Economic subjects | Investments, Stock exchange » Share Repurchases on Trial, Large-Sample Evidence on Share Price Performance, Executive Compensation, and Corporate Investment

Share Repurchases on Trial: Large-Sample Evidence on Share Price Performance, Executive Compensation, and Corporate Investment* Nicholas Guest Cornell University nguest@cornell.edu S.P Kothari Massachusetts Institute of Technology kothari@mit.edu Parth Venkat University of Alabama prvenkat@culverhouse.uaedu January 2023 Abstract Using a large sample of US stocks covering more than three decades, we empirically examine common criticisms of and rationales for stock repurchases. Repurchases account for a tiny fraction of the trading volume in a typical stock, making their price impact too small to generate short-term price manipulation. Price appreciation following repurchases is modest and does not reverse on average, suggesting the small price increases following repurchases signal firms’ good prospects. Also, we find no evidence that CEOs of repurchasing firms are paid excessively or that repurchases crowd out valuable investment opportunities. Because repurchases do not appear to

be systematically abusive, enforcement action should be sufficient to deal with any bad actors, and significant regulation seems unwarranted. * We thank Kathleen Kahle (editor), an anonymous referee, Bruce Dravis, Jeff Hoopes, Anzhela Knyazeva, Simon Lorne, Hamid Mehran, Scott Richardson, and Rene Stulz for helpful comments and suggestions. 1. Introduction Many in politics and the media question the economic efficacy and ethical provenance of share repurchases, a ubiquitous corporate financial activity. Most recently, the federal government’s Inflation Reduction Act of 2022 (H.R 5376) included a one percent excise tax on repurchases,1 senior democrats have shown interest in barring executives from selling shares for three years after a repurchase (see recent New York Times coverage),2 and the SEC has proposed to significantly increase the extent and frequency of repurchase reporting.3 The justification for these regulations affecting repurchase activity often includes either a

deterrent effect on alleged abuses or increased revenues. We examine whether evidence supports the critiques used to justify these and other anti-repurchase initiatives. Prior to enacting regulation, policy makers typically consider the costs to the regulated entities (i.e, US public firms), as well as the benefits to other stakeholders (eg, shareholders, employees, society at large). For the benefits to stakeholders to be large enough to justify costly policies, we expect the drawbacks of repurchases to be widespread and readily observable in public data. To that end, we begin with a brief outline of the common economic rationale for and criticisms of share repurchases. We then document trends in repurchases, and compare trading volume, share price performance, CEO pay, and corporate financial activities (e.g, investment and profitability) of firms that do and do not repurchase shares. We further segment the firms that repurchase shares in two ways: (1) small positive vs. large

positive repurchase amounts (defined as below vs. above median) and (2) infrequent vs infrequent repurchasers (defined as firms that 1 For detailed coverage of the excise tax and its exceptions please see: https://www.davispolkcom/insights/clientupdate/senate-approves-tax-stock-buybacks-and-corporate-minimum-tax 2 https://www.nytimescom/live/2022/03/28/business/business-news-economy-ukraine#bidens-budget-is-set-to-takea-big-swing-at-corporate-buybacks 3 See https://www.secgov/rules/proposed/2021/34-93783-fact-sheetpdf 1 repurchase in one or two quarters vs. three or four quarters of the year) Our goal is not to rehash the numerous conceptual rationales or empirical evidence in defense of share repurchases, nor argue that there are no cases when repurchases could be misused. Instead, we provide a comprehensive, up-to-date, large-sample analysis of several theories and criticisms of share repurchases that should be relevant to policymakers. In particular, we ascertain whether

evidence of alleged abuses from share repurchases is apparent in the data. All payouts to shareholders, i.e, both dividends and share repurchases, can be useful in aligning manager and shareholder incentives (e.g, Easterbrook, 1984), specifically by reducing the potential misuse of free cash flow (Jensen, 1986). Moreover, there are several reasons corporations might prefer using repurchases instead of or in addition to dividends, including (i) maintaining flexibility in determining the amount of cash returned to shareholders, (ii) an ability to award repurchased shares to employees as equity compensation,4 (iii) a modest tax advantage to shareholders (that is less pronounced since the 2003 dividend tax cut),5 and (iv) the ability to send a credible signal of the firm’s (good) prospects.6 In addition to being useful for corporations, multiple safeguards are already in place to limit improper repurchase behavior. For example, firms typically disclose repurchase plans at the time they

are authorized by the board of directors, and securities regulations mandate that firms disclose details about both intended and actual repurchases in their periodic filings.7 Such 4 There is also a literature on whether there is incentive to use repurchases to counteract EPS dilution arising from stock and option grants (Bonaimé, Kahle, Moore, and Nemani, 2020). 5 Traditionally, buybacks have been taxed as capital gains, whereas dividends are subject to ordinary income tax. Jacob and Jacob (2013) provides evidence that firms respond to an increase in what they call “the dividend tax penalty” by decreasing dividends and increasing repurchases. 6 A voluminous body of research in financial economics examines the effects and motivations for payouts via dividends and repurchases. The consensus is that the market reacts favorably to both dividend increases and repurchases (see Farre-Mensa, Michaely, and Schmalz, 2014). There is also new research suggesting that repurchases can

contribute to price stability (Lewis and White, 2021). 7 The SEC details these requirements, and also proposes more frequent reporting of actual repurchases as well as enhanced periodic disclosure, in two recent press releases and accompanying materials. Specifically, SEC Press 2 disclosure enables investors to trade on the basis of their own assessment of the impact of the repurchase on the firm’s economic prospects vis-à-vis managements’ announced intentions. Armed with this knowledge, investors can choose to sell their shares back to the firm, perhaps because they hold a less favorable view of the firm or to satisfy liquidity needs. However, shareholders are not coerced into selling their shares. Share repurchases represent arm’s length transactions at current market prices between willing participants. Ex post, firms must also disclose repurchase activity in quarterly reports. Nonetheless, share repurchases remain a frequent target of critics, who typically make the

following three arguments: (i) Share repurchases enable firms to manipulate the market either by increasing the demand for and therefore the price of shares or by tricking naïve investors through EPS inflation.8 (ii) Share repurchases enable insiders to benefit through compensation contracts or the sale of shares at inflated prices.9 (iii) Share repurchases crowd out investment and thus sacrifice innovation and long-term economic growth.10 To evaluate whether there is evidence to justify these critiques, we analyze the repurchase behavior of thousands of US exchange-listed firms over more than three decades, i.e, 1988-2020 First, we analyze trends in aggregate repurchase activity over time. While repurchases have grown quickly, especially in the early part of the sample period, much of the growth can be attributed to inflation and increases in market cap. In addition, repurchases are now similar in size to dividends Release no. 2021-257 announces Exchange Act Release No

34-93783 and is available at https://www.secgov/news/press-release/2021-257; and SEC Press Release no 2022-216 announces Exchange Act Release No. 34-96458 and is available at https://wwwsecgov/news/press-release/2022-216 8 https://www.cnbccom/2017/05/03/apple-has-been-a-buyback-monsterhtml; https://www.cnbccom/2021/03/02/elizabeth-warren-rips-stock-buybacks-as-nothing-but-paper-manipulationhtml 9 https://hbr.org/2014/09/profits-without-prosperity 10 https://online.wsjcom/public/resources/documents/blackrockletterpdf; https://www.cnbccom/2014/03/26/blackrock-ceo-warns-top-us-firmsdont-overdo-dividends-buybackshtml; https://www.ineteconomicsorg/uploads/papers/WP 60-Lazonick-et-al-US-Pharma-Business-Modelpdf 3 and have not grown faster than dividends for quite some time. However, unlike dividends, repurchases drop precipitously (and temporarily) in times of corporate stress, indicating that they do in fact provide firms with the aforementioned benefit of payout flexibility. Next, we

