Legal Institutions, Ownership Concentration, and Stock Repurchases Around the World: Signal Mimicking?

Available online at www.sciencedirect.com

ScienceDirect The International Journal of Accounting 48 (2013) 427 – 458

Legal Institutions, Ownership Concentration, and Stock Repurchases Around the World: Signal Mimicking?☆ In-Mu Haw a,1,⁎, Simon S.M. Ho b,2 , Bingbing Hu c,3 , Xu Zhang d,4 a

d

M.J. Neeley School of Business, Texas Christian University, Fort Worth, TX 76129, United States b University of Macau, Av. Padre Tomás Pereira, Taipa, Macau c School of Business, Hong Kong Baptist University, Kowloon Tong, Kowloon, Hong Kong Faculty of Business Administration, University of Macau, Av. Padre Tomás Pereira, Taipa, Macau Received 30 November 2012

Abstract One of the central puzzles of signaling theory is how to assess signal quality, in particular the potential for signal mimicking. Our study provides evidence of signal mimicking in the context of stock repurchases. Employing an ex-ante proxy for the likelihood of mimicking stock repurchases and data on open market stock repurchases from 30 countries, we find that long-term operating and market performance following stock repurchases improve less for suspected mimicking firms. This finding contradicts the conventional characterization that managers use stock repurchases to signal undervaluation and enhanced future performance. We find that mimicking firms have smaller capital investments, need greater external financing, buy back fewer shares, and issue more new shares (and/ or resell more treasury shares) in the year of the repurchase. Our analysis further shows that mimicking is more likely in countries with weak investor protections and in firms with higher ownership concentration. Further, mimicking associated with concentrated ownership is mitigated in ☆ We would like to thank A. Rashad Abdel-khalik (Editor), Samir Trabelsi, Gary Biddle (Discussant), the anonymous referee, and the participants at the 2012 Symposium of the International Journal of Accounting for their helpful discussions and comments. We also thank Wayne Yu, Woody Wu, and the seminar participants at Hong Kong Baptist University and the University of Macau for their valuable comments and suggestions. The first author acknowledges that most of his research work was conducted when he was visiting University of Macau. ⁎ Corresponding author. E-mail addresses: [email protected] (I.-M. Haw), [email protected] (S.S.M. Ho), [email protected] (B. Hu), [email protected] (X. Zhang). 1 Tel.: +1 817 257 7563. 2 Tel.: +853 8397 4313. 3 Tel.: +852 3411 5273. 4 Tel.: +853 8397 4186. 0020-7063/$ - see front matter © 2013 University of Illinois. All rights reserved. http://dx.doi.org/10.1016/j.intacc.2013.10.004

428

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

countries with stronger investor protections and by the adoption of International Financial Reporting Standards (IFRS). Altogether, our findings provide evidence of signal mimicking in stock repurchases in international data that is influenced by market, ownership, legal, and financial reporting characteristics of countries. © 2013 University of Illinois. All rights reserved. JEL classification: G14; G35; M41 Keywords: Signaling; Mimicking stock repurchases; Operating and market performance; Legal institutions; Ownership concentration

1. Introduction Open market stock repurchases have become increasingly frequent corporate transactions, especially in the U.S., and this activity has received considerable attention from both academics and practitioners. Grullon and Michaely (2002) report that since 1999, U.S. firms have spent more money annually on stock repurchases than on dividends. Since the late 1990s, an increasing number of countries outside the U.S. have adopted similar laws allowing firms to buy back their shares. As a result, repurchase programs have become more common worldwide. Eije and Megginson (2008) find that while the proportion of European firms paying dividends has declined significantly, the proportion of firms repurchasing their own shares has grown steadily. A similar trend can be observed in East Asia. Following a revision of the Commercial Law, Japanese firms have been able to execute stock repurchases without the approval of a shareholders' meeting since 1997. Despite the growing popularity of stock repurchases as a payout method, little international research has been conducted on the subject. Given the high degree of institutional variation across countries, any conclusions applicable to stock repurchases in the U.S. might not be generalizable elsewhere. Stock repurchases are a flexible and discretionarily temporary payout method. Firms can immediately offset any discretionary capital payout to the degree they choose by selling offsetting treasury shares or issuing new shares, whereas firms treat dividends as longer-term commitments (Jagannathan, Stephens, & Weisbach, 2000). Previous studies identify several motives for stock repurchases, such as to signal undervaluation, distribute free cash flows, achieve an optimal leverage ratio, fund stock options, defend against takeovers, or exploit tax advantages (e.g., Bagwell & Shoven, 1988; Comment & Jarrell, 1991; Gup & Nam, 2001; Ikenberry, Lakonishok, & Vermaelen, 1995; Lie, 2005; Lie & Lie, 1999). Although firms repurchase stocks for various motives, the literature maintains that the signaling of undervaluation is a dominant rationale (Chan, Ikenberry, & Lee, 2004; Dittmar, 2000).5 Brav, Graham, Harvey, and Michaely (2005) show that managers commonly time the market when they believe their stock price is low. The signaling hypothesis suggests that when information asymmetry exists, corporate insiders use repurchases to signal undervaluation or better future performance. It is argued that for this 5

Dittmar (2000) investigates the relation between stock repurchases and distribution, investment, capital structure, corporate control, and compensation policies. She finds that firms repurchase stock to take advantage of potential undervaluation and to distribute excess capital. However, firms also repurchase stock during certain periods to alter their leverage ratio, fend off takeovers, and counter the dilution effects of stock options.

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

429

reason stock repurchase announcements are followed by positive price movements and positive changes in the firm's expected future operating performance. However, a number of recent studies argue that some managers also use stock repurchase announcements as a tool to mislead investors. Hribar, Jenkins, and Johnson (2006) document that some firms use stock repurchases to meet or beat analysts' earnings per share (EPS) forecasts. Chan, Ikenberry, Lee, and Wang (2010) note that for a subset of repurchasing firms, no improvement in operating and market performance follows repurchase programs. Massa, Rehman, and Vermaelen (2007) argue that repurchases are a defensive reaction to avoid potential stock price declines following other firms' repurchase decisions in concentrated industries. Gong, Louis, and Sun (2008) find that managers manipulate earnings downward before stock repurchases, suggesting the coordinated management of EPS. Motivated by this emerging literature on opportunistic or mimicking repurchases, this study has two objectives: first, to identify a subset of repurchasing firms that are likely to use stock repurchases to mislead investors, and second, to investigate the roles played by countries' investor protection institutions, firms' ownership structures, and the adoption of International Financial Reporting Standards (IFRS) in shaping the quality of signaling via repurchase activities in an international setting. As the literature provides little prior evidence on these questions, this study attempts to provide evidence. In particular, we provide evidence on whether repurchases are opportunistic or mimicking and driven by a manager's intention to mislead the market, rather than by the conventional motives of signaling undervaluation or distributing free cash flows. In his seminal research on labor market signals, Spence (1973) shows that educational background can serve as a signal differentiating the quality of job applicants in labor markets. He emphasizes that a signal will not effectively distinguish one applicant from another unless the costs of signaling are reliably and negatively correlated with productive capability; otherwise the signal is subject to manipulation. Furthermore, Spence (2002) observes that high-quality product owners have an incentive to distinguish themselves, while low-quality product owners have an incentive to imitate this signal and thereby mimic the distinction. Breed (2001) argues that honest signals in communication are given when both sender and receiver have the same interest in the result, and deceptive signals appear when the sender can exploit the receiver. In the context of stock repurchases, corporate insiders have incentives to undertake mimicking repurchases if mimicking can bring them private benefits. Examples of such private benefits range from the enhancement of stock-based remuneration to inducements for the conversion of convertible bonds. Stock-based payments account for a large portion of executives' total remuneration, and the long-term cumulative financial gain to CEOs from unexpectedly good stock price performance is positive and significant (Boschen, Duru, Gordon, & Smith, 2003). Moreover, CEO turnover is associated with poor stock returns (DeFond & Hung, 2004). Because stock repurchases are commonly characterized as “good news” signals, and markets react positively to repurchase programs, corporate insiders have incentives to announce repurchases to manipulate EPS (Hribar et al., 2006) and stock prices. Jung, Kim, and Stulz (1996) argue that agency problems may lead managers to ignore the costs of issuing equity with buybacks prior to the issue of new shares as a means of boosting the

430

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

offering price. Anecdotal evidence shows that firms often buy back shares during the conversion period of convertible bonds, and stop buybacks after conversion expiry dates. Although stock repurchase can be costly as a mimicking tool due to the cash consumed, it is less costly to insiders as long as the private benefits obtained are greater than that consumed and the expense is borne by outside investors. Following the approach used by Biddle, Hilary, and Verdi (2009), Gong et al. (2008), and Chan et al. (2010), we construct a proxy for the likelihood of mimicking repurchases based on the ex-ante characteristics of repurchasing firms. Specifically, we focus on (i) the market-to-book ratio; (ii) the operating cash flows to total assets; and (iii) the discretionary accruals prior to stock repurchases. We focus on these factors because previous studies indicate that repurchasing firms are undervalued, have free cash flows, and tend to manipulate their earnings downward prior to stock repurchases. Consequently, we predict that the higher the market-to-book ratio, the lower the operating cash flows to total assets, and the higher the discretionary accruals prior to stock repurchases, the more likely it is that the repurchases are opportunistic or mimicking in nature. Using this approach and data on 11,422 repurchasing firm–year observations between 1998 and 2006 across 30 countries, we identify a subset of firms that are likely to undertake mimicking repurchases. Our empirical results show that for suspected mimicking firms, the long-term operating and market performances do not improve following actual repurchases. In contrast, post-repurchase performance does improve for non-mimicking repurchasing firms. Our analysis indicates that the suspected mimicking firms exhibit poorer operating performance, fewer repurchases, lower capital expenditures, greater need for external financing, and lower levels of intangible assets. Moreover, we find that insiders minimize the costs of false signaling via repurchases by reselling treasury shares, or by issuing new shares in the year of the repurchase. We further test for the effects of country-level legal institutions and firm-level ownership structures on the quality of signaling in the context of repurchases. We find that mimicking repurchases are more likely to occur in countries with weaker investor protections, and in firms with greater ownership concentration. Our results provide a partial explanation for recent findings that stock repurchases are less effective than dividends in mitigating agency conflicts in countries with weaker investor protections (e.g., Haw, Ho, Hu, & Zhang, 2011; Pinkowitz, Stultz, & Williamson, 2006), and that stock price reactions to the repurchases are lower in countries with weak investor protections than those in the U.S. We also report that mimicking behavior induced by ownership concentration is restrained by stronger investor protections. Our additional analyses provide some evidence that the adoption of IFRS helps constrain mimicking repurchases, possibly due to the effect of enhanced transparency in reducing information asymmetry. Our findings are robust to a series of sensitivity tests including alternative measures for mimicking repurchases, legal protections, stock repurchasing activity, and also to the control for endogeneity. Our study contributes to the literature in several ways. First, it extends the scope of the literature on the motives and economic consequences of repurchase programs by enlarging the focus from U.S. firms to the international context, in which ownership structures are more concentrated and country-level institutional environments vary significantly. Our results suggest that a country's legal institutions and a firm's ownership structure both

