Controlling Shareholders and Earnings Informativeness: Evidence from Taiwan

Transcription

1 Asia Pacific Management Review 18(1) (2013) xxxx Controlling Shareholders and Earnings Informativeness: Evidence from Taiwan JeiFang Lew a,*, ShingJen Wu b a Department of Accounting, National Kaohsiung University of Applied Sciences, Taiwan b Department of Accounting, Soochow University, Taiwan Abstract Received 10 December 2010; Received in revised form 5 October 2011; Accepted 30 March 2012 Taiwanese listed corporations, characterized by a high degree of separation between control rights and cash flow rights, are empirically studied in this research. This study investigates how the separation of cash flow rights from board seat control rights, as distinct from voting rights, affects the informativeness of earnings, as measured by the relationship between returns and earnings. This study extends the extant research by investigating the extent to which the level of disparity between these two rights affects the magnitude of earnings informativeness. Based on an empirical sample of Taiwanese listed corporations, the results indicate that earnings are generally less informative when there is a divergence between board seat control rights and cash flow rights. Keywords: Earnings informativeness, ownership structure, board seat control rights. 1. Introduction * Ownership concentration and ineffective corporate governance were two factors that contributed to the Asian financial crisis in 1997 (hereafter referred to as the Crisis) (Prowse, 1998). In particular, areas of concern included concentrated ownership, the dominance of controlling shareholders, separation of control rights and cash flow rights, and the limited protection of minority rights. These problems were particularly acute in countries negatively affected by the Crisis (Claessens et al., 2000). Since the Crisis, there has been increasing investor demand requesting corporate governance reforms in emerging marets (Johnson et al., 2000; Gibson, 2003). Taiwan provides an ideal setting for examining the effectiveness of corporate governance due to the high ownership concentration, wea legal protection for shareholders, deficient law enforcement, and the abundance of family controlled firms (La Porta et al., 1999; Lemmon and Lins, 2003). However, according to Claessens et al. (2002), regarding their opinion in the Journal of Finance, 2002, the empirical data for Asian countries including Hong Kong, Indonesia, and South Korea all suggested that the divergence of ownership and control rights had a negative correlation with firm value, but this trend was not as evident in the empirical results on Taiwan. Moreover, the level of divergence demonstrated by Taiwanese corporations was generally lower than that shown by corporations in Singapore, Japan, and Indonesia. They stated: * Corresponding author: jeifang_lew@hotmail.com ** DOI: /APMR

2 (B) (A) 29.16% 13.9% = 15.67% Deviation between control seats and control holdings (via proxy contests) References Asian Wall Street Journal. (1999) Business transparency in region has worsened. Survey shows, November 2, 5. Atiase, R.K. (1985) Predisclosure information, firm capitalization, and security price behavior around earnings announcement. Journal of Accounting Research, 23, Chaney, P., Jeter, D. (1992) The effect of size on the magnitude of longwindow earnings response coefficients. Contemporary Accounting Research, 8, Claessens, S., Djanov, S., Lang, H.P. (2000) The separation of ownership and control in East Asian corporations. Journal of Financial Economics, 58, Claessens, S., Djanov, S., Fan, J., Lang, H.P. (2002) Disentangling the incentive and entrenchment effects of large shareholders. Journal of Finance, 57(6), Collins, D.W., Kothari, S.P. (1989) An analysis of intertemporal and crosssectional determinants of earnings response coefficient. Journal of Accounting and Economics, 11, Core, J., Guay, W. (1999) The use of equity grants to manage optimal equity incentive levels. Journal of Accounting and Economics, 28, DeAngelo H., DeAngelo, L., Sinner, D. (1992) Dividends and losses. Journal of Finance, 7, Demsetz, H., Lehn, K. (1985) The structure of ownership: causes and consequences. Journal of Political Economy, 93, Dhaliwal, D., Lee, K., Fargher, N. (1991) The association between unexpected earnings and abnormal security returns in the presence of financial leverage. Contemporary Accounting Research, 8, 201. Easton, P., Harris, T. (1991) Earnings as an explanatory variable for returns. Journal of Accounting Research, 29(1), Faccio, M., Lang, L., Young, L. (2001) Dividends and expropriation. American Economic Review, 91(1), 578. Faccio, M., Lang, L. (2002) The ultimate ownership in Western European corporations. Journal of Financial Economics, 65, Fama, E., MacBeth, J. (1973) Ris, return and equilibrium: Empirical tests. Journal of Political Economy, 81, Fan, P.H., Wong, T.J. (2002) Corporate ownership structure and the informativenss of accounting earnings. Journal of Accounting and Economics, 33, Francs, J., Smith, A. (1995) Agency costs and innovation: Some empirical evidence. Journal of Accounting and Economics, 19, Francis J., Schipper, K., Vincent, L. (2005) Earnings and dividend informativeness when cash flow rights are separated from voting rights. Journal of Accounting and Economics, 39, Freeman, R. (1987) The association between accounting earnings and security returns for large and small firms. Journal of Accounting and Economics, 55, Gibson, M.S. (2003) Is corporate governance ineffective in emerging marets? Journal of Financial and Quantitative Analysis, 38, Hartzell, J., Stars, L. (2003) Institutional investors and executive compensation. Journal of Finance, 58, Haw, I.M., Hu, B., Hwang, L.S., Wu, W. (200) Ultimate ownership, income management, and legal and extralegal institutions. Journal of Accounting Research, 2,

