The Role of Institutional Investors in Open-Market Share Repurchase Programs Thomas J. Chemmanur Yingzhen Li February 15, 2015 Professor of Finance, Fulton Hall 330, Carroll School of Management, Boston College, Chestnut Hill, MA 02467. Phone: 617-552-3980. Fax: 617-552-0431. E-mail: chemmanu@bc.edu. The Brattle Group. Phone: 415-217-1000. E-mail: Yingzhen.Li@Brattle.com. For helpful comments and discussions, we thank Sugata Roychowdhury. We are responsible for any remaining errors or omissions.
The Role of Institutional Investors in Open-Market Share Repurchase Programs Abstract Theoretical papers such as Brennan and Thakor (1990) argue that large investors such as institutions produce information about firms around stock repurchases. In this article, we use a large sample of transaction-level institutional trading data and the data on US firms actual repurchases made available by the 2003 revisions to the Exchange Act to test this assumption about open-market repurchase programs (OMRs) for the first time in the literature. We document several results new to the literature. First, institutional trading prior to an OMR announcement has significant predictive power for the magnitude of the abnormal return to the firm s equity around OMR announcements. Second, institutional trading immediately after an OMR announcement has significant predictive power for the firm s subsequent stock return performance: the larger the net buying by institutions, better the subsequent long-run stock returns. Third, institutions are able to realize significant abnormal profits (net of commissions and trading costs) by trading in the equity of firms undergoing open-market repurchases. Fourth, institutional trading immediately after an OMR announcement has significant predictive power for the actual repurchases by the firm: a larger amount of buying of the firm s equity by institutions is associated with a larger actual repurchase by the firm in the subsequent period. Finally, institutional trading has predictive power for the change in the firm s information asymmetry from before the OMR announcement to after: a larger amount of net buying by institutions is associated with a larger decrease in the degree of information asymmetry facing the firm in the equity market. Overall, our results are consistent with the notion that institutional investors are able to generate a significant information advantage for themselves about firms undergoing open-market repurchases. JEL Classification: G23, G35 Keywords: Open-Market Repurchases; Institutional Trading; Information Production
1 Introduction In recent years, the number of firms undertaking stock repurchases has increased dramatically, while the proportion of firms distributing value through cash dividends has declined (see, e.g., Fama and French (2001)). The popularity of share repurchases has not been mitigated even after the passage of the Jobs and Growth Tax Relief Act of 2003 in the U.S. which cut the dividend tax rate to 15%, thus substantially reducing the tax disadvantage of dividend payments to investors (see, e.g., Chetty and Saez (2006)). Open-market repurchases (OMRs) constitute around 90% of the stock repurchases consummated in recent years: see, e.g., Comment and Jarrell (1991) or Grullon and Michaely (2004). An interesting question in this context is regarding the precise economic role played by repurchase programs in general and OMR programs in particular in maximizing shareholder value. The rationale for repurchase programs given by the existing literature is that they serve to signal firm insiders private information about the intrinsic value of the firm to outsiders in the equity market: see, e.g., Vermaelen (1981), Ofer and Thakor (1987), and Constantinides and Grundy (1989). 1 One interesting question in the above context is the role of institutional investors. The theoretical model of Brennan and Thakor (1990) assumes that large investors such as institutions have the ability to produce information about firms in the context of their choice of payment method between open-market repurchases, dividends, and tender o ers. 2 If institutional investors do indeed possess the ability to produce information about firms undergoing OMR programs, how does this information interact with that of firm insiders? To the best of our knowledge, there has been no paper in the literature that empirically analyzes whether institutions indeed have the ability to produce information about firms announcing OMR 1 Most of these signaling models assume that the firm commits to buy a certain number of shares, as in the case in Dutch auction or tender o er repurchase programs. See, however, Oded (2005), who develops a theoretical model demonstrating that, under suitable assumptions, open-market share repurchase programs can signal firm insiders private information even in the absence of a commitment by the firm to repurchase the entire amount of equity authorized to be repurchased by its board. 2 Brennan and Thakor (1990), however, do not assume any private information held by firm insiders about its intrinsic value. 1
programs, and how their information interacts with the private information held by firm insiders (which they may attempt to convey to the equity market through a repurchase program). The objective of this paper is to fill this gap in the literature. The economic setting we consider to develop our empirical analysis can be described as follows. Consider a situation where the insiders of a firm, having private information about its intrinsic value, are considering whether or not to undertake an open-market repurchase program. Let there be two types of outside investors in the equity market. The first type of investors are institutional investors, who have the ability to produce noisy information about the firm (at a cost). The precision of information produced by institutions is lower than that of the private information held by firm insiders, so that, while information production helps institutions reduce their information disadvantage with respect to firm insiders, it does not eliminate it. The second type of investors are retail investors, who do not have any ability to produce information about the intrinsic value of the firm, and are therefore at a disadvantage with respect to both institutions and insiders. Retail investors are essentially liquidity traders in the equity market in the economic setting we study here, similar to their role in market microstructure models such as Kyle (1985). The price of the firm s stock in the equity market is set by a market maker who is uninformed to begin with, but who sets the stock price to break even (after observing the aggregate order flow of trades in the firm s equity), again similar to the price-setting rule in market microstructure models. While the market maker cannot fully separate informed and uninformed trades, the price of the firm s equity will reflect, to some degree, the information held by institutional investors. 3 We address the following research questions in the above economic setting. First, do institutions have the ability to produce information about a firm prior to its announcing an OMR program? We address this question by analyzing whether institutional trading prior 3 The above economic environment that we postulate here is very similar to that of the theoretical model of Chemmanur and Jiao (2011), who, however, focus on the theoretically analyzing on the implications of institutional trading around seasoned equity o erings. Given the absence of any formal theoretical analysis of open-market stock repurchases in an economic environment similar to the above, we choose to adapt the analysis of Chemmanur and Jiao (2011) to the case of open-market repurchases in order to develop testable hypotheses for our empirical analysis. 2
to an OMR program announcement has predictive power for the announcement e ect of such a program. Second, do institutions have the ability to produce information about a firm immediately after its announcing an OMR program, and are they able to profit from this information? We address this question by studying the predictive power of institutional trading immediately after an OMR program for the subsequent long-run performance of the firm s equity and also the profitability of such trading. We also address this research question by analyzing the predictive power of institutional trading immediately after an OMR program announcement for the actual repurchases made by the firm in the subsequent period. Finally, we analyze how institutional trading immediately after an OMR program announcement a ects the change in the information asymmetry faced by the firm from before the announcement of the OMR program to after. As we discuss in detail in Section 3, the last analysis allows us to address the question of how the information produced by institutions interacts with the private information of insiders that is conveyed through the OMR program. We make use of a detailed transaction-level institutional trading database provided by Abel Noser Solutions (formerly Ancerno Ltd., or Abel/Noser Corporation) to address the above research questions. Our data includes transactional-level trading data spanning twelve years from January 2003 to September 2011 originated from 868 di erent institutions, with an aggregate annualized trading principal of around $9 trillion on all U.S. domestic equity. For an average open-market repurchase event, our sample institutions collectively account for about 12% of the CRSP-reported trading volume within the two-year period surrounding the open-market repurchase announcement. With this dataset, we are able to track institutional trading both before and after an open-market repurchase announcement. We are also able to compute realized institutional trading profitability net of explicit trading costs (i.e., brokerage commissions) and implicit trading costs (i.e., market impact). Throughout this paper, we use a variable we call Net Buy to measure institutional trading. We define Net Buy as the number of shares purchased by institutions minus the number of shares sold by 3