investigate the first critique, i.e, whether there is widespread evidence that share repurchases enable firms to manipulate the market. Our first hypothesis is that if firms are using repurchases to create excess demand for shares to drive up stock prices, then firms that repurchase shares should have higher trading volume than those that do not. We find no widespread evidence supporting this hypothesis, in that average trading volumes have largely remained quite similar for firms regardless of whether or how much they repurchase shares. In fact, in recent years the trading volume in firms that do not repurchase shares has greatly exceeded that of repurchasing firms, suggesting any excess volume arising from repurchases is dwarfed by other forces. Our second hypothesis is that if naïve investors are tricked into buying repurchase firms due to EPS inflation, we should observe short term price bumps followed by poor long-term performance. Once again, we find no widespread evidence to

support this hypothesis. That is, regardless of whether or how intensely and frequently firms repurchase shares, we find no evidence that firms significantly outperform in the quarters with repurchases. While we find some marginal evidence that firms that intensely repurchase do outperform slightly in the quarter after repurchases, we find no evidence of future reversals in the short- or long-term. This pattern of outperformance is more consistent with firms using repurchases to signal undervaluation rather than to manipulate the market. The second critique is that repurchases allow insiders to unfairly profit. Our hypothesis is that if using repurchases to provide rents to insiders were a common, systematic abuse, CEOs who use repurchases would receive abnormally high pay. Using a model-generated measure of excess pay that has been vetted by prior academic literature, we estimate that CEOs of firms that make 4 large positive repurchases earn a statistically and economically

insignificant amount of excess pay that is only $51K more than CEOs of firms that do not repurchase shares. The difference is even smaller, $4K, when we compare CEOs of firms that repurchase frequently to firms that do not repurchase. These differences are economically tiny in relation to the average CEO’s pay In addition, these results are not monotonic, as CEOs of firms that repurchase small positive amounts and infrequently earn the least excess pay, $113K and $66K less, respectively, than the CEOs of non-repurchasing firms. None of this evidence is consistent with firms using repurchases to significantly boost CEO pay. The third and final critique we consider is that repurchases reduce firms’ ability to take advantage of investment opportunities. We show that firms that do not repurchase shares invest more, but are also significantly less profitable, than firms that do repurchase shares. The firms that repurchase large amounts and repurchase frequently are highly profitable and

have steadily made investments while at the same time returning capital to shareholders. Consistent with economic theory, this evidence suggests that profitable firms repurchase shares while maintaining a steady level of investment, whereas less profitable or loss-making firms do not (or maybe cannot) repurchase shares but are more investment intensive in the hope of becoming profitable. While some firms may be poorly governed and choose to avoid positive NPV projects and instead return capital to willing shareholders, it is hard to see how directives encouraging or incentivizing a poorly run firm to retain cash would lead to that firm making good investments. Consistent with this reasoning, DeAngelo (2022) points out that managers could invest the retained cash in selfserving initiatives or financial assets instead of the real assets the critics of repurchases desire. In summary, our large-sample evidence refutes critics’ alarming claims. Instead, the evidence shows that repurchases

in the US are a mainstream corporate financial activity that 5 returns several hundred billion dollars of capital to shareholders annually. At an aggregate level, we find that this activity neither creates nor destroys much wealth (i.e, share price changes) In addition, while repurchases are associated with higher past profitability, they are not associated with excessive CEO pay or underinvestment. Each of these findings is consistent with economic theories of payout policy (e.g, Miller and Modigliani, 1961; Brealy, Myers, and Allen, 2018) This large-sample evidence, besides being useful for shining light on economic theories (e.g, signaling vs earnings manipulation), should be particularly relevant in the debate on economic consequences of share repurchases, which often is based on carefully selected anecdotes or small samples. For example, detractors have accused the airline industry (especially American Airlines) of abusing repurchases, specifically for using almost all of

their spare cash to buy back shares at the expense of their financial security, i.e, instead of paying down debt or holding cash (see recent MarketWatch and BBC coverage).11 In rebuttal, proponents could cherry pick Google, which has repurchased tens of billions of dollars of its stock over the past few years, and Amazon, which only recently (in 2022) started repurchasing its stock, as examples of firms that engage in large repurchases and still have plenty of profits and cash left available for investment and related initiatives. However, thousands of firms repurchase their shares, and aggregate repurchases have exceeded $500 billion annually for the past five years. Therefore, any regulation governing repurchases should take into account the potential effects on the entire corporate sector, as opposed to isolated examples of firms identified as potential abusers or examples of good corporate citizens. Repurchases, being frequent and often substantial in magnitude, are the subject of

an extensive finance and accounting literature (see Bonaime and Kahle, 2022 for a review). The questions and arguments we examine have been the subject of many of these studies, which among 11 https://www.marketwatchcom/story/airlines-and-boeing-want-a-bailout-but-look-how-much-theyve-spent-onstock-buybacks-2020-03-18; https://wwwbbccouk/news/business-51903947 6 other things provide evidence on increasing payout trends (Kahle and Stulz, 2021), the contrasting flexibility of repurchases and dividends (Dittmar and Dittmar, 2008; Bliss, Cheng, and Denis, 2015; Floyd, Li, and Skinner, 2015), the (small and temporary) price effects of repurchases (Grullon and Michaely, 2004; Bargeron and Farrell, 2012), CEO behavior and compensation around repurchases (Busch and Obernberger, 2015; Edmans, Fang, and Huang, 2021; Moore, 2022), correlations of repurchases with profitability and investment (Brav, Graham, Harvey, and Michaely, 2005; Fried and Wang, 2019), and the cash management of

poorly-run firms (Dittmar and Mahrt-Smith, 2007). Three additional prior studies are especially related to our work. The first, Edmans (2017), surveys existing academic literature and argues that critiques of share buybacks are “very rarely backed up by large-scale evidence.” The second, DeAngelo (2022), argues that attacks on repurchases will reduce the flow of capital into firms and fail to increase real investment. Unlike our study, neither of these prior papers provide original empirical analysis. The third, Asness, Hazelkorn, and Richardson (2018), examines the empirical evidence relevant to each of the popular critiques and find little evidence of repurchases being systematically abusive. Our study complements and extends Asness et al. (2018) in at least two ways First, they focus on the aggregate trends while we introduce cross-sectional findings by segmenting firms on both repurchase amounts and frequency. Second, we extend the sample period multiple years, which allows for

updated inferences, including about the chaotic early months of the global COVID pandemic. Despite high quality academic work such as Edmans (2017), DeAngelo (2022), and Asness et al. (2018), the repurchase debate persists and the anti-repurchase side of the debate appears to be winning in the court of public opinion and the halls of Washington, DC. Thus, we 7 believe our study is useful because it highlights and updates past literature in a format that is approachable for academics, practitioners, and policymakers alike. Section 2 summarizes recent trends in corporate repurchase behavior. Section 3 presents empirical findings on the association between repurchases and several economic variables of interest. Section 4 concludes 2. Trends in Repurchase Behavior We analyze US exchange-listed firms over more than three decades, i.e, 1988-2020 All NYSE-AMEX-NASDAQ firms with share price greater than one dollar are included.12 We obtain quarterly and annual dividend and repurchase

amounts, share price performance, trading volume, book and market values, investment expenditures, and profitability data from the CRSP and Compustat databases. The average number of firms with available data each year is 3,526, ranging from 2,716 in 2013 to 5,182 in 1997. The aggregate market capitalization of the stocks included in the analysis at the end of the sample period is $25 trillion. In many of our analyses, we segment positive repurchase firms along two dimensions. First, we split based on below (i.e, “Small Positive Repurchase”) vs above (ie, “Large Positive Repurchase”) annual median repurchase amounts, which are calculated following Floyd, Li, and Skinner (2015). That is, we measure net repurchases as the increase in common treasury stock (Compustat item tstkq) if the firm uses the treasury stock method. If the firm uses the retirement method instead (which we infer from the fact that the treasury stock is zero in the current and prior year), we measure

