ties between Directors and CEOs Yield Perverse Effects? The Accounting Review 89: 1545-1563. Securities and Exchange Commission (SEC). Interpretation: Commission Guidance Regarding Management’s Discussion and Analysis of Financial Condition and Results of Operations (2003). Electronic copy available at: https://ssrn.com/abstract=3724612 Watts, R. L., and J. L. Zimmerman. 1986. Positive Accounting Theory. Englewood Cliffs, NJ: Prentice Hall Westfall, P. H., & Henning, K. S. S. 2013. Texts in statistical science: Understanding advanced statistical methods. Taylor & Francis. Yu, F., 2005. Corporate governance and earnings management. Unpublished working paper. University of Chicago, Chicago, IL Yu, F. 2008. Analysts’coverage and Earnings Management. Journal of Financial Economics 88, 245-271. APPENDIX 1 Manipulation Checks A. Based on the case materials, to what extent is the CFO's compensation package influenced by performance-based incentives? (1= Low extent; 10=High extent) B. Based on the case materials, how would you characterize the extent to which analysts cover the firm? (1= Low coverage; 10=High coverage) APPENDIX 2 Monitoring Effect 1. What impact is the extent of analysts covering the firm having on your decision? (l= Low impact; 10=High impact) Electronic copy available at: https://ssrn.com/abstract=3724612 2. To what extent will you as a manager of the firm factor analyst consensus estimates in your decision? (1= Low extent; 10= High extent) 3. To what extent will the level of analyst following the firm help instill discipline in managers when making financial reporting decisions? (1= Not at all; 10= Very Much) APPENDIX 3 Information Environment 1. To what extent will the number of analysts following Gates, Inc. enhance the amount of information available to potential investors to support their investment choices in the company? (1=Not at all; 10=Very Much) Electronic copy available at: https://ssrn.com/abstract=3724612 BIOGRAPHICAL SKETCH I am a Certified Public Accountant (CPA) and hold a Doctor of Business Administration degree (Accounting & Finance Major) from the University of Florida. Currently, I serve as a Controller for JPMorgan Chase bank. Electronic copy available at: https://ssrn.com/abstract=3724612 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png ARN Conferences & Meetings SSRN

Earnings Management and Performance Incentives — The Monitoring Effect of Analysts’ Coverage

Download PDF
31 pages

Loading next page...
 
/lp/ssrn/earnings-management-and-performance-incentives-the-monitoring-effect-b7khdjaYNi

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Copyright
Copyright © 2020 Elsevier Inc.
eISSN
ARCH-4059
DOI
10.2139/ssrn.3724612
Publisher site
See Article on Publisher Site

Abstract

EARNINGS MANAGEMENT AND PERFORMANCE INCENTIVES — THE MONITORING EFFECT OF ANALYSTS’ COVERAGE By ERASMUS ATTAH-GYAMFI Electronic copy available at: https://ssrn.com/abstract=3724612 © 2020 Erasmus Attah-Gyamfi Electronic copy available at: https://ssrn.com/abstract=3724612 ABSTRACT: I examine the effect of analysts’ coverage on managerial behavior to manage earnings to meet or beat analysts’ estimates. I manipulate analysts’ coverage (high versus low) and performance incentives (present versus absent) to investigate their effect on managers’ choices in a revenue recognition scenario. The findings show that when performance incentives are present, managers are likely to manage earnings to meet or beat targets regardless of the level of analysts’ coverage. However, I find that incentivized managers are sensitive to the level of analysts’ coverage, which improves the quality of the information environment and disciplines managerial behavior. This suggests that analysts’ coverage can serve as a useful monitor of managers’ reporting choices. The findings extend accounting literature by using a controlled experiment to examine earnings management, allowing for more reliable causal connections to be inferred between it and the manipulations. The results are potentially useful to regulators and standard setters as they evaluate alternative devices to curtail managerial opportunism in external reporting. Keywords: earnings management, revenue recognition, analysts’ coverage, equity incentives, monitoring. Data availability: Contact the Author. Electronic copy available at: https://ssrn.com/abstract=3724612 I. INTRODUCTION Firms' desire to align managerial performance and shareholder interests, has led to a substantial increase in their use of performance incentives, notably stock-based and option-based executive compensation (Bergstresser and Philippon 2006). Moreover, evidence suggests that “firm performance is positively related to the share of equity held by managers and the share of manager compensation that is equity-based” (see Bergstresser and Philippon 2006; Mehran 1995). Nevertheless, regulators and other stakeholders have raised concerns that performance incentives lead to earnings manipulations, thereby reducing the quality of financial reports with its associated consequences (see Cheng and Warfield 2005; Levitt 1998; Bergstresser and Philippon 2006). Recent research has examined the potential role of governance, regulatory, and external practices to reduce the incidence of earnings management (Hales, Koka, and Venkataraman 2018). This paper extends this literature by examining whether the extent of analysts’ coverage can curb performance-incentivized managers’ tendency to manage earnings. Potentially providing an alternate tool to mitigate opportunistic managerial behavior. Security analysts are legitimate market players whose actions influence the stock price of companies they follow (Graham, Harvey, and Rajpoal 2005). Meeting or missing analysts’ forecasts is value relevant (Lee 2007; Bartov, Givoly, and Hayn 2002). Thus, managers compensations are affected by whether they meet or fail to meet analysts’ expectations (Matsunaga and Park, 2001). Analysts base their forecasts on publicly available information as well as formal presentations by company executives (Payne and Robb 2000; Lees 1981). They are, therefore, seen as knowledgeable professionals with industry expertise. I posit that analysts’ Electronic copy available at: https://ssrn.com/abstract=3724612 coverage improves the information environment of a company, which reduces the attractiveness of earnings management. Moreover, the increased likelihood that analysts will detect and report earnings management to the public serves as a deterrent, thereby providing a monitoring effect. There is reason to believe that analysts’ coverage improves firms’ information environment. First, they are typically the first to detect corporate fraud and thus serve as whistleblowers (Dyck, Morse, and Zingales 2010). Second, archival research shows that analysts' coverage results in fewer earnings management (Degeorge, Ding, Jeanjean and Stolowy 2005; Irani and Oesch 2013). This study uses an experiment to attempt to replicate this finding. Third, analysts serve as monitors, alongside traditional mechanisms of corporate governance, to deter misreporting and to discipline managerial misbehavior (Yu 2008). Fourth, analysts create a stronger and more transparent operating environment, which enhances the efficiency of a firm’s processes, including their investment decisions (Firth, Xie, and Zhang 2016). I use an experiment to test my prediction that analysts’ coverage curbs the negative tendencies of management to manipulate earnings. Thus, the higher the analysts’ coverage, the fewer managers are inclined to manipulate earnings even if they are incentivized with equity- based compensation. I also test the predictions that high analysts’ coverage is associated with an improvement in the quality of the information environment and managers’ perception of being monitored. In the experiment, I manipulate, between participants, performance incentives (present or absent), and analysts’ coverage (high or low). Across all experimental conditions, participants assume the role of a CFO of a hypothetical firm who is asked to recommend an amount of revenue to be recognized for a percentage of the completed contract, juxtaposed against traditional practices. Participants in the performance incentives condition are told that they are Electronic copy available at: https://ssrn.com/abstract=3724612 entitled to equity-based bonuses each year that are very competitive in the industry; this is in addition to an annual salary. The bonuses are tied to the stock price appreciating, and there are no mandatory stock redemptions. Those in the no performance incentives condition are told that they are entitled to only salary and perks commensurate with being a CFO. Participants in the high analysts’ coverage conditions are told that they are heavily followed by 26 high-quality analysts who develop consensus estimates. This level of coverage is significantly above the industry average; Yu (2008) posit that the average analysts’ coverage is 9.44. On the other hand, participants in the low analysts’ coverage conditions are told that the company is followed by only three analysts, whose coverage is often characterized by high forecast dispersion that rarely results in consensus estimates. All other information is consistent across all conditions. I find that when performance incentives are present, managers are likely to manage earnings to meet or beat targets regardless of the level of analysts’ coverage. I also find that when managers are incentivized, high analysts’ coverage improves the quality of the information environment than when coverage is low. In addition, I find that high analysts’ coverage impacts the decisions of incentivized managers than when coverage is low. Further, I find that managers factor analysts’ consensus estimates in their decision making when analysts’ coverage is high, and when performance incentives are present than when they are low and absent, respectively. Similarly, I find that when analysts’ coverage is high, managers are sensitive to their presence, which disciplines their behavior than when analysts’ coverage is low; this is true when performance incentives are present compared to when absent. Taken together, these results suggest that high analysts’ coverage serves as effective monitors of managerial reporting Electronic copy available at: https://ssrn.com/abstract=3724612 activities than when analysts’ coverage is low because managers are sensitive to the presence of analysts, potentially curbing their opportunistic behaviors. This paper contributes to the accounting literature by using a controlled experiment to examine whether analysts’ coverage serves as a transparency tool that mitigates managers’ opportunistic behavior. The use of an experimental approach extends the earnings management literature, given that most results in the arena have used archival data, where controls are difficult to achieve. The results are useful to regulators, and standard setters whose mandated interest is in the propriety of financial information shared with the investment community (e.g., Pozen 2008; SEC 2003; Koonce 2016). The remainder of this paper is organized as follows: In Section II, I review the literature and develop the research hypothesis. Section III discusses the method followed by the results in section IV. The last section discusses the findings and conclusions. II. THEORY AND HYPOTHESIS DEVELOPMENT Earnings Management Prior research shows that managers have a tendency to manage earnings (see Cheng and Warfield 2005; Richardson, Tuna, and Wu 2002). They do so to maximize their interests from explicit contractual arrangements including bonus plans and debt covenants (e.g., Watts and Zimmerman 1986; Richardson et al. 2002; Dechow and Skinner 2000, Healy and Whalen 1998), antitrust/government regulation, and implicit management compensation contracts, including equity incentives tied to stock price (e.g., Healy and Whalen 1998; Dechow and Skinner 2000). In addition to these incentives based rational for earnings management, concentrated ownership, and smaller boards are associated with more earnings management (Yu 2005). In effect, prior Electronic copy available at: https://ssrn.com/abstract=3724612 research consistently shows that performance incentives, which can align managers’ and investors' interests, also have the effect of leading to opportunistic managerial reporting. Analysts’ Coverage Research shows that analysts’ coverage, among other benefits, serves as a robust monitoring mechanism that results in fewer earnings management. Degeorge et al. (2005) posit that analysts’ coverage serves as transparency tools that check earnings management. The effect was amplified when board monitoring was sufficient, and disclosure requirements were high. I expect that when analysts’ coverage is high for a firm, it will deter managers from managing earnings. Based on this, I formally hypothesize that: H1: When equity incentives are present, analysts’ coverage will reduce the opportunistic behavior of managers to manage earnings to meet or beat analysts’ estimates. Expert external financial analysis, a product of analysts’ coverage, enhances the efficiency