institutions, normalized by the number of shares outstanding. Our paper provides a number of new results on the role of institutional investors around open-market repurchases. We organize our empirical tests and results into five parts, corresponding to the five sets of research questions outlined above. First, we study, for the first time in the literature, the informativeness of institutional trading before the announcement of an open-market repurchase program. We find that institutional trading before an openmarket repurchase announcement has considerable predictive power for announcement e ect of programs. Greater selling by institutional investors is significantly associated with a more favorable announcement e ect. This result is robust to controlling for various variables that have been found in the prior literature to be able to predict announcement e ects of openmarket repurchase programs, including prior firm performance and insider trading. This suggests that institutional investors possess private information about the intrinsic value of the firms announcing open-market repurchase programs. Second, we study the predictive power of institutional trading immediately after openmarket repurchase announcements (two quarters) for the firm s subsequent long-run (one year) performance, again for the first time in the literature. We find that institutional trading immediately after an open-market repurchase announcement has considerable predictive power for the firm s subsequent long-run stock performance: a 1% increase in institutional net buying is associated with about 0.4% increase in the firm s abnormal stock return over the subsequent one-year period. This result is robust to controlling for various variables capturing publicly available information, as well as the extent of trading in the firm s equity by insiders. The above result indicates that institutions have a residual information advantage over retail investors, even after the announcement of an open-market repurchase program. Third, we study the realized profitability of institutional trading around open-market repurchase programs, using actual transaction prices and net of brokerage commissions, for the first time in the literature. We find that institutions make positive abnormal profits after the announcement of an open-market program, even after taking commissions and other 4
trading costs into account. This is especially the case when the information conveyed by the announcement of an open-market program is noisier (i.e., when the size of the repurchase program is smaller or when the firm actually repurchases less subsequent to the announcement). For example, over the one-year horizon after an open-market repurchase program announcement, our sample institutions on average realize a risk-adjusted return of 1% when the size of the repurchase program is smaller (i.e., below the sample median), and they realize a risk-adjusted return of 0.8% when the firm actually repurchases less subsequent to the announcement (i.e., below the sample median). These results suggest that the information advantage possessed by institutional investors after an open-market repurchase announcement that we documented earlier translates into real trading profits, especially when the information conveyed by the open-market repurchase announcement made by the firm is noisier (i.e., when the size of the repurchase program is smaller or when the firm actually repurchases less subsequent to the announcement). In summary, we are able to show not only that institutions possess an information advantage after open-market repurchase announcements, but are also able to translate this advantage into real trading profits even after accounting for both explicit (brokerage commissions) and implicit (market impact) trading costs. Fourth, we study the predictive power of institutional trading immediately after openmarket repurchase announcements for the actual repurchases made by the firm in the subsequent period, again for the first time in the literature. We find that institutional trading immediately after an open-market repurchase announcement has considerable predictive power for the subsequent actual repurchases made by the firm: a 1% increase in institutional net buying is associated with about 3% increase in the firm s actual repurchase over the subsequent two-quarter period. This result is robust to controlling for various variables capturing publicly available information, as well as the extent of trading in the firm s equity by insiders. The above result again indicates that institutions have a residual information advantage over retail investors, even after the announcement of an open-market repurchase 5
program. Finally, we examine how institutional trading around open-market repurchase programs a ects the information asymmetry faced by the firm in the equity market. We find that institutional trading over the two-quarter period immediately after an open-market repurchase program announcement has considerable predictive power for the reduction in information asymmetry faced by the firm around the repurchase program announcement (i.e., from before the announcement to after). Greater net buying by institutional investors is associated with more analyst coverage (i.e., more analysts covering the firm) and lower analyst forecasts dispersion (i.e., more precise analyst forecasts), two of the proxies for information asymmetry widely used in the literature. These results indicate that the information conveyed by institutional trading subsequent to open-market repurchase program announcements plays an additional role in reducing the information asymmetry faced by the firm, over and above that of the repurchase announcement itself in reducing this information asymmetry. What do we learn from the above empirical results about the role of institutional investors as information producers around open-market repurchase programs? First, the fact that institutional trading has predictive power for the announcement e ects of open-market repurchase programs, for the long-run stock return performance of the firm subsequent to the announcement of such programs, and for the actual repurchases made by the firm indicates that they have information advantage over retail investors both before and after the announcement of an open-market repurchase program. It should be noted that institutional investors continue to possess this information advantage over retail investors even after an open-market repurchase program announcement, which may convey firm insiders private information as well. Second, our results indicate that institutions are able to use their information advantage to realize abnormal profits by trading in the equity of firms undergoing open-market repurchase programs, especially when the size of the open-market repurchase program is smaller or when the firm makes less actual repurchase subsequent to the announcement of the repurchase program. It is worth noting that we are able to generate the 6
above results mainly because we utilize transaction-level institutional trading data, which also enable us to account for both explicit (brokerage commissions) and implicit (market impact) trading costs of institutional investors. Finally, our results of the e ect of institutional trading on the reduction in the information asymmetry facing the firm in the equity market around open-market repurchase programs indicate that not only do institutional investors have information advantage relative to retail investors around such programs, they are also able to reduce this information asymmetry facing the firms by trading in the equity of firms undergoing such repurchase programs. The rest of the paper is organized as follows. Section 2 relates this paper to the existing literature. Section 3 develops testable hypotheses about institutional trading around openmarket repurchases. Section 4 describes the data and various variables used in this study. Section 5 presents empirical tests and results on institutional trading around open-market repurchases. Section 6 concludes. 2 Relation to the Existing Literature Our paper lies in the intersection of two literatures. The first is the literature on openmarket repurchases. 4 Under an asymmetric information framework, Ofer and Thakor (1987) develop a model in which firms could signal their value through two mechanisms, paying dividends or repurchasing their shares. In their model, the signaling costs through repurchases are higher and thus repurchases are a stronger signal perceived by the equity market. Oded (2005) o ers some theoretical insights on why a price increase accompanies the announcement of an open-market stock repurchase program, even though the announcement of an OMR program is not a commitment to buy back all the shares authorized. In his model, firms face a trade-o between the long-run gains from the informed trading that the option to repurchase shares creates and the short-run costs from the market s accounting for this 4 To the extent that an open-market repurchase program is a form of cash distribution, our paper is also related to the broader literature on payout policy: see Allen and Michaely (2003) for an extensive review on the payout policy literature. 7