repurchases as the difference between stock purchases (prstkcy) and stock issuances (sstky) from the statement of cash flows. Second, we split based on the frequency of 12 The excluded securities amount to a trivial fraction of the total market capitalization of the listed securities and they rarely engage in share repurchases. Those securities tend to be illiquid with large bid-ask spreads, which introduces noise in return calculations. 8 repurchases following Nemani (2020), which follows the repurchase definition of Banyi, Dyl, and Kahle (2008) and the frequency definition of Dittmar and Field (2015) with one exception, i.e, using quarterly instead of monthly repurchase data because monthly data is not available before 2004. We define firms that repurchase in one or two quarters of the year as the “Infrequent Repurchase” group and firms that repurchase in three or four quarters of the year as the “Frequent Repurchase” group. Figure 1 and Table 1 show that aggregate

corporate payouts to shareholders via dividends and repurchases have risen from $253 billion in 1988 to $1.68 trillion in 2020 Share repurchases have grown to be comparable in amount and frequency to dividend distributions. In fact, since about 2011, as many firms have repurchased shares as paid dividends.13 The growth in repurchases is quite dramatic from $63 billion in 1988 to $714 billion in 2020 – an 11-fold increase. Dividend distribution over the corresponding period only rose from $191 billion to $968 billion, i.e, a 5fold increase To examine the extent to which inflation is driving these trends, we also present amounts that are scaled by the Consumer Price Index, with 2015 taken as the reference point. Of course, inflation does explain some of the growth in payouts, e.g, over our sample period repurchases grew 11-fold in nominal terms but only 5-fold in real terms. However, the general trends and takeaways discussed below remain unchanged. Kahle and Stulz (2021) document a

similar trend of increasing payouts. They attribute the increase in repurchases to higher corporate income and higher payout rates. This growth is usually cited in an untestable critique that repurchases have grown inappropriately. We believe this interpretation clouds several key points 13 Many potential factors drive the increase in repurchase behavior. For example, before the global financial crisis and the current stress testing regime, banks were likely hesitant to repurchase shares because they feared doing so would attract the scrutiny of regulators concerned about capital ratios. Perhaps not coincidentally, now that regulators focus more on stress testing, we see more banks repurchase their shares. 9 One noticeable feature of Figure 1 and Table 1 is the uneven growth of repurchases, i.e, there was a short period of rapid growth from 2004 through 2006 in which repurchases caught up to dividends. Since then, dividends and repurchases have grown at similar rates (although

repurchase growth has been more volatile, which we discuss later). In part because of their similar growth rates, but even more so because the paper focuses on repurchases, we do not condition on dividends in the rest of the analyses of the paper. Of course, corporations have also grown significantly over the sample period. To give a sense of the magnitude of growth in repurchases in relation to the market over time, we also plot repurchases as a fraction of the market value (or price) of the firm in Figure 2. This ratio declined in the first few years after 1988 but has remained relatively constant for the past 25 years (see the “All Firms” group). Average repurchases per year during the period are 17% of the firm’s market value. Among the positive repurchase firms, payout as a fraction of equity value has declined slightly in the last twenty years from about 7% to 5%. Even the subset of firms that repurchase the largest amounts (i.e, the above median or “Large Positive

Repurchase” group) have dropped their repurchase-to-price ratio from about 13% to 8%. Similarly, firms that repurchase the most frequently have dropped their repurchase-to-price ratio from about 6% to 3%. These decreases in repurchases relative to market cap contradict claims that firms increasingly remit capital to shareholders at the expense of other firm initiatives, e.g, investing in physical and human capital The volatility in repurchases seen in Figure 1 and Table 1 evinces a key reason companies value repurchases: they are relatively easy to cut or delay in times of corporate stress. For example, consider the precipitous drop in repurchases from $511 billion in 2008 to $201 billion in 2009 as the financial crisis hit the US and global economies. Similarly, in 2018 and 2019, repurchases reached about $1 trillion, surpassing dividends, but retreated severely in 2020 during the COVID 10 pandemic. These large declines in 2009 and 2020 underscore the valuable flexibility

repurchases afford firms to adjust the amount of capital returned to shareholders as market conditions and firm prospects vary. In contrast, dividend payout held steady from $528 billion in 2008 to $553 billion in 2009, and from $972 billion in 2019 to $968 billion in 2020. This contrasting flexibility has been shown previously (Dittmar and Dittmar, 2008; Bliss, Cheng, and Denis, 2015; Floyd et al., 2015) and is due largely to firms’ well-established practice of essentially locking in dividend payments (although this is not contractual), i.e, dividend cuts are generally avoided because they trigger a significant negative market reaction (Jensen and Johnson, 1995). In summary, while the overall amount of repurchases has risen considerably (even in real terms) in recent years, they are on par with dividends in terms of both growth rate and magnitude. Moreover, these trends are slower than the rate of overall stock price appreciation. Finally, repurchase growth is volatile, reflecting

their greater financing flexibility over dividends. None of this supports fears of runaway repurchase growth at the expense of other uses of cash, e.g, investment opportunities. 3. Economic Characteristics Associated with Share Repurchases 3.1 Trading Volume A salient critique alleges that share repurchases inflate share prices by increasing the demand for shares or, similarly, by tricking naïve investors (Jackson, 2018). Focusing first on demand, share repurchases obviously entail buying shares, which may boost prices. However, any share price impact is predicated on an economically meaningful incremental demand for shares from repurchases that is not offset by equivalent amounts of selling. To investigate this possibility, Figure 3 charts average annual trading volume of stocks across various repurchase groups. Trading volume is expressed as a percentage of shares outstanding. Thus, 100% means trading volume in 11 a year equals the number of shares outstanding for the firm in

that year. Average trading volumes have largely been quite similar for all groups over the past few decades. For example, in 2014, average trading volume of zero-repurchase firms was 248% compared to 210% for small positive repurchase and 255% for large positive repurchase firms. Beginning in 2016, the zero-repurchase firms have noticeably higher trading volumes compared to positive repurchase firms. Recall that share repurchases in recent years are on average about 4-6% of shares outstanding (Figure 2 shows this in dollar terms, but the same interpretation extends to the number of shares). Therefore, repurchases account for a small fraction of the trading volume in a typical stock (i.e, 4-6% vs 200-300%) Moreover, even this relatively small trading volume due to share repurchases is likely to be spread over a number of trading days because of SEC Rule 10b-18, which caps daily repurchases at 25% of average daily trading volume (measured in shares, not dollars) with few exceptions (e.g,

a once per week block trade subject to a different set of rules) This regulation likely contributes to firms’ propensity to execute share repurchase programs over an extended period, making them unlikely to overwhelm normal volume on any given day or week. For example, on April 27, 2021, Google announced that their board of directors authorized the company to repurchase up to $50 billion of its own stock, which was about 3% of its market capitalization. By the end of 2021, their 10-K states, “The repurchases are being executed from time to time, subject to general business and market conditions and other investment opportunities.”14 Most repurchasing firms have a similar process in which the firm (ie, the board) approves repurchase amounts, announces the plan to the market, and then repurchases shares over the next year. Taken together, this evidence that repurchases are a small fraction of total trading 14 https://abc.xyz/investor/static/pdf/20220202 alphabet