of a firm's processes by encouraging a stronger and more transparent operating environment. Firth et al. (2016) argue that analysts are responsible for high-quality information gathering and monitoring, which can improve firms’ investment efficiency hence fewer earnings management. For example, Firth et al. (2016) posit that premium-level analysis correlates strongly with optimal investment decisions, ensuring an equilibrium that prevents terrible decisions such as over or under-investment. Irani and Oesch (2013) also suggest that analysts' coverage reduces information asymmetry by making managers afraid that their inflated earnings would be discovered and revealed to the public. Accordingly, I expect that analysts’ coverage improves the information environment. This expectation is formally hypothesized as: Electronic copy available at: https://ssrn.com/abstract=3724612 H2: When analysts’ coverage is high, the quality of the information environment will be higher than when the coverage is low. Yu (2008) suggests that analysts’ coverage serves as a monitoring tool, thus the more the availability of analysts' forecasts, the less the management of earnings. Analysts’ activities play a significant role in serving as whistleblowers by being the first to detect corporate fraud (Dyck et al. 2010). Based on these findings, I predict that when analysts’ coverage is high, managers will be sensitive to their presence, which will discipline their behavior. Based on this prediction, I formally hypothesize that: H3: When analysts’ coverage is high, the monitoring of the reporting activities of incentivized managers will be better than when coverage is low. III. METHODS Participants The participants are 162 managers recruited by using Qualtrics. The sample consists of 74 females (45.6 percent) and 88 males (54.3 percent); 59 (36.4 percent) participants have at least a Master's degree, and 67 (41.3 percent) possess a Bachelor's degree. All participants are mid-upper level managers with a mean age of 40.98 years (Standard Deviation = 10.05, Minimum = 25.00, Maximum = 75.00, Skewness = 0.71, Kurtosis = 0.02). Participants have 10.48 mean years of experience in finance and accounting (Standard Deviation = 6.00, Skewness = 0.36, Kurtosis = -1.01). Table 1 summarizes these demographic statistics. The skewness and kurtosis suggest that the presence of outliers is unlikely. While incentives are likely to induce When the skewness is greater than 2 in absolute value, the variable is considered to be asymmetrical about its mean. When the kurtosis is greater than or equal to 3, then the variable's distribution is markedly different from a normal distribution in its tendency to produce outliers (Westfall & Henning, 2013). Electronic copy available at: https://ssrn.com/abstract=3724612 participants to exert additional effort on the experimental task (Camerer and Hogarth 1999). I, nevertheless, incentivized each individual with a $30 visa gift card. Institutional Review Board (IRB) approval was obtained for the experiment. All study participants provided informed consent for their participation, and they were assured of confidentiality. This is consistent with the IRB requirements. Task and Materials Participants are asked to assume the role of a CFO of a hypothetical firm — Gates, Inc. Their hypothetical firm is structured exactly as a brick and mortar firm would be, including standard operation procedures, history of gains and losses, and examples of employer/employee interaction. The first part of the case materials provides the participants with information about the company, including comparative financial statements, its operations, the board of directors, and internal control mechanisms. The company’s stock is traded on NYSE. Internal controls are described as typical for a company of its size and standing. The financial information suggests that Gates Inc. has performed well over the past five years, consistently meeting or beating analysts’ earnings per share (EPS) estimates. They are advised that the firm faces competitive pressures from a resurgent market player; as a result, profits of the firm have been affected. They are also informed that they have won a contract which will be 75 percent complete by year-end. The case materials include the current FASB guidance on Revenue Recognition for Contracts (ASU 2014-09; Revenue from Contracts with Customers (Topic 606)). Participants are told that, in accordance with GAAP, while managers can recognize contract revenue when each performance of a deliverable is complete, historically, the managers Electronic copy available at: https://ssrn.com/abstract=3724612 at Gates, Inc. have been very conservative and recognize revenues related to each contract when it is 100 percent complete (all performance deliverables are complete). Table 1 Demographic Descriptive Statistics Panel A: Frequencies Variable N % Sex Male 88 54.32 Female 74 45.68 Missing 0 0 Education High school graduate (high school diploma or equivalent including GED) 6 3.70 Some college but no degree 9 5.56 Associate degree in college (2-year) 11 6.79 Bachelor's degree in college (4-year) 67 41.36 Master's degree 59 36.42 Doctoral degree 5 3.09 Professional degree (JD, MD) 5 3.09 Missing 0 0 Professional Credential If Yes, state credential 52 32.10 No 110 67.90 Missing 0 0 Panel B: Summary Statistics Table for Age and Years of Experience Variable Mean Std. Deviation N Min Max Skewness Kurtosis Age 40.98 10.05 162 25.00 75.00 0.71 0.02 Years of Experience 10.48 6.00 162 0.00 20.00 0.36 -1.01 Participants are then informed that they are in the fourth quarter, for the year-end financial results, and that the results of Gates, Inc., are expected to fall short of analysts' Participants are new in their roles even though they have prior experience. Electronic copy available at: https://ssrn.com/abstract=3724612 consensus EPS estimate for the year-end (by 10 cents). Participants are informed that the analysts’ EPS consensus estimate is $3.15, whereas the internal year-end analysis yields an estimate of $3.05. An inability to meet the $3.15 will probably cause the value of the stock to drop by 7%. Historically, values have dropped by 7% for such forecast misses. The task for participants will be deciding how to deal with the shortfall by recognizing an amount of revenue to meet or beat the analysts' estimate. Participants then respond to additional questions that relate to the information environment, the monitoring effect of analysts’ coverage, and manipulation checks. Appendices 1-3 shows the respective scales used for these