3 Hayn, C. (1995) The information content of losses. Journal of Accounting and Economics, 20, Imhoff, E., Lobo, G. (1992) The effect of ex ante earnings uncertainty on earnings response coefficients. The Accounting Review, 67, Jensen, M.C., Mecling, W.H. (1976) Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, Jensen, M.C. and Rubac, R.S. (1983) The maret for corporate control: Empirical evidence. Journal of Financial Economics, 11(1), 550. Johnson, S., Boone, P., Breach, A., Friedman, E. (2000) Corporate governance in the Asian financial crisis. Journal of Financial Economics, 58, La Porta, R., LopezdeSilanes, F., Shleifer, A. (1999) Corporate ownership around the world. Journal of Finance, 5, La Porta, R., LopezdeSilanes, F., Shleifer, A., Vishny, R. (2000) Agency problems and dividend policies around the world. Journal of Finance, 55, 133. Lemmon, M.L., Lins, K.V. (2003) Ownership structure, corporate governance, and firm value: Evidence from the East Asian financial crisis. Journal of Finance, 58, Morc, R., Shleifer, A, Vishny W. (1988) Management ownership and maret valuation: An empirical analysis. Journal of Financial Economics, 20(12), Prowse, S. (1998) Corporate governance: Emerging issues and lessens from East Asia. Responding to the global financial crisis. World Ban memo. Shleifer, A., Vishny, R. (1997) A survey of corporate governance. Journal of Finance, 52, Subramanyam, K.R., Wild, J. (1996) Goingconcern status, earnings persistence, and informativeness of earnings. Contemporary Accounting Research, 13, Teoh, S., Wong, T.J. (1993) Perceived auditor quality and the earnings response coefficient. The Accounting Review, 68, Warfield, T., Wild, J., Wild, K. (1995) Managerial ownership, accounting choices and informativeness of earnings. Journal of Accounting and Economics, 20, Yeh, Y.H. (2005) Do controlling shareholders enhance corporate value? Corporate Governance: An International Review, 13, Yeh, Y.H., Ko, C.E., Su, Y.H. (2003) Ultimate control and expropriation of minority shareholders: New evidence from Taiwan. Academic Economic Papers, 31(3), Yeh, Y.H., Woidte, T. (2005) Commitment or entrenchment? Controlling shareholders and board composition. Journal of Baning and Finance, 29(7), Yeo, G.., Tan, P., Ho, K., Chen, S. (2002) Corporate ownership structure and the informativeness of earnings. Journal of Business, Finance and Accounting, 29, Panel A: Sample selection procedure Table 1. Sample selection and distribution 250