10Kpdf?cache=fc81690 12 volume and that they are typically spread over an extended period suggest they are unlikely to substantially boost share prices. Moreover, any share price increases due to repurchases would make the firm less attractive to other investors and encourage sales (or short-sales). Thus, theory predicts price increases, if any, to reverse within a short time period. Consistent with the theory, Bargeron and Farrell (2021) find small demand-effects on prices to the tune of 0.4 to 07%, which reverse quickly They use a clever test based on a sample of 19 firms from 2004-2019 with dual-class shares and different payout policies. The event study with a tiny sample over a 15-year period precludes generalizability of the findings to the economy-wide phenomena of repurchases. Moreover, in supplementary materials one of the authors of the study concludes that the “small, short-lived increase in the stock price leaves little scope for CEOs to benefit from repurchases

motivated by self-interest.”15 We also examine the potential price effects and benefits to CEOs in the next two sections. 3.2 Share Price Performance Critics argue that repurchases might manipulate share prices by boosting EPS, which could trick naïve investors into forming overly optimistic beliefs about firm performance. This concern is based on two observations: (i) investors care about EPS and (ii) share repurchase programs mechanically increase EPS (a ratio) by reducing the number of shares outstanding (the denominator) but not the firm’s earnings (the numerator). An alternative version of this critique maintains that firms with poor prospects could engage in repurchases to herd with firms with good prospects that also use repurchases to signal. To evaluate this criticism, we compare the share price performance of the portfolio of firms that do not repurchase shares to the portfolios of firms that repurchase the largest amounts (“Large 15

https://uwm.edu/business/farrell-research/ 13 Positive Repurchase”) and the most frequently (“Frequent Repurchase”). Table 2 reports share price performance of these three portfolios in the quarter before (Quarter -1), quarter of (Quarter 0), and two quarters after (Quarter +1 and +2) the payout. In Panel A, performance is raw returns, defined as average return minus the risk-free rate (i.e, 3-month Treasury Bill return) In Panel B, we report Fama-French alphas, i.e, the intercept from the regression of raw returns on the five Fama-French factors plus the momentum factor (Fama and French, 1993; Fama and French, 2015).16 Portfolio returns are market-value weighted17 Although the Panel A and B results are quite similar, we focus the remaining discussion on Panel A because our paper is establishing a non-result, which makes it important to ensure the result is not driven by our inclusion or choice of control variables. The performance of the three portfolios during Quarter 0 is

statistically indistinguishable. Contrary to concerns that repurchases artificially boost share prices, returns of zero-repurchase stocks are actually slightly higher. In terms of economic magnitude, the zero-repurchase portfolio return is 3.05% compared to 284% for the large-positive repurchase portfolio and 298% for the frequent repurchase portfolio, a difference of only 0.20% and 007%, respectively This means their performance is essentially the same, so investors do not appear to be systematically fooled by repurchases. We observe a similar pattern in the prior quarter (Quarter -1). This means poor performance, on average, does not seem to motivate firms to repurchase shares. In individual cases, however, it is possible that a firm might choose to repurchase shares in the face of a price drop that the management believes is unwarranted. But if this phenomenon were widespread, we would have 16 Consistent with prior research, this six-factor model explains much of the variation in

portfolio returns, with R 2 values of 53-96% in the quarterly analysis and 49-66% in the annual analysis. 17 We also estimated performance using equal-weighted returns. While the numbers are different, the tenor of the results is unchanged. 14 observed poor performance for the large-positive and frequent repurchase portfolios in quarter -1 relative to their repurchase announcements. Turning to future quarters, the large-positive repurchase portfolio (3.46% return) marginally outperforms the zero-repurchase portfolio (2.76% return) in Quarter +1 The difference of 0.70% return is marginally statistically significant at the 10% level However, the same is not true for the frequent repurchase portfolio (i.e, 292% and 276% are statistically indistinguishable) The three portfolios perform statistically similarly in Quarter +2. While economically small, the statistically superior performance of the large-positive repurchase portfolio in Quarter +1 can be interpreted as the market

responding to those firms’ management signaling good future prospects and/or a belief that their firms are undervalued. If this is the case, we expect the price increase to be permanent, i.e, the result of investors’ new higher (and well-founded) valuation of the firm. Under the alternative possibility that the repurchases inappropriately manipulate investor beliefs, we predict the stock price bump would eventually reverse as the firm performs worse than expected and investors recognize they have been tricked. Of course, the major concern is that insiders could opportunistically sell the stock while the stock price is temporarily inflated and thereby extract rents from other shareholders. Our analysis of Quarter +2 does not show any evidence of reversal, so the signaling explanation appears more likely than price manipulation thus far. However, it may take longer than one or two quarters for the manipulated price to revert. Thus, we next examine annual share price performance.

Specifically, Table 3 shows that in the year contemporaneous with share repurchases (Year 0), the large-positive repurchases portfolio underperforms the zero-repurchase portfolio by a statistically and economically significant 3.11% The frequent repurchase portfolio also underperforms the zero-repurchase portfolio, but by a lesser 15 amount (1.11%) that is not statistically significant These differences are consistent with stock price underperformance as a motivator to engage in repurchases and signal to the investment community that the firm might be undervalued.18 In Year +1 the large-positive repurchase and frequent repurchase portfolios perform similarly to the zero-repurchase portfolio. The difference between their performances is less than one percent and not statistically significant. Thus, regardless of the managers’ motivation for share repurchases, the evidence shows they neither generate a noticeable, long-term price appreciation nor do they dissipate firm value.19

The preceding analysis examines return performance of firms sampled from more than three decades. Because repurchase activity has intensified in recent years, we also examine whether the alleged motive of boosting share prices via repurchases is manifested more recently. In particular, we examine return performance across groups of repurchase firms over 2001-2010 and 2011-2020. The sub-period analysis, which we tabulate in the internet appendix, is consistent with our main analysis. That is, only a few of the differences are economically and marginally statistically significant, and none of them suggest a pattern of aggregate share price behavior consistent with price manipulation, i.e, a price increase with share repurchase activity, which reverses subsequently. 18 It is also possible that the management possesses information that their stock is undervalued and that they exploit their knowledge to buy stock on the cheap. This would generate profits from “insider trading” for the

management because they exploited material, non-public information to their advantage. Corporate policies and SEC regulation serve as deterrent and SEC enforcement and private litigation are means of recovering the managers’ gains from such trades. In addition, managers face civil and criminal penalties Notwithstanding these safeguards against illegal managerial actions, we do observe incidence of such activities, albeit infrequently (see De Cesari, Espenlaub, Khurshed, and Simkovic, 2012). 19 The quarterly and annual return analysis leaves open the possibility that repurchases cause price inflation that lasts only a few days, which managers could exploit by selling shares (see next section). Recall, however, that our goal is to examine aggregate trends to ascertain whether the widespread activity of share repurchases generates aggregate trends consistent with price manipulation that would generate profit opportunities to insiders, including via excessive compensation. Besides, prior

research has already examined short-term pricing effects of repurchase announcements, finding a small (albeit significant and permanent) price increase in the 1-3% range (Grullon and Michaely, 2004). Notwithstanding these inferences from large-sample analyses, we expect and observe individual instances of price manipulation and insider selling, which are subjects of SEC enforcement and/or private litigation. 16 Overall, the pattern of return performance surrounding repurchases does not indicate large, lasting price effects that would be consistent with firms artificially boosting share prices via repurchases. In contrast, we find some evidence that underperformance (or perceived undervaluation) motivates firms to repurchase shares to signal firms’ good prospects and/or perceived undervaluation. These findings are consistent with our theoretical understanding of financial markets. For the EPS/naïve investor channel to be a systematic problem, a myriad of analysts and

institutional investors would need to be unaware that the effect of share repurchases on EPS is mechanical and not economic. While some might be ignorant of the effect, many are well versed in it (eg, the effect is discussed in standard textbooks in financial economics and accounting). More broadly, firms must provide extensive disclosures to the market, the financial press and analysts routinely discuss corporate share repurchase programs, and private litigation as well as SEC enforcement focus on rooting out cases of temporary price inflation and fraud.20 Therefore, it is implausible that the marginal investor – likely a sophisticated trader – would be fooled by the EPS boost. It is even more implausible that this effect, if driven by naïve investors, would be persistent given the existence of many sophisticated investors with the incentive and means to eliminate mispricing. 3.3 CEO Compensation The second common critique is that repurchases allow insiders to profit. This could

