Dividend Policy and Earnings Management: Based on Discretionary Accruals and Real Earnings Management

Transcription

1 , pp Dividend Policy and Earnings Management: Based on Discretionary Accruals and Real Earnings Management 1 Chae Chang Im, 2 Jeong Ho Kim, 3 Min Kyung Choi 1 Ph. D. Student, College of Business Administration, Hankuk University of Foreign Studies, Seoul, Korea ( imcee@naver.com) 2 Assistant Professor, Department of taxation and accounting, Silla University, Busan, Korea, taxskim@naver.com 3 M.D. Student, College of Business Administration, Hankuk University of Foreign Studies, Seoul, Korea ( jamk.choi@naver.com) Abstract. In this paper, we attempt to identify the effect of high dividend payouts on earnings management by examining discretionary accruals and real earnings management. Firms experience pressure to payout dividends from institutional and foreign investors who pursue investments in firms with a high dividend policy. In such cases, managers execute high dividend payouts to meet expectations. However, in the face of lower income or losses, it will be more difficult for managers to justify the high payout of dividends to stakeholders. Therefore, firms will prefer to payout dividends when there is sufficient income. In turn, managers are likely to depend on earnings management as a means to increase the income from which dividends are paid. Under this rationale, we study the effect of high dividend payouts on earnings management. Key Words: Dividend Payout Ratio, Discretionary Accruals, Real Earnings Management 1 Introduction Firms around the world establish various programs for risk management, and this has been highlighted after the financial crisis in Before the 1990s, Korean firms implemented policies that focused on new investment opportunities to stimulate growth, but this trend has rapidly changed since the Asian financial crisis when many firms were completely disassembled and liquidated. Therefore, during the past 10 years, the cash holdings of large Korean firms have been maintained at a level higher than the prior decade. However, this trend is slowly faltering as various papers have suggested that dividends are a primary determinant for a firm s future growth and value. In particular, Baker and Powell (2000) and Baker et al. (2001) assert that strong intentions to maintain a sustainable dividend payout is the most important factor for a firm s ISSN: ASTL Copyright 2015 SERSC

2 dividend policy and report that most firms exert efforts to provide consistent dividend payouts. This shows that a decrease in dividends can lead to negative consequences for the market value of the firm. Thus, we attempt to identify the effect of high dividend payouts on earnings management by examining discretionary accruals and real earnings management. Firms experience pressure to payout dividends from institutional and foreign investors who pursue investments in firms with high dividend payouts. Managers consider these anticipations, which may lead them to choose to payout high dividends to live up to expectations on dividend policy decisions. However, if firms opt for high dividend payouts when faced with lower income or losses, stakeholders are not likely to approve this decision. Therefore, firms will prefer to execute dividend payouts from sufficient income, which may ultimately provide incentives to managers to depend on earnings management as a means to aggrandize their income. With this rationale, we attempt to study the effect of high dividend payouts on earnings management. 2 Literature Review & Hypotheses Many prior studies have focused on reporting results of the effect of individual dividend policies, which have consistently found to be related to the firm s market value. Thus, mangers are required to make a tough decision closely considering firm value among many factors. Lintner (1956) and Brav et al. (2005) report that managers seem to hesitate to cut off dividends and that they dispose their assets, lay off employees, and raise external funds before executing their dividend policies, as an attempt to protect the firm from radical future stock price crashes. In addition, Daniel et al. (2008) assert that firms that are imposed the common debt contract covenants consistently manage their earnings to meet dividend restrictions set by the debtors. The results are likely to support the argument that firms attempt to conduct earnings management in order to sustain a certain level of dividend payouts. Firms have been found to use discretionary accruals to sustain a certain level of dividend payouts, as in Deniel et al. (2008) which suggest that managers are likely to attempt to increase earnings when they realize that their firms are not likely to meet the earnings level required for dividend payouts. Cohen and Zarowin (2010) and Zang (2012) describe that real earnings management activity is associated with earnings management through discretionary accruals, implying that in general the two types of earnings management can substitute each other. Thus, we suggest our main hypotheses based on prior research evidence as follows: H1: Firms with high dividend payouts are likely to conduct earnings management through discretionary accruals. H2: Firms with high dividend payouts are likely to conduct earnings management through real earnings management. 34 Copyright 2015 SERSC