4 Initial Sample (Taiwanese listed firmyear observations) Year Subtract: Baning and insurance industries Sales or boo value of equity are less zero Missing data Outliers Final sample 5,570 ( 6) ( ) (862) ( 6),59 Panel B: Sample distribution by year and industry Industry Total Number Percent 11 Cement Food Plastics Textiles Electric and machinery Electrical appliance and cable Chemical Glass and ceramics Paper and pulp Steel and iron Rubber Automobile Electronics Construction Transportation Tourism Department stores Other Total % Table 2. Descriptive statistics (60 firms, 59 firmyear observations) Variables Mean Std. dev. Q1 Median Q3 Min Max Panel A: Related corporate governance variables 251

5 DEV (Deviation) Voting rights (%) Board seat control rights (%) Cash flow rights (%) Panel B: Other variables in the models R (%) Earn Earn Size Growth Loss Lev Net Worth (in millions) 13, ,912 1,826 3,665 8, ,138 Maret Value (in millions) 18,70 76,979 1,562 3,510 9, ,523,30 8 Total Assets (in millions) 25,3 0 6,516 3,318 6,80 16, ,021,9 5 Variables definition: DEV is the calculated multiple of board seat control rights over cash flow rights. Voting rights is the procedure of identifying ultimate owners used in La Porta et al. (1999). Board seat control rights are the ratio of board members controlled by the ultimate owners to board size. Cash flow rights are the cash flow right held by the controlling shareholder. R is measured by the cumulative 12month raw stoc return until four months after the end of fiscal year t. Earn is earning per share (before extraordinary items) for fiscal year t. Earn is the change in Earn. Size is the size of firm which is measured as natural log of total assets at the end of year t1. Growth is M/B ratio computed as the maret value of equity/boo value of equity. Loss is loss incidence which equals 1 if firm s earnings before extraordinary items are less than 0, and 0 otherwise. Lev is measured as total liability divided by total equity at the end of year t. Table 3. Relationship between earnings and returns dependent on the level of Divergence between board seat control rights and cash flow rights Board seat control right/cash flow rights (DEV) Notes: # of firmyear observations Correlation between earnings and returns All > All correlations (Pearson) between annual accounting earnings per share (EPS) and stoc returns, and the earnings coefficients from the regression of stoc returns on accounting earnings per share, are significant at the level or better. 2. The six divergence groups are categorized according to the percentile of 5%, 25%, 50%, 75% and 95% of DEV distribution. 252

6 Table. Pearson correlation analysis Variables R 2. Earn *** 3. Earn 0.35 *** ***. DEV 0.06 *** Size *** *** 6. Growth *** *** *** 0.05 *** 7. Loss *** *** 0.21 *** ** *** *** 8. Lev * *** *** *** *** Notes: 1. Variables definition: R is measured by cumulative 12month raw stoc return until four months after the end of fiscal year t. Earn is earning per share (before extraordinary items) for fiscal year t. Earn is the change in Earn. DEV is the calculated multiple of board seat control rights over cash flow rights. Size is the size of firm which is measured as natural log of total assets at the end of year t1. Growth is M/B ratio computed as the maret value of equity/boo value of equity. Loss is loss incidence which equals 1 if firm s earnings before extraordinary items are less than 0, and 0 otherwise. Lev is measured as total liability divided by total equity at the end of year t. 2. ***, **, and * indicate significance at the 1%, 5%, and 10% levels for the twotailed tests, respectively. Table 5. The Effect of Divergence between Board Seat Control Rights and Cash Flows Right on the Earnings Informativeness (n=59) Variables Pred. Sign Model (1) Model (2) Earn *** ( 2.68) (0.08) Earn DEV ** * (2.00) (1.3) ΔEarn *** (17.3) ΔEarn DEV ** (1.73) Earn Size? (1.15) (0.39) ΔEarn Size? (5.66) *** Earn Growth? *** (0.1) ( 3.29 ) ΔEarn Growth? *** (5.67) Earn Loss *** *** (.13) ( 2.50 ) ΔEarn Loss + (.81) *** Earn Lev? *** * ( 3.21) ( 1.86 ) ΔEarn Lev? ( 0.8 ) COEF. Test (Ftest) Earn DEV+ΔEarn DEV Fvalue=6.08 ** Highest VIF Adj. R 2 3.3% 15.27% 253