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

431

influence insiders' mimicking behavior in repurchase activities. This study also complements the work of Massa et al. (2007), which identifies mimicking repurchases in concentrated industries in the U.S. Second, our work augments signaling theory addressing the signal quality (e.g., Breed, 2001; Dawkins & Guilford, 1991; Spence, 2002). This study provides empirical evidence that legal institutions and ownership structures jointly affect signal quality in the context of stock repurchase programs. Furthermore, as Spence (2002) argues, low-quality product owners have an incentive to imitate the signals of high-quality products. Our study provides empirical evidence of this in capital markets and further evidence on factors that can mitigate mimicking behavior. Third, our study provides a partial explanation for the lower effectiveness of stock repurchases in mitigating agency conflicts and the lower valuation of stock repurchases in countries with weak investor protections (Haw, Ho, Hu, & Zhang, 2011; Haw, Ho, & Li, 2011, Pinkowitz et al., 2006). The remainder of this study is organized as follows. In Section 2, we develop our major hypotheses. The research design and methodology are described in Section 3, with the data and sample selection presented in Section 4. Section 5 reports the empirical results and Section 6 concludes the paper. 2. Hypotheses development Signaling theory suggests that when valuation is noisy, firms that view themselves as undervalued may engage in activities to signal their quality. However, to be effective, the costs of signaling must be negatively correlated with productive capability. In other words, the signaling costs for mimickers must be higher than those for non-mimickers; otherwise, the signals are subject to manipulation by the signal senders (Spence, 1973). Signaling theory also suggests that lower-quality entities have an incentive to imitate the signals of higher-quality entities (Spence, 2002). In capital markets, if investors have difficulty in distinguishing the quality of signals, and if corporate insiders gain large private benefits from controlling their firms, the insiders may have incentives to send false signals to mislead investors. By doing so, they can obtain private benefits even if their firms are not undervalued. This argument suggests that repurchases can be used as a misleading signal by corporate insiders, given the considerable managerial flexibility in determining the timing, amount, execution, and offset of stock repurchases. Breed (2001) argues that honest signals in communication are given when both sender and receiver have the same interest in the result, and that deceptive signals appear when the sender can exploit the receiver. As strong corporate governance mechanisms can mitigate the conflict of interests between corporate insiders and outsiders, we explore the role of governance mechanisms in shaping the quality of signals in the context of repurchases. 2.1. The effect of investor protections on mimicking behavior La Porta, Lopez-de-Silanes, and Shleifer (1998) argue that fundamentally important corporate governance mechanisms depend on the extent to which a country's laws protect investor rights, and the extent to which those laws are enforced. In countries with strong investor

432

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

rights protections, insiders have an incentive to conceal their private benefits from outsiders, because if these benefits are detected, the outsiders are likely to take disciplinary actions against them (e.g., Shleifer & Vishny, 1997; Zingales, 1994). Previous studies find that external governance has a significant effect on the incentives and behavior of corporate insiders. In countries with weak investor protections, corporate insiders can enjoy the large private benefits arising from control, given the low costs to extract such benefits (Dyck & Zingales, 2004). Leuz, Nanda, and Wysocki (2003) argue that insiders have the incentive to manage reported earnings, to mask true performance, and to conceal their private benefits of control from outsiders. These authors indicate that such earnings management decreases with legal protections. DeFond, Hung, and Trezevant (2007) find that annual earnings announcements are more informative in countries with strong investor protections due to the underlying mechanisms of higher-quality earnings and better-enforced insider trading laws. The findings of these studies suggest that earnings are less credible in countries with weak investor protections. In a recent study, Haw, Hu, Lee, and Wu (2012) show that stock price informativeness about future earnings is lower in countries with weak investor protections due to the lower quality of earnings, lower level of financial disclosure, and weaker infrastructure for information dissemination. In countries with weaker investor protections, information asymmetry between corporate insiders and outside investors is more severe (Bushman, Piotroski, & Smith, 2004). When corporate insiders undertake repurchase programs, outside investors may not have sufficient information to assess the true value of the firm, which makes mimicking repurchases efficacious. This argument is consistent with findings of Morck, Yeung, and Yu (2000), which indicate that stock price fluctuations are less correlated with corporate fundamentals in emerging markets. All of these findings suggest that repurchase programs might be less credible in countries with weaker investor protections than in countries with stronger investor protections. Another stream of literature shows that companies can use stock repurchases as a tool to manipulate EPS (Bens, Nagar, Skinner, & Wong, 2003; Hribar et al., 2006). A stock repurchase is a temporary payout to shareholders, because firms can cancel this payout by reselling treasury shares or by issuing new shares. Stock repurchase announcements may thus offer a low-cost way to obtain private benefits at the expense of investors, with weak legal environments more conducive to deceiving or misleading investors rather than signaling undervaluation or distributing cash to shareholders. The above intuition leads to our first hypothesis. Hypothesis 1. Mimicking stock repurchases are more likely to occur in countries with weaker investor protections, ceteris paribus. 2.2. The effect of ownership concentration on mimicking behavior In countries outside the U.S., firms are typically controlled by a few major shareholders who possess control rights exceeding their cash-flow rights (Claessens, Djankov, & Lang, 2000; Faccio & Lang, 2002; La Porta, Lopez-de-Silanes, & Shleifer, 1999). The fundamental agency problem produces a conflict of interests between the controlling owners and the minority shareholders (Shleifer & Vishny, 1997). Ownership concentration has both an entrenchment effect and an alignment effect. On the one hand, concentrated control is considered detrimental to minority shareholders

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

433

because it induces insider expropriation and distorts management decision-making (Bebchuk, Kraakman, & Triantis, 2003; Shleifer & Vishny, 1997). On the other hand, the presence of controlling owners helps alleviate the traditional agency problem between owners and managers. However, the literature suggests that the alignment effect is subordinated to the entrenchment effect in concentrated ownership structures (e.g., Claessens, Djankov, Fan, & Lang, 2002; Lins, 2003). La Porta et al. (1999) investigate the prevalence of ownership concentration around the world, finding that weak legal and institutional environments are associated with highly concentrated ownership. Similarly, Dyck and Zingales (2004) find that higher private benefits of control are associated with more concentrated ownership. Thus, previous studies find that corporate governance is weak in firms with concentrated ownership whereby corporate insiders can obtain large private benefits from control. Because controlling shareholders typically have a larger proportion of a firm's control rights than cash flow rights, they have both the incentive and the ability to extract the private benefits of control from the minority shareholders. Furthermore, the desire of such owners to hide insider expropriation induces them to mask their control benefits, which they do by exploiting firm opacity through earnings management and higher cash dividend payments (Faccio, Lang, & Young, 2001; Haw, Hu, Hwang, & Wu, 2004). Such behavior in turn leads investors to perceive the firm's financial reporting as less credible and informative (e.g., Francis, Schipper, & Vincent, 2005; Haw, Ho, & Li, 2011; Leuz et al., 2003). We thus argue that the decisions of self-interested managers to undertake repurchases may arise from the desire to increase their private benefits of control, rather than to signal undervaluation or distribute cash to shareholders. The preceding discussion leads to the second hypothesis. Hypothesis 2. Mimicking stock repurchases are more likely to occur in firms with higher ownership concentration, ceteris paribus. 2.3. The joint effect of investor protections and ownership structure on mimicking behavior It is well established that legal protection is the primary constraint on agency costs and insider expropriation (Bebchuk et al., 2003; Doidge, Karolyi, & Stulz, 2007; Jensen & Meckling, 1976). When agency problems are embedded in a concentrated ownership structure, well-functioning external institutions become an important means of protecting minority shareholders. Furthermore, previous studies suggest that country-level institutions dominate firm-level governance mechanisms in mitigating agency conflicts (Dittmar, Mahrt-Smith, & Servaes, 2003; Doidge et al., 2007; Harford, Mansi, & Maxwell, 2008; La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 2000). These studies indicate that true entrenchment of insider control requires weak legal shareholder rights. To the extent that strong legal investor protections restrain the corporate insiders' incentive to pursue their own benefits, strong monitoring curbs insiders' intention to mimic, as induced by concentrated ownership. The preceding discussion leads to the third hypothesis. Hypothesis 3. The propensity for mimicking stock repurchases in firms with concentrated ownership is mitigated in countries with stronger investor protections, ceteris paribus.

434

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

3. Research design 3.1. Measuring the likelihood of mimicking repurchases Although previous studies document a subset of firms that appear to undertake repurchases to mimic undervalued firms, it is challenging to empirically distinguish mimicking repurchases from genuine repurchases, because motivation is not directly observable. Nor are attestations reliable, with most repurchasing firms claiming that their shares are undervalued, and that buybacks will enhance shareholder value. Massa et al. (2007) investigate mimicking repurchases in the U.S. market, finding that firms in highly-concentrated industries buy back their shares to mimic the behavior of other repurchasing firms in their industries and thereby avoid negative effects on their own share prices. However, it is more difficult to identify mimicking behaviors in countries outside the U.S., due to the less complete disclosures in these markets. Following the approach used by Biddle et al. (2009), Gong et al. (2008), and Chan et al. (2010),6 we identify three ex-ante firm-specific characteristics that are likely to capture managers' incentives to undertake repurchases. Specifically, we focus on (i) the market-to-book ratio, measured as the ratio of total assets minus book value of equity plus market capitalization, divided by total assets; (ii) the cash flows from operating activities, scaled by total assets; and (iii) the level of discretionary accruals in the year prior to a repurchase. We use the market-to-book ratio as a partitioning variable, based on the rationale that the signaling of undervaluation is the dominant motive for stock repurchases (Chan et al., 2004; Dittmar, 2000). Previous studies argue that firms repurchase their shares to distribute free cash flows to shareholders (e.g., Easterbrook, 1984; Jensen, 1986). Thus, we consider the ratio of cash flows from operating activities to assets as another means to identify mimicking repurchases. Our third partitioning variable, discretionary accruals, is chosen because earnings quality may reflect the managerial intention to mislead investors (Chan, Chan, Jegadeesh, & Lakonishok, 2006). Biddle et al. (2009) find that lower financial reporting quality relates to investment inefficiency, which highlights the importance of reporting quality as a gauge of managerial incentives and ability.7 Gong et al. (2008) show that firms tend to manipulate their earnings downward before undertaking stock repurchases. Chan et al. (2010) use discretionary accruals to identify repurchases that are potentially used to mislead investors. The discussion above indicates that the quality of reported earnings serves as a credible tool to signal the quality of managerial decisions. In particular, Biddle et al. (2009) measure the likelihood of over/under investment based on two ex-ante firmspecific characteristics: cash balance and leverage ratio. They first rank firms into deciles based on their cash balance and their leverage, and then create a composite score measure of over/under investment as the average of the ranked values of the two partitions' variables. 7 Hribar et al. (2006) document that firms use stock repurchases to meet or beat the EPS forecasts of analysts, which suggests that stock repurchases can be used as an alternative tool (secondary to accrual management) for misleading the market. Because accruals management does not consume cash, it is less costly and is presumably preferable to manipulating underlying business activities (Black, Sellers, & Manly, 1998). However, when managers face constraints in their ability to increase earnings via accruals manipulation, they are more likely to attempt increasing their earnings through stock repurchases (Xu & Taylor, 2007). 6