adverse selection. Under this trade-o, only good firms announce open-market repurchase programs, so that the announcement of an OMR program acts as a credible signal. On the empirical side, Dann (1981) notes that the announcement of open market share repurchase programs tends to be preceded by a recent decline in the announcing firms stock price. Comment and Jarrell (1991) study the relative signaling power of Dutch-auction, selftender o ers and open-market repurchases. Ikenberry, Lakonishok, and Vermaelen (1995) study long-run stock returns following OMR announcements. Nevertheless, there has been no study in the literature analyzing the role of institutional investors as information producers around open market share repurchase programs. 5 The second literature our paper is related to is the theoretical and empirical literature on institutional trading and information production by institutional investors around corporate events. Chemmanur and Jiao (2011) develop a model of institutional trading and information production around SEOs. Gibson, Safieddine, and Sonti (2004) and Chemmanur, He, and Hu (2009) empirically analyze institutional trading around SEOs. Ours is, however, the only paper to study institutional trading around open-market repurchases, and the only one to make use of transaction-level institutional trading data. While this is not the primary focus, our paper also contributes to the broader literature on the determinants of the extent of information production by outsiders about a firm. Anumberofauthorshavedevelopedtheoreticalanalysesabouttheincentivesofoutsiders to acquire information about a firm (see, e.g., Grossman (1976), Grossman and Stiglitz (1980), Diamond and Verrecchia (1981), Verrecchia (1982), Admati and Pfleiderer (1986), and Bhushan (1989a)). There have also been a number of empirical analyses regarding the extent of information production by analysts about a firm (see, e.g., Bhushan (1989b), O Brien and Bhushan (1990), Lang and Lundholm (1996)) and the informativeness or accu- 5 Two other contemporaneous papers also study trading by institutional investors around stock repurchases. DeLisle, Morscheck, and Nofsinger (2014) document that institutional investors are net sellers during share repurchases. In a similar spirit, Huang and Zhang (2013) show that institutions sell after share repurchase announcements. Neither of these papers analyze and test hypotheses regarding the information production role of institutional investors that is our focus here. 8
racy of the information produced (see, e.g., Frankel, Kothari, and Weber (2006)). Much of the above empirical literature has focused on the cross-sectional variation in firm characteristics that lead to di erences in information production by analysts and others across firms. In contrast to the above information production literature, our paper focuses on information production by institutional investors around a specific corporate event, namely, open-market repurchases. 3 Theory and Development of Hypotheses In this section we will briefly discuss the underlying theoretical motivation and develop the hypotheses we test in our empirical analysis. An important theoretical model where institutional investors produce information around payout (dividends or repurchases) is that of Brennan and Thakor (1990). The authors build on the notion that the share price is not a perfect aggregator of the private information of investors about the prospects of the firm and that the collection of information by investors is costly. Under this assumption, it follows that share repurchases oblige shareholders either to incur information collection costs, or to run the risk of partial expropriation by trade with better informed investors. As a result, uninformed shareholders (retail investors) prefer dividends to repurchases, while informed investors (large investors such as institutional investors) will prefer repurchases because this allows them to profit at the expense of the uninformed. Thus, the focus of Brennan and Thakor (1990) is on a firm s choice between cash dividends, open-market repurchases, and tender o er repurchases. Unlike the above theoretical analysis, the focus of our empirical analysis is not on alternative methods of cash distribution. However, we empirically test whether institutions indeed have the ability to produce valuable information about firms undergoing open-market repurchase programs. The economic setting we consider to develop testable hypotheses for our empirical tests can be described as follows. Consider a situation where the insiders of a firm, having private 9
information about its intrinsic value, are considering whether or not to undertake an openmarket repurchase program. Outsiders in the equity market consist of two types of investors. The first type of investors are institutional investors, who have the ability to produce noisy information about the firm (at a cost). The precision of information produced by institutions is lower than that of the private information held by firm insiders, so that, while information production helps institutions reduce their information disadvantage with respect to firm insiders, it does not eliminate it. The second type of investors are retail investors, who do not have any ability to produce information about the intrinsic value of the firm, and are therefore at a disadvantage with respect to both institutions and insiders. Retail investors are essentially liquidity traders in the equity market in the economic setting we study here, similar to their role in market microstructure models such as Kyle (1985). The price of the firm s stock in the equity market is set by a market maker who is uninformed to begin with, but who sets the stock price to break even (after observing the aggregate order flow of trades in the firm s equity), again similar to the price-setting rule in market microstructure models. While the market maker cannot fully separate the informed and uninformed trades, the price of the firm s equity will reflect, to some degree, the information held by institutional investors. The above economic environment that we postulate here is very similar to that of the theoretical model of Chemmanur and Jiao (2011), who, however, focus on the theoretically analyzing on the implications of institutional trading around seasoned equity o erings. Given the absence of any formal theoretical analysis of open-market stock repurchases in an economic environment similar to the above, we choose to adapt the analysis of Chemmanur and Jiao (2011) to the case of open-market repurchases in order to develop testable hypotheses for our empirical analysis. 6 In the following, we develop testable hypotheses for analyzing the 6 Apart from Brennan and Thakor (1990), we are not aware of any theoretical model of stock repurchases which incorporates the role of institutional investors and their ability to produce information about the firm. In the interest of conserving space, we choose not to develop a formal theoretical model here to analyze institutional trading around stock repurchases, but instead adapt the theoretical analysis of Chemmanur and Jiao (2011) to the stock repurchase setting. 10
e ect of institutional trading around open-market repurchases for the announcement e ect of a stock repurchases; actual shares repurchased (as against the authorized repurchase); the long-run return performance of the firm s equity subsequent to the announcement of an open-market repurchase; and finally, the abnormal profits realized by institutional investors (net of all transaction costs) by trading in the equity of firms subsequent to their announcement of an open-market repurchase program. We also develop testable hypotheses for the relationship between institutional trading immediately after an OMR program announcement and the change in the information asymmetry faced by the firm from before the announcement of the OMR program to after. Our first hypothesis deals with the relationship between institutional trading and the announcement e ects of open-market repurchase programs. Before the announcement of the repurchase program, ordinary (uninformed) investors in the equity market will have a prior valuation of the firm, which will be reflected in the pricing of the firm s equity before the repurchase announcement. Consider now the scenario where institutional investors trade in the firm s equity prior to the repurchase announcement, and the extent (number of shares) and direction (buy or sell) is publicly disclosed subsequent to their trading. In this case, assuming that all investors are aware that institutions have more precise information compared to other outsiders, the stock market will reflect the additional information contained in the trading by institutions, with the price of the firm s equity falling lower if the net buy by institutions (number of shares bought minus number of shares sold) is negative, while it will rise higher if their net buy is positive. At this point, if the firm announces an open-market repurchase program, stock market investors will further positively update the value of the firm s equity, knowing that the decision to repurchase or not is made by firm insiders who have information that is even more precise than that of institutional investors about the intrinsic value of the firm. 7 This means that the announcement e ect, which will reflect the 7 Here we are assuming that an open-market repurchase program credibly conveys information about the firm from firm insiders to outside investors in the equity market. Some readers may wonder how an open-market repurchase program can convey such information, given that the firm does not commit to repurchase the entire number of shares authorized by the board. It should be noted in answer to this 11