7 Model F Notes:1. The HuberWhite heteroscedasticityconsistent robust standard errors (Rogers, 1993, generalizing White 1980) adjustment procedure is used to estimate the reported tstatistics (in parentheses). 2. Model specification: Rit, 0 1 Earnit, 2 Earnit, DEVit, Earnit, Xit, it, Model (1) 1 Model (2) Rit, 0 1 Earn it, 2 Earn it, DEVit, 3 Earn it, Earn it, DEVit, Earn it, Xit, Earn it, Xit, it, Variables definition: R i, t = firm i s cumulative 12month raw stoc return until four months after the end of fiscal year t Earn i, t = firm i s earning per share (before extraordinary items) for fiscal year t DEV i, t = The ratio of board seat control rights over cash flows rights of the largest ultimate owner of firm i = vector of control variables, =1,2, 3, and : Sizei,t = size of the firm, natural log of firm i s total assets at the end of year t1 : Growthi,t= M/B ratio computed as the maret value of equity/boo value of equity : Lossi,t = 1 if firm i s earnings before extraordinary items are less than 0, and 0 otherwise : Levi,t = firm i s total liability divided by total equity at the end of year t. ***, ** and * respectively indicate significance levels at the 1%, 5%, and 10% (onetailed for the coefficients with predicted sign, and twotailed otherwise). 25

8 Asia Pacific Management Review 18(1) (2013) xxxx 233

9 Asia Pacific Management Review 18(1) (2013) xxxx Table 6. Additional Test: Endogeneity Concern when Controlling Shareholders Control exceeds Ownership (2SLS) Panel A: The empirical results of Stage 2 (Model (1) and Model (2)) Panel B: The empirical results of the Stage 1 (Model (3)) Variables Pred. Sign Model (1) Model (2) Variables Pred. Sign vs. Model (1) vs. Model (2) Earn *** *** *** *** Pyramid + ( 6.67) ( 3.91) (.63) ( 10.11) Earn DEV *** *** ** *** Cross + (3.9) (3.2) ( 2.19) ( 3.8) ΔEarn *** *** *** Size? (17.29) (5.72) (11.60) ΔEarn DEV *** *** Growth? (0.15) ( 5.28) ( 2.79) Earn Size? SE? ( 0.0) ( 0.11) ( 0.88) ( 0.11) ΔEarn Size? *** SE ? (5.90) ( 0.17) (0.33) Earn Growth? ** (2.11) (0.10) ΔEarn Growth? (.1) *** Earn Loss *** *** (.88) (.5) ΔEarn Loss + (6.78) *** Earn Lev? *** *** (.73) ( 3.76) ΔEarn Lev? (0.08) COEF. Test (Ftest) Earn DEV+ΔEarn DEV Fvalue=28.62 *** Adj. R % 50.12% 12.39% 1.12% Model F