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

435

We estimate accrual quality using performance-matched discretionary accruals. First, we use the following Jones model (1991) to estimate unadjusted discretionary accruals,8 where total accruals (TACC) are defined as the total net income minus the total net cash flows from operations, scaled by lagged total assets9: TACC it ¼ β 0 þ β 1 ð1=T A−1 Þ þ β 2 ΔSalesit þ β 3 GPPEit þ εit ;

ð1Þ

where ΔSales is the change in sales and GPPE is the gross property, plant, and equipment, both scaled by lagged total assets, TA− 1. The Jones model is estimated cross-sectionally for each country–year using all of the observations with the same one-digit SIC code.10 We require a minimum of 20 firms in each country–year–industry combination. Consistent with previous studies, we define non-discretionary accruals as the fitted values from the Jones model for a given firm. Unadjusted discretionary accruals are then defined as the residual for a given case that departs from its expected value. Following Kothari, Leone, and Wasley (2005), we then sort the observations in the same country– year–industry into five quintiles, based on each firm's return on assets (ROA), which is defined as operating income divided by average total assets. The performance-matched discretionary accrual measure adjusts a firm's discretionary accruals by subtracting the median of the corresponding unadjusted discretionary accruals in the same country–year– industry–ROA combination. We then rank all firms (including repurchasing and non-repurchasing firms) based on each of the three partitioning variables into five quintiles. In particular, we rank the market-to-book ratio for each country–year–industry combination with similar market value of equity, and the ratio of the operating cash flows to total assets and the discretionary accruals for each country–year–industry combination. We multiply net operating cash flows by − 1 before ranking so that a higher rank indicates a lower level of cash flows and a greater likelihood of mimicking repurchase. Industry is defined according to a one-digit SIC code. We then compute a composite score measure, Mimic, as the average of the ranked values of the three partitioning variables. As a result, Mimic ranges from 1 to 5. We use a composite score because the aggregation of multiple measures can reduce measurement errors contained in individual measures. By construction, a greater value of Mimic indicates that the firm is potentially overvalued, short of cash, and/or reports a greater income-increasing accrual in the year prior to the actual repurchases, and therefore has a greater likelihood of mimicking a share repurchase. 3.2. Measuring abnormal long-term operating performance Although accrual-based performance is a conventional measure for operating performance, it may lead to an erroneous conclusion with regard to the post-repurchase performance due to 8

We include a constant in the model, as Kothari et al. (2005) suggest. As the cash flows from operations reported in Worldscope include extraordinary items, we use bottom-line net income rather than net income before extraordinary items. 10 Although Kothari et al. (2005) document that within-industry estimation is less important, we also estimate the Jones model cross-sectionally, using country–year observations with the same two-digit SIC code and the same performance-matching method. The results are very similar, although the sample size decreases by 30%. 9

436

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

pre-repurchase accruals management. Barber and Lyon (1996) suggest that in certain sampling situations where sample firms are motivated to manipulate their reported earnings, a cash-based rather than the accrual-based performance measure provides a better indicator of whether an erosion in operating performance is the result of declining performance or of a reversal in pre-event accruals. Because previous studies show that managers tend to manipulate accruals before making stock repurchases (Chan et al., 2010; Gong et al., 2008), we use cash-based performance to measure operating performance in the period following repurchases.11 Specifically, operating performance is measured by net cash flows from operating activities scaled by the average of cash-adjusted assets (i.e., book value of assets less cash and short-term investments) at the beginning and end of the fiscal year. Abnormal operating performance is defined as the difference between the operating performance of our sample firms and that of the control firms. Following Lie (2005), we select control firms from the non-repurchasing firms in the same industry that have similar performance characteristics and market-to-book ratios. Specifically, for each repurchasing firm, we identify a control firm in the same country and with the same two-digit SIC code that has an operating cash flow performance within ± 20% (or within ± 0.01) of the performance of the repurchasing firm in the year of the repurchase, and a market-to-book value of assets within ± 20% (or within ± 0.1) of the repurchasing firm in the year prior to repurchase. If no firms meet these criteria, we relax the industry criterion to a one-digit SIC code.

3.3. Measuring abnormal long-term market performance Following Barber and Lyon (1997), we use buy-and-hold abnormal returns to measure long-term market performance after the repurchase events. Using monthly windows, buy-and-hold returns (BHRs) are estimated up to one, two, and three years following stock repurchase programs. Abnormal BHRs are computed as the differences between the BHRs of the sample and those of the matched control firms using the following equation, where Rits denotes the stock return of sample firm i, and Ritc denotes the stock return of the control firm i in month t: τ
 τ
 AbnormalBHRit ¼ ∏ 1 þ Rsit − ∏ 1 þ Rcit : t¼1

t¼1

A sample firm is matched to a control firm by first identifying all firms from the same country and with the same two-digit SIC code that have a market value of equity between 70% and 130% of the market value of the sample firm's equity. From this set of firms, we choose the firm with the market-to-book ratio that is closest to that of the sample firm. If no firms meet these criteria, we relax the industry criterion to a one-digit SIC code. 11

For comparison, we also calculate the post-operating performance following repurchases based on operating income in Panels C and D of Table 5.

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

437

3.4. Testing the influence of legal institutions and ownership concentration on mimicking repurchases We develop the following ordered logistic regression model to test our three hypotheses (firm i, time t, and t − 1 subscripts are suppressed), where Mimic is a ranked variable measuring the likelihood of a mimicking repurchase and Legal is a dummy variable that equals one if a country's company law or commercial code originates from common law, and zero otherwise: Mimic ¼ β 0 þ β 1 Legal þ β 2 LOwn þ β 3 Legal  LOwn þ β 4 CFO þ β 5 Rep þ β 6 Issue þ β 7 DVD þ β 8 CPTEXP þ β 9 ITA þ β 10 ExtFin þ β 11 LSize þ β 12 LLev þ β 13 LRet þ β 14 HHI þ β 15 LnGDP þ β 16 SMGDP X X þ ηi YearDummies þ γi IndustryDummies þ εt :

ð2Þ

La Porta et al. (1998) find that common law countries generally have the strongest legal protections for investors. Bushman et al. (2004) report that governance transparency is higher in countries with a common law legal origin. This suggests that information between insiders and outsiders is less asymmetric, and therefore the costs of mimicking are higher in common law countries. Hypothesis 1 predicts that β1 will be negative. We measure ownership concentration (LOwn) using data on the closely held shares extracted from the Worldscope database, following Stulz (2005)12 and Armstrong, Barth, Jagolinzer, and Riedl (2010). This item represents the sum of shareholdings by all block holders who hold 5% or more of outstanding shares, which may not precisely capture the incentive of the controlling owner. Although this ownership measure is far from perfect, Stulz (2005) reports similar results when he uses the family-controlled ownership data (based on Claessens et al., 2000; La Porta et al., 1999) or the data from the Worldscope on fractional cash-flow ownership by insiders in explaining the expropriation index and anti-director rights index. However, the direction of the effect of insiders' ownership on mimicking is unclear, as the cash flow rights may affect the controlling shareholders' incentives to pay out cash, and the empirical evidence is mixed (Farinha, 2003; Gugler, 2003). Although some studies find that block shareholders play an effective monitoring role (e.g., Karamanou & Vafeas, 2005), others show that block shareholders behave as insiders who extract private benefits of control from the minority shareholders (e.g., Ajinkya, Bhojraj, & Sengupta, 2005). La Porta et al. (2000) and Stulz (2005) indicate that firms with voting rights exceeding cash flow rights in countries outside the U.S. commonly have an ownership structure dominated by family control. To the extent that ownership by corporate insiders represents the controlling rights of controlling shareholders, we predict a positive relationship between insider ownership and mimicking behavior (i.e., positive β2), if Hypothesis 2 holds. This is consistent with Armstrong et al. (2010), who use closely-held shares to measure information asymmetry and suggest that the agency problem is more severe in 12 Stulz (2005) uses the same ownership data and gives a detailed discussion of the weaknesses and advantages of insider ownership as determinants of agency problems. He obtains similar results using alternative ownership measures.

438

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

firms with concentrated ownership. To test Hypothesis 3, we add an interaction between legal institutions and ownership concentration (Legal ∗ LOwn) to the model. Hypothesis 3 predicts β3 to be negative. We also include a number of firm-specific variables to control for different firm characteristics. We include profitability (CFO) and cash dividends (DVD), because stock repurchasing is regarded as a method of corporate payouts. To explore the costs and benefits of mimicking, we add cash paid for repurchases (Rep), and the proceeds from new shares issued or treasury shares resold (Issue) during the year of the repurchase. As repurchases consume cash resources, mimicking firms are likely to minimize these costs by issuing new shares or reselling treasury stocks. We incorporate capital expenditure (CPTEXP) and the need for external financing (ExtFin) to capture the effects of investment opportunities and external financing on mimicking behavior. Lagged firm size (LSize) and intangible assets (ITA) are included to control for information asymmetry between inside and outside investors (Barth & Kasznik, 1999). As previous studies document that firms repurchase shares to achieve optimal leverage ratios (Bagwell & Shoven, 1988; Opler & Titman, 1996), lagged leverage ratio (LLev) is used as a target leverage ratio, below which firms are likely to repurchase shares. The annual stock return in the prior year (LRet) is added because stock repurchases are negatively related to prior stock price performance (Stephens & Weisbach, 1998). Because Massa et al. (2007) find that firms in concentrated industries are more likely to mimic the repurchase decisions of other firms in the same industry, we include the Herfindahl– Hirschman index (HHI) to control for industry competition. Following La Porta et al. (2000), we include two country-level variables to control for cross-country differences in economic and financial market development, the logarithm of GDP per capita (LnGDP), and the stock market capitalization to GDP (SMGDP), both of which are obtained from the World Bank website. We present detailed descriptions of all of the variables in Table 1. To control for industry and year correlations, we incorporate industry and year dummies in Model (2). As legal institutions are country-level variables, we cluster the standard errors at the country level to obtain conservative results.

4. Sample selection and descriptive statistics We employ corporate financial data from the Worldscope, and stock price data from the DataStream for 40 countries and economies within a sample period from 1998 through 2006, and delete all financial companies (SIC code 6000–6999). We require all country– year–one digit SIC code combinations to have at least 20 observations, among which at least one observation involves a stock repurchase. This requirement reduces the number of countries in our sample to 30. We then estimate the performance-matched discretionary accruals and rank the market-to-book ratio, cash flows from operating activities to total assets, and discretionary accruals in each country–year–industry combination, respectively, to calculate the proxy for mimicking repurchases (Mimic). This reduced sample is further used to identify control firms and calculate post-repurchase performance. Our final sample consists of 11,422 repurchase firm–year observations (4142 unique firms) for the 1998 to 2006 period.