happen if (1) insiders sell a portion of their shares at prices that are inflated due to a buyback or (2) insiders’ compensation is tied to the firm’s share price and/or EPS that are inflated due to a buyback. Evidence in the prior section that repurchases do not distort prices helps rule out possibility (1), so in this section we focus on (2). 20 SEC disclosure rules do not confer immunity to corporations from possible 10b-5 liability where a firm fails to disclose material, non-public information prior to engaging in repurchases. 17 CEO compensation in the form of cash bonus and equity awards is commonly tied to meeting share price and EPS targets (Bettis, Bizjak, Coles, and Kalpathy, 2018). Therefore, the potential increase in share price and EPS due to repurchases, and the resulting compensation benefit, might motivate managers to repurchase shares. Moore (2022) finds evidence that CEOs may time equity sales around repurchases but argues there is no negative impact on

shareholder value. Alternatively, Edmans et al (2021) provides some evidence that repurchases are associated with vesting equity (albeit poor long-term returns), but staunchly defends buybacks overall while supporting more disclosure to mitigate this concern.21 Relatedly, Busch and Obernberger (2015) “find no evidence that managers use share repurchases to manipulate stock prices when selling their equity holdings or exercising stock options.” In summary, prior literature documents small abuses at most and focuses on the exercise of options and/or sale of equity stakes by executives. We examine a related hypothesis; if using repurchases to provide rents to insiders were a common, systematic abuse, we would expect that CEOs who use repurchases would receive excessive (i.e, abnormally high) overall pay To measure excess pay, we calculate the difference between actual pay and a model-generated prediction of normal pay. Actual pay, which we collect from S&P ExecuComp, is the sum of

the CEO’s salary, bonus, other non-equity incentives, stock and option awards, and any other annual pay. Following prior research, our estimate of predicted pay is based on a regression of CEO pay on economic variables such as market capitalization, revenue, book-to-market, stock returns, and return on assets.22 Specifically, we estimate the following regression at the firm i year t level: log(Total CEO Pay)i,t = β0 + β1 x log(Market Cap)i,t-1 + β2 x log(Revenue)i,t-1 21 https://www.secgov/comments/s7-21-21/s72121-20128261-290078pdf An extensive literature identifies determinants of CEO pay, which includes several firm characteristics and firm performance measures (e.g, Core, Holthausen, and Larcker, 1999; Smith and Watts, 1992; Murphy, 2013) 22 18 + β3 x Book-to-Marketi,t-1 + β4 x Stock Returni,t-1 + β5 x ROAi,t-1 + εi,t. While CEO compensation might be more complicated than our simple model suggests, we believe it captures the major factors well enough that we should

see CEOs of firms with more repurchases receive higher levels of excess compensation if repurchases are used to systematically overpay CEOs.23 Consistent with prior research, Panel A of Table 4 shows that market cap, revenue, and past stock returns are all positively correlated with CEO pay. The negative coefficient on return-on-assets (ROA) is counterintuitive but is actually in line with prior studies estimating this model (e.g, Table 4 of Core, Guay, and Larcker, 2008) and may be due to high collinearity with stock returns. Also, the model estimation results in high R2 values (ie, above 50%) that are comparable to recent research using a similar approach (see Guest, Kothari, and Pozen, 2022), suggesting the model captures a non-trivial portion of the variation in CEO pay. Panel B of Table 4 reports actual and estimated excess CEO pay and several firm characteristics for various repurchase portfolios.24 Note that ExecuComp data is not available for all firms, and we restrict the

analysis to 2006-2020 because regulation required firms to disclose comparable compensation data during this period. As a result, the sample is skewed toward larger stocks in recent years. However, repurchases are similarly concentrated among larger firms and have gained more attention in recent years, both of which mitigate the influence of these data availability issues on our inferences. Most notably, average CEO pay is highest for the large-positive repurchase firms and the frequent repurchase firms, at about $8.16 and $796 million annually, respectively (The pay figures are in nominal dollars over the 2006-20 period.) It is well known that CEO pay increases 23 In econometric terms, this means the omitted complexities from the compensation model are unlikely to mask the systematic excessive compensation, if any, associated with repurchases. 24 Because the dependent variable of the regression equation is in logs, we subtract the exponentiated (and bias corrected) predicted value

from total CEO pay to calculate excess pay in dollar terms. 19 in firm size and revenues and that bonuses are tied to accounting performance (Healy, 1985; Core et al., 1999; Murphy, 2013) Therefore, because Table 4 also shows that ROA is higher for largepositive and frequent repurchase firms, it is not surprising that these firms’ CEOs earn more pay Whether this difference represents excess pay for the large-positive and frequent repurchase firms can be evaluated using the pay model described above. As reported in Table 4, the estimated excess pay for the CEOs of the large-positive (frequent) repurchase firms is $71,000 ($24,000). These amounts are economically small, only about 0.9% (03%) of the average CEO’s total pay, and statistically insignificant. Compared to the zero-repurchase firms’ CEOs, the large-positive (frequent) firms’ CEOs earn $51,000 ($4,000) more excess pay on average (i.e, $71,000 - $20,000 and $24,000 - $20,000). These differences are not statistically

significant either, and they are economically small in relation to the average CEO’s pay. Collectively, these small differences suggest that repurchases are not associated with excessive CEO pay. This evidence on CEO pay supports our findings in prior sections that insider sales after repurchases are consistent with arm’s length transactions at prevailing market prices and not suggestive of an inflated (or opportunistic) profit opportunity for managers. Nonetheless, some managers might still exploit share repurchases to earn excessive compensation.25 However, as seen from the large-sample evidence here, the majority of insider sales instead must be a result of something else, perhaps such as managers’ liquidity needs, portfolio diversification considerations, or the belief that share values will decline in future.26 Logically, the economic magnitude of the effect of repurchases on compensation cannot be large. First, because repurchases typically reduce the number of shares

outstanding by only a Indeed, evidence does suggest managers’ opportunistic behavior to influence bonuses via share repurchases, but the frequency and magnitude of the effects are relatively small (Cheng, Harford, and Zhang, 2015). 26 Dravis (2019) discusses how equity compensation and share repurchases can be complementary, e.g, share repurchases help offset the dilution that occurs when additional shares are issued and given to employees. 25 20 couple of percentage points, any EPS boost would be relatively small.27 Second, for a large number of firms, repurchases are a recurring phenomenon and therefore the EPS targets already factor in the ongoing effects of share repurchases. Third, compensation committees are knowledgeable about the effect of share repurchases on EPS, which can be helpful in setting EPS targets as well as in determining CEO compensation. Overall, our results build on prior literature, such as the following conclusions of Busch and Obernberger (2017):

“Share repurchases help to maintain accurate stock prices by providing price support at fundamental values. We find no evidence that managers use share repurchases to manipulate stock prices when selling their equity holdings or exercising stock options.” (see also McNally et al, 2006, and Hillert et al, 2016) 3.4 Investment and Profitability The final critique we consider is that repurchases reduce firms’ ability to take advantage of investment opportunities, thus hindering economic growth, including innovation, capital expenditures, employment, etc. Following prior research, we measure investment intensity as the sum of capital expenditures and research and development outlays, scaled by total assets. We compare the investment of zero-repurchase firms, positive repurchase firms, large- vs. smallpositive repurchase firms, and frequent vs infrequent repurchase firms Panel A of Figure 4 shows that all groups of firms with positive repurchases have steadily invested for the past

two decades, despite a rise in repurchases. This initial evidence is not suggestive of repurchases increasingly crowding out investment over time. However, Panel A of Figure 4 also clearly shows that from the mid-1990s onward firms that repurchase shares have invested less than those that do not. Based on this evidence, it is 27 We acknowledge that it is possible for a small increase in EPS to trigger a nonlinear increase in compensation, depending on the EPS thresholds employed in the compensation structure (Healy, 1985). However, such features in compensation contracts are rare. 21 tempting to conclude that repurchases crowd out investment, but this would be premature. In fact, both investment and repurchases may be driven by underlying omitted firm characteristics such as growth and investment opportunities. To distinguish between these alternative explanations, we next investigate profitability and book-to-market ratios. A low book-to-market ratio is indicative of the market

