Effect of Real Earnings Management on Firm Performance: Evidence from India

Abstract

This study investigates the effect of real earnings management (REM) on firms’ future performance in India. The sample comprises a balanced panel of 108 non-financial firms belonging to 21 industries (as per the two-digit NIC classification code) from 2006 to 2018. The proxy for REM given by Roychowdhury (2006) is used to measure REM, and the firm’s performance is measured through return on assets (ROA), return on equity (ROE) and price-to-earnings (PE) ratio. While ROA and ROE are measures of accounting performance, PE captures market performance. To explore the impact of REM activities on firms’ future performance, a Generalized method of moments (GMM) estimator is used in a dynamic panel setting. Since the proxy variables for performance is measured on a lead year basis, the analysis is restricted to the period 2006–2017. We also control for firm size, financial stability and growth. Our research reflects that Indian firms usually manipulate earnings by reducing discretionary expenditures. Regression results indicate that REM activities affect both accounting and market performance negatively.

Keywords

Earnings Management Real Earnings Management Firm Performance India

Executive Summary

There are two means of managing earnings that are documented in literature—Accrual-based earnings management (AEM) and real earning management (REM). Although AEM does not have any direct impact on cash flows of the firm, REM affects firms’ cash flow directly, and thereby effecting firm’s operating performance (Cohen & Zarowin, 2010). While REM enhances the current firm performance, scholars have a mixed view about the consequence of REM on future performance.

Roychowdhury considers three different aspects of REM: (1) abnormally low cash flows from operations, (2) abnormally high production costs and (3) abnormally low discretionary expenditures relative to sales. To avoid the problem of double counting (Cohen & Zarowin, 2010, 9; Zang, 2012, 682) along with multicollinearity, this study has dropped the abnormal cash flows as a proxy for REM and have used the remaining two aspects viz. overproduction and discretionary expenditures.

This study aims to address two research questions: 1. What is the type of REM adopted by managers of publicly listed non-financial firms in India? 2. What is the effect of REM on the future performance of these firms?

In this study, accounting performance is captured using return on assets (ROA) and return on equity (ROE), and the market performance by price-to-earnings (PE) ratio. Since we are interested on firm’s future performance, ROA, ROE and PE are measured on a lead year basis. This study used a GMM estimator in a dynamic panel setting to examine the impact of REM activities on firms’ future performance.

Empirical evidence obtained in this study indicates that Indian firms, to report better earnings, prefer reducing discretionary expenditures rather than overproducing. This addresses our first research question. In addition, this study finds that the future performance of the firms is adversely affected by REM, which addresses the second research question.

Introduction

Earnings management (EM) is an activity that simultaneously influences multiple aspects of business, and therefore attracts numerous questions. These questions emerge in the context of the effect of EM on financial reporting, stakeholders’ decision-making ability, managerial incentives and the firm’s future performance. For example, Li et al. (2014), in their study, have shown significant and systemic signs of misreporting of financial information in major emerging countries, which affects the quality of the decision made by stakeholders. In studies concerning the managerial motivation behind EM, existing literature indicates that they come in the form of expectation of bonus incentives (Healy, 1985), avoidance of debt-covenant violations (DeFond & Jiambalvo, 1994) or to meet and/or beat earnings thresholds (Bartov et al., 2002).

REM activities usually involve cost-cutting measures like reducing or postponing discretionary expenditures, overproduction to reduce the cost of goods sold (COGS), delayed/non-commencement of a profitable project to save present cash outlay. These measures are undertaken to either meet the earnings benchmark or beat the analyst’s expectations. Although it may serve the purpose in the short-term, the consequences on firm performance may be adverse in the long run.

Most research in EM and firm performance is carried out in developed nations and are limited to AEM. Real earnings management (REM) practices in emerging economies like India are yet to be fully explored. This study aims to explore the type of REM commonly adopted by Indian firms listed on National Stock Exchange (NSE), and investigate the impact of REM on their future performance.

The rest of the article is divided into four sections. The second section includes a review of related literature and hypothesis development. The third section describes the data set used for the study, along with the detailed methodology. In the fourth section, the results are reported, and finally, the fifth section concludes the article.

Literature Review and Hypothesis Development

Accrual-based Earnings Management (AEM)

Studies on EM have majorly focused on accruals management. Prior studies in this area are on (1) estimation of discretionary accruals as a proxy for EM (Dechow & Dichev, 2002; Dechow et al., 1995; Jones, 1991; Kothari et al., 2005); (2) managerial motivation for EM (Bartov et al., 2002; Charoenwong & Jiraporn, 2009; Chen et al., 2010; DeFond & Jiambalvo, 1994; Habib & Hossain, 2008; Healy, 1985); (3) the role of corporate governance in curbing EM (Iqbal et al., 2016; Jaiswall, 2012; J. Sarkar et al., 2008); and (4) impact on the subsequent market performance of a firm indulged in EM, either before an IPO (DuCharme et al., 2001; Teoh, Wong, et al., 1998) or seasoned equity offering (Rangan, 1998; Teoh, Welch, et al., 1998; Yang et al., 2013).

In addition to the above areas, recent studies have also explored the relationship of accruals management with financially distressed firms (Agrawal & Chatterjee, 2015), the global financial crisis of 2008 (Dimitras et al., 2015; Filip & Raffournier, 2014; Kumar & Vij, 2017) and executives’ compensation (Zhou et al., 2018).

Activities of EM are often driven by managerial pressure. Li et al. (2014) studied EM in BRICs nation and found that Indian listed firms owing to tremendous performance pressure manipulate their earnings more than the unlisted firms. It is thus likely that such stress may tempt managers to indulge in EM practices (Trejo-Pech et al., 2014) by inflating the reported earnings in the current period (Mizik & Jacobson, 2007). In another study, Bartov et al. (2002) found that firms manage earnings to meet or beat analysts’ expectations (MBE). Studies have also shown that firms engage in such practices to report zero/positive profits (Charoenwong & Jiraporn, 2009) or avoid reporting losses/decreased profits (Burgstahler & Dichev, 1997).