10 Notes: 1. The variables of Model (1) and Model (2) are the same as for Table 5. The variables definition in Model (6) is as follows: Pyramid is equal to 1 if firm has stoc pyramids, and 0 otherwise. Cross is equal to 1 is firm has crossholding, and 0 otherwise. SE is measured as the standard deviation of monthly stoc return of lagged fiveyears. SE2 is measured as squared term of SE. 2. ***, ** and * respectively indicate significance levels at the 1%, 5%, and 10% (onetailed for the coefficients with predicted sign, and twotailed otherwise). Table 7. Additional Test: Tests for Subsample Depends on the Ultimate Control Type Variables Pred. Sign Earn + Earn DEV ΔEarn + ΔEarn DEV Earn Size? ΔEarn Size? Earn Growth? ΔEarn Growth? Earn Loss + ΔEarn Loss + Earn Lev? ΔEarn Lev? Group 1: Family Group 2: Mutual Group 3: Manager Group : Government Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) *** * ( 2.97) ( 0.30) ( 1.51) ( 0.9) (0.35) (0.80) (0.2) (0.7) *** ** ** ** (3.1) (2.05) ( 0.37) ( 0.88) (1.97) (1.82) ( 0.21) ( 0.22) *** *** *** (12.57) (.9) ( 9.70) ( 1.09) *** * (2.55) (0.7) (1.9) (1.02) (1.31) (0.33) ( 1.08) ( 0.12) ( 0.22 ) ( 0.18) ( 0.99) ( 0.67) *** *** (3.7) ( 0.68) (.6) (0.50) *** * *** ** (0.9) ( 2.55) ( 0.16) (0.51) ( 1.58 ) ( 2.50) ( 0.50) ( 1.82) *** * *** (.2) (1.35) (5.07) (0.30) *** ** * * ( 3.53) ( 1.85) ( 1.33) ( 0.89) ( 1.0 ) ( 0.35) (0.69) ( 0.81) *** *** (3.79) (3.32) (0.82) (1.28) ** * ( 2.56) ( 1.36) (0.15) ( 1.8) ( 0.35 ) ( 0.58) ( 1.81) (0.67) (0.61) (0.02) ( 0.86) ( 1.23) COEF. Test (Ftest) Earn DEV+ΔEarn DEV (8.70) *** (0.15) (5.0) ** (0.80) 23

11 # of obs Highest VIF Adj. R 2.37% 16.33% 1.80% 12.91% 0.83% 13.9%.0%.22% Model F (p=0.067) (p=0.039) (p=0.221) 1.22 (p=0.2712) Notes: 1. The variables are the same as for Table ***, ** and * respectively indicate significance levels at the 1%, 5%, and 10% (onetailed for the coefficients with predicted sign, and twotailed otherwise). 235

12 Asia Pacific Management Review 18(1) (2013) xxxx Table 8. Agency Problem for Controlling Shareholders and Dividends Informativeness Variables Pred. Sign Model () Model (5) Div * ( 1.39) (0.63) Div DEV ** * (1.98) (1.9) ΔDiv + (11.21) *** ΔDiv DEV ** (1.67) Div Size? *** (3.03) (1.1) ΔDiv Size? (.02) *** Div Growth ** ( 0.66) ( 1.9) ΔDiv Growth + (6.69) *** Div Loss *** ( 0.07) ( 1.2) ΔDiv Loss (0.10) Div Lev? *** *** ( 5.08) ( 2.86) ΔDiv Lev? ( 6.15) *** COEF. Test (Ftest) Earn DEV+ΔEarn DEV Fvalue=3.30 * Highest VIF Adj. R % 11.21% Model F Notes: 1. The HuberWhite heteroscedasticityconsistent robust standard errors (Rogers, 1993, generalizing White 1980) adjustment procedure is used to estimate the reported tstatistics (in parentheses). 2. Model specification: Rit, 0 1 Divit, 2 Divit, DEVit, Divit, X it, it, Model (1) 1 Model(2) Rit, 0 1 Divit, 2 Divit, DEV it, 3 Div it, Divit, DEV it, Divit, Xit, Div it, Xit, it, Variables definition: R i, t = firm i s cumulative 12month raw stoc return until four months after the end of fiscal year t Div i, t = firm i s common stoc dividends in year t, scale by maret value of equity at the end of year t1 DEV i, t = The ratio of board seat control rights over cash flows rights of the largest ultimate owner of firm i = vector of control variables, =1,2, 3, and 233

13 1 2 3 : Sizei,t = size of the firm, natural log of firm i s total assets at the end of year t1 : Growthi,t= M/B ratio computed as the maret value of equity/boo value of equity : Lossi,t = 1 if firm i s earnings before extraordinary items are less than 0, and 0 otherwise : Levi,t = firm i s total liability divided by total equity at the end of year t. ***, ** and * respectively indicate significance levels at the 1%, 5%, and 10% (onetailed for the coefficients with predicted sign, and twotailed otherwise). 23