di erence in the firm s stock price from before the repurchase to immediately after, will be decreasing in the net buy of institutional investors. This will be the first hypothesis that we test here (H1). We now turn to trading by institutions subsequent to the announcement of an openmarket repurchase program. If the announcement of an open-market repurchase conveys firm insiders private information to the equity market, the price of the firm s stock immediately after the repurchase announcement will reflect this information. However, given that, in an open-market repurchase program, the firm does not make a commitment to repurchase all the shares authorized by the board, the repurchase announcement may only be a noisy signal of firm insiders private information. If this is indeed the case, institutions may have aresidualinformationadvantageoverretailinvestorsevenaftertheannouncementofan open-market repurchase program. If such an information advantage persists, then trading by institutions after an open-market repurchase announcement will have predictive power for future firm performance. This is the second hypothesis we test here (H2). Further, in this scenario, institutions will be able to make abnormal profits using their information advantage by trading in the equity of firms announcing repurchase programs (subsequent to the repurchase). This is therefore the third hypothesis we test here (H3). 8 If indeed, the repurchase announcement is only a noisy signal of firm insiders private information, then firm insiders themselves may have residual private information relative to outsiders about their firm s intrinsic value as well. Further, the more favorable this residual private information held by insiders (i.e., the more undervalued insiders believe their firm s question that an open-market repurchase program can indeed credibly convey such information as long as there is a monotonic relationship between firm insiders private information and the size (number of shares) of the repurchase program authorized (which, in turn, can be shown to be the case if firm insiders will incur a cost of falsely announcing a larger program than that warranted by insiders private information). See Oded (2005) for such a theoretical model of firms credibly signaling insiders private information through an open-market repurchase program, even in the absence of a commitment by the firm to repurchase all the shares authorized. 8 The residual information advantage of institutional investors over retail investors is likely to be greater as the information conveyed by the repurchase announcement itself becomes more noisy: e.g., for repurchase programs that are smaller (as a fraction of total shares outstanding). This, in turn, implies that the abnormal profits made by institutions from trading in the firm s equity after an open-market repurchase program announcement will also be greater for smaller repurchase programs. 12
equity is even after the repurchase announcement), the larger the amount of shares the firm is likely to actually repurchase (normalized by the total number of shares outstanding in the firm), in an attempt to eliminate this residual undervaluation. At the same time, as discussed above, the noisier the signal conveyed by the announcement of the repurchase program by the firm, the greater the information advantage held by institutional investors relative to retail investors. This means that trading by institutions after an open-market repurchase announcement will be positively associated with the amount of shares actually repurchased by the firm (normalized by the total number of shares outstanding). Thus, a greater net buy of the firm s equity by institutions after an open-market repurchase announcement will be positively related to the actual repurchases made by the firm in the subsequent period (H4). Finally, it is interesting to examine how institutional trading around open-market repurchase programs a ects the information asymmetry faced by the firm. Clearly, if open-market repurchase announcements serve as noisy signals of firm insiders private information to outsiders in the equity market, then the information asymmetry faced by the firm will be reduced: i.e, the extent of information asymmetry faced by the firm in the equity market subsequent to an open-market repurchase announcement will be lower than that before the repurchase announcement. The question we examine here, however, is the e ect of the interaction between the signal conveyed by the repurchase program announcement and the information conveyed by institutional trading immediately after the announcement of the repurchase on the information asymmetry facing the firm. In particular, the reduction in the information asymmetry facing the firm will be greater when the noisy signal conveyed by the repurchase announcement and the information conveyed to the equity market by institutional trading reinforce each other (which will be the case when the institutional net buy immediately after the repurchase announcement is positive). On the other hand, the reduction in information asymmetry facing the firm will be smaller when the noisy signal conveyed by the repurchase announcement and the information conveyed to the equity market by in- 13
stitutional trading oppose each other (which will be the case when the institutional net buy immediately after the repurchase announcement is negative). In summary, the reduction in information asymmetry facing the firm following the announcement of an open-market repurchase program will be positively associated with the institutional net buy immediately after the repurchase announcement (H5). 4 Data and Summary Statistics 4.1 Open-Market Repurchases Data The data on open-market repurchase programs (OMRs) in this study come from several sources. Our initial sample of open-market repurchase announcements from January 2004 to December 2010 comes from the SDC Platinum Database of Mergers and Acquisitions. We then exclude announcements such that the repurchase may be executed through tender o er, private negotiation, or Dutch auction. 9 If a firm makes multiple repurchase announcements in the same calendar year, we only keep the first announcement. We also require that accounting information from Compustat and stock return information from CRSP are available for the firms in our OMR data. Our data on U.S. firms actual share repurchases come from Quarterly Compustat, which are made availably by the regulatory changes to Rule 10b-18 of the Securities Exchange Act of 1934 in 2003. 10 Beginning from 2004, U.S. firms are required to make quarterly disclosures of actual share repurchases and average prices paid. We then match this actual repurchase data from Quarterly Compustat with the data on OMR announcements we obtained from SDC. 9 The purpose is to eliminate repurchase programs that may be executed through a combination of methods (e.g., open-market and private negotiation). 10 Earlier studies (e.g., Stephens and Weisbach (1998), Fama and French (2001), and Grullon and Michaely (2002)) have used a variety of other CRSP- and Compustat-based measures to estimate actual share repurchases by U.S. firms. These estimations invariably su er from di erent measurement biases. For a detailed discussion of these measures, see Banyi, Dyl, and Kahle (2008). 14
Table 1 reports summary statistics of our OMR data. We have about 3,000 open-market repurchase programs announced from January 2004 to December 2010. The average OMR program size, defined as the dollar amount value of the OMR program normalized by the market capitalization of the firm, is 7.94%, consistent with prior studies in the literature (e.g., Peyer and Vermaelen (2009)). We find a significant 1.74% average abnormal return in the 5-day window around an OMR announcement, which is also consistent with the empirical findings in the literature (e.g., Babenko, Tserlukevich, and Vedrashko (2012)). Over the 4 fiscal quarters following an OMR announcement, our sample firms actual repurchases on average account for about 86% of the OMR program size announced. This is largely consistent with prior findings that firms complete a significant portion of the repurchase programs within the one-year period after the announcement (e.g., Stephens and Weisbach (1998)). 4.2 Institutional Trading Data We obtain transaction-level institutional trading data from Abel Noser Solutions, a leading execution quality measurement service provider for institutional investors. The data are similar to those used by several microstructure studies on institutional trading costs, for example, Keim and Madhavan (1995), Conrad, Johnson, and Wahal (2001), and Jones and Lipson (2001). To the best of our knowledge, this is the first paper to use institutional trading data to study institutional investors trading behavior around open-market repurchase programs. The data cover equity trading transactions by a large sample of institutions from January 2003 to September 2011. For each transaction, the data include the date of the transaction, the stock traded, the number of shares traded, the dollar principal traded, commissions paid by the institution, and whether it is a buy or sell by the institution. The data are provided to us under the condition that the names of all institutions are removed from the data. However, identification codes are provided enabling us to separately identify all institutions. 15
Sample institutions are either investment managers or plan sponsors. Investment managers are mutual fund families such as Fidelity Investments, Putnam Investments, and Lazard Asset Management. Examples of pension plan sponsors include United Airlines pension plan and the Commonwealth of Virginia. Table 2 reports summary statistics of our institutional data. We have 868 institutions in our sample, with 372 of them being investment managers and 496 of them being plan sponsors. In aggregate, these institutions have an annualized trading volume of around 304 billion shares and an annualized trading principal of around $9 trillion. In association with these trading activities, our sample institutions in aggregate incur an annualized commission expense of about $7.5 billion. If we consider a two-year trading horizon surrounding the OMR announcement dates in our OMR data, on average (for each OMR event), our sample institutions in aggregate execute about 24,510 transactions, with a trading principal of about $3.5 billion, and account for about 12% of the trading volume reported by CRSP. 5 Empirical Tests and Results 5.1 Pattern of Institutional Trading before OMR Announcements As the first step of our empirical analysis, we examine the pattern of institutional trading before open-market repurchase announcements. To the best of our knowledge, this has not been investigated in the literature before. We focus on institutional trading during the oneyear period before OMR announcements. Table 3 present the results of our analysis. We aggregate trading from institutional investors over di erent trading horizons before OMR announcements. For ease of disposition, Net Buy, our main variable to measure institutional trading, is expressed in basis points in Table 3. The general pattern is that Net Buy is significantly positive from minus quarter 2 to minus quarter 4. For example, during the 4th quarter before an OMR announcement, institutional investors in our sample on average net buy 20.85 basis points of the firm s shares outstanding. Nevertheless, during the quarter 16