AEM and Performance

The relationship between EM and performance is studied in several contexts. Tang & Chang (2015) examined the role of a firm’s governance mechanism on the relationship between EM and firm performance. They found that in the presence of a weak (or strong) corporate governance system, discretionary accruals have a negative (or positive) impact on the future performance of firms.

In studies regarding earnings manipulation around IPO issue, Teoh, Wong, et al. (1998) found evidence of high positive earnings and abnormal accruals in firms around IPO issue to artificially inflate the issue price of the stock (Chen et al., 2013). In another study, Teoh, Wong, et al. (1998) provides evidence of poor post-issue performance for firms that manage accruals in the year of IPO issue. Similar findings are reported in Morsfield & Tan (2006). In Indian context, Das & Jena (2016) reported that Indian firms also engage in earnings manipulation at the time of IPO issue. During their investigation, they found higher EM in firms issuing equity domestically than those going for an overseas issue. The consequence of raising a higher amount through IPO issue than its actual worth is reflected through the post-issue underperformance of such firms.

Real Earnings management

Managers are more likely to manipulate earnings through real actions (like delaying or cutting down on R&D, advertising expenditures or even forgoing projects with profitable opportunities) rather than accounting manipulation (Graham et al., 2005). Cohen and Zarowin (2010) show that firms choose between REM and AEM based on their ability to engage in AEM and its associated cost. Further, stringent accounting regulations, transparent corporate governance policies, etc. have made AEM more challenging. Hence, the managers have either entirely shifted to REM (Cohen et al., 2008) or considered REM as supplementary to AEM (Cohen & Zarowin, 2010).

Kim & Sohn (2013) examined whether the REM activities of a firm influence its cost of equity capital. They conclude that investors consider REM activities detrimental to the information quality of reported earnings and thus demand higher risk premiums over and above the risk premium for AEM. Zhang and He (2013) found that average performing Chinese firms reduce R&D expenditures to meet the profit benchmark. Mizik & Jacobson (2007) found that managers cut down marketing expenditures to inflate current profits, which temporarily appreciate the current stock prices, but is detrimental to firms’ earnings in the long run.

REM and Performance

Some of the existing studies found that the use of REM does not harm firms’ future operating performance, while others reported its negative implications on performance.

Taylor and Xu (2010) found that firms engaged in REM activities manage their operating activities occasionally and do not experience poor operational performance in the future. In yet another study, Gunny (2010) provided evidence of better subsequent operating performance for firms suspected of being engaged in REM activities in comparison to their counterparts that do not engage in REM. In contrast, Tan and Jamal (2006) found operational EM (REM) adopted by high foresighted managers to be detrimental to the long-term growth of the firm. In a more recent study, Wang and Zheng (2020) also found REM to be negatively associated with the future operating performance of the firm.

Cohen and Zarowin (2010) examined the impact of REM and AEM on the post-SEO operating performance. They found that although the subsequent operating performance declines for firms engaging in either form of EM, the impact is more severe for firms using REM. Bhojraj et al. (2009) compared the future performance of firms that engage in EM activities to beat analysts’ forecast with those that did not manage their earnings and missed the analysts’ forecasts. They found that firms managing their accruals or deliberately reducing the discretionary expenditures exhibit temporary appreciation in the stock price. However, in the long run, their performance is relatively lower than firms that maintain discretionary spending at the usual level.

From the above discussion, it is evident that EM through real activities of the firm is relatively unexplored, especially in countries with emerging economies like India. Further, there is no consensus among researchers on the relationship between REM activities of the firm and its future performance. We, therefore, hypothesize that REM has some implications on the future performance of the Indian listed firms.

H₁: REM affects the future performance of the Indian listed firm.

Methodology

Data and Sample

The data is collected from the Prowess database maintained by the Centre for Monitoring Indian Economy (CMIE) spanning over the period 2006–2018. Since the proxy variables for firm performance used in the study is on the lead year basis, the analysis is restricted to 2006–2017. Our sample is drawn out of 1,919 firms listed on NSE as of 31 March 2016. Since, the REM model of Roychowdhury (2006) is applicable for non-financial firms only, 223 firms were excluded leaving 1,696 non-financial NSE listed firms belonging to 63 industries as per two-digit National Industrial Classification (NIC) code. The requirement to compute REM measures is that at least 20 observations from an industry must be available. This resulted in further of 303 firms, leaving behind 1,393 firms across 25 industries. Finally, 1,285 firms were further excluded for non-availability of data for the entire study period (2006–2018). Thus, our final sample comprises a balanced panel of 108¹ non-financial firms belonging to 21 industries as per the two-digit NIC code. The sample selection process is summarized in Table 1, and Table 2 shows the industry composition of sample firms. It can be observed from Table 2 that around 92% of the firms belong to the manufacturing sector, and the remaining 8% belong to other industries.

Table 1.

Sample Selection Process

NSE listed firms as of 31 March 2016	1,919
Less: Firms from financial sectors as per two-digit NIC classification	223
Less: Firms eliminated due to infeasibility to compute REM measure	303
Less: Non-availability of data for the complete study period (2006–2018)	1,285
Final sample firms taken for the study	108
Number of firm year observations (balanced panel 108 firms for 12 years)	1,296

Source: Authors’ Compilation.

Table 2.