14 Asia Pacific Management Review 18(1) (2013) xxxx Table 9. Yearbyyear Regression with Interactions for Ownership Structure (for the Test of Model (1)) Variables 2000 (n=252) 2001 (n=297) 2002 (n=352) 2003 (n=387) 200 (n=36) 2005 (n=75) (n=579) 2007 (n=603) 2008 (n=613) 2009 (n=600) FamaMacBeth tstatistic Mean (tvalue) Earn * *** *** * *** ** ( 1.03) (1.1) ( 5.01) (0.12) (.81) ( 1.62) (0.69) ( 0.67) ( 2.37) (0.63) ( 2.1) Earn DEV * ** *** ** * ** ** *** *** (1.36) (2.05) (3.08) ( 0.18) (2.12) ( 0.78) (1.56) (1.67) (1.87) (3.0) (2.99) Sign(+/ ) + + 8/10 Earn Size ** *** ** (0.58) ( 1.61) (1.3) (2.39) (0.33) (2.91) (1.69) (0.39) ( 0.79) ( 1.23) (1.02) Earn Growth * *** ** * ** * (0.02) (1.31) (3.58) ( 1.0) (1.10) ( 1.03) (2.10) ( 1.53) (2.07) (0.89) (1.55) Earn Loss * *** *** *** ** ( 1.53) (1.10) ( 3.65) ( 0.77) (1.07) ( 0.30) ( 3.77) ( 1.13) ( 0.38) ( 3.27) ( 1.92) Earn Lev *** ** * ** *** *** ( 1.02) ( 3.99) (0.3) (1.57) ( 2.38) ( 1.29) ( 1.37) ( 1.95) ( 2.38) ( 2.62) ( 2.75) Highest VIF Adj.R % 9.57% 36.37% 7.7% 32.7% 17.25% 12.5% 25.66% 16.15% 8.59% Model F Notes: 1. The HuberWhite heteroscedasticityconsistent robust standard errors (Rogers, 1993, generalizing White 1980) adjustment procedure is used to estimate the reported t statistics (in parentheses). 2. Model specification: Rit, 0 1 Earnit, 2 Earnit, DEVit, Earnit, Xit, it, Model (1) 1 3. Variables definition: R i, t = firm i s cumulative 12month raw stoc return until four months after the end of fiscal year t Earn i, t = firm i s earning per share (before extraordinary items) for fiscal year t DEV i, t = The ratio of board seat control rights over cash flows rights of the largest ultimate owner of firm i = vector of control variables, =1,2, 3, and