immediately before OMR announcements, Net Buy is significantly negative, indicating that institutional investors on average net sell over this period. This sell-o is consistent with the large decline in stock price during the one-quarter period before OMR announcements. Further, the trading pattern over the one-year period before OMR announcements and over di erent sub-periods is largely similar for both investment managers and plan sponsors. 5.2 Institutional Trading and OMR Announcement E ects Hypothesis H1 predicts that Net Buy from institutions is inversely related with OMR announcement e ects. In this subsection, we make use of institutional trading data and examine the relationship between institutional trading and the announcement e ects of openmarket repurchase programs. Table 4 presents the results of our OLS analysis. The dependent variable is the cumulative abnormal return over a 5-day period around OMR announcements (i.e., the announcement return described in Table 1). Following the literature in open-market repurchases, we calculate the announcement e ects based on a market model, where the market beta is estimated with daily returns over the 250-day period ending 31 trading days before OMR announcements. The variable-of-interest is Net Buy from institutions. For ease of interpretation, from this subsection onwards, Net Buy is expressed in percentage rather than basis points. In Models (1) - (3), Net Buy is aggregated over the 6-month period before OMR announcements, whereas in Models (4) - (6), Net Buy is aggregated over the 12-month period before OMR announcements. From Model (1) we can see that the coe cient on Net Buy is negative and statistically significant. This is consistent with H1, suggestingthatinstitutionaltradingpriortoannouncements of open-market repurchase programs indeed conveys private information held by institutions about the intrinsic value of the firms. We control for usual suspects that have been found to be able to explain OMR announcement e ects in the literature, such as the size of the OMR programs and the past return of the firm. In a recent paper, Babenko, 17
Tserlukevich, and Vedrashko (2012) find that insider trading and insider holdings provide additional explanatory power regarding OMR announcement returns. In Models (2) and (3), we incrementally control for these variables and the coe cient on Net Buy remains negative and statistically significant. In Models (4) - (6), we aggregate institutional net buying over the 12-month period before OMR announcements and perform similar multivariate analysis as in Models (1) - (3). Our results remain qualitatively the same in the sense that the coe cients on Net Buy are all negative and statistically significant. To summarize this subsection, we find evidence that is consistent with hypothesis H1. Institutional trading before OMR announcements has predictive power for the announcement e ects, in the sense that smaller Net Buy is associated with more favorable OMR announcement e ects. This results holds even after we control for publicly available information such as size of the OMR program, prior firm performance, and insider trading. This evidence suggests that institutional trading prior to an open-market repurchase program announcement indeed reflects the private information held by institutional investors about the intrinsic value of the firm. 5.3 Institutional Trading and Subsequent Stock Return Performance In the previous subsection, we examine the relationship of institutional trading before OMR announcements and the OMR announcement e ects. From this subsection onwards, we focus on institutional trading immediately after OMR announcements and examine the informativeness of such trading. We first investigate whether institutional trading immediately after OMR announcements predicts subsequent stock return performance of the firm (H2). Table 5 reports the results of our multivariate analysis. The variable-of-interest is institutional trading, measured in Net Buy, after OMR announcements and we focus on institutional 18
trading over the two fiscal quarters after an OMR announcement. The independent variable is the buy-and-hold abnormal return over the subsequent 12-month period. 11 Several recent studies in the open market repurchases literature use the returns from matching firms to adjust long-run stock returns of the OMR announcing firms (e.g., Chan, Ikenberry, and Lee (2007) and Babenko, Tserlukevich, and Vedrashko (2012)). We follow these papers and employ this approach to calculate the buy-and-hold abnormal stock returns. Specifically, for each OMR announcing firm, we start with a list of firms that belong to the same industry as defined by the 2-digit SIC code but do not make OMR announcements within 3 years around the OMR announcement date of the announcing firm. For each firm in this group, we then calculate the sum of absolute deviations from the announcing firm s book-to-market ratio and market capitalization. The five firms with the smallest sum of absolute deviations are chosen as matching firms. If this procedure produces less than five matching firms, we relax the industry requirement by defining industries using the 1-digit SIC code. For each OMR announcing firm, the benchmark return is the buy-and-hold return of an equally-weighted portfolio of the corresponding five matching firms. As we can see in Model (1) in Table 5, the coe cient on Net Buy is positive and statistically significant, which is consistent with hypothesis H2. Wecontrolforthesizeofthe open-market repurchase program, the amount of shares actually repurchased by the firm during the two fiscal quarters after the announcement, and various variables capturing different aspects of the firm s characteristics. In Models (2) - (4), we separately control for contemporaneous insider trading, industry fixed e ects, and year fixed e ects. In Model (5), we control for these variables altogether. The coe cients on Net Buy remain positive and statistically significant. To summarize this subsection, we find evidence that institutional trading after OMR announcements has predictive power regarding the subsequent stock return performance of the OMR announcing firm. This is consistent with hypothesis H2, suggesting that insti- 11 We focus on buy-and-hold abnormal returns (BHARs) rather than cumulative abnormal returns (CARs) because using CARs over a long period may yield positively biased test statistics (Barber and Lyon (1997)). 19
tutional investors possess a residual information advantage over retail investors about the firm s intrinsic value, even after the announcement of an open-market repurchase program. 5.4 Profitability of Institutional Trading around Open-Market Repurchases In the previous subsection, we presented evidence that institutional trading after openmarket repurchase announcements has predictive power regarding the subsequent long-run stock return performance of the firm. This result suggests that institutions possess private information about the intrinsic value of firms undergoing open-market repurchases, even after the announcement of such programs which may convey firm insiders private information as well. In this subsection, we investigate hypothesis H3 by analyzing whether institutions make use of this private information and earn abnormal trading profits after open-market repurchase announcements and, if yes, whether institutional investors make larger abnormal profits when the information conveyed by an open-market repurchase program is more noisy (i.e., when the size of the open-market repurchase program is smaller or when the firm makes less actual repurchase). We make use of our transaction-level institutional trading data to calculate trading profits, capital committed, and investment returns earned by sample institutions. Following Chemmanur, He, and Hu (2009), we consider a raw measure as well as a risk-adjusted measure, where we use the cumulative returns from the Fama-French 25 portfolios matched on size and book-to-market to discount profits and capital committed back to the first day of the trading horizons. We focus on institutional trading over the 4 fiscal quarters immediately after OMR announcements. Table 6 presents the results of our analysis. We split our data by the level of asymmetric information about OMR announcing firms. In Panel A, we split by the cumulative actual repurchases made by the firm during the first two fiscal quarters after an OMR announcement. If a firm repurchases more, we expect that more information about firm value is 20
impounded into the stock prices. Therefore, we expect that institutional investors have less informational advantage compared to retail investors (and hence make less abnormal trading profits) when the firm actually repurchases more. In Panel B, we split by the OMR program size. We expect that a larger OMR program size sends a stronger signal to the market, which induces more information production by market participants (both institutional and retail investors). Therefore, we expect that institutional investors have less informational advantage compared to retail investors (and hence make less abnormal trading profits) when the OMR program size is larger. As we can see in Table 6, institutional investors make positive abnormal trading profits, net of brokerage commissions paid, when the firm actually repurchases less (Panel A) and when the OMR program size is smaller (Panel B). Under these circumstances, we expect that institutional investors have a larger informational advantage over retail investors. On the other hand, when the firm actually repurchases more (Panel A) and when the OMR program size is larger (Panel B), institutional investors make, on average, negative dollar amount trading profits. We next examine the investment amount committed to trading by our sample institutions after OMR announcements. We consider two measures here. The first one is buy principal, which is the total dollar amount of all buying transactions plus commissions paid. The second measure is maximum investment, which is the maximum dollar amount committed to trading sample firms shares during the trading horizon. This measure takes into account the fact that institutions can use selling proceeds for later buying transactions. We again adjust these two measures by discounting them with returns of the Fama-French 25 portfolios matched on size and book-to-market. As we can see in Table 7 Panel A, institutional investors in our sample commit less trading capital in the subsample where the firm actually repurchases less and commit more trading capital when the firm actually repurchases more. The di erence between these two is also statistically significant. In Panel B where we split the sample by OMR program size, we can see that institutional investors commit essentially 21