Industry Composition of Sample Firms for Panel Analysis

NIC Codes (2-digits)	Industry	No. of Firms
1	Crop and animal production	1
10	Food products	3
11	Beverages	3
13	Textiles	2
14	Wearing apparel	1
17	Paper and paper products	2
20	Chemicals and chemical products	14
21	Pharmaceuticals, medicinal chemical and botanical products	17
22	Rubber and plastics products	7
23	Other non-metallic mineral products	7
24	Basic metals	5
25	Fabricated metal products, except machinery and equipment	5
26	Computer, electronic and optical products	2
27	Electrical equipment	7
28	Machinery and equipment	9
29	Motor vehicles, trailers and semi-trailers	14
34	Diversified	3
41	Construction of buildings	1
42	Civil engineering	2
46	Wholesale trade	2
62	Computer programming, consultancy and related activities	1
Total		108

Source: Authors’ Compilation.

REM Proxy

To measure REM, we adopt the estimation model given by Roychowdhury (2006) used widely in studies (Cohen et al., 2008; Cohen & Zarowin, 2010; Zang, 2012). The estimated model of Roychowdhury considers three different aspects of the deviation from regular business practices of a firm: (1) abnormally low cash flows from operations, (2) abnormally high production costs and (3) abnormally low discretionary expenditures relative to sales. However, the present study does not incorporate abnormal cash flows as a proxy for REM because earlier studies suggest that activities that cause operating cash flows to be abnormally low also lead production costs to be unusually high. Thus, including proxies of operating cash flows and production costs in the same regression would lead to double counting of the REM (Cohen & Zarowin, 2010, 9; Zang, 2012, 682) besides the problem of multi-collinearity.

We model the normal level of production cost (PROD) as a linear function of sales (S) and change in sales (∆S) as given by (Roychowdhury, 2006).

\begin{matrix} \frac{P R O D_{i, t}}{T A_{i, t - 1}} = α_{0} + α_{1} (\frac{1}{T A_{i, t - 1}}) + α_{2} (\frac{S a l e s_{i, t}}{T A_{i, t - 1}}) + \\ α_{3} (\frac{Δ S a l e s_{i, t}}{T A_{i, t - 1}}) + α_{4} (\frac{Δ S a l e s_{i, t - 1}}{T A_{i, t - 1}}) + ε_{i, t} \end{matrix}

(1)

where PROD_t is the sum of the COGS and change in inventory from year t – 1 to t; TA_t_–1 is total assets in year t – 1; Sales_t is the sales in the year t; ΔSales_t is the change in sales from year t – 1 to t; and ΔSales_t_–1 is the change in sales from year t – 2 to t – 1. Eq. 1 is regressed cross-sectionally for each industry year with at least 20 observations from BSE listed companies at the two-digit level of the NIC code. These coefficients are then used in the equation to estimate the normal level of the production costs for the sample firm years, and the abnormal level of the production cost (RM_PROD) is given by the estimated residuals. Higher residuals indicate a more massive amount of inventory overproduction, reducing the COGS to report higher earnings.

Similarly, to compute the abnormal level of discretionary expenditures, we model equation (2) as given by (Roychowdhury, 2006)

\frac{D I S X_{i, t}}{T A_{i, t - 1}} = α_{0} + α_{1} (\frac{1}{T A_{i, t - 1}}) + α_{2} (\frac{S a l e s_{i, t - 1}}{T A_{i, t - 1}}) + ε_{i, t}

(2)

where DISX_i,t discretionary expenditure is the sum of R&D, advertising and selling and general and administrative expenses. Other variables are as used in Eq. 1. The above regression is estimated cross-sectionally for each industry year with at least 20 observations. The estimated residuals from the above regression give the abnormal level of discretionary expenditure. The negative value for residuals would mean abnormal cutting down of discretionary spending to report more significant earnings. For consistency, the residuals are multiplied by ‘−1’ and denoted as RM_DISX, so that higher value of RM_DISX means a greater cut in the discretionary expenditure by the firms to report higher earnings. To capture the total effect of REM, a comprehensive metric is a sum of RM_DISX and RM_PROD (Cohen & Zarowin, 2010; Zang, 2012), denoted by RM_AGG in this study.

Dependent Variables

There are several measures of financial performance that had been used in the literature relating to corporate financial performances. Kyereboah-Coleman (2007) used ROA and ROE. Chakravarthy (1986) used return on sales (ROS), return on total capital (ROTC) and ROE as profitability measure of firms’ performance and market-to-book ratio (MTB) and Altman’s Z factor as a measure of financial market performance and comprehensive measure of firm performance, respectively. In another study, Zeitun and Tian (2007) used ROA, ROE and earnings before interest and tax (EBIT) as accounting measures of performance and PE ratio, MTB, and Tobin’s Q as market-based performance measures.

Prior studies investigating the effect of earnings management on the future performance of firms have used ROA (Cohen & Zarowin, 2010; Gong et al., 2008; Huang & Sun, 2017; Tabassum et al., 2015; Tang & Chang, 2015; Taylor & Xu, 2010); cash flow from operating activities (CFOA) scaled by total assets (Huang & Sun, 2017; Taylor & Xu, 2010); size-adjusted stock returns (SAR) (Taylor & Xu, 2010); Tobin’s Q ratio (Tang & Chang, 2015); and abnormal stock returns (Gong et al., 2008; Louis & Robinson, 2005) as measures of firm performance.

We use three performance measures, viz. ROA, ROE and PE ratio, in this study, to investigate the impact of REM on a firm’s future performance. To explore the effects of REM on firms’ future performance, the following dynamic panel model is used:

\begin{matrix} F i r m P e r f_{i, t + 1} = β_{0} + β_{1} R M_{i, t} + β_{2} S I Z E_{i, t} + \\ β_{3} Z S C O R E_{i, t} + β_{4} G R O W T H_{i, t} \\ + β_{5} F i r m P e r f_{i, t} + β_{6} Y e a r D u m m i e s + ε_{i, t + 1} \end{matrix}

(3)

The key explanatory variable in the model, RM_i,t has three different proxies for REM: (i) RM_PROD, the residual from Eq. 1, (ii) RM_DISX, the residual from Eq. 2 multiplied by minus 1 and (iii) RM_AGG, the comprehensive measure of REM. The dependent variable FirmPerf_i,t₊₁ is measured by two accounting-based firm performance measures viz. ROA and ROE and one market-based firm performance measure viz. PE ratio.