tests. Participants are also asked to provide demographic information. Experimental Design I use a 2x2 between-subjects research design to test my predictions. The two independent variables, performance incentives, and analysts’ coverage are manipulated between subjects. I manipulate performance incentives at two levels: present versus absent. Participants in the “present” performance incentives condition are told that they are entitled to equity-based bonuses each year in addition to an annual salary. The bonuses are tied to Gates Inc.’s equity (stock) price change of +/-5%. Participants are advised that the company has no policy of mandatory stock redemptions; as such meeting, the set target is the way to cashing in on the bonuses. The equity-based bonus serves as the proxy for performance incentives. According to Cornet et al. (2008), “equity-based compensation has been used as a proxy for incentives to manage earnings in several papers” (see, for example, Bergstresser and Philippon 2006; Cheng and Warfield 2005; Cohen, Dey, and Lys 2005). Participants in the “absent” performance incentives condition are told that they are only entitled to industry-standard competitive annual Electronic copy available at: https://ssrn.com/abstract=3724612 salary and perks commensurate with being a CFO. There is no provision for bonuses or additional rewards tied to equity performance. I manipulate analysts’ coverage at two levels: high versus low. Participants in the “high” analysts’ coverage condition are told that they are heavily followed by 26 analysts who develop consensus estimates. This level of coverage is significantly above the industry average; Yu (2008) posit that the average analysts’ coverage is 9.44. On the other hand, participants in the “low” analysts’ coverage condition are told that the company is followed by only three analysts, whose coverage is often characterized by high forecast dispersion that rarely results in consensus estimates. I validate my manipulations by conducting a pilot experiment where Amazon Mechanical Turk (MTurk) workers rate the extent of analysts’ coverage and the extent of CFO compensation package influenced by performance incentives on 10-point scales (see, Asay, Elliot and Rennekamp 2017; Henry and Peytcheva 2020, for a similar out of sample manipulation check validations). Appendix 1 shows the scales used in obtaining the manipulation checks. Non- tabulated results from the pilot experiment reveal that both the analysts’ coverage (F (1,158) =15.178, p< 0.001), and the performance incentives (F (1,158) =10.39, p= 0.002) significantly impacts the responses to the manipulation checks. A comparison of the means shows that participants in the high analysts’ coverage condition rated the extent of analysts’ coverage at a mean of 7.60 (2.14) compared to those in low coverage conditions who rated it at a mean of 5.56 (2.60) (t= 3.28, p=.002). Similarly, participants in performance incentives present condition rated the extent of CFOs compensation influenced by performance incentives at a mean of 7.40 (1.80) compared to a mean of 5.50 (2.54) by those in absent performance incentives condition (t= 3.311, p= 0.004). These results suggest that the manipulations were successful Electronic copy available at: https://ssrn.com/abstract=3724612 Dependent Variables There are three dependent variables in this study. First is the amount of revenue recognized. I ask participants to record how much of the $12 million potential revenue they would recognize for the percentage of contract completed by year-end. I assume that no earnings management is consistent with sticking with the company’s conservative posture and historical practices of not recognizing revenue until the contract is completed. Conversely, the degree of earnings management is in increasing the amount of revenue recognized. The second dependent variable is managers’ rating of the quality of the information environment. Participants are asked to rate on a 10-point scale, how much the level of analysts’ coverage enhanced the quality of information available to potential investors to support their investment decisions. High (low) ratings are indicative of managers’ belief that high (low) analysts’ coverage improves (diminishes) the quality of information available to investors. The last dependent variable is managers’ perception of the effect of analysts’ coverage on their reporting activities. I ask participants to rate on a 10-point scale, three questions related to the impact of analysts’ coverage on the revenue recognition decision they make, the impact of analyst consensus estimates on their choices, and how much the presence of analysts enhanced discipline in their reporting activities. I infer that high (low) ratings imply a strong (weak) effect of analysts’ coverage on managerial reporting activities. IV. RESULTS Manipulations Checks After making revenue recognition decisions and answering questions about the monitoring effect of analysts’ coverage and the information environment, I ask participants to Electronic copy available at: https://ssrn.com/abstract=3724612 characterize the extent to which analysts cover the firm and to rate the extent to which the CFO compensation is influenced by performance incentives; on 10-point scales (see appendix 1). I test the effectiveness of the manipulations with a 2 x 2 multiple analysis of variance (MANOVA) that employs analysts’ coverage and performance incentives as independent variables. The dependent variables are the responses to the manipulation checks. Table 2 reveals that both the analysts’ coverage (F (2, 158) = 151.93, p < .001) and the performance incentives (F (2, 158) = 111.73, p < .001) significantly impact the responses to the manipulation checks. Univariate comparison of the means reveals that participants in the high analysts’ coverage condition rate the level of analysts’ coverage at a mean of 7.95 (1.50) compared to a mean of 3.49 (1.66) by those in low coverage condition. Participants in the present performance incentives condition rate the level of CFOs compensation influenced by performance incentives at 7.67 (1.54) compared to a mean of 3.33 (2.02) by those in the no performance incentives condition. Table 3 shows descriptive statistics. Overall, these results provide evidence of the success of the manipulations. Table 2. 