anticipating growth as the firm exploits investment opportunities. Panel B of Figure 4 shows that over the past two decades, firms that do not repurchase shares have been less profitable and have on average incurred losses. In addition, Panel C of Figure 4 shows that their book-to-market ratios were slightly lower than repurchasing firms during the 1990s and 2000s, and about the same in the 2010s. The largepositive and frequent repurchase groups are both highly profitable and have steadily made investments while at the same time returning capital to shareholders. Overall, this evidence suggests that profitable firms repurchase shares while maintaining a steady level of investment, whereas less profitable or loss-making firms do not (or maybe cannot) repurchase shares but are more investment intensive in the hope of becoming profitable. This behavior is consistent with economic theory, i.e, that firms return capital to investors when they run out of positive net present value investment

projects (see, e.g, Brealey et al, 2018) Relatedly, Fried and Wang (2018) emphasize corporations’ considerable R&D expenditures and cash balances as evidence that repurchases are not crowding out investment and innovation. The unanswered question is whether profitable companies might have invested more instead of distributing dividends or repurchasing shares. While we will never know the counterfactual, the earlier evidence that we report as well as evidence from the literature that 22 repurchases did not impart a boost to share prices is helpful in answering the question.28 With no systematic, sustained share-price or compensation benefit from repurchases, it is implausible that managers would forgo valuable investment opportunities in favor of repurchasing shares. Stated differently, because profitable investments increase share prices, repurchases would have to increase prices at least as much; instead, repurchases barely move prices. Thus, firms are unlikely to forego

valuable investments in favor of repurchases. Almeida et al (2016) use a regression discontinuity design (RDD) around near earnings misses to argue that the counterfactual of less repurchases is more hiring and investment. The key tradeoff with RDDs is a better estimate of counterfactuals in exchange for less external validity. That is, RDDs may (or may not) provide convincing evidence that a subset of firms would invest more if they did not repurchase shares, but they cannot inform on whether this effect would exist for the broader set of firms that are not near an earnings miss. This limitation is precisely why our broader sample analysis is valuable The crowding-out argument further ignores the extensive evidence on the well-known agency conflict that motivates CEOs to retain rather than distribute funds back to shareholders, often referred to as “empire building” (Jensen, 1986). That is, if such CEOs had instead retained cash and invested it in the firm, those investments

potentially tainted by the managers’ agency conflicts might have been value-destroying negative-NPV projects (e.g, wasteful mergers and acquisitions, perquisite consumption, etc.) It is hard to see how directives encouraging or incentivizing a poorly run firm to retain cash would lead to that firm making good investments (DeAngelo, 2022). Indeed, there is substantial evidence that firms that payout more cash are better governed (La Porta, Lopez-de-Silanes, Shleifer, and Vishny, 2000; Alzahrani and Lasfer, 2012). 28 Everyone, not just CEOs and insiders, benefits from a permanent boost to share prices as a result of a firm credibly signaling superior prospects or the market correcting a firm’s undervaluation, especially because mandated disclosures level the informational playing field for insider and outsider investors. 23 4. Conclusion This study reexamines the major arguments for and against share repurchases. Our main contribution is in providing up to date large-sample

evidence at both the aggregate and firmspecific levels that should help academics, practitioners, and regulators alike to evaluate the merits of each argument. In addition, we examine a period of several decades, up to and including the first year of the recent COVID pandemic. This helps us emphasize the importance of designing legislation and regulation for the entire corporate sector based on large-sample evidence over a long period as opposed to small subsets of firms and/or interest groups. Overall, our findings refute the popular criticisms of repurchases, namely, that they have become excessive, that they manipulate share prices and EPS, that they enrich insiders at the expense of outsiders, or that they stifle investment. Instead, the results suggest that share repurchases are a mainstream corporate activity that for the most part do not harm the overall market, but do provide various benefits to firms such as a reduction in the agency costs of free cash flow, additional

financial flexibility, and the ability to signal good prospects. The setting of our study examining the validity of the critiques of share buybacks is quite different from the typical setting in which a researcher seeks to draw causal inferences. In a typical study, a researcher observes some correlations and then sets out to test whether they represent certain causal phenomenon. In these settings, strong correlations generally exist, but causality cannot be inferred from the correlations alone. In contrast, in our paper we do not provide any identification strategies to produce causal stories because the evidence points to a lack of correlation, not to mention causality, between repurchases and the variables of interest. In particular, we do not observe much if any correlation between share buybacks and the alleged 24 malpractices, e.g, excessive CEO compensation and return reversals suggesting price manipulation. In summary, the absence of correlation between share repurchases

and price manipulation, return reversals, excess CEO compensation, and underinvestment, makes it highly implausible that economically significant causal effects of share repurchases still underlie the data. It seems even more implausible that policies designed to reduce repurchases will reduce alleged but unobserved malpractices. 25 References Almeida, H., Fos, V, Kronlund, M, 2016 The real effects of share repurchases Journal of Financial Economics 119, 168-185. Alzahrani, M., Lasfer, M, 2012 Investor protection, taxation, and dividends Journal of Corporate Finance 18, 745–762. Asness, C., Hazelkorn, T, Richardson, S, 2018 Buyback derangement syndrome Journal of Portfolio Management 44, 50-57. Banyi, M., Dyl, E, Kahle, K, 2008 Errors in estimating share repurchases Journal of Corporate Finance 14, 460–474. Bargeron, L., Farrell, M, 2021 The price effect of stock repurchases: Evidence from dual class firms Management Science 67, 6568-6580. Bettis, J., Bizjak, J, Coles, J,

Kalpathy, S, 2018 Performance-vesting provisions in executive compensation. Journal of Accounting and Economics 66, 194-221 Bliss, B., Cheng, Y, Denis, D, 2015 Corporate payout, cash retention, and the supply of credit: Evidence from the 2008-2009 credit crisis. Journal of Financial Economics 115, 521-540 Bonaime, A., Kahle, K, 2022 Share repurchases Working paper, University of Arizona Bonaimé, A., Kahle, K, Moore, D, Nemani, A, 2020 Employee Compensation Still Impacts Payout Policy. Critical Finance Review, Forthcoming Brav, A., Graham, J, Harvey, C, Michaely, R, 2005 Payout policy in the 21st century Journal of Financial Economics 77, 483-527. Brealey, R., Myers, S, Allen, F, 2018 Principles of Corporate Finance, 12/e, McGraw-Hill Education, Chapter 16: 410-429. Busch, S., Obernberger, P, 2017 Actual share repurchases, price efficiency, and the information content of stock prices. Review of Financial Studies 30, 324–362 Cheng, Y., Harford, J, Tianming, Z, 2015 Bonus-driven

repurchases Journal of Financial and Quantitative Analysis 50, 447–475. Core, J., Guay, W, Larcker, D, 2008 The power of the pen and executive compensation Journal of Financial Economics 88, 1-25. Core, J., Holthausen, R, Larcker, D, 1999 Corporate governance, chief executive officer compensation, and firm performance. Journal of Financial Economics 51, 371-406 De Cesari, A., Espenlaub, S, Khurshed, A, Simkovic, M, 2012 The effects of ownership and stock liquidity on the timing of repurchase transactions. Journal of Corporate Finance 18, 1023-1050 DeAngelo, H., 2022 The attack on share buybacks European Financial Management, forthcoming 26 Dittmar, A., Dittmar, R, 2008 The timing of financing decisions: An examination of the correlation in financing waves. Journal of Financial Economics 90, 59-83 Dittmar, A., Field, L, 2015 Can managers time the market? Evidence using repurchase price data Journal of Financial Economics 115, 261–282. Dittmar, A., Mahrt-Smith, J, 2007 Corporate

governance and the value of cash holdings Journal of Financial Economics 83, 599-634. Dravis, B., 2019 Dilution, disclosure, equity compensation, and buybacks The Business Lawyer 74, 631658 Easterbrook, F., 1984 Two agency-cost explanations of dividends American Economic Review 74, 650659 Edmans, A., Fang, V, Huang, A H, 2021 The long‐term consequences of short‐term incentives Journal of Accounting Research 60, 1007-1046. Edmans, A., 2017 The case for stock buybacks Harvard Business Review September 15 Fama, E., French, K, 1993 Common risk factors in the returns on stocks and bonds Journal of Financial Economics 33, 3–56. Fama, E., French, K, 2015 A five-factor asset pricing model Journal of Financial Economics 116, 1-22 Farre-Mensa, J., Michaely, R, Schmalz, M, 2014 Payout policy Annual Review of Financial Economics 6, 75–134. Floyd, E., Li, N, Skinner, D, 2015 Payout policy through the financial crisis: The growth of repurchases and the resilience of dividends. Journal of