The variables ‘size’, ‘z-score’ and ‘growth’, respectively, are included in the model to control the effect of a firm’s size, financial stability and sales growth. While larger firms, owing to its large pool of resources, market opportunities, superior negotiating powers and economies of large scale induce positive performance, an excessive size can override cost differentials and reduce profitability (Lopez-Valeiras et al., 2016). Thus, size is expected to have either a positive or negative effect on firm performance. Growing sales ensures achievement of the firm’s present financial objectives and scope for potential future expansion, which boosts performance (Kaplan & Norton, 1992, 1993, 1996, as cited in Brush et al., 2000). Therefore, ‘growth’ is expected to affect a firm’s performance positively. Similarly, financially stable firms are expected to exhibit better performance, and hence ‘z-score’ is expected to influence firm performance positively. Both the dependent and control variables used in the model are defined in Table 3.

Table 3.

Variable Definitions

Variables	Definition
Dependent Variables
ROA_t+1	Profit after tax net of prior period and extraordinary transactions divided by average of total assets at the beginning and at the end of the year
ROE_t+1	Profit after tax net of prior period and extraordinary transactions divided by net worth
PE_t+1	Market price of company’s share divided by earnings per share (EPS)
Control Variables
Size	Natural logarithm of total assets
Growth	Difference of sales revenue in the year t and t–1 divided by the sales revenue in the year t–1
Z-score	Altman’s Z-score measured as 3.3 × [EBIT /total assets] + 1.0 × [sales /total assets] + 1.4 × [retained earnings/total assets] + 1.2 × [working capital/total assets]

Source: Authors’ Compilation.

The inclusion of a lagged dependent variable in panel data regression causes endogeneity violating the assumption of the ordinary least square (OLS) method. In the presence of endogeneity, the OLS estimators are biased and inconsistent, and the fixed effect (FE) within estimators suffer from Nickell (1981) downward bias (Baltagi, 2013, 155). To overcome endogeneity in panel data models, Arellano and Bond (1991) propose a first-difference generalized method of moments (FD-GMM) estimator. FD-GMM uses additional lag values of regressors as instruments to give consistent estimates. Blundell and Bond (1998) suggest system GMM estimator as an improvement over the FD-GMM model and consider it as more efficient. Bond et al., (2001) suggest a two-step procedure² for the choice between standard FD-GMM estimator and system GMM estimator to be used for a dynamic panel model. Following Bond et al. (2001), the study found that Blundell–Bond one-step system GMM estimator is more appropriate for estimating Eq. 3

Results and Discussion

The descriptive statistics of all the variables, after winsorizing at 1 and 99 percentiles to reduce the impact of outliers, are reported in Table 4. The mean of RM_AGG is −0.047, implying that, on an average, firms in the sample manage earnings through real activities to the extent of 4.7% of their total assets. The mean values of RM_PROD and RM_DISX are −0.037 and −0.009, respectively. Average z-score value of 5.366 indicates that firms in the sample are financially healthy.³ The average annual sales growth rate is 14.7%. The average return on total assets for firms is 9.5%, while the ROE is 17.8%.

Table 4.

Descriptive Statistics

Variables	N	Mean	Median	Standard Deviation	Minimum	Maximum
ROA	1,296	0.095	0.082	0.062	0.003	0.305
ROE	1,296	0.178	0.164	0.100	0.009	0.504
PE	1,296	17.584	14.435	13.597	2.450	82.210
RM_PROD	1,296	−0.037	−0.031	0.122	−0.436	0.447
RM_DISX	1,296	−0.009	0.002	0.052	−0.212	0.088
RM_AGG	1,296	−0.047	−0.032	0.158	−0.614	0.505
Size	1,296	9.428	9.364	1.409	6.023	13.296
Z-score	1,296	5.366	3.632	4.780	1.208	28.268
Growth	1,296	0.147	0.125	0.187	−0.289	0.870

Source: Authors’ Compilation.

Note: All variables are winsorized at 1 and 99 percentiles to reduce the impact of outliers.

Table 5 portrays pairwise Pearson’s correlation coefficient of all the variables. The bivariate analysis depicts a negative and significant correlation between ROA and REM proxies—RM_PROD, RM_DISX and RM_AGG. Similarly, the other two performance measures (ROE and PE) also exhibit a negative correlation with the REM proxies. There is a significant positive correlation between the dependent variables (ROA and ROE) and the control variables ‘size’, ‘z-score’, ‘growth’, which suggests that large-sized firms, financially healthy firms and firms with growing sales reflect better accounting performance. The dependent variable PE is also positively correlated with ‘size’ and ‘z-score’; however, it is significantly negatively correlated with ‘growth’. This suggests that although large firms and financially well-off firms show better market performance, growth firms report poor market performance. This could be because investors typically prefer to invest in well-established firms to a relatively new entrant struggling to expand its market share. The correlation coefficients are found to be significant for almost every pair of variables at normal levels, but are not too high $(| r | > 0.7)$ among the explanatory variables, and therefore the existence of multicollinearity among the explanatory variables can be safely ruled out (Dormann et al., 2013).

Table 5.