3 3 Methodology 3.1 Discretionary Accruals This paper adopts the modified Jones Model suggested by Kothari et al. (2005) to estimate a firm s earnings management behavior through discretionary accruals as follows: DisAcc i,t = (TA i,t /A ) (β 0 ( 1 ) + β A 1 [ RET i,t REC i,t ] + β A 2 ( PPE i,t ) + A β 3 lagged(roa ) Model (1) 3.2 Real Earnings Management This paper uses Roychowdhury s (2006) model to estimate a firm s real earnings management behavior as follows: INV it 1 Sales it Sales = k Assets 1t + k Assets 2t + k Assets 3t + ε Assets it Prod it 1 = k Assets it + k Assets 2t + ε it Sales it + k Assets 3t Model (2) Sales it Sales + k Assets 4t Assets DiscExp it 1 = k Assets 1t + k Assets 2t Sales Assets + ε it Model (3) Model (4) The measures for models (2), (3), and (4) are calculated by subtracting an estimated value from a real value, yielding measures that are defined as abnormal operating cash flows, abnormal production costs and abnormal discretionary costs. In this paper, we use the sum of all three measures as the dependent variable to identify our main hypotheses. 3.3 Research Model We set the models to identify the relation between a firm s payout ratio and the firm s earnings management as follows: Copyright 2015 SERSC 35

4 DisAcc i,t ( TREM i,t ) = 0 + 1HPA OUT + 2 I E + 3M + 4R A + 5 E + 6 IG i,t + 7 MAR ET i,t + 8 H D it + 9 FH D i,t + 10I D i,t + 9 EAR i,t + it Model (5) Our main variables included in the models are as follows: first, we use the DisACC it as the dependent variable to show the relation of earnings management and the dividend policy at time t. H A T i,t represents the level of dividend payouts, which is our main independent variable of interest. In addition, we add control variables ( SIZE t, MB t, LEV t, ROA t BIG t, OHOLD t, FHOLD t, MARKET t, INDUS t and EAR t ), which are regarded to be determinants of earnings management. SIZE t represents the size of a firm at time t, and is included because large firms are likely to have a stronger monitoring system than small firms that deters managers from conducting earnings management. We expect SIZE t to show a negative ) relation with earnings management. MB t is the market to book ratio, in that a higher MB t represents higher growth. In general, growth firms are more likely to manage their performance. Therefore, we expect MB t to be positively related to earnings management. LEV t represents the firm s financial stability (leverage) at time t. Since higher leverage is likely to lead to higher financial risk, we expect LEV t to be positively associated with the firm s earnings management. ROA t is the firm s profitability at time t. With higher profitability related to stable financial conditions, we expect the control variable to exhibit a negative relation with earnings management. Both OHOLD t and FHOLD t represent the ratio of large shareholders, which is related to the firm s governance and regarded to be one of the determinants of earnings management. We also suggest BIG t as the proxy of financial transparency. Finally, we add INDUS t and EAR t to control for firm fixed-effects and time effects. 4 Sample and Results 4.1 Sample Our research sample consists of listed firms on KOSPI and KOSDAQ in the Korean stock market from the year 2001 to Observations of firms from financial industries are excluded from the sample, primarily because financial institutions substantially differ from firms of other industries. In addition, since the regressions require consistency in the point of measurement, we only include firms whose fiscal year ends in December. Furthermore, we only include sample firms that provide dividends during the firm s fiscal year. The final total sample consists of 6,744 firmyear observations (N=6,744). 36 Copyright 2015 SERSC