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

439

Table 1 Definition of variables. Variable

Code

Description

Country-level variables Legal origins

Legal

Anti-self-dealing

ASD

Log GDP per capita

LnGDP

Market capitalization to GDP

SMGDP

Dividend tax preference

DVDTax

Equals one if the origin of the company law or commercial code is common law, and otherwise zero. An index that measures the protections of minority shareholders against self-dealing transactions benefiting controlling shareholders, constructed by Djankov et al. (2008). The logarithm of the per capita Gross Domestic Product (in U.S. dollars) as recorded in the World Bank database. Ratio of total stock market capitalization over GDP, from the World Bank database. The difference between taxes on income and capital gain from La Porta et al. (2000).

Firm-level variables Likelihood of mimicking repurchase

Mimic

Lagged insider ownership

LOwn

Cash flow performance Share repurchase

CFO Rep

Share issuance

Issue

Cash dividends Capital expenditure

DVD CPTEXP

Intangible assets

ITA

Needs of external financing

ExtFin

Lagged firm size

LSize

Lagged leverage ratio Lagged annual stock return

LLev LRet

Industry-level variable Industry concentration

HHI

The average of the sum of the ranks for market-to-book ratio, cash flow from operating activities to total assets (multiplied by −1), and performance-matched discretionary accruals at the beginning of the year. The rank of market-to-book ratio is obtained by classifying the country–year–industry–market value of equity portfolio into five quintiles, and the ranks of operating cash flows to total assets and discretionary accruals are obtained by classifying the country–year– industry combination into five quintiles. Industry is defined by a one-digit SIC code. The percentage of insider ownership at the beginning of the repurchase year, where insider ownership is measured by the closely held block holdings incorporated in the Worldscope database. Net cash flows from operating activities, scaled by average total assets. Cash paid for repurchases, scaled by lagged market valuation of equity. The proceeds from new share issuances or treasury shares resale during the repurchase year, scaled by lagged market valuation of equity. Cash dividends paid to shareholders, scaled by average total assets. Funds used to acquire fixed assets other than acquisitions, scaled by average total assets. The assets not having a physical existence, scaled by average total assets. An indicator variable that equals one if capital expenditure is less than cash flow from operating activities, and otherwise, zero. The logarithm of total market capitalization in U.S. dollars at the beginning of the year. Total liabilities divided by total assets at the beginning of the year. Dividends adjusted raw annual stock return in the year prior to stock repurchases. The Herfindahl–Hirschman index, which is the sum of the squared market share of each firm in the same industry during a year. Market share is defined as the total sales of the firm in a given year divided by the total sales of the industry in the same year. The industry is defined at the two-digit SIC code level.

440

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

However, because of constraints imposed by control firms and the requirement for three-year post-repurchase performance data, there are only 5390 firm–year observations (2733 unique firms) in the operating cash flow performance analysis, 5449 firm–year observations (2715 unique firms) in the operating income performance analysis, and 5947 firm–year observations (2832 unique firms) in the market performance analysis. We focus on actual stock repurchases rather than repurchase announcements for two reasons. First, most of our sample countries do not require the disclosure of repurchase announcements, with data infrequently reported and unreliable.13 Second, repurchase announcements do not commit the firm to actually buying back shares. Lie (2005) points out that without the likelihood of a commitment to follow up announcements with an actual repurchase, the signaling models unravel due to the absence of a signaling cost. Stephens and Weisbach (1998) and Allen and Michaely (2003) both suggest that the dollar amounts spent on repurchases as reported in cash flow statements are likely to yield the least biased estimates of the actual dollar amount spent on such repurchases. Therefore, we estimate actual stock repurchases in a given year using Worldscope data item #04751 (Common/Preferred Redeemed, Retired, Converted, etc.). This item is an aggregation of many other types of transactions besides stock repurchases, including purchases of treasury stock and conversions of preferred stock into common stock. As this item may overestimate actual repurchases, we attempt to reduce the noise associated with Worldscope item #04751 through a sensitivity analysis — i.e., we estimate actual repurchases by excluding firms the volume of whose preferred shares decreased during the fiscal year of repurchases.14 Table 2 presents the number of firm–year observations across countries, the country-level average of likelihood for mimicking repurchase activities (Mimic), and descriptive statistics for other country-level variables. There is a significant variation in the number of firm–year observations between countries. Stock repurchases are most common in Japan, with 5608 observations, which account for 50% of our sample.15 On average, Spain exhibits the highest likelihood of mimicking repurchases, and Brazil has the lowest. Table 2 shows substantial cross-country variations in GDP per capita, stock market capitalization over GDP (SMGDP), and dividend tax preference (DVDTax). Table 3 provides summary statistics for the firm-level independent variables. To mitigate the effect of outliers, the firm-level data are winsorized at the 1st and 99th percentile. We classify the sample into four intervals based on the value of Mimic and report the mean value of each variable by intervals. Cash flow performance (CFO), cash paid for repurchases (Rep), and capital expenditure (CPTEXP) monotonically decrease

13 For example, listed firms in Hong Kong must obtain authorization to buy back their own stock at their shareholders' meetings (which become announcement events), but we find that firms securing such authorization frequently do not undertake repurchases within the fiscal year. 14 To ascertain the accuracy of the estimated actual repurchases, we manually collect actual repurchases from the websites of four sample economy stock exchanges: Hong Kong, South Korea, Malaysia, and Taiwan. We find that the correlation coefficient of the amount of repurchases between the manually collected repurchases for these four countries and their estimated repurchase data on Worldscope is 0.98. This suggests that the repurchase data based on Worldscope are reasonably reliable. 15 We perform sensitivity tests without Japanese firms and report the results in Table 8.

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

441

Table 2 Country-level variables. N refers to the number of firm–year observations in the 1998–2006 sample period. All variables are defined in Table 1. Country

N

Mimic

Legal

ASD

GDP per capita

SMGDP

DVDTax

Australia Austria Belgium Brazil Canada Chile Demark Finland France Germany Hong Kong India Indonesia Israel Italy Japan Korea (South) Malaysia Mexico Netherlands Norway Philippines Singapore South Africa Spain Sweden Switzerland Taiwan Thailand UK Total Mean Median

354 18 5 17 468 11 120 57 360 230 374 79 12 24 53 5608 888 299 11 147 56 26 94 141 6 92 265 381 16 1210 11,422 381 93

2.58 2.70 2.40 2.08 2.61 3.06 2.70 2.74 2.84 2.78 2.80 2.89 3.17 2.72 2.65 2.87 2.82 2.63 2.58 2.76 2.92 2.53 2.74 2.56 3.72 2.55 2.87 2.69 2.79 2.66

1 0 0 0 1 0 0 0 0 0 1 1 0 1 0 0 0 1 0 0 0 0 1 1 0 0 0 0 1 1

0.76 0.21 0.54 0.27 0.64 0.63 0.46 0.46 0.38 0.28 0.96 0.58 0.65 0.73 0.42 0.5 0.47 0.95 0.17 0.2 0.42 0.22 1 0.81 0.37 0.33 0.27 0.56 0.81 0.95

36,427 35,971 37,230 4653 33,571 7149 45,973 36,450 33,628 33,175 22,583 719 1117 19,721 30,230 34,882 16,239 5162 7724 37,769 63,699 1073 25,866 5020 26,448 39,079 49,048 15,255 2615 35,796

1.16 0.24 0.88 0.47 1.08 1.03 0.60 1.36 0.83 0.50 3.28 0.53 0.21 0.79 0.48 0.79 0.57 1.38 0.24 1.17 0.51 0.44 2.00 1.87 0.94 1.08 2.48 1.30 0.61 1.00

0.90 0.78 0.74

0.77 0.70 0.79 0.68 1.00 0.40 1.08 1.05 0.96 0.85 0.72 1.03 0.56 0.60 0.90 0.83

2.75 2.73

0.33 0

0.52 0.46

24,809 28,339

1.01 0.86

0.81 0.78

0.89 0.67 1.07 0.63 0.86 1.00 0.58 0.76

with Mimic, but the proceeds from the issuance of new shares and/or the resale of treasury shares (Issue) monotonically increase with Mimic. Interestingly, in the two highest intervals of Mimic, the mean values of Issue (2.1% and 2.6% of market value of equity, respectively) exceed those of Rep (1.9% and 1.5% of market value of equity, respectively), indicating that these firms potentially benefit from mimicking repurchases in terms of net cash inflows. The need for external financing (ExtFin) generally increases with Mimic, suggesting that firms that need external financing are more likely to undertake mimicking repurchases. Lagged leverage ratio (LLev) and lagged annual stock market return (LRet) increase with Mimic, suggesting that mimicking firms have higher leverage ratios and experience increases in stock prices prior to stock repurchases. Table 3 also demonstrates considerable cross-sectional variation in each variable. For example, the maximum insider ownership (LOwn) is 87.6%, and the

442

Variables

LOwn CFO Rep Issue DVD CPTEXP ITA ExtFin LSize LLev LRet HHI

Mean 1 ≤ Mimic b 2

2 ≤ Mimic b 3

3 ≤ Mimic b 4

4 ≤ Mimic ≤ 5

0.386 0.091 0.026 0.014 0.014 0.061 0.055 0.296 12.321 0.488 0.144 0.194

0.379 0.089 0.022 0.015 0.017 0.053 0.064 0.252 12.894 0.504 0.190 0.200

0.384 0.068 0.019 0.021 0.015 0.042 0.055 0.317 12.665 0.518 0.167 0.181

0.420 0.046 0.015 0.026 0.012 0.036 0.046 0.394 12.236 0.540 0.213 0.176

Min

Median

Max

Std. dev.

0.000 − 0.169 0.000 0.000 0.000 0.001 0.000 0.000 8.841 0.075 − 0.734 0.000

0.382 0.072 0.006 0.000 0.008 0.036 0.011 0.000 12.462 0.516 0.732 0.117

0.876 0.329 0.300 0.617 0.121 0.262 0.613 1.000 17.579 1.019 2.701 1.000

0.200 0.078 0.037 0.076 0.021 0.046 0.114 0.457 1.804 0.211 0.552 0.191

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Table 3 Descriptive statistics of firm-level independent variables. All variables are defined in Table 1. The sample contains 11,422 firm–year observations.