2x2 MANOVA for Manipulation Check Questions by Experimental Conditions Pillai’s Residual p- Variable F df Trace df value Analysts’ Coverage 0.66 151.93 2 158 < .001 0.66 Performance Incentives 0.59 111.73 2 158 < .001 0.59 Analysts’ Coverage x Performance 0.004 .304 2 158 .738 0.004 Incentives Table 3. Mean (Standard Deviation) for Manipulation Checks Analysts' Coverage Manipulation Checks Low High Total Performance Incentives Absent 3.03 (1.96) n=32 3.65 (2.06) n=31 3.33 (2.02) n=63 Present 4.00 (1.51) n=31 8.15(1.49) n=68 7.67 (1.54) n=99 Total 3.49 (1.66) n=63 7.95 (1.50) n=99 Electronic copy available at: https://ssrn.com/abstract=3724612 Tests of Hypotheses Hypothesis 1 H1 states that when equity incentives are present, analysts’ coverage will reduce the opportunistic behavior of managers to manage earnings to meet or beat analysts' estimates. I test the hypothesis with an analysis of variance (ANOVA) that employs analysts’ coverage and performance incentives as two levels between subjects' factors. The dependent variable is the amount of revenue recognized. Table 4 presents the ANOVA model, which shows that there is a significant main effect for performance incentives (F (1, 158) = 6.77, p = .010). The reported mean (standard deviation) = 7.44 (3.13) of the present condition is significantly larger than the absent condition 5.80(3.28). Contrary to expectations, the interaction effect is not significant (F (1, 158) = 2.50, p = .116). The main effect for analysts’ coverage is also not significant (F (1, 158) = 1.33, p = .251). Thus, hypothesis 1 is not supported. Table 5 presents the descriptive statistics which show that managers in high incentive conditions are likely to report a higher revenue than those in the low incentive conditions, regardless of the level of analysts’ coverage. Table 4. Analysis of Variance for Revenue Recognized Term SS df F p-value η 2 Analysts’ Coverage 13.28 1 1.33 0.251 0 Performance Incentives 67.66 1 6.77 0.01 0 Analysts’ Coverage x Performance Incentives 25 1 2.50 0.116 0 Residuals 1579.72 158 Table 5. Mean (Standard Deviation) for Revenue Recognized Analysts' Coverage Revenue Recognized Low High Total Performance Incentives Absent 5.92 (3.23) n=32 5.69 (3.39) n=31 5.80 (3.28) n=63 Present 6.45 (3.08) n=31 7.89 (3.06) n=68 7.44 (3.13) n=99 Total 6.18 (3.14) n=63 7.20 (3.31) n=99 Electronic copy available at: https://ssrn.com/abstract=3724612 Hypothesis 2 H2 proposes that when analysts’ coverage is high, the quality of the information environment will be higher than when the coverage is low. I test the hypothesis with an analysis of variance (ANOVA) that employs analysts’ coverage and performance incentives as two levels between subjects' factors. The dependent variable is the managers' ratings of the quality of the information environment. Table 6 presents the results of the ANOVA model. The main effect for analysts’ coverage is significant (F (1, 158) = 25.39, p < .001). The main effect for performance incentives is not significant (F (1, 158) = 0.29, p = .589). However, because of the significant interaction described below, the main effects cannot be given a normal interpretation. The interaction effect between analysts’ coverage and performance incentives is significant (F (1, 158) = 9.96, p = .002). To further understand the nature of the significant interaction, I conduct a simple effects analysis. Table 6. Analysis of Variance for Information Environment Quality Term SS df F p-value η 2 Analysts’ Coverage 87.42 1 25.39 < .001 0.1 Performance Incentives 1.01 1 0.29 0.589 0 Analysts’ Coverage x Performance Incentives 34.29 1 9.96 0.002 0.1 Residuals 544.08 158 Table 7 presents results from the tests of the simple effects, which reveal that for incentivized managers, high analysts’ coverage led to a stronger information environment rating than low analysts’ coverage (F (1, 158) = 8.04, p = .005). For non-incentivized managers, analysts’ coverage had no effect (F (1, 158) = 2.97, p = .087). A comparison of the means indicates that participants in high analysts’ coverage and performance incentives present conditions reported a mean (standard deviation) of 7.72 (1.68) compared to low analysts’ coverage and performance Electronic copy available at: https://ssrn.com/abstract=3724612 incentives present conditions 5.19 (2.04). Participants in high analysts’ coverage and performance incentives absent conditions reported 6.58 (1.78) compared to those in low analysts’ coverage and performance incentives absent conditions 6.00 (2.08). These results suggest that when analysts’ coverage is high, the quality of the information environment is more robust than when coverage is low. The robustness is likely amplified when performance incentives are present compared to when absent. However, when analysts’ coverage is low, performance incentives likely reduces the quality of the information environment compared to when absent. Figure1 illustrates this observation. Esti mated Margi nal Means of Informati on Envi ronment Pe rformance Ince ntive s Condition 8.0 Absent Present 7.5 7.0 6.5 6.0 5.5 High Low Analys ts ' Cove rage Condition Figure 1. Information Environment Quality Means by factors levels of Analysts' Coverage and Performance incentives Electronic copy available at: https://ssrn.com/abstract=3724612 Estimated Marginal Means Table 8 presents the descriptive statistics, which shows that when analysts’ coverage is high, managers are more likely influenced to improve the information environment than when coverage is low. This result provides support for hypothesis 2. Table 7. Simple Effect Test for Information Environment Quality Analysts' Coverage SS df F p-value η 2 High Contrast 27.67 1 8.04 0.005 0.048 Error 544.08 158 Low Contrast 10.24 1 2.97 0.087 0.018 Error 544.08 158 Table 8. Mean (Standard Deviation) for Information Environment Quality Analyst Coverage Information Environment Low High Total Performance Incentives Absent 6.00 (2.08) n=32 6.58 (1.78) n=31 6.29 (1.95) n=63 Present 5.19 (2.04) n=31 7.72(1.68) n=68 6.93 (2.14) n=99 Total 5.60 (2.08) n=63 7.36 (1.79) n=99 Hypothesis 3 H3 predicts that when analysts’ coverage is high, the reporting activities of incentivized managers are effectively monitored compared to when coverage is low. To test this hypothesis, I ask participants to rate (1) the impact of analysts’ coverage on the revenue recognition decision they had made, (2) the impact of analysts’ consensus estimates on their choices, and (3) how much the presence of analysts enhanced discipline in their reporting activities. The dependent variables are the managers’ ratings to these questions. To obtain an overall score for the managers’ ratings, I calculate the aggregate mean score for all three questions. This score represents the overall monitoring effect of analysts’ coverage on managerial, financial reporting Each F tests the simple effects of Performance Incentives Condition within each level combination of the other effects shown. These tests are based on the linearly independent pairwise comparisons among the estimated marginal means. Electronic copy available at: https://ssrn.com/abstract=3724612 decisions. I test the hypothesis by first conducting a multiple analysis of variance (MANOVA) that employs analysts’ coverage and performance incentives as independent variables. Table 9 presents the MANOVA results which reveal that the main effects for analysts’ coverage (F (3, 156) = 8.51, p < .001) and performance incentives (F (3, 156) = 9.45, p < .001) are significant . Similarly, the interaction effect between analysts’ coverage and performance incentives (F (3, 156) = 4.25, p = .006) is significant. Table 9. 