the same amount of trading capital in the two subsamples. Overall, there is no evidence that institutions commit more capital in trading in the subsample where there is a larger degree of information asymmetry (i.e., when the firm actually repurchases less or when the OMR program size is smaller), suggesting the larger trading profits institutions earn under those circumstances are not due to more trading capital committed. Lastly, we examine the realized investment return by institutions after OMR announcements. We calculate two return measures, i.e., return on buy principal and return on maximum investment. For each return measure, we calculate a raw return measure without risk adjustment and a risk-adjusted return measure by discounting using benchmark returns of the Fama-French 25 portfolios matched on size and book-to-market. For example, raw return on buy principal is defined as raw profit divided by buy principal and risk-adjusted return on maximum investment is defined as risk-adjusted profit divided by risk-adjusted maximum investment. As we can see in Table 7 Panel A, when the firm actually repurchases less, institutions earn positive and statistically significant investment returns. For example, institutions on average realize a risk-adjusted return on maximum investment of 1.03%. On the other hand, when the firm actually repurchases more, institutions make negative and sometimes statistically significant investment returns. The di erence between the two is also statistically significant. In Panel B where we split our sample by the OMR program size, we find similar results when there is a larger degree of information asymmetry: institutions make positive and statistically significant investment returns when the OMR program size is smaller. In summary, we find that institutional investors realize abnormal investment returns in trading the stock of OMR announcing firms when the size of the open-market repurchase program is smaller and when the firm makes less actual repurchase subsequent to the openmarket repurchase announcement (H3), but not otherwise. This suggests that institutional investors enjoy significant information advantage over retail investors and are able to translate this information advantage into realized trading profits, when the information conveyed 22
by an open-market repurchase announcement itself is more noisy. 5.5 Institutional Trading and Subsequent Actual Repurchases by the Firm In this subsection, we investigate the relationship between institutional trading after OMR announcements and the firm s actual repurchase activities. Specifically, we test hypothesis H4 by examining whether institutional trading after an open-market repurchase announcement is positively associated with the actual repurchases made by the firm subsequent to the announcement. Table 7 presents the results of our multivariate analysis. We aggregate institutional trading over the first fiscal quarter after an OMR announcement. The dependent variable is the firm s actual repurchases during the subsequent two fiscal quarters, which is defined as the number of shares repurchased by the firm normalized by the number of shares outstanding. As we can see from Model (1), the coe cient on Net Buy is positive and statistically significant, indicating that a higher Net Buy after an OMR announcement is followed by more actual repurchases by the firm. This is consistent with hypothesis H4, suggesting that institutional investors possess private information about firm value, based on which the management of the firm decides on the actual repurchase activity in the subsequent period. We control for variables capturing di erent firm characteristics. In particular, we control for several variables that a ect the firm s financial ability to repurchase. Specifically, we control for the firm s cash holdings, capital expenditures, R&D expenses, and whether the firm is paying dividends. All of the coe cients on these variables have the expected signs and, most of the time, are statistically significant. In Models (2) - (10), we additionally control for di erent combinations of contemporaneous insider trading, insider holdings, industry fixed e ects, and year fixed e ects. The coe cients on Net Buy remain positive and statistically significant. In fact, the magnitude and statistical significance of the coe cients on Net Buy remain similar across all the mod- 23
els we consider here, suggesting that institutional trading possesses additional predictive power that is not captured in these control variables regarding the firm s subsequent actual repurchases. In summary, in this subsection we find evidence that institutional trading after an openmarket repurchase announcement has predictive power for the subsequent actual repurchases made by the firm, in the sense that more Net Buy by institutional investors after an OMR announcement is associated with more actual repurchases by the firm in the subsequent period. This is consistent with hypothesis H4, suggestingthatinstitutionalinvestorspossessaresidual information advantage over retail investors about the intrinsic value of firms undergoing open-market repurchase programs, even after the announcement of such programs. 5.6 Institutional Trading and the Change in the Extent of Asymmetric Information around OMRs The notion that a change in payout policy has informational content about firm value dates back to Miller and Modigliani (1961). Ofer and Thakor (1987) present a model in which firms can signal their value through repurchasing shares as well as paying dividends. Brennan and Thakor (1990) and Oded (2005) argue that there is likely more informed trading of the firm s equity subsequent to an OMR announcement. These and other theoretical models imply that information asymmetry faced by the firm should decrease after an open-market repurchase program. In this subsection, we empirically examine the relationship between institutional trading after OMR announcements and the change in the firm s information asymmetry from before the OMR announcement to after. We hypothesize that a greater Net Buy by institutional investors is associated with a larger decrease in the degree of information asymmetry facing the firm in the equity market (H5), with the notion that more Net Buy impounds more favorable information into the stock prices, complementary to the positive signal conveyed by an OMR announcement. Table 8 presents the results of our multinomial analysis. We make use of the analyst 24
earnings forecast data from I/B/E/S and construct our measures of information asymmetry. Specifically, for each OMR announcement, we retrieve analyst earnings forecasts for the previous and the next fiscal year-end. If the next annual earnings announcement is within 6monthsoftheOMRannouncementdate,wejumpaheadtothefollowingfiscalyear. If the previous annual earnings announcement is within 6 months of the OMR announcement date, we jump backward to the preceding fiscal year. We consider two measures of change in information asymmetry and treat them as our dependent variables in separate panels. The first one is change in the number of analysts following the firm (Panel A). The second one is the change in the analyst forecasts dispersion, defined as the standard deviation of analyst forecasts normalized by the stock price. The variable-of-interest is Net Buy, which is the aggregate institutional trading over the two fiscal quarters immediately after an OMR announcement. As we can see from Model (1) from Table 8 Panel A, the coe cient on Net Buy is positive and statistically significant, suggesting that a larger Net Buy is associated with more analyst coverage of the firm. This result is robust to the inclusion of several variables capturing di erent firm characteristics. In Models (2) - (6), we additionally control for di erent combinations of contemporaneous insider trading, insider holdings, industry fixed e ects, and year fixed e ects. The coe cients on Net Buy remain positive and statistically significant. In Table 8 Panel B, we measure change in information asymmetry from the quality perspective. Specifically, we use change in analyst forecasts dispersion as our dependent variable. A decrease in this variable is considered as a decrease in information asymmetry. As we can see from Model (1), the coe cient on Net Buy is negative and statistically significant, which indicates that a larger Net Buy is associated with lower analyst forecasts dispersion (and hence information asymmetry). This result is also robust to the inclusion of several variables capturing di erent firm characteristics. In Models (2) - (6), we again additionally control for di erent combinations of contemporaneous insider trading, insider 25
holdings, industry fixed e ects, and year fixed e ects. The coe cients on Net Buy remain negative and statistically significant across all models considered. Both the results in Panel A and Panel B are consistent with hypothesis H5, suggesting that a larger Net Buy is associated with more analyst coverage and lower analyst forecasts dispersion. To summarize this subsection, we find evidence that is consistent with hypothesis H5. Institutional trading after OMR announcements has predictive power regarding the change in the firm s information asymmetry from before the OMR announcement to after, in the sense that a larger amount of net buying by institutions is associated with a larger decrease in the degree of information asymmetry facing the firm in the equity market. This suggests that institutional investors play an additional role in reducing the information asymmetry facing the firm, over and above that of the open-market repurchase program announcement itself. 6 Conclusion Using a large sample of transaction-level institutional trading data and the data on US firms actual repurchases made available by the 2003 revisions to the Exchange Act, we have empirically analyzed, for the first time in the literature, the role of institutional investors as producers of information about firms undergoing open-market repurchases. We made use of institutional net buy (number of shares bought minus shares sold, normalized by shares outstanding) as our measure of institutional trading. We documented several results new to the literature. First, institutional trading prior to an OMR announcement has significant predictive power for the magnitude of the abnormal return to the firm s equity around OMR announcements. Second, institutional trading immediately after an OMR announcement has significant predictive power for the firm s subsequent stock return performance: the larger the net buying by institutions, better the subsequent long-run stock returns. Third, institutions are able to realize significant abnormal profits (net of commissions and trading 26