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

443

minimum is 0.0%. The median of stock repurchases is 0.6% of the market value of equity, with a maximum amount of 30.0%.16 The Pearson correlation coefficients in Table 4 show that the likelihood of mimicking (Mimic) displays the predicted signs of correlation with most of the independent variables. For example, Mimic is negatively correlated with the two measures for legal institutions (Legal and ASD) and positively correlated with lagged ownership concentration (LOwn), which is consistent with our hypotheses. Mimic is also negatively correlated with CFO, Rep, DVD, CPTEXP, ITA, and LSize, and positively correlated with Issue, LLev, and LRet, which is in general consistent with the patterns shown in Table 3. 5. Empirical results 5.1. Long-term performance following stock repurchases The number of observations in each Mimic interval varies, because Mimic is obtained by ranking the market-to-book ratio, operating cash flows to total assets, and discretionary accruals relative to an all-stocks universe based on country–year–industry combinations. As a higher value of Mimic represents a greater likelihood of mimicking, we predict that repurchasing firms that fall into the lower intervals of Mimic (i.e., which are more likely to be genuine) will exhibit improvements in long-term operating and market performance following repurchases, but those that rank in the higher intervals of Mimic (i.e., which are more likely to be mimickers) may not. 5.1.1. Operating performance Table 5 presents both levels of, and changes in, long-term operating performance following stock repurchases for the full sample and for each of the four intervals. Panel A shows that during the year of repurchase (Year 0), the mean and median values of the abnormal operating cash flows are 0.002 and 0.000, respectively for the “all” sample and both are statistically different from zero at the 1% level. Given the small absolute magnitudes, these levels of statistical significance arise from low standard deviations of differences between the performances of repurchasing firms and the performance-matched control firms in the years of repurchase.17 The levels of abnormal operating cash flow performance monotonically decrease with Mimic intervals in each of the three years following repurchases. More importantly, repurchasing firms ranked in the first two intervals (1 ≤ Mimic b 2 and 2 ≤ Mimic b 3) outperform their control counterparts, but repurchasing firms in the fourth interval (4 ≤ Mimic b 5) do not. Panel B of Table 5 shows that changes in abnormal operating cash flows from year 0 to each of the three following years exhibit significant improvements for the whole sample, consistent with Lie (2005). All years in the first two intervals exhibit significant improvements, but those in the fourth interval do not. In particular, the means (medians) of 16 As Worldscope item #04751 may include repurchases other than open market repurchases (such as tender offers), we exclude repurchases larger than 10% of market value of equity from our sample in a sensitivity test. 17 Lie (2005) reports similar results.

444

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Table 4 Pearson correlations between variables. This table presents the Pearson correlation coefficients between variables. All variables are defined in Table 1. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively.

Legal ASD LOwn CFO Rep Issue DVD CPTEXP ITA ExtFin LSize LLev LRet HHI LnGDP SMGDP

Mimic

Legal

ASD

LOwn

CFO

Rep

Issue

− 0.094*** − 0.070*** 0.047 − 0.197*** − 0.081*** 0.058*** − 0.052*** − 0.176*** − 0.038*** 0.084*** − 0.044*** 0.077*** 0.019** − 0.048*** 0.029*** − 0.042***

0.883 − 0.133 0.177*** 0.177 0.059*** 0.368*** 0.164*** 0.273*** − 0.023** − 0.065*** − 0.052*** 0.010 0.230*** − 0.243*** 0.541***

− 0.115*** 0.107*** 0.118*** 0.026*** 0.333*** 0.083*** 0.130*** − 0.016* − 0.090*** − 0.072*** − 0.013 0.050*** − 0.201*** 0.532***

− 0.034*** − 0.072*** − 0.023** − 0.062*** − 0.064*** − 0.117*** 0.021** − 0.239*** − 0.109*** 0.034*** − 0.063*** − 0.047*** − 0.020**

0.064*** − 0.107*** 0.388*** 0.348*** 0.094*** − 0.513*** 0.225*** − 0.096*** 0.097*** 0.157*** − 0.026*** 0.106***

0.087*** 0.071*** 0.006 0.053*** − 0.020** − 0.076*** 0.010 − 0.049*** 0.149*** − 0.100*** 0.064***

− 0.087*** 0.040*** 0.078*** 0.112*** − 0.151*** 0.145*** 0.020** 0.071*** − 0.096*** 0.016*

the changes in abnormal operating cash flows from year 0 to year + 1 for the four intervals are 0.017 (0.012), 0.013 (0.008), 0.001 (− 0.001), and − 0.002 (− 0.006), respectively, revealing a monotonic decrease across the Mimic intervals. The figures for the first two intervals are significant at the 1% level, but those of the third and fourth intervals are insignificant at conventional levels. The changes in abnormal operating cash flows from year 0 to year + 2 (and year + 3) exhibit similar patterns. Although operating income may not be reliable for performance evaluation due to potential accrual management prior to repurchase transactions, we report the levels of and changes in abnormal operating income performance in Panels C and D of Table 5 for comparison. In general, the patterns are consistent with those in Panels A and B. Taken together, the results in Table 5 suggest that, unlike genuine repurchases, potential mimickers do not demonstrate improvements in operating performance following their stock repurchases. 5.1.2. Market performance Table 6 reports market performance following repurchases. Panel A shows the annual abnormal buy-and-hold returns (BHRs), and Panel B shows the cumulative BHRs. For the whole sample, the repurchasing firms outperform their control firms in each of the three years following repurchases (except for the third year), consistent with previous studies (e.g., Ikenberry et al., 1995; Stephens & Weisbach, 1998). Panel A shows that, similar to operating performance in Table 5, market performance following repurchases decreases monotonically with the Mimic intervals in each year. Panel B shows significant improvements in cumulative BHRs for the first three Mimic intervals, but not for the fourth interval in any of the three years following repurchases. For example, the means (medians) of cumulative abnormal BHRs for the first three intervals from year 0 to year + 1 are 0.507 (0.343), 0.187 (0.170), and 0.112 (0.105), respectively,

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

DVD

CPTEXP

ITA

ExtFin

LSize

LLev

LRet

HHI

445

LnGDP

0.113*** 0.138*** − 0.040*** − 0.181*** 0.256*** − 0.098*** 0.178*** 0.136*** 0.173*** − 0.160*** − 0.184*** 0.006 0.045*** 0.072*** 0.049*** 0.078*** 0.081*** 0.027*** − 0.017* 0.104*** 0.018* 0.195*** 0.151*** 0.181*** − 0.028*** 0.061*** 0.047*** 0.020** − 0.068*** − 0.069*** 0.100*** − 0.045*** 0.187*** 0.008 − 0.016* − 0.098*** 0.292*** 0.062*** 0.112*** − 0.034*** 0.001 − 0.105*** 0.073*** 0.172*** − 0.019**

most of which are significant at the 1% level. However, the corresponding mean and median for the fourth interval (− 0.119 and − 0.100) are not positively significant. The cumulative abnormal BHRs from year 0 to year + 2, and year + 3 exhibit similar patterns. Taken together, these performance results suggest that repurchasing firms in the highest Mimic intervals show little or no improvement in post-repurchase performance in contrast to signaling of undervaluation. Interestingly, Tables 5 and 6 indicate that operating and market performance following repurchases decrease monotonically with Mimic intervals, suggesting that Mimic appears to be an effective ex-ante proxy for capturing managerial intention to conduct opportunistic or mimicking repurchases. 5.2. Regression results for the effects of legal institutions and ownership concentration on mimicking repurchases Table 7 reports regression results. In column 1, the coefficient on legal origins (Legal) is − 0.167, significant at the 5% level, suggesting that in countries with civil law origins, repurchasing firms are more likely to demonstrate higher market-to-book ratios, lower operating cash flows to total assets, and/or higher levels of discretionary accruals (on average) before actual stock repurchases. This result contradicts conventional characterizations of stock repurchases. Thus, we conclude that repurchasing firms in civil law countries are more likely to use repurchasing as a tool to mislead investors, consistent with our first hypothesis. Further, the coefficient on the lagged corporate insider ownership (LOwn) is 0.388, significant at the 1% level. To the extent that corporate insider ownership represents a capacity for expropriation, this result indicates that entrenched insiders undertaking stock repurchases are more likely to mimic undervalued firms, consistent with Hypothesis 2. Column 1 also demonstrates that the coefficient is significantly negative for Rep, but

446

Panel A: Levels of abnormal operating cash flow Mimic intervals

All 1≤ 2≤ 3≤ 4≤

Mimic Mimic Mimic Mimic

b2 b3 b4 ≤5

N

Year 0

5390 698 2320 1821 551

Year 1

Year 2

Year 3

Mean

Median

Mean

Median

Mean

Median

Mean

Median

0.002*** 0.002*** 0.002*** 0.001*** 0.000

0.000*** 0.000*** 0.000*** 0.000*** 0.000

0.010*** 0.018*** 0.015*** 0.003 − 0.002

0.006*** 0.015*** 0.009*** − 0.001 − 0.005

0.012*** 0.015*** 0.017*** 0.009*** − 0.002

0.005*** 0.014*** 0.007*** 0.002* − 0.005

0.013*** 0.016*** 0.018*** 0.009*** − 0.004

0.007*** 0.016*** 0.013*** 0.003* − 0.006*

Panel B: Changes in abnormal operating cash flow Mimic intervals

N

All 1≤ 2≤ 3≤ 4≤

5390 698 2320 1821 551

Mimic Mimic Mimic Mimic

b2 b3 b4 ≤5

Year 0 to +1

Year 0 to + 2

Year 0 to + 3

Mean

Median

Mean

Median

Mean

Median

0.008*** 0.017*** 0.013*** 0.001 − 0.002

0.004*** 0.012*** 0.008*** − 0.001 − 0.006

0.019*** 0.030*** 0.029*** 0.010** − 0.005

0.009*** 0.023*** 0.013*** 0.003 − 0.007

0.033*** 0.045*** 0.049*** 0.019*** − 0.008

0.017*** 0.036*** 0.026*** 0.007* − 0.017

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Table 5 Levels and changes of operating performance following stock repurchases. This table reports the levels of and changes in abnormal operating performance following actual stock repurchases. Operating performance is measured as either net operating cash flow or operating income scaled by the average cash-adjusted assets (i.e., book value of assets less cash and short-term investments) at the beginning and end of the fiscal year. Year 0 is the fiscal year of actual repurchases. Abnormal performance is defined as the paired difference between the performance of sample firms and that of matched control firms. Control firms are constructed based on a market-to-book ratio at the beginning of the year and the operating performance during the repurchase year. Mimic is a proxy measure for the likelihood of mimicking repurchases. N is the number of sample observations with the available data. We partition the sample into four equal intervals based on the value of Mimic. Because Mimic is obtained by ranking the market-to-book ratio, the operating cash flows, and the discretionary accruals relative to an all-stock universe based on a country–year–industry combination, the number of observations in each interval varies. To mitigate the effect of outliers, the data have been trimmed at the 1st and 99th percentile. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively.

