Article 14 of the Securities and Exchange Act, revised and effective in Taiwan on January 1, 2007, requires that firms satisfying certain qualifications must appoint independent directors of no less than two in number and no less than
38 In Model (1), the fiscal firms with small annual positive earnings but without corresponding variables of real activities manipulations or accrual-based earnings management were eliminated, reducing the number of samples to 773.
one-fifth of the total number of directors.39 Three additional variables were added to Model (1) to examine the effect of this regulation regarding independent directors on the empirical results of this study: the dummy variable of the regulation (one if the sample is from 2007 or later, zero otherwise), the product of the dummy variable and the percentage of independent directors (INDit), and the product of INDit and the dummy variable of the regulation as well as the nonfiscal group dummy (FYit). The empirical results for the sample before 2007 were consistent with Model (1) in this study. In other words, in the fiscal group, the percentage of independent directors is significantly positively correlated with manipulating earnings around the zero-earnings threshold (coefficient= 3.1448), supporting Hypothesis 2. Conversely, in the nonfiscal group, the percentage of independent directors causes a significant negative incremental effect on manipulating earnings around the zero-earnings threshold (coefficient= -2.3487);
however, the ratio is not significantly correlated with the manipulation. After 2007, i.e., when the independent director regulation became effective, the positive correlation between the percentage of independent directors and earnings manipulations around the zero-earnings threshold decreased significantly (incremental coefficient= -2.8775) in the fiscal group. Thus, the regulation on independent director seats affected the correlation between the percentage of independent directors and earnings manipulations around the zero-earnings threshold.
Measurement changes of some of the variables in Model (1) were tested to determine their effect on the empirical results. When the natural logarithm of the number of independent directors plus one was used to conduct a sensitivity analysis, δ = .1587, thus the result is still significant. The predicted signs and 6 levels of significance of the other variables remained unchanged. Thus, the results of the sensitivity test are consistent with the empirical conclusion of this study. In
39 The Securities and Exchange Act authorizes the competent authority to require financial institutions (holding companies, banks, and ticketing, insurance, securities investment trust, integrated securities, and listed futures companies), which offer stocks publicly in accordance with the law, and nonfinancial listed companies with paid-up capitals of up to NT$10 billion to appoint independent directors of no less than two in number and no less than one-fifth of the total number of directors, in accordance with the articles.
addition, if the institutional ownership of shares is replaced with the foreign shareholding ratio, then the coefficient corresponding to the foreign shareholding ratio is .0078, which yields a slightly higher value than that of the institutional ownership of shares but a slightly lower significance (p= .1000). The predicted signs and levels of significance of the other variables remained unchanged.
Finally, in the additional test of the annual earnings histogram, after tax net incomes (Figure 2-1 and 2-2) and incomes from continuing operation (Figure 3-1 and 3-2) deflated by market capitalization were used to plot the earnings histogram. Because of the space limitations of this paper, only histograms combining the fiscal and nonfiscal annual earnings are displayed. Discontinuity still occurs around the zero-earnings threshold in the histograms, consistent with the results based on EPS.40
Figure 2
40 The earnings histogram plotted according to net income after subtracting total assets (not displayed in this paper) also corresponds with the conclusion of this paper.
Figure 2-1
Earnings Histograms (After Tax Net Income Deflated by Market Capitalization) for Nonfiscal Year
Ending at the End of the First, Second, and Third Quarters
Figure 2-2
Earnings Histograms (After Tax Net Income Deflated by Market Capitalization) for fiscal Year
Figure 3
6. Conclusion
The manipulation of small amounts around the zero-earnings threshold differs from conventional earnings management. According to prospect theory and cost-benefit considerations of transactions, real activity manipulations entail actual enterprise transactions, and manipulations of small amounts around the zero-earnings threshold, which involve discretionary accrual estimates that fall within the acceptable range of GAAP, are difficult or impossible to detect.
Therefore, in addition to a perspective that emphasizes conventional corporate governance and audit quality effectively inhibiting earnings management, we examined behaviors involving manipulating earnings around the zero-earnings threshold from a perspective that emphasizes the relationship between stakeholders and firm decisions, thereby providing a broader and more dynamic analysis of corporate governance.