Financial Economics 118, 299-316 Fried, J., Wang, C, 2018 Are buybacks really shortchanging investment? Harvard Business Review 96: 88–95. Fried, J., Wang, C, 2019 Short-termism and capital flows The Review of Corporate Finance Studies 8, 207–233. Grullon, G., Michaely, R, 2004 The information content of share repurchase programs Journal of Finance 59, 651-680. Guest, N., Kothari, SP, Pozen, R, 2022 Why do large positive non-GAAP earnings adjustments predict abnormally high CEO pay? The Accounting Review 97, 297-326. Healy, P., 1985 The effect of bonus schemes on accounting decisions Journal of Accounting and Economics 7, 85-107. Hillert, A., Maug, E, Obernberger, S, 2016 Stock repurchases and liquidity Journal of Financial Economics 119, 186-209. Jackson, R., 2018 Stock buybacks and corporate cashouts Transcript of speech delivered on June 11, 2018 retrieved from https://www.secgov/news/speech/speech-jackson-061118# ftn16 27 Jacob, M., Jacob, M, 2013 Taxation, dividends, and

share repurchases: Taking evidence global Journal of Financial and Quantitative Analysis 48, 1241-1269. Jensen, G. R, Johnson, J M, 1995 The dynamics of corporate dividend reductions Financial Management 24, 31–51. Jensen, M., 1986 Agency costs of free cash flow, corporate finance, and takeovers American Economic Review 76, 323-329. Kahle, K., Stulz, R, 2021 Why are corporate payouts so high in the 2000s? Journal of Financial Economics 142, 1359-1380. La Porta, R., F Lopez-de-Silanes, A Shleifer, and RW Vishny, 2000 Agency problems and dividend policies around the world. Journal of Finance 55, 1–33 Lewis, C., White, J, 2021 Corporate liquidity provision & share repurchase programs Working paper, Vanderbilt University. McNally, W., Smith, B, Barnes, T, 2006 The price impact of open market repurchase trades Journal of Business Finance & Accounting 35, 735-752. Miller, M., Modigliani, F, 1961 Dividend policy, growth, and the valuation of shares Journal of Business 34, 411-433.

Moore, D., 2022 Strategic repurchases and equity sales: Evidence from equity vesting schedules Available at SSRN 3014462. Murphy, K., 2013 Executive compensation: Where we are, and how we got there, in George Constantinides, Milton Harris, and René Stulz (eds.), Handbook of the Economics of Finance, Chapter 4: 211-356. Nemani, A., 2020 Cross-sectional variation in repurchase motives: Evidence from repurchase frequency Working paper, Bentley University. Smith, C., Watts, R, 1992 The investment opportunity set and corporate financing, dividend, and compensation policies. Journal of Financial Economics 32, 263-292 28 Figure 1. Aggregate Payouts The top panel of this figure shows aggregate US corporate dividend and repurchase payouts ($ in billions) from 1988 to 2020. The bottom panel is inflation-adjusted using the Consumer Price Index, with 2015 taken as the reference point. 29 Figure 2. Payout-to-Price Ratios This figure shows average payout-to-price ratios across repurchase

portfolios from 1988 to 2020. The “Positive Repurchase” portfolio includes all firms with positive repurchase amounts, and the “All Firms” portfolio includes all firms. The “Small Positive Repurchase” and “Large Positive Repurchase” portfolios are the result of splitting the firms with positive amounts of repurchases based on the median (i.e, below median and above median) for the year The “Infrequent Repurchase” and “Frequent Repurchase” portfolios are the result of splitting the firms with positive amounts of repurchases based on whether they repurchase during one or two quarters (i.e, infrequently) or three or four quarters (i.e, frequently) of the year 30 Figure 3. Trading Volume This figure shows average annual trading volume as a percentage of shares outstanding across repurchase portfolios from 1988 to 2020. The “All Firms” portfolio includes the entire sample We then split the entire sample into firms that do not repurchase shares (i.e, the

“Zero Repurchase” portfolio) and firms with positive repurchase amounts (ie, the “Positive Repurchase” portfolio). We further split the positive repurchase firms in two ways First, we split them into two groups that are above and below the median (i.e, the “Small Positive Repurchase” and “Large Positive Repurchase” portfolios) repurchase amount each year. Second, we split them into two groups based on whether they repurchase during one or two quarters (i.e, “Infrequent Repurchase”) or three or four quarters (ie, “Frequent Repurchase”) of the year. 31 Figure 4. Investment, Profitability, and Book-to-Market This figure shows average annual investment (Panel A), profitability (Panel B), and book-to-market (Panel C) across repurchase portfolios from 1988 to 2020. Investment is the sum of capital expenditures and research and development outlays, scaled by total assets. Profitability is net income, scaled by total assets The “All Firms” portfolio includes

the entire sample. We then split the entire sample into firms that do not repurchase shares (ie, the “Zero Repurchase” portfolio) and firms with positive repurchase amounts (i.e, the “Positive Repurchase” portfolio) We further split the positive repurchase firms in two ways. First, we split them into two groups that are above and below the median (ie, the “Small Positive Repurchase” and “Large Positive Repurchase” portfolios) repurchase amount each year. Second, we split them into two groups based on whether they repurchase during one or two quarters (i.e, “Infrequent Repurchase”) or three or four quarters (i.e, “Frequent Repurchase”) of the year Panel A: Investment 32 Panel B: Profitability 33 Panel C: Book-to-Market 34 Table 1. Aggregate Payouts Panel A of this table shows the number of US public firms making dividend and repurchase payouts as well as aggregate dividend and repurchase payouts ($ in billions) from 1988 to 2020. We also present

the number of firms in our sample as well as the aggregate equity market capitalization of the firms in our sample for comparison. Panel B shows the amounts after inflation-adjusting using the Consumer Price Index, with 2015 taken as the reference point. Panel A: Nominal Values Year 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Dividends Aggregate # of firms $ in billions 1322 191 1388 174 1356 181 1321 176 1379 189 1423 200 1474 208 1495 212 1466 231 1458 243 1481 270 1342 263 1216 285 1063 264 942 269 972 280 1110 333 1190 473 1168 426 1150 495 1066 528 927 553 929 491 996 533 1049 636 1026 697 1086 758 1118 855 1086 850 1061 865 1054 933 1065 972 1099 968 Repurchases Aggregate # of firms $ in billions 695 63 604 44 687 36 504 22 458 24 493 27 563 38 681 73 736 69 900 133 1110 190 1193 189 1053 181 841 130 666 126 681 138 562 219 722 352 817 467 890 597 1073 511

723 201 738 351 946 558 951 474 912 549 990 724 1136 806 1114 702 1098 640 1191 1070 1264 953 1211 715 35 Market Cap Aggregate # of firms $ in billions 3370 2122 3437 2135 3254 2454 3321 2460 3672 2993 4023 3328 4403 3703 4592 4052 4879 5373 5182 6384 5127 8541 4790 10033 4567 13562 4097 12278 3635 9565 3528 7934 3511 9934 3479 10886 3458 11641 3393 12760 3158 13071 2891 8995 2904 10176 2836 12180 2751 12260 2716 14219 2787 17552 2806 18956 2760 18366 2736 20052 2712 22992 2737 24136 2835 25002 Panel B: Inflation-Adjusted Values Year 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Dividends Aggregate 2015 $ in # of firms billions 1322 382 1388 332 1356 329 1321 306 1379 320 1423 329 1474 332 1495 329 1466 349 1458 359 1481 392 1342 374 1216 392 1063 353 942 355 972 361 1110 418 1190 574 1168 501 1150 565 1066 581 927 611 929 533 996 562 1049 656 1026 709 1086 758