Correlation Matrix

	ROA	ROE	PE	RM_PROD	RM_DISX	RM_AGG	Size	Z-Score
ROE	0.849^***
PE	−0.007	−0.113^***
RM_PROD	−0.381^***	−0.347^***	−0.112^***
RM_DISX	−0.223^***	−0.208^***	−0.213^***	0.537^***
RM_AGG	−0.371^***	−0.339^***	−0.157^***	0.961^***	0.745^***
Size	0.125^***	0.082^***	0.247^***	−0.082^***	−0.211^***	−0.133^***
Z-Score	0.588^***	0.323^***	0.400^***	−0.294^***	−0.227^***	−0.302^***	0.156^***
Growth	0.311^***	0.404^***	−0.063^**	−0.084^***	−0.093^***	−0.097^***	−0.056^**	−0.001

Source: Authors’ Compilation.

Note: *, ** and *** indicate significance at 10%, 5% and 1% level or stronger.

The results of one-step GMM (Blundell–Bond) estimators for the relationship between REM and the three performance measures viz. ROA, ROE and PE are given in Table 6. The t-statistics reported in parentheses are computed using robust heteroscedasticity and autocorrelation corrected (HAC) standard errors.

The results of two critical specification test, which are (i) no second-order serial correlation in the disturbance term and (ii) instruments validity, are consistent with the requirement of (Blundell–Bond) system GMM estimators. The Arellano–Bond test statistic for no first-order serial correlation [AR (1)] is significant, while the test statistic for no second-order serial correlation [AR (2)] is insignificant. Thus, the requirement of no AR (2) is satisfied.

Both Hansen J-statistics and Sargan test for instruments over-identification restriction are also statistically insignificant, suggesting the validity of the instruments used to estimate the one-step system GMM estimators. The coefficients of the lag of dependent variables viz. ROA_t, ROE_t and PE_t are positive and significant at less than 1% in the respective models. Thus, all the requirements of the one-step system GMM estimator are met.

In Table 6, the coefficients of all the three REM proxies are negative and significant for Model 2 through Model 9. The coefficient of RM_PROD in Model 1 with the dependent variable ROA_t₊₁ is negative but insignificant. The results provide evidence of negative impact of REM on future ROA, ROE and PE ratio, suggesting poor performance for the firm in the subsequent period. The finding of negative association between firm’s subsequent year ROA (ROA_t₊₁) and current year REM is consistent with (Gunny, 2005; Tabassum et al., 2015) while it is inconsistent with (Gunny, 2010; Taylor & Xu, 2010). Our finding of adverse impact of firm’s current year REM on its subsequent period ROE (ROE_t₊₁) and PE (PE_t₊₁) ratio is consistent with the study by Tabassum et al. (2015).

Table 6.

Results of One-Step System GMM Estimates

Independent Variable	Expected Sign	Dependent Variable
		Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7	Model 8	Model 9
		ROA _t+1	ROA _t+1	ROA _t+1	ROE_t+1	ROE_t+1	ROE_t+1	PE_t+1	PE _t+1	PE _t+1
Intercept	(+/−)	0.0390^***	0.0420^***	0.0406^***	0.077^***	0.080^***	0.077^***	−2.575	−0.879	−1.944
		(4.31)	(4.71)	(4.40)	(4.62)	(5.01)	(4.76)	(−0.97)	(−0.32)	(−0.74)
RM_PROD	(−)	−0.0216			−0.037^*			−8.509^**
		(−1.62)			(−1.84)			(−2.38)
RM_DISX	(−)		−0.0594^**			−0.126^***			−29.946^***
			(−2.26)			(−2.56)			(−3.39)
RM_AGG	(−)			−0.0211^**			−0.038^**			−8.715^***
				(−2.12)			(−2.41)			(−3.12)
Size	?	−0.0009	−0.0013	−0.001	0.000	−0.001	0.000	1.634^***	1.454^***	1.569^***
		(−0.95)	(−1.37)	(−1.11)	(0.22)	(−0.32)	(0.05)	(5.29)	(4.60)	(5.17)
Z-Score	(+)	0.0019^***	0.0019^***	0.0019^***	0.001^**	0.001^*	0.001^*	0.309^***	0.317^***	0.293^***
		(3.05)	(2.82)	(2.99)	(2.04)	(1.66)	(1.77)	(2.84)	(3.08)	(−2.78)
Growth	(+)	0.0062	0.0045	0.0057	0.026^*	0.023	0.025^*	−3.135	−3.502^*	−3.388^*
		(0.73)	(0.53)	(0.68)	(1.83)	(1.59)	(1.77)	(−1.63)	(−1.84)	(−1.76)
ROA	(+)	0.6407^***	0.6533^***	0.6390^***
		(7.61)	(8.24)	(7.76)
ROE	(+)				0.616^***	0.623^***	0.613^***
					(11.01)	(11.45)	(11.18)
PE	(+)							0.236^***	0.231^***	0.236^***
								(6.70)	(6.56)	(6.72)
Year Dummies		Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N		1296	1296	1296	1296	1296	1296	1296	1296	1296
Group		108	108	108	108	108	108	108	108	108
No. of Instruments		28	28	28	28	28	28	28	28	28
No AR (1) – z-stat		−6.53^***	−6.55^***	−6.55^***	−7.11^***	−7.15^***	−7.12^***	−5.64^***	−5.67^***	−5.63^***
Prob. > z		0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
No AR (2) – z-stat		−0.48	−0.45	−0.48	−0.77	−0.73	−0.76	0.40	0.33	0.41
Prob. > z		0.633	0.65	0.634	0.43	0.468	0.444	0.686	0.741	0.681
Sargan – $χ_{d f}^{2} = 11$		17.23	17.41	17.34	18.02	18.14	18.25	9.14	8.73	8.89
Prob. > χ²		0.101	0.096	0.098	0.081	0.078	0.076	0.609	0.647	0.632
Hansen J-statistic		15.97	16.13	15.99	16.66	16.88	16.73	18.22	16.18	17.47
Prob. >J		0.142	0.136	0.141	0.118	0.112	0.116	0.077	0.135	0.095
F (16, 107)		123.23	116.46	121.09	63.21	63.69	63.97	34.85^***	38.83^***	35.96^***
Prob. > F		0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000

Source: Authors’ Compilation.