2x2 MANOVA for Overall Monitoring Effect Pillai’s Residual Variable F df p η 2 Trace df Analysts’ Coverage 0.14 8.51 3 156 < .001 0.14 Performance Incentives 0.15 9.45 3 156 < .001 0.15 Analysts’ Coverage x Performance 0.08 4.25 3 156 0.006 0.08 Incentives Given that the main and interaction effects from the MANOVA are all significant, I proceed to conduct individual ANOVAs on the three questions to determine the nature of these effects. Analysts’ Impact on Managerial Decision-making: The first question for the monitoring effect of analysts relates to whether analysts’ coverage impacts managerial decision making. An ANOVA that employs analysts’ coverage and performance incentives as two levels between subjects' factors is conducted. The dependent variable is managers’ rating of the impact of analysts' coverage on their managerial decisions. Table 10 presents the ANOVA model. Univariate comparison of means shows that participants in high analysts’ coverage conditions reported a mean (standard deviation of 6.39 (1.99) compared to low conditions 5.08 (1.91) p=0.005. Participants in performance incentives present conditions reported 6.97(1.93) compared to absent conditions 5.02 (1.95) p=0.005. Electronic copy available at: https://ssrn.com/abstract=3724612 The main effect for analysts’ coverage is significant (F (1, 158) = 22.84, p < .001). The main effect for performance incentives is also significant (F (1, 158) = 18.41, p < .001). However, because of the significant interaction described below, the main effects cannot be given a normal interpretation. The interaction effect between analysts’ coverage and performance incentives is significant (F (1, 158) = 8.25, p = .005). To further understand the nature of the significant interaction effect, I conduct a simple effects analysis. Table 10. Analysis of Variance for Analysts’ Impact Term SS df F p-value η 2 Analysts’ Coverage 90.36 1 22.84 < .001 0.1 Performance Incentives 72.85 1 18.41 < .001 0.1 Analysts’ Coverage x Performance Incentives 32.66 1 8.25 0.005 0.1 Residuals 625.1 158 Table 11 presents results from the tests of the simple effects which reveal that for incentivized managers, high analysts’ coverage led to higher monitoring scores than low analysts’ coverage (F (1, 158) = 30.18, p < .001). For non-incentivized managers, analysts’ coverage had no effect (F (1, 158) = 0.87, p = .351). A comparison of the means indicates that participants in high analysts’ coverage and present performance incentives conditions reported a mean (standard deviation) of 7.53 (1.87) compared to low analysts’ coverage and performance incentives present conditions 5.00 (2.14). Participants in high analysts’ coverage and performance incentives absent conditions reported a mean (standard deviation) of 5.16 (2.18) compared to low analysts’ coverage and performance incentives absent conditions 4.53 (1.88). These results suggest that the impact of analysts’ coverage on managerial decisions is stronger when analysts’ coverage is high, and performance incentives are present. Electronic copy available at: https://ssrn.com/abstract=3724612 Table 12 presents descriptive statistics which show that incentivized managers are sensitive to analysts' coverage, which likely impacts their financial reporting decisions. This result supports hypothesis 3. Table 11. Simple Effect Test for Analysts' Impact Analysts' Coverage SS df F p-value η 2 High Contrast 119.41 1 30.18 0.000 0.2 Error 625.10 158 Low Contrast 3.46 1 0.87 0.351 0.0 Error 625.10 158 Table 12. Mean (Standard Deviation) for Analysts’ Impact Analyst Coverage Analysts’ Impact Low High Total Performance Incentives Absent 4.53 (1.88) n=32 5.16 (2.18) n=31 4.84 (2.04) n=63 Present 5.00 (2.14) n=31 7.53(1.87) n=68 6.74(2.28) n=99 Total 4.76 (2.01) n=63 6.79 (2.25) n=99 Impact of Analysts’ Consensus Estimates: The second question for the monitoring effect of analysts is whether managers factored analysts’ consensus estimates in their decision making. An ANOVA that employs analysts’ coverage and performance incentives as two levels between subjects' factors is conducted. The dependent variable is managers' rating of the effect of analysts' consensus estimates on their managerial decisions. Table 13 presents the ANOVA model which shows that the main effect for analysts’ coverage is significant (F (1, 158) = 14.86, p < .001), indicating that the analysts’ consensus estimate effect is higher when analysts’ coverage is high (mean (standard deviation) = 6.93(2.40) compared to when analysts’ coverage is low 5.11(2.20)). The main effect for performance incentives is also significant (F (1, 158) = 24.43, p < .001), indicating that the analysts’ consensus estimate effect is higher when performance incentives are present (mean (standard deviation) = 7.05(2.09) compared to when performance incentives are absent 4.92(2.50)). The interaction effect between analysts’ coverage and performance incentives is not significant (F (1, 158) = 2.66, p = .105). Electronic copy available at: https://ssrn.com/abstract=3724612 Table 14 presents the descriptive statistics which provide evidence that managers in high analysts’ coverage conditions, as well as those in performance incentives present conditions, are more likely to factor analysts' consensus estimates in their decisions. This result supports hypothesis 3 Table 13. Analysis of Variance for Analysts’ Consensus Estimate Term SS df F p-value η 2 Analysts’ Coverage 68.13 1 14.86 < .001 0.1 Performance Incentives 112.01 1 24.43 < .001 0.1 Analysts’ Coverage x Performance Incentives 12.2 1 2.66 0.105 0 Residuals 724.35 158 Table 14. Mean (Standard Deviation) for Analysts' Consensus Estimates Analyst Coverage Analysts' Consensus Low High Total Estimates Performance Incentives Absent 4.53 (2.27) n=32 5.32 (2.69) n=31 4.92(2.50) n=63 Present 5.71 (1.99) n=31 7.66(1.85) n=68 7.05(2.09) n=99 Total 5.11 (2.20) n=63 6.93(2.40) n=99 Disciplining managers: The last question relating to the monitoring effect is whether analysts’ coverage instilled discipline in managers. This is tested using an ANOVA that employs analysts’ coverage and performance incentives as two levels between subjects' factors. The dependent variable is the managers' rating of the level of discipline analysts’ coverage instilled in managers. Table 15 shows the ANOVA model. The main effect for analysts’ coverage is significant (F (1, 158) = 15.82, p < .001) indicating that discipline among managers is higher when analysts’ coverage is high (mean (standard deviation) = 7.08 (2.20) compared to when analysts’ coverage is low 5.37 (2.37)). The main effect for performance incentives is also significant (F (1, 158) = 18.82, p < .001) Electronic copy available at: https://ssrn.com/abstract=3724612 indicating that discipline among managers is also higher when performance incentives are present (mean (standard deviation) = 7.12 (2.11) compared to when performance incentives are absent 5.30 (2.45)). The interaction effect between analysts’ coverage and performance incentives is not significant (F (1, 158) = 0.04, p = .833). Table 16 presents descriptive statistics that provide evidence that when analysts’ coverage is high, and when performance incentives are present, managers are more likely to be disciplined than when analysts’ coverage is low, and performance incentives are absent. This result provides support for hypothesis 3. Table 15. Analysis of Variance for Managerial Discipline Term SS df F p-value η 2 Analysts’ Coverage 73.66 1 15.82 < .001 0.1 Performance Incentives 87.59 1 18.82 < .001 0.1 Analysts’ Coverage x Performance Incentives 0.21 1 0.04 0.833 0 Residuals 735.45 158 Table 16. Mean (Standard Deviation) for Managerial Discipline Analyst Coverage Managerial Discipline Low High Total Performance Incentives Absent 4.56 (2.45) n=32 6.06 (2.25) n=31 5.30(2.45) n=63 Present 6.19 (2.01) n=31 7.54(2.03) n=68 7.12(2.11) n=99 Total 5.37 (2.37) n=63 7.08(2.20) n=99 V. DISCUSSION AND CONCLUSION The experimental results show that when performance incentives are present, managers are likely to manage earnings to meet or beat targets. This is in spite of the level of monitoring of analysts, contrary to my prediction that analysts’ coverage will curb such opportunistic behavior. Electronic copy available at: https://ssrn.com/abstract=3724612 Nevertheless, managers believe analysts’ coverage significantly impacts their financial decisions, even though that impact does not seem to translate into affecting their opportunistic behavior. This phenomenon could be attributable to the effect of performance incentives being more potent than the impact of analysts’ monitoring when it comes to earnings management. The findings reveal that when analysts’ coverage is high, the presence of performance incentives improves the quality of the information environment. Conversely, when analysts’ coverage is low, the presence of performance incentives potentially reduces the quality of the information environment. This supports the assertion that analysts’ coverage provides a robust and more transparent information environment that supports investment decisions (Firth et al. 2016) and reduce information asymmetry (Irani and Oesch 2013). This study also shows that managers are sensitive to the presence of analysts, even when performance incentives are present, suggesting that analysts potentially serve as useful monitors of the reporting activities of managers. Specifically, managers are found to rate the impact of analysts’ coverage highly as well as factor analysts’ consensus estimates highly in their decision making; this is confirmed when both analysts’ coverage is high, and performance incentives are present. Further, I find that when analysts’ coverage is high, performance incentives produced stronger discipline among managers. Taken together, these suggest that high analysts’ coverage serves as effective monitors of managerial reporting activities, potentially mitigating their opportunistic behaviors- this result buttresses prior research findings that analysts' coverage disciplines managerial misbehavior (Yu, 2008). These findings enhance the understanding of managerial behavior in the face of incentives and the role of analysts in curbing earnings management. The results will be useful to regulators and standard setters as well as parties interested in corporate governance and the Electronic copy available at: https://ssrn.com/abstract=3724612 quality of financial reporting (e.g., external auditors, audit committee members, regulators). Specifically, for regulators, and standard setters whose mandated interest is in the propriety of financial information shared with the investment community (e.g., Pozen 2008; SEC 2003; Koonce 2016); the findings explicitly imply that analysts are critical to the quality of information environment and serve as useful monitoring tools; hence regulators should encourage a considerable level of analysts’ coverage for firms. This study is among the first papers, if not the first, to test the interaction of analysts’ coverage, performance incentives, and earnings management using an experimental approach. This paper contributes to the accounting literature by using a controlled experiment to examine whether analysts' coverage serves as a transparency tool that mitigates management's opportunistic financial reporting. The use of an experimental approach is a critical addition to the earnings management literature, given that most results in the arena have used archival techniques, where control is difficult to achieve. The findings show that motivated managers are sensitive to analysts’ coverage. Nevertheless, they are likely to engage in opportunistic behavior even in the face of heavy analysts’ monitoring. This suggests that analysts’ coverage alone might not be enough to curb earnings management. Therefore, together with analysts’ coverage, firms should consider other corporate governance mechanisms like board monitoring and strong audit committees to minimize the incidence of earnings management. This is an avenue for further research; indeed, the focus on improving corporate governance in our current dispensation provides numerous avenues for future research. Electronic copy available at: https://ssrn.com/abstract=3724612 REFERENCES Asay, H. S., W. B. Elliott, and K. M. Rennekamp. 