5 4.2 Empirical Results The empirical results confirm our main hypothesis that a firm s dividend policy is likely to be related to earnings management behavior. Table 3. The Result of Main Regression (DA) Independent variable Dependent variable: absolute value of discretionary accruals DisAcc i,t Model 1 Model 2 Model 3 Intercept (5.15) (2.67) (3.06) HPA OUT (1.61) (1.84) (2.07) I E ) (0.76) (0.56) M (8.31) (6.47) (6,25) R A (4.84) (4.64) (4.85) E (0.74) (0.59) IG i,t (0.97) (1.18) MAR ET i,t ) H D i,t FH D it Year Dummy Included Included Included Industry Dummy Included Included Included FValue Adj R-Sq N.obs 6,744 6,735 6, ) 3.22) 0.73) <Table 3> suggests the effect of a firm s dividend policy on the firm s earnings management during the sample period to verify for our first hypothesis that asserts that a firm s high payout policy behavior drives earnings management behavior with discretionary accruals. From Column (1) to Column (3) in <Table 3>, a positive coefficient is observed on HPA OUT t (i> Coef: 0.007, t-value: 1.61, ii> Coef: 0.008, t-value: 1.84, iii> Coef: 0.009, t-value: 2.07). These results indicate that a firm s high payout policy behavior drives discretionary earnings management. Copyright 2015 SERSC 37

6 Table 4. The Result of Main Regression(REM) Dependent variable: real earnings management activity TREM i,t Independent variable Model 1 Model 2 Model 3 Intercept (4.99) (1.50) (0.41) HPA OUT (0.85) (1.25) (1.38) I E ) (1.27) (2.09) M ) ) ) R A ) ) ) E (1.77) (1.72) IG i,t (3.25) (3.31) MAR ET i,t ) ) H D i,t (0.56) FH D it ) Year Dummy Included Included Included Industry Dummy Included Included Included F Value Adj R-Sq Number of observations 6,774 6,765 6,759 Furthermore, <table 4> reports the regression result for the regression of the firm s real earnings management behavior on HPA OUT t. The positive coefficient of HPA OUT t shows a positive relation with TREM t (real earnings management) (i> Coef: 0.004, t-value: 0.85, ii> Coef: 0.005, t-value: 1.25, iii> Coef: 0.006, t-value: 1.38). Unfortunately, these results show a weak relation between a firm s dividend policy and real earnings management behavior. Overall, the results of <Table 3> and <Table 4> indicate that a firm s dividend policy is likely to increase earnings management. This result supports our hypotheses that a firm s high payout dividend policy is likely to drive earnings management, since firms have significant incentives to engage in earnings management, especially by using discretionary accruals to create retained earnings from which managers can pay dividends to shareholders. 38 Copyright 2015 SERSC

7 5 Conclusion In this study, we attempt to contribute to the literature by examining the effects of a firm s dividend policy on the firm s earnings management. The results of our tests suggest that firms with a high dividend payout policy are likely to conduct more earnings management through both discretionary accruals and real earnings management activities. This supports our assertion that managers are likely to conduct earnings management to increase retained earnings from which dividends are paid out, as the pressure on the firm is to maintain consistency in the payout of dividends, even when facing decreased earnings or losses. References 1. Baker, H., and G. Powell.: Determinants of corporate dividend policy: A survey of NYSE firms. Financial Practice and Education 10: (2000) 2. Baker, H., E. Veit, and G. Powell.: Factors influencing dividend policy decisions of NASDAQ firms. The Financial Review 36: (2001). 3. Brav, A., J. Graham, C. Harvey, and R. Michaely.: Payout policy in the 21st century. Journal of Financial Economics 77: (2005) 4. Daniel, N., D. Denis, and L. Naveen.: Do firms manage earnings to meet dividend thresholds? Journal of Accounting & Economics 45: 2 26 (2008) 5. Dechow, P., S. Kothari, and R. Watts.: The relation between earnings and cash flows. Journal of Accounting & Economics 25: (1998) 6. Lintner, J.: Distribution of incomes of corporations among dividends, retained earnings, and taxes. American Economic Review 46: (1956) 7. Roychowdhury, S.: Earnings management through real activities manipulation. Journal of Accounting & Economics 42: (2006) 8. Zang, A.: Evidence on the tradeoff between real activities manipulation and accrual-based earnings manipulation. The Accounting Review 87 (2): (2012) Copyright 2015 SERSC 39