Notes: *, * and * indicate significance at 10%, 5% and 1% level or stronger.

Model: $F i r m P e r f_{i, t + 1} = β_{0} + β_{1} R M_{i, t} + β_{2} S I Z E_{i, t} + β_{3} Z S C O R E_{i, t} + β_{4} G R O W T H_{i, t} + β_{5} F i r m P e r f_{i, t} + β_{6} Y e a r D u m m i e s + ε_{i, t + 1}$

Among the control variables, the coefficient of ‘z-score’ is consistently positive and significant throughout all the models. The statistically significant and positive coefficients for ‘z-score’ indicate that financially strong firms tend to perform better in the subsequent period. The coefficients of the control variable ‘size’ are insignificant in the models with accounting performance measures (ROA_t₊₁ and ROE_t₊₁) as the dependent variables. However, ‘size’ is statistically significant and positively associated with the lead year PE ratio implying that large-sized firms depict better market performance in lead years. This positive relationship between ‘size’ and PE ratio in the lead year may be partly attributed to the investors’ confidence in big firms with larger assets as against in small firms with fewer assets. Further, this result is also consistent with the argument that large firms command better market reputation and investors' preference, thus leading to improved market value. ‘Growth’ is found to be positively associated with firms’ lead period accounting performance measures (ROA_t₊₁ and ROE_t₊₁)—although the relationship is significant only in Model 4 and Model 5—while it is negatively associated with (PE_t₊₁).

Overall, the results from Table 6 indicate a statistically significant negative relationship between REM and the lead period firms’ performance measures. All the REM proxies viz. RM_PROD, RM_DISX, and RM_AGG are negatively related to the future performance of firms. This would suggest that firms engaged in real activity manipulations might perform poorly in the future. The results are consistent across all the performance measures and REM proxies and thus robust. Further, since the coefficient of RM_DISX is found to be consistently higher in magnitude than the coefficient of RM_PROD, it may be inferred that REM through discretionary expenditures (i.e., cutting down on R&D expenses, selling and advertising expenses, etc.) has a more detrimental effect on a firm’s future performance in comparison to RM_PROD.

Conclusion

Understanding the implication of REM on firm performance is of particular interest to managers as their future incentives will be in line with the firm’s future performance. It is equally significant for the shareholders and creditors to learn about REM implications to make informed decisions.

This study examines the impact of REM on firms’ future performance using two accounting measures of performance viz. ROA and ROE, and one market measure of performance viz. PE ratio. Consistent with prior studies (Cohen & Zarowin, 2010; Zang, 2012), we concentrate on the abnormal level of production cost and discretionary expenditures to identify the level of REM. REM is seen to have a negative effect on a firm’s future performance. This indicates that the market adjusts the stock price appropriately considering the accounting performance of the firms.

Upon reviewing the literature, we found that past studies are mostly concentrated around AEM in developed nations. There is limited literature on REM in emerging countries. Further, there is no consensus among researchers on the relationship between REM and firms’ future performance. Some authors find REM detrimental to firms’ future performance, while others found that REM does not lead to poor subsequent operating performance; there are still others who found that REM improves future firm performances. Lack of unanimity accompanied by a dearth of literature in the Indian context was the primary motivation behind this study. This is one of the first studies to explore the relationship between REM and future performance in the context of Indian listed firms.

Footnotes

Declaration of Conflicting Interests

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

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

References

Agrawal

, & Chatterjee

(2015). Earnings management and financial distress: Evidence from India. Global Business Review, 16(5S), 140S–154S. https://doi.org/10.1177/0972150915601928

Arellano

, & Bond

(1991). Some tests of specification for panel data: Monte Carlo application and an application to employment equations. Review of Economic Studies, 58(2), 277–297.

Baltagi

B. H.

(2013). Econometric analysis of panel data (5th ed.). John Wiley & Sons Ltd.

Bartov

, Givoly

, & Hayn

(2002). The rewards to meeting or beating earnings expectations. Journal of Accounting and Economics, 33(2), 173–204. https://doi.org/10.1016/S0165-4101(02)00045-9

Bhojraj

, Hribar

, Picconi

, & McInnis

(2009). Making sense of cents: An examination of firms that marginally miss or beat analyst forecasts. Journal of Finance, 64(5), 2361–2388. https://doi.org/10.1111/j.1540-6261.2009.01503.x

Blundell

, & Bond

(1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115–143. https://doi.org/10.1016/S0304-4076(98)00009-8

Bond

, Hoeffler

, & Temple

(2001). GMM estimation of empirical growth models (Economics Papers, Economics Group No. 2001-W21). CEPR Discussion Paper. https://doi.org/10.1016/S0165-1765(02)00083-6

Brush

T. H.

, Bromiley

, & Hendrickx

(2000). The free cash flow hypothesis for sales growth and firm performance. Strategic Management Journal, 21(4), 455–472. https://doi.org/10.1002/(SICI)1097-0266(200004)21:4<455::AID-SMJ83>3.0.CO;2-P

Burgstahler

, & Dichev

(1997). Earnings management to avoid earnings decreases and losses. Journal of Accounting and Economics, 24(1), 99–126.

10.

Chakravarthy

B. S.

(1986). Measuring strategic performance. Strategic Management Journal, 7(5), 437–458.

11.

Charoenwong

, & Jiraporn

(2009). Earnings management to exceed thresholds: Evidence from Singapore and Thailand. Journal of Multinational Financial Management, 19(3), 221–236. https://doi.org/10.1016/j.mulfin.2008.12.001

12.