costs) by trading in the equity of firms undergoing open-market repurchases when the firm actually repurchases more and when the OMR program size is smaller (i.e., when there is less new information impounded into the stock prices). Fourth, institutional trading immediately after an OMR announcement has significant predictive power for the actual repurchases by the firm: a larger amount of buying of the firm s equity by institutions is associated with a larger actual repurchase by the firm in the subsequent period. Finally, institutional trading has predictive power for the change in the firm s information asymmetry from before the OMR announcement to after: a larger amount of net buying by institutions is associated with a larger decrease in the degree of information asymmetry facing the firm in the equity market. Overall, our results are consistent with the notion that institutional investors are able to generate a significant information advantage for themselves about firms undergoing open-market repurchases. 27
References Admati, Anat R, and Paul Pfleiderer, 1986, A monopolistic market for information, Journal of Economic Theory 39 (2), 400 438. Allen, Franklin, and Roni Michaely, 2003, Handbook of the Economics of Finance (Elsevier North-Holland). Babenko, Ilona, Yuri Tserlukevich, and Alexander Vedrashko, 2012, The credibility of open market share repurchase signaling, Journal of Financial and Quantitative Analysis 47 (5), 1059 1088. Banyi, Monica L., Edward A. Dyl, and Kathleen M. Kahle, 2008, Errors in estimating share repurchases, Journal of Corporate Finance 14 (4), 460 474. Barber, Brad M., and John D. Lyon, 1997, Detecting long-run abnormal stock returns: The empirical power and specification of test statistics, Journal of Financial Economics 43 (3), 341 372. Bhushan, Ravi, 1989a, Collection of information about publicly traded firms: Theory and evidence, Journal of Accounting and Economics 11 (2-3), 183 206., 1989b, Firm characteristics and analyst following, Journal of Accounting and Economics 11 (2-3), 255 274. Brennan, Michael J., and Anjan V. Thakor, 1990, Shareholder preferences and dividend policy, Journal of Finance 45 (4), 993 1018. Chan, Konan, David L. Ikenberry, and Inmoo Lee, 2007, Do managers time the market? Evidence from open-market share repurchases, Journal of Banking & Finance 31 (9), 2673 2694. Chemmanur, Thomas J., Shan He, and Gang Hu, 2009, The role of institutional investors in seasoned equity o erings, Journal of Financial Economics 94 (3), 384 411. 28
Chemmanur, Thomas J., and Yawen Jiao, 2011, Institutional trading, information production, and the SEO discount: A model of seasoned equity o erings, Journal of Economics and Management Strategy 20 (1), 299 338. Chetty, Raj, and Emmanuel Saez, 2006, The e ects of the 2003 dividend tax cut on corporate behavior: Interpreting the evidence, American Economic Review, Papers and Proceedings 96 (2), 124 129. Comment, Robert, and Gregg A. Jarrell, 1991, The relative signalling power of dutch-auction and fixed-price self-tender o ers and open-market share repurchases, Journal of Finance 46 (4), 1243 1271. Conrad, Jennifer S., Kevin M. Johnson, and Sunil Wahal, 2001, Institutional trading and soft dollars, Journal of Finance 56 (1), 397 416. Constantinides, George M., and Bruce D. Grundy, 1989, Optimal investment with stock repurchase and financing as signals, Review of Financial Studies 2(4),445 465. Dann, Larry Y., 1981, Common stock repurchases: An analysis of returns to bondholders and stockholders, Journal of Financial Economics 9(2),113 138. DeLisle, Jared, Justin Morscheck, and John R. Nofsinger, 2014, Share repurchases and wealth transfer among shareholders, Working Paper. Diamond, Douglas W., and Robert E. Verrecchia, 1981, Information aggregation in a noisy rational expectations economy, Journal of Financial Economics 9(3),221 235. Fama, Eugene F., and Kenneth R. French, 2001, Disappearing dividends: changing firm characteristics or lower propensity to pay?, Journal of Financial Economics 60 (1), 3 43. Frankel, Richard, S.P. Kothari, and Joseph Weber, 2006, Determinants of the informativeness of analyst research, Journal of Accounting and Economics 41 (1-2), 29 54. 29
Gibson, Scott, Assem Safieddine, and Ramana Sonti, 2004, Smart investments by smart money: Evidence from seasoned equity o erings, Journal of Financial Economics 72 (3), 581 604. Grossman, Sanford, 1976, On the e ciency of competitive stock markets where trades have diverse information, Journal of Finance 31 (2), 573 585. Grossman, Sanford J., and Joseph E. Stiglitz, 1980, On the impossibility of informationally e cient markets, American Economic Review 70 (3), 393 408. Grullon, Gustavo, and Roni Michaely, 2002, Dividends, share repurchases, and the substitution hypothesis, Journal of Finance 57 (4), 1649 1684.,2004,Theinformationcontentofsharerepurchaseprograms,Journal of Finance 59 (2), 651 680. Huang, Sheng, and Joe Zhang, 2013, How do institutional investors trade when firms are buying back shares?, Working Paper. Ikenberry, David, Josef Lakonishok, and Theo Vermaelen, 1995, Market underreaction to open market share repurchases, Journal of Financial Economics 39 (2-3), 181 208. Jones, Charles M, and Marc L Lipson, 2001, Sixteenths: direct evidence on institutional execution costs, Journal of Financial Economics 59 (2), 253 278. Keim, Donald B., and Ananth Madhavan, 1995, Anatomy of the trading process empirical evidence on the behavior of institutional traders, Journal of Financial Economics 37 (3), 371 398. Kyle, Albert S., 1985, Continuous auctions and insider trading, Econometrica 53 (6), 1315 1335. Lang, Mark H., and Russell J. Lundholm, 1996, Corporate disclosure policy and analyst behavior, The Accounting Review 71 (4), 467 492. 30
Miller, Merton H., and Franco Modigliani, 1961, Dividend policy, growth, and the valuation of shares, Journal of Business 34 (4), 411 433. O Brien, Patricia C., and Ravi Bhushan, 1990, Analyst following and institutional ownership, Journal of Accounting Research 28, 55 76. Oded, Jacob, 2005, Why do firms announce open-market repurchase programs?, Review of Financial Studies 18 (1), 271 300. Ofer, Aharon R., and Anjan V. Thakor, 1987, A theory of stock price responses to alternative corporate cash disbursement methods: Stock repurchases and dividends, Journal of Finance 42 (2), 365 394. Peyer, Urs, and Theo Vermaelen, 2009, The nature and persistence of buyback anomalies, Review of Financial Studies 22 (4), 1693 1745. Stephens, Cli ord P., and Michael S. Weisbach, 1998, Actual share reacquisitions in openmarket repurchase programs, Journal of Finance 53 (1), 313 333. Vermaelen, Theo, 1981, Common stock repurchases and market signalling: An empirical study, Journal of Financial Economics 9(2),139 183. Verrecchia, Robert E., 1982, Information acquisition in a noisy rational expectations economy, Econometrica 50 (6), 1415 1430. 31
Table 1: Summary Statistics of Open-Market Repurchases Data (2004-2010) This table presents summary statistics of the open-market repurchase programs (OMRs) data from January 2004 to December 2010. OMR Program Size is the value of the OMR program (in dollar amount), normalized by the market capitalization of the firm as of the most recent month-end before the announcement date. Announcement Effect is measured as the five-day ([-2,2]) abnormal return, where date 0 is the announcement date. Abnormal returns are calculated based on a market model, where the market beta is estimated using returns over 250 trading days ending 31 trading days before the announcement (that is, [-280, -31]). Actual Completion is the actual repurchase by the firm during the four fiscal quarters after the announcement date as a percentage of OMR Program Size. Log Total Assets is the natural logarithm of total assets at the most recent fiscal year end before the announcement; similarly, Log B/M is the natural logarithm of Book-to-Market ratio; Cash Holdings is the firm s cash holdings normalized by total assets; R&D Expenses is the firm s R&D expenses normalized by total assets; Dividend Paying Dummy is a dummy variable which equals one if the firm pays out dividends and zero otherwise. Prior Quarter Stock Return is the market adjusted stock return during the one-quarter period before the announcement. Prior Year Stock Return is the market adjusted stock return during the one-year period before the announcement. Variable N Mean Median Std. Err. OMR Program Size 2988 7.94% 6.19% 0.12% Announcement Effect [-2, 2] 2988 1.74% 1.39% 0.18% Actual Completion [FQ1, FQ4] 2988 86.29% 48.80% 11.69% Log Total Assets 2988 7.15 7.05 0.04 Log B/M 2988-0.87-0.81 0.01 Cash Holdings 2988 0.18 0.10 0.00 R&D Expenses 2988 0.03 0.00 0.00 Dividend Paying Dummy 2988 0.52 1.00 0.01 Prior Quarter Market-adj. Return 2988-6.18% -5.49% 0.31% Prior Year Market-adj. Return 2988-2.58% -7.58% 0.73%