2017. Disclosure readability and the sensitivity of investors’ valuation judgments to outside information. The Accounting Review 92 (4): 1–25. Bartov, E., Givoly, D., Hayn, C., 2002. The Rewards to Meeting or Beating Earnings Expectations. Journal of Accounting and Economics 33, 173-204. Bergstresser, D., and Philippon, T. (2006). CEO incentives and earnings management. Journal of Financial Economics, 80(3), 511–529. Camerer, C.F., and Hogarth, R.M. (1999). The Effects of Financial Incentives in Experiments: A Review and Capital-Labor-Production Framework. Journal of Risk and Uncertainty 19, 7–42. Cheng, Q. and T. Warfield. 2005. Equity incentives and earnings management. The Accounting Review 80: 441-476. Clor-Proell, S., and L. Maines. 2014. The impact of recognition versus disclosure on financial information: A preparer's perspective. Journal of Accounting Research 52: 671-701 Cohen, D.A., Dey, A., Lys, T.Z., 2005. Trends in earnings management and informativeness of earnings announcements in the pre- and post-Sarbanes Oxley periods. Available at SSRN Cornett, M., Marcus, A., and Tehranian, H. 2008. Corporate governance and pay-for- performance: The impact of earnings management. Journal of Financial Economics, 87, pp.357– Dechow, P. M., & Skinner, D. J. 2000. Earnings Management: Reconciling the Views of Accounting Academics, Practitioners, and Regulators. SSRN Electronic Journal. Degeorge, F., Ding, Y., Jeanjean, T., and Stolowy, H. 2005. "Does Analyst Following Curb Earnings Management?," HEC Research Papers Series 810, HEC Paris. Dyck, A., Morse, A., Zingales, L. 2010. Who Blows the Whistle on Corporate Fraud? Journal of Finance 65, 2213-2253. Firth, M., Xie, L., and Zhang, Y. (2016). How do analysts' forecast characteristics relate to investment efficiency? Paper. Lingnan University Graham, J. R., C. R. Harvey, and S. Rajgopal. 2005. The economic implications of corporate financial reporting. Journal of Accounting and Economics 40: 3-73. Hales, J., Koka, B., & Venkataraman, S. 2018. Curbing Earnings Management: Experimental Evidence on How Clawback Provisions and Board Monitoring Affect Managers Use of Discretion. SSRN Electronic Journal. Electronic copy available at: https://ssrn.com/abstract=3724612 Healy, P.M and Wahlen, J.M. 1998. A Review of the Earnings Management Literature and Its Implications for Standard Setting. Accounting Horizons: December 1999, Vol. 13, No. 4, pp. 365-383 Henry. E., and Peytcheva, M. 2020. Joint Effects of Boilerplate and Text Markup on the Judgments of Novice and Experienced Users of Financial Information. Behavioral Research in Accounting: Spring 2020, Vol. 32, No. 1, pp. 1-20. https://www.fasb.org/jsp/FASB/Page/ImageBridgePage&cid=1176169257359#section_2 Irani, R., and Oesch, D. 2013. Analyst Coverage and Earnings Management: Quasi-Experimental Evidence. 2013 Accounting Conference. Lee.2007. Earnings Management to Just Meet Analyst Forecast. Thesis. Northwestern University. Lees, F. A. 1981. "Public Disclosure of Corporate Earnings Forecasts." New York: Conference Board. Levitt, A. 1998. The numbers game. The CPA Journal 68 (12): 14-18. Koonce, L., Seybert N, and Smith, J. 2016. Management Speaks, Investors Listen: Are Investors Too Focused on Managerial Disclosures? Journal of Behavioral Finance, 17:1, 31-44 Matsunaga, S. R., Park, C. W. 2001. The Effect of Missing a Quarterly Earnings Benchmark on the CEO's Annual Bonus. The Accounting Review 76, 313-332. Mehran, H., 1995. Executive compensation structure, ownership, and firm performance. Journal of Financial Economics 38, 163–184. Payne, J. L., & Robb, S. W. G. 2000. Earnings Management: The Effect of Ex Ante Earnings Expectations. Journal of Accounting, Auditing & Finance, 15(4), 371–392. Pozen, Robert C. “Final Report of the Advisory Committee on Improvements to Financial Reporting to the United States Securities and Exchange Commission, August 1.” Securities and Exchange Commission Advisory Committee, Washington, DC (2008). Richardson, S., Tuna, A., and Wu, M. 2002. Predicting Earnings Management: The Case of Earnings Restatements. SSRN Electronic Journal. Rose, J., A.M. Rose, C.s. Norman and C.R. Mazza. 2014. Will Disclosure of Friendship Ties between Directors and CEOs Yield Perverse Effects? The Accounting Review 89: 1545-1563. Securities and Exchange Commission (SEC). Interpretation: Commission Guidance Regarding Management’s Discussion and Analysis of Financial Condition and Results of Operations (2003). Electronic copy available at: https://ssrn.com/abstract=3724612 Watts, R. L., and J. L. Zimmerman. 1986. Positive Accounting Theory. Englewood Cliffs, NJ: Prentice Hall Westfall, P. H., & Henning, K. S. S. 2013. Texts in statistical science: Understanding advanced statistical methods. Taylor & Francis. Yu, F., 2005. Corporate governance and earnings management. Unpublished working paper. University of Chicago, Chicago, IL Yu, F. 2008. Analysts’coverage and Earnings Management. Journal of Financial Economics 88, 245-271. APPENDIX 1 Manipulation Checks A. Based on the case materials, to what extent is the CFO's compensation package influenced by performance-based incentives? (1= Low extent; 10=High extent) B. Based on the case materials, how would you characterize the extent to which analysts cover the firm? (1= Low coverage; 10=High coverage) APPENDIX 2 Monitoring Effect 1. What impact is the extent of analysts covering the firm having on your decision? (l= Low impact; 10=High impact) Electronic copy available at: https://ssrn.com/abstract=3724612 2. To what extent will you as a manager of the firm factor analyst consensus estimates in your decision? (1= Low extent; 10= High extent) 3. To what extent will the level of analyst following the firm help instill discipline in managers when making financial reporting decisions? (1= Not at all; 10= Very Much) APPENDIX 3 Information Environment 1. To what extent will the number of analysts following Gates, Inc. enhance the amount of information available to potential investors to support their investment choices in the company? (1=Not at all; 10=Very Much) Electronic copy available at: https://ssrn.com/abstract=3724612 BIOGRAPHICAL SKETCH I am a Certified Public Accountant (CPA) and hold a Doctor of Business Administration degree (Accounting & Finance Major) from the University of Florida. Currently, I serve as a Controller for JPMorgan Chase bank. Electronic copy available at: https://ssrn.com/abstract=3724612

Journal

ARN Conferences & Meetings – SSRN

Published: Jun 18, 2020

Keywords: earnings management, revenue recognition, analysts’ coverage, equity incentives, monitoring

There are no references for this article.