Chen

S. K.

, Lin

B. X.

, Wang

, & Wu

(2010). The frequency and magnitude of earnings management: Time-series and multi-threshold comparisons. International Review of Economics and Finance, 19(4), 671–685. https://doi.org/10.1016/j.iref.2010.02.005

13.

Chen

S. S.

, Lin

W. C.

, Chang

S. C.

, & Lin

C. Y.

(2013). Information uncertainty, earnings management, and long-run stock performance following initial public offerings. Journal of Business Finance & Accounting, 40(9–10), 1126–1154. https://doi.org/10.1111/jbfa.12046

14.

Cohen

D. A.

, Dey

, & Lys

T. Z.

(2008). Real and accrual-based earnings management in the pre-and post-Sarbanes-Oxley periods. The Accounting Review, 83(3), 757–787.

15.

Cohen

D. A.

, & Zarowin

(2010). Accrual-based and real earnings management activities around seasoned equity offerings. Journal of Accounting and Economics, 50(1), 2–19. https://doi.org/10.1016/j.jacceco.2010.01.002

16.

Das

R. C.

, & Jena

S. K.

(2016). Earnings management and equity issue firms: A study in Indian context. Jindal Journal of Business Research, 5(1), 64–85. https://doi.org/10.1177/2278682116680926

17.

Dechow

P. M.

, & Dichev

I. D.

(2002). The quality of accruals and earnings: The role of accruals estimation errors. The Accounting Review, 77(s-1), 35–59. https://doi.org/10.2308/accr.2002.77.s-1.61

18.

Dechow

P. M.

, Sloan

R. G.

, & Sweeney

A. P.

(1995). Detecting earnings management. The Accounting Review, 70(2), 193–225. https://doi.org/10.2307/248303

19.

DeFond

M. L.

, & Jiambalvo

(1994). Debt covenant violation and manipulation of accruals. Journal of Accounting and Economics, 17(1–2), 145–176. https://doi.org/10.1016/0165-4101(94)90008-6

20.

Dimitras

A. I.

, Kyriakou

M. I.

, & Iatridis

(2015). Financial crisis, GDP variation and earnings management in Europe. Research in International Business and Finance, 34, 338–354. https://doi.org/10.1016/j.ribaf.2015.02.017

21.

Dormann

C. F.

, Elith

, Bacher

. (2013). Collinearity: A review of methods to deal with it and a simulation study evaluating their performance. Ecography, 36(1), 27–46. https://doi.org/10.1111/j.1600-0587.2012.07348.x

22.

DuCharme

L. L.

, Malatesta

P. H.

, & Sefcik

S. E.

(2001). Earnings management: IPO valuation and subsequent performance. Journal of Accounting, Auditing & Finance, 16(4), 369–396. https://doi.org/10.1177/0148558X0101600409

23.

Filip

, & Raffournier

(2014). Financial crisis and earnings management: The European evidence. The International Journal of Accounting, 49(4), 455–478. https://doi.org/10.1016/j.intacc.2014.10.004

24.

Gong

, Louis

, & Sun

A. X.

(2008). Earnings management and firm performance following open-market repurchases. The Journal of Finance, 63(2), 947–986. https://doi.org/10.1111/j.1540-6261.2008.01336.x

25.

Graham

J. R.

, Harvey

C. R.

, & Rajgopal

(2005). The economic implications of corporate financial reporting. Journal of Accounting and Economics, 40(1–3), 3–73. https://doi.org/10.1016/j.jacceco.2005.01.002

26.

Gunny

K. A.

(2005). What are the consequences of real earnings management? Working Paper, Haas School of Business, University of California. http://w4.stern.nyu.edu/accounting/docs/speaker_papers/spring2005/Gunny_paper.pdf

27.

Gunny

K. A.

(2010). The relation between earnings management using real activities manipulation and future performance: Evidence from meeting earnings benchmarks. Contemporary Accounting Research, 27(3), 855–888. https://doi.org/10.1111/j.1911-3846.2010.01029.x

28.

Habib

, & Hossain

(2008). Do managers manage earnings to “just meet or beat” analyst forecasts? Evidence from Australia. Journal of International Accounting, Auditing and Taxation, 17(2), 79–91. https://doi.org/10.1016/j.intaccaudtax.2008.07.004

29.

Healy

P. M.

(1985). The effect of bonus schemes on accounting decisions. Journal of Accounting and Economics, 7(1–3), 85–107. https://doi.org/10.1016/0165-4101(85)90029-1

30.

Huang

, & Sun

(2017). Managerial ability and real earnings management. Advances in Accounting, 39, 91–104. https://doi.org/10.1016/j.adiac.2017.08.003

31.

Iqbal

, Zhang

, & Jebran

(2016). Corporate governance and earnings management: A case of Karachi Stock Exchange listed companies. Indian Journal of Corporate Governance, 8(2), 103–118. https://doi.org/10.1177/0974686215602367

32.

Jaiswall

, & Banerjee

(2012, July 2012). Exploring the relationship between earnings management and corporate governance characteristics in the Indian context [Working Paper No. 704]. Indian Institute of Management, Calcutta. Retrieved from https://stafflive.iimcal.ac.in/sites/all/files/pdfs/wps_704.pdf

33.

Jones

J. J.

(1991). Earnings management during import relief investigations. Journal of Accounting Research, 29(2), 193–228. https://doi.org/10.2307/2491047

34.

Kim

J.-B.

, & Sohn

B. C.

(2013). Real earnings management and cost of capital. Journal of Accounting and Public Policy, 32(6), 518–543. https://doi.org/10.1016/j.jaccpubpol.2013.08.002

35.