15 1 2 3 : Sizei,t = size of the firm, natural log of firm i s total assets at the end of year t1 : Growthi,t= M/B ratio computed as the maret value of equity/boo value of equity : Lossi,t = 1 if firm i s earnings before extraordinary items are less than 0, and 0 otherwise : Levi,t = firm i s total liability divided by total equity at the end of year t. ***, ** and * respectively indicate significance levels at the 1%, 5%, and 10% (onetailed for the coefficients with predicted sign, and twotailed otherwise). Table 10. Yearbyyear Regression with Interactions for Ownership Structure (for the Test of Model (2)) Variables 2000 (n=252) 2001 (n=297) 2002 (n=352) 2003 (n=387) 200 (n=36) 2005 (n=75) 2006 (n=579) 2007 (n=603) 2008 (n=613) 2009 (n=600) FamaMacBeth tstatistic Mean (tvalue) Earn ** *** *** ** ** *** ( 2.00) (0.89) ( 3.2) (0.66) ( 3.8) ( 2.26) (0.08) ( 0.81) ( 2.16) ( 0.02) ( 2.55) Earn DEV *** *** ** *** ** *** (0.03) (2.58) (2.61) (0.78) ( 0.5) (2.06) (2.70) ( 0.17) (1.91) (0.10) (2.86) Sign(+/ ) + + 8/10 ΔEarn * *** *** *** *** *** *** *** *** *** 0.5 *** ( 1.3) ( 5.0) ( 5.97) (.56) (.37) ( 6.05) ( 7.22) ( 3.76) ( 6.08) ( 6.81) (10.33) ΔEarn DEV *** ** * ** * ** ** ** ** *** (0.9) (2.95) (1.73) (1.36) ( 1.79) (1.65) (1.8) (1.99) (2.13) (2.02) (8.95) Sign(+/ ) 10/10 Earn Size ** (0.3) ( 0.32) (0.82) (1.0) ( 0.03) (2.55) ( 0.88) (0.35) (0.98) ( 0.93) (1.35) ΔEarn Size * ** *** ** *** *** (0.16) (0.52) (1.86) (2.02) ( 0.06) ( 1.02) (2.80) (0.72) (2.39) (2.69) (3.10) Earn Growth *** ** * ** ( 0.6) (0.96) (2.6) ( 1.73) ( 0.3) (0.21) (0.2) ( 1.57) (1.86) (0.71) (1.11) ΔEarn Growth ** * *** ** *** *** (2.13) (0.01) (1.39) (0.7) ( 0.99) (0.16) (1.15) (2.59) (2.30) (3.27) (.21) Earn Loss *** *** *** * ( 0.70) (1.05) (.32) ( 0.62) ( 0.6) ( 0.68) ( 3.56) ( 0.7) (0.80) ( 2.76) ( 1.63) ΔEarn Loss * *** ** ** *** ** *** *** (0.72) (1.62) (.33) (1.05) ( 1.66) (2.0) (.71) (1.65) ( 0.0) (3.22) (.51) Earn Lev *** * ** ** * (0.26) ( 3.59) (0.20) (1.81) ( 2.38) ( 1.53) (0.3) ( 1.61) ( 2.35) ( 0.1) ( 1.79) ΔEarn Lev ** *

16 ( 1.9) (1.63) (0.09) (1.12) ( 1.17) (2.30) ( 1.83) (0.89) (0.6) ( 0.55) (0.55) Ftest:(Fvalue) Earn DEV *.86 ** **.59 ** 3.19 * 3.31 * 5.12 ** ΔEarn DEV Highest VIF Adj.R % 16.39% 3.9% 15.9% 37.21% 25.86% 25.82% 31.1% 2.83% 16.62% Model F Notes: 1.ΔEarni,t is the change in Earni,t between year t1 and t. The remaining variables in Table 8 are the same as Table ***, ** and * respectively indicate significance levels at 1%, 5%, and 10% (onetailed for the coefficients with predicted sign, and twotailed otherwise). 235

17 Asia Pacific Management Review 18(1) (2013) xxxx Table 11. Additional Test: Use of Alternative Measure of Cumulative Abnormal Returns (CARs) Variables Pred. Sign Earn + Earn DEV ΔEarn + ΔEarn DEV Earn Size? ΔEarn Size? Earn Growth? ΔEarn Growth? Earn Loss + ΔEarn Loss + Earn Lev? ΔEarn Lev? Model (1) Model (2) *** ( 3.22 ) ( 1.01 ) ** *** (1.76 ) (2.71 ) *** ( 9.37 ) *** (3.13 ) *** (3.20 ) (1.1 ) *** (.76 ) *** *** (.08 ) (3.02 ) *** (0.59 ) *** *** ( 3.0 ) ( 2.86 ) *** (3.8 ) *** *** (.73 ) (.02 ) (1.35) COEF. Test (Ftest) Earn DEV+ΔEarn DEV Fvalue=21.1 *** Highest VIF Adj. R 2.60% 7.71% Model F Notes: 1. The HuberWhite heteroscedasticityconsistent robust standard errors (Rogers, 1993, generalizing White 1980) adjustment procedure is used to estimate the reported tstatistics (in parentheses). 2. Model specification: CARit, 0 1 Earnit, 2 Earnit, DEVit, Earnit, X it, it, Model (1) 1 Model (2) CARit, 0 1 Earn it, 2 Earn it, DEVit, 3 Earnit, Earnit, DEV it, Earn it, Xit, Earnit, Xit, it, Variables definition: CAR i, t =firm i s cumulative 12month abnormal stoc return until four months after the end of fiscal year t Earn i, t = firm i s earning per share (before extraordinary items) for fiscal year t DEV i, t = The ratio of board seat control rights over cash flows rights of the largest ultimate owner of firm i = vector of control variables, =1,2, 3, and 233