Panel C: Levels of abnormal operating income N

Year 0 Mean

Median

Mean

Median

Mean

Median

Mean

Median

All 1≤ 2≤ 3≤ 4≤

5449 708 2333 1806 602

0.001*** 0.001*** 0.001*** 0.001*** 0.001

0.000*** 0.000** 0.000*** 0.000*** 0.000

0.007*** 0.005** 0.008*** 0.007*** 0.004

0.003*** 0.003* 0.004*** 0.003*** 0.002

0.010*** 0.006** 0.013*** 0.011*** 0.004

0.005*** 0.003* 0.005*** 0.005*** 0.007*

0.010*** 0.003 0.014*** 0.012*** − 0.001

0.006*** 0.005 0.006*** 0.007*** 0.003

Mimic Mimic Mimic Mimic

b2 b3 b4 ≤5

Year 1

Year 2

Year 3

Panel D: Changes in abnormal operating income Mimic intervals

N

Year 0 to +1 Mean

Median

Mean

Median

Mean

Median

All 1≤ 2≤ 3≤ 4≤

5449 708 2333 1806 602

0.006*** 0.004* 0.007*** 0.006*** 0.003

0.002*** 0.002 0.004*** 0.002*** 0.001

0.017*** 0.011** 0.020*** 0.018*** 0.008

0.008*** 0.003* 0.009*** 0.008*** 0.006

0.026*** 0.014* 0.033*** 0.030*** 0.005

0.012*** 0.010** 0.016*** 0.011*** 0.007

Mimic Mimic Mimic Mimic

b2 b3 b4 ≤5

Year 0 to +2

Year 0 to + 3

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Mimic intervals

447

448

Panel A: Annual abnormal BHRs Mimic intervals

N

All 1≤ 2≤ 3≤ 4≤

5947 782 2408 1991 766

Mimic Mimic Mimic Mimic

b2 b3 b4 ≤5

Year 0

Year 1

Year 2

Year 3

Mean

Median

Mean

Median

Mean

Median

Mean

Median

0.037*** 0.144*** 0.048*** 0.021 − 0.062**

0.031*** 0.130*** 0.046*** 0.008* − 0.038**

0.030*** 0.081*** 0.029* 0.022 − 0.001

0.036*** 0.045*** 0.051*** 0.026*** − 0.001

0.023*** 0.048** 0.020 0.021 0.016

0.027*** 0.069*** 0.017** 0.037*** − 0.007

0.011 0.006 0.002 0.019 0.026

0.006 0.009 0.000 0.006 0.024

Panel B: Cumulative abnormal BHRs Mimic intervals

All 1≤ 2≤ 3≤ 4≤

Mimic Mimic Mimic Mimic

b2 b3 b4 ≤5

N

5947 782 2408 1991 766

Year 0 to +1

Year 0 to + 2

Year 0 to +3

Mean

Median

Mean

Median

Mean

Median

0.165*** 0.507*** 0.187*** 0.112** − 0.119

0.129*** 0.343*** 0.170*** 0.105*** − 0.100*

0.497*** 1.373*** 0.546*** 0.355*** − 0.182

0.373*** 1.120*** 0.357*** 0.359*** − 0.270*

1.198*** 3.197*** 1.143*** 1.031*** − 0.232

0.760*** 2.545*** 0.626*** 0.706*** − 0.605

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Table 6 Levels and changes of market performance following stock repurchases. This table shows the levels of and changes in abnormal market performance following actual stock repurchases. Market performance is measured as buy-and-hold stock returns (BHRs). Year 0 is the fiscal year of actual repurchases. Abnormal performance is defined as the paired difference between the performance of sample firms and that of matched control firms. Control firms are constructed based on the market value of equity and the market-to-book ratio at the beginning of the year of the repurchase. Mimic is a proxy measure for the likelihood of mimicking repurchases. N is the number of sample observations with the available data. We partition the sample into four equal intervals based on the value of Mimic. Because Mimic is obtained by ranking the market-to-book ratio, the operating cash flows, and the discretionary accruals relative to an all-stock universe based on country–year–industry combination, the number of observations in each interval varies. To mitigate the effect of outliers, the data have been trimmed at the 1st and 99th percentile. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively.

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

449

Table 7 Effects of legal institutions and ownership concentration on mimicking repurchases. This table presents the ordered logistic regression results from Model (2). The dependent variable is Mimic or the likelihood of mimicking repurchases. All variables are defined in Table 1. The z-statistics are reported in parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. (1)

(2)

(3)

(4)

− 0.167** (− 2.23) 0.388*** (3.86)

− 0.190*** (− 3.31) 0.381*** (3.68)

− 2.600*** (− 11.60) − 3.498*** (− 4.27) 1.147*** (7.55) 5.358*** (4.40) − 6.784*** (− 12.06) − 0.543*** (− 4.25) 0.300*** (4.92) 0.012 (0.91) 0.679*** (3.30) 0.001*** (2.87) − 0.032 (− 0.25) 0.034 (0.57) − 0.015 (− 0.42)

− 0.055 (− 0.68) 0.516*** (5.18) − 0.334** (− 2.48) − 2.606*** (− 11.83) − 3.476*** (− 4.21) 1.151*** (7.51) 5.293*** (4.53) − 6.790*** (− 12.12) − 0.551*** (− 4.20) 0.301*** (4.84) 0.012 (0.91) 0.677*** (3.27) 0.001*** (2.87) − 0.023 (− 0.18) 0.022 (0.37) − 0.003 (− 0.09)

− 2.603*** (− 11.60) − 3.510*** (− 4.18) 1.128*** (7.16) 5.322*** (4.29) − 6.793*** (− 11.92) − 0.511*** (− 3.99) 0.299*** (4.89) 0.013 (1.06) 0.680*** (3.33) 0.001*** (2.80) − 0.043 (− 0.35)

− 0.057 (− 0.76) 0.522*** (5.21) − 0.358*** (− 2.63) − 2.609*** (− 11.87) − 3.485*** (− 4.15) 1.140*** (7.11) 5.281*** (4.49) − 6.800*** (− 12.00) − 0.534*** (− 4.14) 0.301*** (4.81) 0.012 (1.02) 0.677*** (3.28) 0.001*** (2.82) − 0.027 (− 0.22)

Year & industry dummies Clustered by

Yes Country

Yes Country

Yes Country

Yes Country

Pseudo R2 N

2.02% 11,422

2.02% 11,422

2.12% 11,422

2.13% 11,422

Legal LOwn Legal ∗ LOwn CFO Rep Issue DVD CPTEXP ITA ExtFin LSize LLev LRet HHI LnGDP SMGDP

positive for Issue, indicating that mimicking firms buy back fewer shares, but issue more shares. Because stock issuance or treasury stock resale can compensate for the cash consumed on stock repurchases, this finding suggests that mimicking repurchases could be less costly than genuine repurchases. This result is corroborated by the positive coefficient on the needs for external financing (ExtFin), which suggests that firms needing external financing are more likely to undertake

450

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

mimicking repurchases. These findings potentially explain the phenomenon described by Fama and French (2005), that many firms that are net issuers in a given year repurchase stock in the same year, and that many of the firms making net repurchases have financing deficits. The coefficients on CFO, CPTEXP, and ITA are significantly negative, indicating that mimicking repurchases are related to poor operating performance, lower investment in capital expenditure, and fewer intangible assets. The coefficient on dividend payment (DVD) is significantly positive, suggesting that mimicking firms pay higher dividends. This finding is consistent with Faccio et al. (2001), who report that the controlling shareholders pay higher dividends. The positive coefficients on LLev and LRet indicate that mimicking firms exhibit both higher leverage ratios and greater stock returns prior to repurchases. As one component of our proxy for mimicking repurchase, market-to-book ratio, could be interpreted as a growth measure, one potential concern is that our findings may be driven by growth firms. For example, both the descriptive statistics in Table 3 and the regression analysis in Table 7 reveal that firms with a higher likelihood of mimicking are characterized by a greater amount of new share issuances, a smaller amount of stock repurchase, and a greater need for external financing — all of which are typical characteristics of growth firms. Although we acknowledge such a possibility, Table 7 shows that mimicking firms have lower capital expenditures, which is unlikely to be a feature of growth firms.18 To test Hypothesis 3, we add an interaction term between legal origins and lagged closely held block shareholdings (Legal ∗ LOwn). As column 2 reveals, the coefficient on Legal ∗ LOwn is − 0.334 and significant at the 5% level, which suggests that the positive effect of ownership concentration on mimicking repurchases is constrained in common law countries, supporting our third hypothesis. The coefficients on the control variables are similar to those in column 1. The GDP per capita and the stock market capitalization over GDP highly correlate with legal origins (as shown in Table 4) and may affect the coefficient on legal origins. We exclude LnGDP and SMGDP in columns 3 and 4, and the regression results are not sensitive to the exclusion of these variables. 5.3. Robustness tests In this section, we perform a set of sensitivity tests to assure that the documented results are robust. 5.3.1. Excluding Japanese firms As Japanese firms account for the largest portion of our sample, and thus may drive the results, we repeat the regressions in Table 7 after excluding Japanese firms from the sample, with results reported in columns 1 and 2 of Table 8. Coefficients on Legal, LOwn, and Legal ∗ LOwn are significant with predicted signs, indicating that our main results are robust to the exclusion of Japanese firms. 18

We thank the discussant for pointing this out.

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

451

5.3.2. Refining the measure of stock repurchase As we mentioned earlier, the Worldscope data item #04751 (Common/Preferred Redeemed, Retired, Converted, etc.) is an aggregation of different types of buybacks that may introduce measurement error. Previous studies document that firms may buy back shares to avoid EPS dilution from the exercise of stock options. However, such data are unavailable in most countries outside the U.S. To control for the effect of this motive, we delete observations with repurchase amounts less than 0.01% of a firm's market value, as the repurchase amounts for stock option programs are assumed to be smaller. As repurchase data from Worldscope do not distinguish between different types of stock repurchase, we also delete observations with repurchase amounts greater than 10% of a firm's market value to rule out the effects of tender offers. The regression results, after excluding the large and small amounts of repurchases, are reported in columns 3 and 4 of Table 8. The results are virtually identical to those in Table 7. Furthermore, as the data on actual stock repurchases from Worldscope may incorporate repurchases of preferred shares, we exclude observations on firms whose preferred shares decreased during the repurchase year. The results, shown in columns 5 and 6 of Table 8, remain unchanged. 5.3.3. Alternate measure of mimicking repurchases Previous studies show that discretionary accruals are influenced by legal institutions and ownership structures (Haw et al., 2004; Leuz et al., 2003). Because our proxy for mimicking repurchases contains discretionary accruals, our findings may reflect relationships between earnings quality and investor protections and/or ownership structures, rather than mimicking behavior. To address this concern, we remove the accrual component from our composite measure of Mimic. Thus, we re-compute Mimic based on the average of rankings between the market-to-book ratio and the cash flows from operating activities to total assets (multiplied by − 1) prior to stock repurchases. We rerun Model (2) and present the regression results in columns 7 and 8 of Table 8. Coefficients on Legal, LOwn, and Legal ∗ LOwn are significant with predicted signs, and the pseudo R2s are somewhat larger than those in Table 7. 5.3.4. Alternate measure of legal institutions We use an alternate measure of investor protections, the anti-self-dealing index (ASD), to test the effect of legal institutions on mimicking behavior. As proposed by Djankov, La Porta, Lopez-de-Silanes, and Shleifer (2008), ASD measures the level of legal protections available to minority shareholders against expropriation by corporate insiders. ASD focuses on private enforcement mechanisms (such as disclosure, approval, and litigation) that govern a specific self-dealing transaction. Djankov et al. (2008) suggest that this theoretically grounded index predicts a variety of stock market outcomes, and generally works better than the previously introduced index of anti-director rights. We replace Legal with ASD in Model (2) and report the regression results in columns 9 and 10 of Table 8. The coefficient on ASD is − 0.291 (significant at the 5% level) in column 9, which suggests that mimicking repurchases are less likely to occur in countries where self-dealing transactions are more strongly controlled. The coefficient on LOwn is 0.803 (significant at the 1% level), and the coefficient on ASD ∗ LOwn is − 0.682 (significant at