Kothari

S. P.

, Leone

A. J.

, & Wasley

C. E.

(2005). Performance matched discretionary accrual measures. Journal of Accounting and Economics, 39(1), 163–197. https://doi.org/10.1016/j.jacceco.2004.11.002

36.

Kumar

, & Vij

(2017). Earnings management and financial crisis: Evidence from India. Journal of International Business and Economy, 18(2), 84–101. https://doi.org/10.51240/jibe.

37.

Kyereboah-Coleman

(2007). The impact of capital structure on the performance of microfinance institutions. Journal of Risk Finance, 8(1), 56–71. https://doi.org/10.1108/15265940710721082

38.

, Park

S. H.

, & Bao

R. S.

(2014). How much can we trust the financial report?: Earnings management in emerging economies. International Journal of Emerging Markets, 9(1), 33–53. https://doi.org/10.1108/IJoEM-09-2013-0144

39.

Lopez-Valeiras

, Gomez-Conde

, & Fernandez-Rodriguez

(2016). Firm size and financial performance: Intermediate effects of indebtedness. Agribusiness, 32(4), 454–465. https://doi.org/10.1002/agr

40.

Louis

, & Robinson

(2005). Do managers credibly use accruals to signal private information? Evidence from the pricing of discretionary accruals around stock splits. Journal of Accounting and Economics, 39(2), 361–380. https://doi.org/10.1016/J.JACCECO.2004.07.004

41.

Mizik

, & Jacobson

(2007). Myopic marketing management: Evidence of the phenomenon and its long-term performance consequences in the SEO context. Marketing Science, 26(3), 361–379. https://doi.org/10.1287/mksc.1060.0261

42.

Morsfield

S. G.

, & Tan

C. E. L.

(2006). Do venture capitalists influence the decision to manage earnings in initial public offerings? The Accounting Review, 81(5), 1119–1150. https://doi.org/10.2308/accr.2006.81.5.1119

43.

Nickell

(1981). Biases in models with fixed effects. Econometrica, 49(6), 1417–1426.

44.

Rangan

(1998). Earnings management and the performance of seasoned equity offerings. Journal of Financial Economics, 50(1), 101–122. https://doi.org/10.1016/S0304-405X(98)00033-6

45.

Roychowdhury

(2006). Earnings management through real activities manipulation. Journal of Accounting and Economics, 42(3), 335–370. https://doi.org/10.1016/j.jacceco.2006.01.002

46.

Sarkar

, Sarkar

, & Sen

(2008). Board of Directors and opportunistic earnings management : Evidence from India. Journal of Accounting, Auditing & Finance, 23(4), 517–551. https://doi.org/10.1177/0148558X0802300405

47.

Tabassum

, Kaleem

, & Nazir

M. S.

(2015). Real earnings management and future performance. Global Business Review, 16(1), 21–34. https://doi.org/10.1177/0972150914553505

48.

Tan

H. C.

, & Jamal

(2006). Effect of accounting discretion on ability of managers to smooth earnings. Journal of Accounting and Public Policy, 25(5), 554–573. https://doi.org/10.1016/j.jaccpubpol.2006.07.007

49.

Tang

H. W.

, & Chang

C. C.

(2015). Does corporate governance affect the relationship between earnings management and firm performance? An endogenous switching regression model. Review of Quantitative Finance and Accounting, 45(1), 33–58. https://doi.org/10.1007/s11156-013-0427-z

50.

Taylor

G. K.

, & Xu

R. Z.

(2010). Consequences of real earnings management on subsequent operating performance. Research in Accounting Regulation, 22(2), 128–132. https://doi.org/10.1016/j.racreg.2010.07.008

51.

Teoh

S. H.

, Welch

, & Wong

T. J.

(1998). Earnings management and the underperformance of seasoned equity offerings. Journal of Financial Economics, 50(1), 63–99. https://doi.org/10.1016/S0304-405X(98)00032-4

52.

Teoh

S. H.

, Wong

T. J.

, & Rao

G. R.

(1998). Are accruals during initial public offerings opportunistic? Review of Accounting Studies, 3(1–2), 175–208. https://doi.org/10.1023/a:1009688619882

53.

Trejo-Pech

C. J. O.

, Weldon

R. N.

, & Gunderson

M. A.

(2014). Earnings management through specific accruals and discretionary expenses: Evidence from U.S. agribusiness firms. Canadian Journal of Agricultural Economics, 64(1), 89–118. https://doi.org/10.1111/cjag.12063

54.

Wang

, & Zheng

(2020). Real earnings manipulation and future performance: A revisit using quarterly data of firms with debt covenants. Review of Financial Economics, 38(1), 76–96. https://doi.org/10.1002/rfe.1070

55.

Yang

Ben, Hsu

, & Yang

T. H.

(2013). Earnings management, institutional shareholdings, and performance of SEO firms. Managerial Finance, 39(6), 528–549. https://doi.org/10.1108/03074351311322834

56.

Zang

A. Y.

(2012). Evidence on the trade-off between real activities manipulation and accrual-based earnings management. The Accounting Review, 87(2), 675–703. https://doi.org/10.2308/accr-10196

57.

Zeitun

, & Tian

G. G.

(2007). Capital structure and corporate performance: Evidence from Jordan. Australasian Accounting, Business and Finance Journal, 1(4), 40–61. https://doi.org/10.2139/ssrn.2496174

58.

Zhang

, & He

(2013). R&D-based earnings management, accounting performance and market return. Chinese Management Studies, 7(4), 572–585. https://doi.org/10.1108/CMS-09-2013-0176

59.

Zhou

, Wang

, Zhang

, & An

(2018). The impacts of accrual-based and real earnings management on executive compensation: evidence from Chinese public firms in the private sector. Asia-Pacific Journal of Accounting and Economics, 25(1–2), 128–144. https://doi.org/10.1080/16081625.2016.1222296