nature of the decision, as well as its close relation to most of the financial and investment
decisions firms make (Allen & Michaely, 2003). Theories of optimal cash holdings, financing,
capital structure decisions, mergers and acquisitions, and employee compensation all have
conceptual and mechanical links to payout levels. Understanding the patterns and
determinants of payout policy could also help us to better understand the other pieces in this
puzzle. The time trend of dividend payouts has attracted significant research attention,
including authors such as Baker and Wurgler (2004a, 2004b), Bradford, Chen, and Zhu (2017),
DeAngelo and DeAngelo (2007), DeAngelo, DeAngelo, and Skinner (2004), DeAngelo,
DeAngelo, and Stulz (2006), Fama and French (2001), Farre‐Mensa et al. (2014), and Hoberg
and Prabhala (2009). There have been fewer studies on stock repurchase time trends, however.
Compared with dividends, stock repurchases have become the most popular form of corporate
payout since 1997, and Farre‐Mensa et al. (2014) and Floyd, Li, and Skinner (2015) have shown
them becoming more volatile over time. The aggregate dollar amount of repurchases increased
dramatically after 2002, peaking in 2007 and then sharply declining during the 2008–2009
financial crisis. These time‐variant characteristics of repurchases motivate us to explore the
reasons for stock repurchase changes over time.
We examine the time trends in stock repurchases through the lens of earnings management.
Stock repurchases can be a device of real activities earnings management (Almeida, Fos, &
Kronlund, 2016; Bens, Nagar, Skinner, & Wong, 2003; Brav, Graham, Harvey, & Michaely,
2005; Farrell, Unlu, & Yu, 2014; Hribar, Jenkins, & Johnson, 2006; Myers, Myers, & Skinner,
Firms’stock repurchases can improve their earnings per share (EPS) by reducing the
amount of shares outstanding, that is, lowering the denominator of the relation.
We take advantage of the macroeconomic event of the passage of the Sarbanes–Oxley Act
(SOX) to test our conjecture. Did it affect stock repurchases? Passed by US Congress in July
2002, SOX aimed to improve the quality of financial reporting and enhance investor confidence
by curbing earnings management and accounting fraud. Cohen, Dey, and Lys (2008) and Cohen
and Zarowin (2010) show that accrual‐based earnings management has declined significantly
since the passage of SOX, and they also note that this decline was concurrent with increases in
real activities earnings management.
The reason for this is the heightened costs of using
accrual‐based earnings management, given the additional scrutiny of auditors and regulators
and potential litigation penalties after the legislation's passage (Graham, Harvey, & Rajgopal,
2005; Zang, 2012). Managers’choices of particular earnings management methods depend on
the trade‐off between the benefits and costs. We therefore expect firms to be more likely to
engage in stock repurchases to boost their EPS post‐SOX.
We start by examining the impact of SOX on stock repurchases. We split the sample period
into two segments: the period from 1987 through 2001 (the pre‐SOX period) and the period
from 2002 through 2017 (the post‐SOX period). Consistent with our hypothesis, following the
passage of SOX, stock repurchases increased significantly. At the same time, the decline in
accrual‐based earnings management caused firms to repurchase more shares. This finding
suggests that firms switched to managing earnings using stock repurchases, a manipulation of
Other uses include signaling undervaluation (Bessler, Drobetz, Seim, & Zimmermann, 2016; Chan, Ikenberry, & Lee, 2004; Ikenberry, Lakonishok, & Vermaelen, 1995;
Ikenberry, Lakonishok, & Vermaelen, 2000; Lie, 2005), the distribution of excess cash (Brennan & Thakor, 1990; Denis & Denis, 1993; Grullon & Michaely, 2004), shifting
toward optimal financial leverage (Dittmar, 2000), the expropriation of creditors (Maxwell & Stephens, 2003), the funding of employee stock option plans (Kahle, 2002),
takeover defense (Billett & Xue, 2007), and the enhancement of investor–management agreements (Huang & Thakor, 2013).
Real activities earnings management is defined as actions that deviate from normal business practices, such as the acceleration of the timing of sales through
increased price discounts or more lenient credit terms, the reporting of lower costs of goods sold through increased production, and reductions in discretionary
expenses (e.g., Cohen et al., 2008; Cohen & Zarowin, 2010; Roychowdhury, 2006).
HSU AND HUANG EUROPEAN