18 1 2 3 : Sizei,t = size of the firm, natural log of firm i s total assets at the end of year t1 : Growthi,t= M/B ratio computed as the maret value of equity/boo value of equity : Lossi,t = 1 if firm i s earnings before extraordinary items are less than 0, and 0 otherwise : Levi,t = firm i s total liability divided by total equity at the end of year t. ***, ** and * respectively indicate significance levels at 1%, 5%, and 10% (onetailed for the coefficients with predicted sign, and twotailed otherwise) Table 12. The Alternative Measures of BoardCash Divergence Panel A: Descriptive Statistics Variables of Divergence Mean Std. dev. Q1 Median Q3 Min Max BC Natural log of B/C Natural log of BC Panel B: The Empirical Results Variables Pred. Sign Earn + Earn Divergence ΔEarn + ΔEarn Divergence Earn Size? ΔEarn Size? Earn Growth + ΔEarn Growth + Earn Loss + ΔEarn Loss + Earn Lev? ΔEarn Lev? COEF. Test (Ftest) Earn Divergence + ΔEarn Divergence Alternative Measures of Divergence Natural log of B/C BC Natural log of BC Model (1) Model (2) Model (1) Model (2) Model (1) Model (2) *** *** *** ( 3.3) (0.9) ( 2.56) (0.72) ( 2.30 ) (0.05) ** ** ** ** * (1.93) (1.97) (1.68) (1.79) (1.29) (0.88) 0.11 *** *** *** (1.1) (1.88) (11.69) * ** ** (1.51) (2.27) (2.33) ** * * (0.09) ( 1.65) (0.11) ( 1.70) ( 0.03 ) ( 1.83) *** *** *** (5.36) (5.65) (5.9) *** *** *** ( 0.59) ( 3.50) ( 0.1) ( 3.23) (0.13) ( 2.87) *** *** *** (5.78) (5.75) (5.86) *** *** *** *** *** *** ( 3.77) ( 3.3) ( 3.72) ( 3.2) (.09 ) ( 3.3) *** *** *** (3.93) (3.95) (3.79) * * * * ( 1.7) ( 0.52) ( 1.89) ( 0.6) ( 1.8 ) ( 0.20) ( 0.95) ( 1.0) ( 1.15) Fvalue= Fvalue= Fvalue= 6.09 ** 9.72 *** 6.19 ** Highest VIF Adj. R % 10.72% 2.78% 10.77% 2.69% 10.1% Model F

19 Notes: 1. The HuberWhite heteroscedasticityconsistent robust standard errors (Rogers, 1993, generalizing White 1980) adjustment procedure is used to estimate the reported tstatistics (in parentheses). 2. Model specification: Model (1) Rit, 0 1 Earnit, 2 Earnit, DEVit, Earnit, X it, it, 1 Model (2) R Earn Earn DEV Earn Earn DEV Earn X Earn X it, 0 1 it, 2 it, it, 3 it, it, it, it, it, it, it, it, Variables definition: R i, t = firm i s cumulative 12month raw stoc return until four months after the end of fiscal year t Earn i, t = firm i s earning per share (before extraordinary items) for fiscal year t DEV i, t = The ratio of board seat control rights over cash flows rights of the largest ultimate owner of firm i = vector of control variables, =1, 2, 3, and : Size i, t = size of the firm, natural log of firm i s total assets at the end of year t1 : Growth i,t= M/B ratio computed as the maret value of equity/boo value of equity : Loss i,t = 1 if firm i s earnings before extraordinary items are less than 0, and 0 otherwise : Lev i,t = firm i s total liability divided by total equity at the end of year t B = firm i s board seat control rights for fiscal year t C = firm i s cash flows rights for fiscal year t. ***, ** and * respectively indicate significance levels at 1%, 5%, and 10% (onetailed for the coefficients with predicted sign, and twotailed otherwise). 235