452

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Table 8 Robustness tests. This table presents the ordered logistic regression results from Model (2). All variables are defined in Table 1. The z-statistics are reported parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. Excluding Japan

Excluding large and small amount of repurchases

Excluding obs. with decreased preferred shares

(1)

(2)

(3)

(4)

(5)

(6)

− 0.162* (− 1.88)

− 0.016 (− 0.16)

− 0.150** (− 2.06)

− 0.052 (− 0.66)

− 0.164** (− 1.96)

0.034 (0.38)

0.299*** (2.66)

0.532*** (3.16) − 0.432** (− 2.43)

0.332*** (3.66)

0.449*** (4.34) − 0.295** (− 2.00)

0.427*** (2.86)

0.637*** (4.37) − 0.586*** (− 3.18)

LnGDP and SMGDP Other control variables Year & industry dummies Clustered by

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Pseudo R2/R2 N

2.43% 5814

2.45% 5814

2.11% 9720

2.11% 9720

2.17% 10,763

2.19% 10,763

Legal ASD LOwn Legal ∗ LOwn ASD ∗ LOwn DVDTax

the 5% level) in column 10, which indicates that corporate insiders' incentives for expropriation are constrained in countries with stronger self-dealing controls. 5.3.5. Endogeneity issue Our regression specification assumes that our dependent variable (Mimic), the country-level legal institutions, and firm-level ownership structures are exogenous. There is a possibility that these factors are simultaneously determined. To alleviate this concern, we conduct a two-stage least square (2SLS) analysis. We use the anti-self-dealing index (ASD) to proxy for legal institutions, and use English, French, German, and Scandinavian legal origins and LnGDP in 1995 as instrumental variables for ASD and LOwn. The 2SLS estimation results shown in columns 11 and 12 of Table 8 are consistent with those reported in Table 7, suggesting that endogeneity is not a serious threat. 5.3.6. Incorporating dividend tax preference (DVDTax) Previous studies note that stock repurchases act as a substitute for dividends in distribution of capital, especially when the dividend tax rate is higher than the capital gains tax rate (Grullon & Michaely, 2002; Lie & Lie, 1999). Therefore, we include dividend tax preference (DVDTax) in our regression analysis. DVDTax is measured as the difference

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Alternative measure of mimicking

Alternative measure of legal institutions

2SLS

(7)

(8)

(9)

(11)

− 0.425*** (− 3.83)

− 0.278** (− 2.11)

0.604** (2.57)

0.749*** (3.23) − 0.433* (− 1.79)

− 0.291** (− 2.14) 0.390*** (3.96)

(10)

− 0.056 (− 0.41) 0.803*** (4.37)

− 0.134* (− 1.88) 0.720*** (2.81)

− 0.682** (− 2.43)

453

Including dividend tax preference (12)

3.235 (1.64) 7.243* (1.81)

(13)

(14)

− 0.134* (− 1.68)

− 0.031 (− 0.38)

0.390*** (3.77)

0.514*** (4.90) − 0.325** (− 2.27)

− 0.224 (− 0.89)

− 0.195 (− 0.77)

− 8.996* (− 1.74)

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

Yes Yes Yes Country

3.80% 11,422

3.81% 11,422

2.02% 11,422

2.03% 11,422

8.00% 11,422

8.00% 11,422

2.02% 11,370

2.02% 11,370

between taxes on income and on capital gains, and the data are collected from La Porta et al. (2000). Due to limited data availability, the inclusion of DVDTax reduces our sample countries from 30 to 27. The results as given in columns 13 and 14 of Table 8 show that coefficients on Legal, LOwn, and Legal ∗ LOwn are significant with predicted signs, respectively. This finding indicates that legal institutions and ownership concentration play an additional role beyond dividend tax preference in shaping mimicking behavior. 5.4. The effect of IFRS adoption on mimicking repurchases During our sample period, some countries adopted IFRS. For instance, the European Union (EU), Australia, and South Africa have adopted IFRS since 2005.19 The adoption of IFRS may enhance the credibility of a firm's information environment. Armstrong et al. (2010) find that markets react positively to firms with lower pre-adoption IFRS information disclosure, and higher pre-adoption information asymmetry. In particular, they find positive market reactions to IFRS adoption for firms with higher concentrations of ownership and negative market reactions for firms domiciled in civil law countries. 19

Information on IFRS adoption is obtained from the IFRS website.

454

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Hope, Jin, and Kang (2006) show that the scope for managerial rent extraction is reduced after the adoption of IFRS due to the greater financial reporting transparency. In this section, we use a sub-sample of firms that adopted IFRS during our sample period and test whether the adoption of IFRS affects mimicking repurchase behavior. We expect that the enhanced transparency resulting from the adoption of IFRS will mitigate corporate insiders' incentives to mimic. Moreover, as financial reporting standards are less stringently enforced in civil law countries (Ball, Robin, & Wu, 2003), we predict that the benefits from the convergence in accounting standards will be greater in common law countries than in civil law countries. We rerun the regressions for pre- and post-IFRS adoption periods separately and report results in Table 9. In columns 1 and 2, the coefficient on LOwn is 0.777 and is significant at the 1% level for the pre-adoption period, but is statistically insignificant for the post-adoption period. This is consistent with our prediction that the adoption of IFRS reduces information asymmetry and managerial rent extraction for firms with previously low information quality, and hence constrains corporate insiders' abilities to engage in mimicking repurchases. The coefficient on Legal is − 0.257 for the pre-adoption period and − 0.419 for the post-adoption period, both significant at the 5% level. The difference of the coefficients for pre- and post-adoption is also significant at the 5% level. This suggests that mimicking activities have fallen more in common law countries, consistent with their stricter enforcement of IFRS compliance. We then incorporate the interaction term, Legal ∗ LOwn, in columns 3 and 4. The coefficient on Legal ∗ LOwn is significantly negative for the pre-adoption period. However, none of the hypothesized variables are significant for the post-adoption period. A possible explanation for the insignificant interaction terms in the post-adoption Table 9 The effect of IFRS adoption on mimicking repurchases. This table compares the effect of IFRS adoption using a subset of the sample of companies that adopted IFRS in our sample period and presents the ordered logistic regression results from Model (2). The dependent variable is Mimic or the likelihood of making mimicking repurchases. All variables are defined in Table 1. The z-statistics are reported in parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. Pre-IFRS

Post-IFRS

Pre-IFRS

Post-IFRS

(1)

(2)

(3)

(4)

− 0.257** (− 2.03) 0.777*** (3.27)

− 0.419** (− 2.06) 0.023 (0.12)

− 0.052 (− 0.30) 1.090*** (3.34) − 0.625** (− 2.09)

− 0.297 (− 1.05) 0.194 (0.61) − 0.366 (− 1.04)

Control variables Year & industry dummies Clustered by

Yes Yes Country

Yes Yes Country

Yes Yes Country

Yes Yes Country

Pseudo R2 N

3.30% 1993

5.42% 1121

3.33% 1993

5.43% 1121

Legal LOwn Legal ∗ LOwn

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

455

period is that IFRS adoption reduces corporate insiders' abilities to mimic in both common and civil law countries. Overall, our results provide evidence that the adoption of IFRS helps constrain mimicking repurchases, possibly due to reduced information asymmetry and enhanced transparency.20 6. Conclusions Previous studies identify the intent to signal undervaluation and/or distribute free cash flows as major motives for undertaking stock repurchases. They generally find that long-term operating and market performance improve following stock repurchases. However, previous empirical and theoretical findings suggest that other motives can arise, which encourage false signaling or mimicking. In augmenting this interpretation, our study has two aims. First, by using a composite measure to proxy for the likelihood of mimicking repurchases, this study identifies a subset of repurchasing firms that are likely to use stock repurchases as a tool to mislead the markets for the private benefits of corporate insiders. Second, we use international data to investigate the respective roles of country-level investor protections, ownership structures, and the adoption of IFRS in mitigating signal mimicking and thus improving signal quality in the context of stock repurchases. We construct a composite measure of the likelihood of mimicking repurchases, based on the ex-ante characteristics of repurchasing firms. Using this measure on a sample of repurchasing firms across 30 countries outside the U.S. between 1998 and 2006, we find that both the levels of, and changes in, operating and market performance decrease monotonically with the likelihood of mimicking repurchases during the three years following actual stock repurchases. Instead of signaling motives of undervaluation and free cash distribution, suspected mimicking firms exhibit lower capital investment and greater need for external financing. They repurchase fewer shares, and issue more new shares (or resell more treasury shares) in the year of the repurchase. These phenomena indicate a lower cost associated with mimicking versus genuine repurchases. Considered together, these findings provide evidence that firms engage in mimicking repurchases internationally. Further findings reveal that mimicking is more likely in countries with weak investor protections and in firms with highly concentrated insider ownership. We also find that stronger legal institutions and the adoption of IFRS enhance the signaling credibility of stock repurchases by curbing mimicking behavior by corporate insiders. Consistent with the Spence (2002) argument that low-quality product owners have an incentive to imitate the signals of high-quality products, our study provides empirical evidence of it in the context of the capital markets and further evidence on the factors that can mitigate mimicking behavior. Given the limited international research to date on stock repurchases, our study highlights the importance of a country's legal institutions, firm ownership structures, and the adoption of IFRS in enhancing signal quality in the context of stock repurchase programs. Our empirical findings are subject to two caveats. First, our proxy for mimicking repurchase (Mimic) is subject to remaining measurement error. Second, we use closely held 20

We thank a referee for pointing out this important issue.