1118 855 1086 839 1061 837 1054 881 1065 901 1099 886 Repurchases Aggregate 2015 $ in # of firms billions 695 126 604 84 687 65 504 38 458 41 493 44 563 61 681 114 736 104 900 196 1110 277 1193 268 1053 249 841 174 666 166 681 177 562 275 722 428 817 549 890 683 1073 563 723 222 738 382 946 588 951 490 912 559 990 725 1136 806 1114 693 1098 618 1191 1010 1264 884 1211 654 36 Market Cap Aggregate 2015 $ in # of firms billions 3370 4252 3437 4082 3254 4452 3321 4280 3672 5056 4023 5461 4403 5921 4592 6303 4879 8119 5182 9426 5127 12419 4790 14275 4567 18666 4097 16436 3635 12603 3528 10222 3511 12465 3479 13212 3458 13687 3393 14587 3158 14390 2891 9938 2904 11061 2836 12834 2751 12656 2716 14466 2787 17573 2806 18956 2760 18137 2736 19389 2712 21702 2737 22376 2835 22897 Table 2. Quarterly Returns This table shows average quarterly value-weighted returns across repurchase portfolios. Panel A shows raw returns, i.e, we subtract the risk-free rate (ie, Retrf, 3-month Treasury Bill

return) from portfolio returns, Retp Panel B shows Fama-French alphas, i.e, the intercept from the regression of Retp - Retrf on the five Fama-French factors and the momentum factor. *, , and indicate significance at the 1, 5, and 10 percent level, respectively. The sample period is 1988 to 2020. Quarter 0 is the quarter in which we measure repurchases Panel A: Raw Returns Quarter -1 2.63 Quarter 0 3.05 Small Positive Repurchase Large Positive Repurchase Large Positive - Zero 2.01 2.76 0.13 2.34 2.84 -0.20 2.80 3.46 0.70* 3.44 3.30 0.46 Infrequent Repurchase Frequent Repurchase Frequent - Zero 1.79 2.75 0.12 2.33 2.98 -0.07 2.52 2.92 0.16 3.95 2.88 0.04 Quarter -1 0.66 Quarter 0 0.88 Quarter +1 0.80 Quarter +2 0.96 Small Positive Repurchase Large Positive Repurchase Large Positive - Zero -0.09 0.79 0.13 -0.13 0.60 -0.28 0.43 1.34 0.54 1.01 1.10 0.14 Infrequent Repurchase Frequent Repurchase Frequent - Zero -0.25 0.75 0.09 0.03 0.73 -0.15 0.46 0.69 -0.11 1.51

0.64 -0.32 Zero Repurchase Quarter +1 2.76 Quarter +2 2.84 Panel B: Fama-French Alphas Zero Repurchase 37 Table 3. Annual Returns This table shows average annual value-weighted returns across repurchase portfolios. Panel A shows raw returns, ie, we subtract the risk-free rate (i.e, Retrf, 3-month Treasury Bill return) from portfolio returns, Retp Panel B shows Fama-French alphas, i.e, the intercept from the regression of Retp - Retrf on the five Fama-French factors and the momentum factor. *, , and indicate significance at the 1, 5, and 10 percent level, respectively. The sample period is 1988 to 2020. Year 0 is the year in which we measure repurchases Panel A: Raw Returns Zero Repurchase Year 0 15.74 Year +1 12.88 Small Positive Repurchase Large Positive Repurchase Large Positive - Zero 12.81 12.64 -3.11* 11.79 13.64 0.76 Infrequent Repurchase Frequent Repurchase Frequent - Zero 11.75 14.63 -1.11 11.80 12.19 -0.69 Panel B: Fama-French Alphas Zero Repurchase

Year 0 7.01 Year +1 3.69 Small Positive Repurchase Large Positive Repurchase Large Positive - Zero 4.58 3.95 -3.05 4.43 4.99 1.30 Infrequent Repurchase Frequent Repurchase Frequent - Zero 2.76 6.64 -0.37 4.46 3.00 -0.69 38 Table 4. CEO Pay This table shows average annual CEO total and excess pay across repurchase portfolios from 2006 to 2020. The time period examined is shorter than in prior analyses due to data availability and limitations. Excess CEO pay is estimated as the difference between total pay and the predicted value from the following model: log(Total CEO Pay)i,t = β0 + β1 x log(Market Cap)i,t-1 + β2 x log(Revenue)i,t-1 + β3 x Book-to-Marketi,t-1 + β4 x Stock Returni,t-1 + β5 x ROAi,t-1 + εi,t. Panel A shows the estimated parameters from this model, both with and without year fixed effects. t-statistics are reported in parentheses below coefficients and are based on standard errors that are clustered by firm. *, , and indicate significance at the 1, 5,

and 10 percent level, respectively. In Panel B, we present means across repurchase portfolios for total CEO pay and excess CEO pay, as well as the economic determinants of CEO pay included in the model. N is the number of observations in each portfolio The “Zero Repurchase” portfolio includes firms that do not repurchase shares. The “Positive Repurchase” portfolio includes all firms with positive repurchase amounts The “Small Positive Repurchase” and “Large Positive Repurchase” portfolios are the result of splitting the firms with positive amounts of repurchases based on the median (i.e, below median and above median) for the year The “Infrequent Repurchase” and “Frequent Repurchase” portfolios are the result of splitting the firms with positive amounts of repurchases based on whether they repurchase during one or two quarters (i.e, infrequently) or three or four quarters (i.e, frequently) of the year The number of observations in the “Infrequent Repurchase”

and “Frequent Repurchase” portfolios does not sum to exactly the number of observations in the “Positive Repurchase” portfolio because a slightly different definition of repurchases is used to identify repurchase frequency, following prior literature (e.g, Banyi, Dyl, and Kahle, 2008) Finally, the “All Firms” portfolio includes all firms Panel A: Regression Estimates Independent Variable log(Market Cap) log(Revenue) Book-to-Market Stock Return ROA Year Fixed Effects? R2 N Y = log(Total CEO Pay) (1) (2) 0.31* 0.29* (21.91) (20.47) 0.17* 0.17* (14.78) (14.82) -0.00 -0.01 (-0.25) (-0.32) 0.16* 0.15* (14.48) (12.25) -0.67* -0.61* (-7.58) (-6.85) No Yes 0.5274 0.5392 21114 21114 39 Panel B: Variation across Repurchase Portfolios Repurchases Portfolio Zero Repurchase Positive Repurchase Small Positive Repurchase Large Positive Repurchase Infrequent Repurchase Frequent Repurchase All Firms Repurchases Portfolio Zero Repurchase Positive Repurchase Small Positive Repurchase

Large Positive Repurchase Infrequent Repurchase Frequent Repurchase All Firms N Total CEO Pay ($ in thousands) Excess CEO Pay ($ in thousands) Market Cap ($ in millions) 8788 12326 5515 6811 4532 9836 21114 5120 7458 6592 8159 5867 7956 6485 20 -2 -93 71 -46 24 7 5950 11862 9547 13736 7268 13333 9401 Revenue ($ in millions) Book-to-Market Stock Return ROA 1111 2040 1626 2376 1409 2196 1654 58.8% 56.3% 58.9% 54.2% 59.9% 51.9% 57.4% 19.0% 11.4% 11.9% 11.0% 12.4% 12.9% 14.6% 2.1% 5.9% 4.3% 7.2% 3.8% 6.6% 4.3% 40