The empirical results of this study indicate that the perspective that focuses on the relationship between stakeholders and firm decisions explains behaviors involving earnings manipulations around zero-earnings threshold more effectively than does the perspective that focuses on conventional corporate governance and
Figure 3-1
Earnings Histograms (Income from Continuing Operating Deflated by
Market Capitalization) for Nonfiscal Year Ending at the End of
the First, Second, and Third Quarters
Figure 3-2
Earnings Histograms (Income from Continuing Operating Deflated by
Market Capitalization) for fiscal Year
audit quality. The results are listed as follows: (a) Independent directors and the institutional ownership of shares are significantly positively correlated with manipulating earnings around the zero-earnings threshold. However, audit quality (measured by Big Four audit firms) and CEO–chairperson duality are not significantly correlated with such manipulations. The empirical results differ from the perspective that emphasizes conventional corporate governance and supports the perspective that emphasizes the relationship between stakeholder and firm decisions. (b) The results of the additional test, conducted using the research design of Jacob and Jorgensen (2007), indicate that discontinuities around the zero-earnings threshold in the fiscal annual earnings histogram of the Taiwanese listed firms are not caused by scaling. These results confirmed that the firms manipulated their earnings into small gains when facing small losses, thereby improving on the flaws of studies that adopted the model of Burgstahler and Dichev (1997). In addition, the means of the real activity manipulations and abnormal discretionary accruals by the firms that crossed the zero-earnings thresholds did not deviate significantly from zero, which is consistent with our argument that manipulating earnings around the zero-earnings threshold involves manipulating small earnings amounts.
In summary, this study shows that manipulating earnings around the zero-earnings threshold, a subtopic of earnings management, differs from other forms of earnings management and confirms that the threshold is the reference point of the valuation function for stakeholders (e.g., executives, corporate boards, investors, and accountants). Therefore, understanding firm behaviors enables the users of financial statements to interpret corporate income information and studies on accounting to accumulate multifaceted empirical evidence.
References
Al-Fayoumi, N., Abuzayed, B. and Alexander, D. (2010), “Ownership Structure and Earnings Management in Emerging Markets: The Case of Jordan,”
International Research Journal of Finance and Economics, 38, 28-47.
Becker, C. L., DeFond, L. M., Jiambalvo, J. and Subramanyam, K. R. (1998),
“The Effect of Audit Quality on Earnings Management,” Contemporary
Accounting Research, 159(1), 1-24.
Bradshaw, M., Richardson, S. and Sloan, R. (2001), “Do analysts and Auditors as Information in Accruals?” Journal of Accounting Research, 39(1), 45-74.
Brenner, S.N. and Cochran, P. (1991), “The Stakeholder Theory of the Firm:
Implications for Business and Society Theory and Research”, In Mahon, J.F.
(Ed.), Proceedings of the International Association for Business and Society
Second Annual Conference, Sundance, UT, 449-467.
Burgstahler, D. and Dichev. (1997), “Earnings Management to Avoid Earnings Decreases and Losses,” Journal of Accounting and Economics, 24(1), 99-126.
Bushee, B. (1998), “Institutional Investors, Long Term Investment, and Earnings Management,” The Accounting Review, 73(3), 305-33.
Chang, W. J., Chou, L. T. and Lin, W. H. W. (2001), “The Effect of Audit Quality on Earnings Management to Sustain Recent Performance,” Review of
Securities and Futures Markets, 13(2), 31-70.
Chen, J. T. and Yeh, Y. H. (2002), “The Relationship between Corporate Restructuring, Corporate Governance and Earning Management,” The
International Journal of Accounting Studies, 34, 1-29.
Chen, K. Y. and Liu, Jo-L. (2010), “Earnings Management, CEO Domination, and Growth Opportunities: Evidence from Taiwan,” International Journal of
Public Information Systems, 6(1), 43-69.
Chung, R., Firth, M. and Kim, J. B. (2002), “Institutional Monitoring and Opportunistic Earnings Management,” Journal of Corporate Finance, 8(1), 29-48.
Clarkson, B. E. (1995), “A Stakeholder Framework for Analyzing and Evaluating Corporate Social Performance,” Academy of Management Review, 20(1), 92-117.
Cohen D., Dey, A. and Lys, T. (2008), “Real and Accrual Based Earnings Management in the Pre and Post Sarbanes Oxley Periods,” The Accounting
Review, 83(3), 757-787.
Cornett, M. M., McNutt, J. J. and Tehranian, H. (2009), “Corporate Governance and Earnings Management at Large U.S. Bank Holding Companies,” Journal