456

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

blocks of shareholdings to proxy for corporate insider ownership. One weakness of this measure is that it includes all block holdings, some of which may be unrelated to controlling shareholders, so this measure may not accurately capture the degree of control rights concentration. References Ajinkya, B., Bhojraj, S., & Sengupta, P. (2005). The association between outside directors, institutional investors and the properties of management forecasts. Journal of Accounting Research, 43(3), 343–376. Allen, F., & Michaely, R. (2003). Payout policy. In G. Constantinides, M. Harris, & R. Stulz (Eds.), Handbook of economics (pp. 337–430). Amsterdam: North-Holland. Armstrong, C. S., Barth, M. E., Jagolinzer, A. D., & Riedl, E. J. (2010). Market reaction to the adoption of IFRS in Europe. The Accounting Review, 85(1), 31–61. Bagwell, L. S., & Shoven, J. B. (1988). Share repurchase and acquisition: An analysis of which firms participate. In Alan J. Auerbach (Ed.), Corporate takeover: Causes and consequences. Chicago: University of Chicago Press. Ball, R., Robin, A., & Wu, J. (2003). Incentives versus standards: Properties of accounting income in four East Asian countries. Journal of Accounting and Economics, 36, 235–270. Barber, B. M., & Lyon, J. D. (1996). Detecting abnormal operating performance: The empirical power and specification of test statistics. Journal of Financial Economics, 41(3), 359–399. Barber, B. M., & Lyon, J. D. (1997). Detecting long-term abnormal stock returns: The empirical power and specification of test statistics. Journal of Financial Economics, 43(3), 341–372. Barth, M. E., & Kasznik, R. (1999). Share repurchases and intangible assets. Journal of Accounting and Economics, 28(2), 211–241. Bebchuk, L., Kraakman, R., & Triantis, G. (2003). Stock pyramids, cross-ownership, and dual class equity in concentrated corporate ownership. In R. K. Morck (Ed.), Concentrated corporate ownership (pp. 295–318). Chicago: National Bureau of Economic Research 2000. Bens, D. A., Nagar, V., Skinner, D. J., & Wong, M. H. F. (2003). Employee stock options, EPS dilution, and stock repurchases. Journal of Accounting and Economics, 36(1–3), 51–90. Biddle, G. C., Hilary, G., & Verdi, R. S. (2009). How does financial reporting quality relates to investment efficiency? Journal of Accounting and Economics, 48(2–3), 112–131. Black, E. L., Sellers, K. F., & Manly, T. S. (1998). Earnings management using asset sales: An international study of countries allowing noncurrent asset revaluation. Journal of Business Finance & Accounting, 25(9–10), 1287–1319. Boschen, J. F., Duru, A., Gordon, L. A., & Smith, K. J. (2003). Accounting and stock price performance in dynamic CEO compensation. The Accounting Review, 78(1), 143–168. Brav, A., Graham, J. R., Harvey, C. R., & Michaely, R. (2005). Payout policy in the 21st century. Journal of Financial Economics, 77(3), 483–527. Breed, M. D. (2001). Animal behavior (An online textbook, at www.Animalbehavioronline.com/deceit.html) Bushman, R. M., Piotroski, J. D., & Smith, A. J. (2004). What determines corporate transparency? Journal of Accounting Research, 42(2), 207–251. Chan, K., Chan, L. K. C., Jegadeesh, N., & Lakonishok, J. (2006). Earnings quality and stock returns. Journal of Business, 79(3), 1041–1082. Chan, K., Ikenberry, D. L., & Lee, I. (2004). Economic sources of gain in stock repurchases. Journal of Financial and Quantitative Analysis, 39(3), 461–480. Chan, K., Ikenberry, D. L., Lee, I., & Wang, Y. (2010). Share repurchases as a potential tool to mislead investors. Journal of Corporate Finance, 16(2), 137–158. Claessens, S., Djankov, S., Fan, J. P. H., & Lang, L. H. P. (2002). Disentangling the incentive and entrenchment effects of large shareholdings. The Journal of Finance, 57(6), 2741–2771. Claessens, S., Djankov, S., & Lang, L. H. P. (2000). The separation of ownership and control in East Asian corporations. Journal of Financial Economics, 58(1–2), 81–112. Comment, R., & Jarrell, G. A. (1991). The relative signaling power of Dutch-auction and fixed-price self-tender offers and open market share repurchases. The Journal of Finance, 46(4), 1243–1271.

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

457

Dawkins, M. S., & Guilford, T. (1991). The corruption of honest signaling. Animal Behavior, 41(5), 865–873. DeFond, M. L., & Hung, M. (2004). Investor protection and corporate governance: Evidence from worldwide CEO turnover. Journal of Accounting Research, 42(2), 269–312. DeFond, M. L., Hung, M., & Trezevant, R. (2007). Investor protection and the information content of annual earnings announcements: International evidence. Journal of Accounting and Economics, 43(1), 37–67. Dittmar, A. K. (2000). Why do firms repurchase stock? Journal of Business, 73(3), 331–355. Dittmar, A. K., Mahrt-Smith, J., & Servaes, H. (2003). International corporate governance and corporate cash holdings. Journal of Financial and Quantitative Analysis, 38(1), 111–133. Djankov, S., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2008). The law and economics of self-dealing. Journal of Financial Economics, 88(3), 430–465. Doidge, C., Karolyi, G. A., & Stulz, R. M. (2007). Why do countries matter so much for corporate governance? Journal of Financial Economics, 86(1), 1–39. Dyck, A., & Zingales, L. (2004). Private benefits of control: An international comparison. The Journal of Finance, 59(2), 537–600. Easterbrook, F. H. (1984). Two agency-cost explanations of dividends. The American Economic Review, 74, 650–659. Eije, H. V., & Megginson, W. L. (2008). Dividends and share repurchases in the European Union. Journal of Financial Economics, 89(2), 347–374. Faccio, M., & Lang, L. H. P. (2002). The ultimate ownership of Western Europe corporations. Journal of Financial Economics, 65(3), 365–395. Faccio, M., Lang, L. H. P., & Young, L. (2001). Dividends and expropriation. American Economic Review, 91, 54–78. Fama, E. F., & French, K. R. (2005). Financing decision: Who issues stock? Journal of Financial Economics, 76(3), 549–582. Farinha, J. (2003). Dividend policy, corporate governance and the managerial entrenchment hypothesis: An empirical analysis. Journal of Business Finance & Accounting, 30(9–10), 1173–1209. Francis, J., Schipper, K., & Vincent, L. (2005). Earnings and dividends informativeness when cash flow rights are separated from voting rights. Journal of Accounting and Economics, 39(2), 329–360. Gong, G., Louis, H., & Sun, A. X. (2008). Earnings management and firm performance following open-market repurchases. The Journal of Finance, 63(2), 947–986. Grullon, G., & Michaely, R. (2002). Dividends, share repurchases, and the substitution hypothesis. The Journal of Finance, 57(4), 1649–1684. Gugler, K. (2003). Corporate governance, dividend payout policy, and the interrelation between dividends, R&D, and capital investment. Journal of Banking & Finance, 27(7), 1297–1321. Gup, B. E., & Nam, D. (2001). Stock buybacks, corporate finance, and EVA. Journal of Applied Corporate Finance, 14(1), 99–110. Harford, J., Mansi, S., & Maxwell, W. (2008). Corporate governance and firm cash holdings in the US. The Journal of Financial Economics, 87(3), 535–555. Haw, I. -M., Ho, S., Hu, B., & Zhang, X. (2011). Contribution of stock repurchases to the value of firm and cash holding around the world. Journal of Corporate Finance, 17(1), 152–166. Haw, I. -M., Ho, S., & Li, A. (2011). Corporate governance and earnings management by classification shifting. Contemporary Accounting Research, 28(2), 517–553. Haw, I. -M., Hu, B., Hwang, L. S., & Wu, W. (2004). Ultimate ownership, income management, and legal and extra-legal institutions. Journal of Accounting Research, 42(2), 423–462. Haw, I. -M., Hu, B., Lee, J., & Wu, W. (2012). The investor protection and price informativeness about future earnings: International evidence. Review of Accounting Studies, 17(2), 389–419. Hribar, P., Jenkins, N. T., & Johnson, W. B. (2006). Stock repurchases as an earnings management device. Journal of Accounting and Economics, 41(1), 3–27. Hope, O., Jin, J., & Kang, T. (2006). Empirical evidence on jurisdictions that adopt IFRS. Journal of International Accounting Research, 5(2), 1–20. Ikenberry, D. L., Lakonishok, J., & Vermaelen, T. (1995). Market under-reaction to open market share repurchases. Journal of Financial Economics, 39(2–3), 181–208. Jagannathan, M., Stephens, C. P., & Weisbach, M. S. (2000). Financial flexibility and the choice between dividends and stock repurchases. Journal of Financial Economics, 57(3), 355–384.

458

In-Mu Haw / The International Journal of Accounting 48 (2013) 427–458

Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance and takeovers. American Economic Review, 76, 323–339. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360. Jung, K., Kim, Y., & Stulz, R. M. (1996). Timing, investment opportunities, managerial discretion, and the security issue decision. Journal of Financial Economics, 42(2), 159–185. Karamanou, I., & Vafeas, N. (2005). The association between corporate boards, audit committees, and management earnings forecasts: An empirical analysis. Journal of Accounting Research, 43(3), 453–486. Kothari, S. P., Leone, A. J., & Wasley, C. E. (2005). Performance matched discretionary accrual measures. Journal of Accounting and Economics, 39(1), 163–197. La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (1998). Law and finance. Journal of Political Economy, 106(6), 1113–1155. La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (1999). Corporate ownership around the world. The Journal of Finance, 54(2), 471–517. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (2000). Agency problems and dividend policies around the world. The Journal of Finance, 55(1), 1–33. Leuz, C., Nanda, D., & Wysocki, P. D. (2003). Earnings management and investor protection: An international comparison. Journal of Financial Economics, 69(3), 505–527. Lie, E. (2005). Operating performance following open market share repurchase announcements. Journal of Accounting and Economics, 39(3), 411–436. Lie, E., & Lie, H. J. (1999). The role of personal taxes in corporate decisions: An empirical analysis of share repurchases and dividends. Journal of Financial and Quantitative Analysis, 34(4), 533–552. Lins, K. L. (2003). Equity ownership and firm value in emerging market. Journal of Financial and Quantitative Analysis, 38(1), 159–184. Massa, M., Rehman, Z., & Vermaelen, T. (2007). Mimicking repurchases. Journal of Financial Economics, 84(3), 624–666. Morck, R., Yeung, B., & Yu, W. (2000). The information content of stock markets: Why do emerging markets have synchronous stock price movements? Journal of Financial Economics, 58(1–2), 215–260. Opler, T., & Titman, S. (1996). The debt–equity choice: An analysis of issuing firms. Working paper. Columbus: Ohio State University. Pinkowitz, L., Stultz, R., & Williamson, R. (2006). Does the contribution of corporate cash holdings and dividends to firm value depend on governance? A cross-country analysis. The Journal of Finance, 61(6), 2725–2751. Shleifer, A., & Vishny, R. W. (1997). A survey of corporate governance. The Journal of Finance, 52(2), 737–779. Spence, M. (1973). Job market signaling. The Quarterly Journal of Economics, 87(3), 355–374. Spence, M. (2002). Signaling in retrospect and the informational structure of markets. The American Economic Review, 92(3), 434–459. Stephens, C. P., & Weisbach, M. S. (1998). Actual share re-acquisitions in open-market repurchase programs. The Journal of Finance, 53(1), 313–333. Stulz, R. M. (2005). The limits of financial globalization. The Journal of Finance, 60(4), 1595–1638. Xu, R. Z., & Taylor, G. K. (2007). Economic cost of earnings management through stock repurchases. : Working paper. Zingales, L. (1994). The value of voting rights: A study of the Milan stock exchange experience. Review of Financial Studies, 7(1), 125–148.