Table 2: Summary Statistics of Institutional Trading Data (January 2003 September 2011) This table presents summary statistics of the institutional trading sample from January 2003 to September 2011. Annualized number of transactions, annualized trading volume, annualized principal traded, and annualized commission expense are computed based on all U.S. domestic equity traded by sample institutions from January 2003 to September 2011. Sample mean, median, and total are presented. Trading around OMRs is the aggregated trading by sample institutions during the two-year period surrounding OMR announcement dates in our OMR data. T-tests on sample means and Mann-Whitney tests on sample medians are performed and p-values are reported in the last column. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. All Institutions Investment Managers Plan Sponsors Test of Diff. (p-value) Number of Institutions 868 372 496 Annualized Number of Transactions (thousands) Mean 75.42 134.64 31.01 <0.0001*** Median 7.59 26.73 4.77 <0.0001*** Total 65,464.75 50,085.37 15,379.38 Annualized Trading Volume (millions) Mean 350.69 600.57 163.29 <0.0001*** Median 37.75 133.74 12.96 <0.0001*** Total 304,402.64 223,412.34 80,990.29 Annualized Principal Traded ($ millions) Mean 10,318.32 17,910.06 4,624.51 <0.0001*** Median 1,060.36 3,614.01 374.07 <0.0001*** Total 8,956,297.68 6,662,542.45 2,293,755.23 Annualized Commission Expense ($ millions) Mean 8.65 15.55 3.47 <0.0001*** Median 0.87 3.67 0.40 <0.0001*** Total 7,506.46 5,786.06 1,720.40 Trading around OMRs ([-4Q, 4Q]) Number of Institutions trading around OMRs 849 360 489 Number of Transactions per OMR (thousands) 24.51 21.49 3.02 <0.0001*** Trading Volume per OMR (millions) 100.02 86.01 14.61 <0.0001*** Principal Traded per OMR ($ millions) 3,501.76 3,000.47 522.43 <0.0001*** Commission Expense per OMR ($ millions) 2.58 2.26 0.33 <0.0001***
Table 3: Trading Pattern before OMR Announcements (2004-2010) This table reports the pattern of institutional trading before announcements of open-market repurchase programs (OMRs) between January 2004 and December 2010. Net Buy is expressed in basis points of the number of shares outstanding of a company. Day 0 is the announcement date, or the closest future trading day if the announcement date is not a trading day. T-tests are performed on the significance of Net Buy. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Mean Net Buy (in bps) All Institutions Investment Managers Plan Sponsors #. Obs. 3362 3362 3362 [-1,1] -3.66*** -2.75** -1.14*** [-2,2] -4.47*** -3.21** -1.51*** Day 0-1.38* -1.07-0.46* Lag Day 1-1.22*** -0.99** -0.34*** Lag Day 2-0.50-0.23-0.38*** Lag Days [-2,-1] -1.64** -1.16* -0.62*** Lag Week 1-3.45*** -2.47** -1.18*** Lag Week 2-4.00*** -3.02*** -1.19*** Lag Week 3-5.44** -5.04** -0.58** Lag Week 4-2.11* -1.27-0.98*** Lag Month 1-14.98*** -11.74*** -3.74*** Lag Month 2-3.25-2.44-0.92 Lag Month 3 0.37 0.67-0.32 Lag Month 4 3.70 3.60 0.16 Lag Month 5 2.63 1.27 1.52* Lag Month 6 8.58*** 7.32*** 1.51** Lag Month 7 9.82*** 9.31*** 0.69 Lag Month 8 7.95*** 7.28*** 0.86 Lag Month 9 13.70*** 12.32*** 1.67** Lag Month 10 10.71*** 10.56*** 0.34 Lag Month 11 5.41* 3.77 1.87*** Lag Month 12 5.53* 4.43 1.28** Lag Quarter 1-17.21*** -13.02** -4.70*** Lag Quarter 2 14.42** 11.76** 3.03** Lag Quarter 3 30.34*** 27.73*** 3.04*** Lag Quarter 4 20.85*** 18.02*** 3.28*** Lag Year 1 46.34*** 42.63*** 4.32*
Table 4: OMR Announcement Effects and Institutional Trading before OMR Announcements This table presents OLS regression analysis of the relation between Net Buy from institutional investors and the announcement effect of open-market repurchase programs (OMRs) between 2004 and 2010. Announcement Effect, the dependent variable, is measured as the five-day ([-2,2]) abnormal return, where date 0 is the announcement date. Abnormal returns are calculated based on a market model, where the market beta is estimated using returns over 250 trading days ending 31 trading days before the announcement (that is, [-280, -31]). Net Buy i, expressed in percentage points of the number of shares outstanding of firm i, is the aggregate Net Buy from institutional investors during the 6-month and one-year periods before the announcement date; OMR Program Size is the value of the OMR program (in dollar amount), normalized by the market capitalization of the firm as of the most recent monthend before the announcement date; Log Total Assets is the natural logarithm of total assets of firm i at the most recent fiscal year end before the announcement; Log B/M is the natural logarithm of Book-to-Market ratio of firm i at the most recent fiscal year end before the announcement; Industry Adj. ROA is firm i's EBIT/Total Asset minus 2- digit SIC industry's median EBIT/Total Asset, at the most recent fiscal year end before the announcement; Cash Holdings is firm i's cash holdings normalized by total assets, at the most recent fiscal year end before the announcement; Prior Stock Return is the market adjusted stock return during the one-year period before the announcement; Leverage is firm i's total liabilities normalized by total assets, at the most recent fiscal year end before the announcement; Dividend Paying Dummy is a dummy variable which equals one if firm i pays out dividends during the most recent fiscal year before the announcement, and equals zero otherwise; Insider Net Buy is the aggregate net buy from top level insiders of firm i during the 6-month (or one-year period) before the announcement; Insider Holding is the aggregate stock holding of top level insiders of firm i. Year fixed effects are included. T-statistics are in parentheses. ***, **, and * represent statistical significance at the 1, 5, and 10 percent levels. Net Buy [-126, -1] -0.0008** -0.0008** -0.0007** (-2.25) (-2.27) (-2.20) Dependent Variable: Announcement Returns over [-2,2] (1) (2) (3) (4) (5) (6) Net Buy [-252, -1] -0.0005* -0.0004** -0.0005** (-1.84) (-2.06) (-2.35) OMR Program Size 0.0012*** 0.0010*** 0.0010*** 0.0012*** 0.0011*** 0.0011*** (4.37) (4.14) (4.13) (4.45) (4.39) (4.31) Log Total Assets -0.0028** -0.0021** -0.0020** -0.0030*** -0.0024** -0.0025** (-2.58) (-2.14) (-2.00) (-2.78) (-2.54) (-2.51) Log B/M -0.0015 0.0034 0.0030-0.0014 0.0033 0.0022 (-0.50) (1.25) (1.12) (-0.45) (1.22) (0.80) Industry-Adj. ROA -0.1461*** -0.0352* -0.0384** -0.1472*** -0.0356* -0.0347* (-7.11) (-1.87) (-2.02) (-7.17) (-1.89) (-1.83) Cash Holdings -0.0115-0.0109-0.0116-0.0118-0.0103-0.0113 (-0.97) (-1.02) (-1.07) (-1.00) (-0.96) (-1.04) Past Return -0.0143*** -0.0138*** -0.0135*** -0.0153*** -0.0178*** -0.0160*** (-3.12) (-3.36) (-3.28) (-3.35) (-4.36) (-3.90) Leverage -0.0048 0.0046 0.0031-0.0044 0.0067 0.0052 (-0.51) (0.53) (0.35) (-0.46) (0.78) (0.60)
Div. Paying Dummy 0.0053 0.0020 0.0022 0.0050 0.0014 0.0017 (1.30) (0.54) (0.60) (1.23) (0.39) (0.46) Insider Net Buy [-126, -1] 0.0081*** 0.0082*** (2.70) (2.75) Insider Holding as of Date -126 0.0000 (0.18) Insider Net Buy [-252, -1] 0.0036** 0.0045*** (2.22) (2.78) Insider Holding as of Date -252-0.0000 (-0.14) Constant 0.0315*** 0.0207** 0.0212** 0.0337*** 0.0233** 0.0244** (3.06) (2.24) (2.20) (3.28) (2.51) (2.53) Year Fixed Effects Yes Yes Yes Yes Yes Yes Observations 3017 2943 2924 3035 2972 2936 R-squared 0.052 0.038 0.038 0.054 0.042 0.040
Table 5: Subsequent Long-run Stock Return Performance and Institutional Trading after OMR Announcements This table presents OLS regression analysis of the effects that Net Buy from institutional investors after an OMR announcement has on the subsequent one-year stock return performance of a firm. Net Buy i is the aggregate Net Buy from institutional investors during the two fiscal quarters after the OMR announcement. Stock Return, the dependent variable, is measured over the one-year period subsequent to the trading horizon of Net Buy i. The buyand-hold abnormal return is adjusted based on the returns of an equally-weighted portfolio of matching firms. The matching firms do not make repurchase announcements and are matched based on industry, market capitalization, and book-to-market value. OMR Program Size is the value of the OMR program (in dollar amount), normalized by the market capitalization of the firm as of the most recent month-end before the announcement date. Log Total Assets is the natural logarithm of total asset of firm i at the most recent fiscal year end before the one-year stock performance measurement horizon; similarly, Log B/M is the natural logarithm of firm i's Book-to-Market ratio; Industry Adjusted ROA is the industry-adjusted EBIT/Total Asset ratio, where industries are defined by the first 2- digit of SIC codes; Cash Holdings is firm i's cash holdings normalized by total assets; Cash Flow is firm i's cash flow normalized by total assets. Prior Stock Performance is the market adjusted stock return during the one-year period before the stock performance horizon. S&P 500 is an indicator variable that equals one for firms in the S&P 500 Index and zero otherwise. Actual Repurchase is the amount of shares actually repurchased by the firm during the two fiscal quarters after the OMR announcement, normalized by the number of shares outstanding. Insider Net Buy is the aggregate Net Buy from top level insiders of firm i during the two fiscal quarters after the OMR announcement. Insider Holding is the aggregate stock holding of top level insiders of firm i. Industry and Year fixed effects are included. T-statistics are in parentheses. ***, **, and * represent statistical significance at the 1, 5, and 10 percent levels. Dependent Variable: Buy-and-Hold Abnormal Returns over the Subsequent 12 Months (1) (2) (3) (4) (5) Net Buy [FQ1,FQ2] 0.0034** 0.0039** 0.0032* 0.0033* 0.0038** (1.97) (2.23) (1.87) (1.94) (2.18) OMR Program Size -0.0007-0.0010-0.0012-0.0006-0.0014 (-0.41) (-0.55) (-0.73) (-0.36) (-0.77) Past Stock Return -0.0438-0.0571-0.0424-0.0585-0.0637 (-1.15) (-1.47) (-1.12) (-1.52) (-1.62) Log Total Assets 0.0344*** 0.0366*** 0.0303*** 0.0322*** 0.0314*** (4.12) (4.24) (3.63) (3.57) (3.34) Industry-Adj. ROA 0.2191 0.2590 0.2931-0.3336-0.3043 (0.58) (0.67) (0.77) (-0.71) (-0.65) Cash Holdings 0.1465** 0.1652** 0.1272* 0.0893 0.0960 (2.24) (2.48) (1.94) (1.20) (1.27) Cash Flow -0.0840-0.1490-0.1800 0.4524 0.3875 (-0.23) (-0.41) (-0.50) (0.98) (0.83) S&P 500 Dummy -0.0591* -0.0606* -0.0602* -0.0672* -0.0682* (-1.69) (-1.72) (-1.73) (-1.83) (-1.83) Log B/M -0.0400** -0.0351** -0.0348* -0.0467** -0.0370* (-2.30) (-1.98) (-1.93) (-2.57) (-1.93)
Actual Repurchase [FQ1,FQ2] -0.0005-0.0006 0.0004 0.0003 0.0011 (-0.13) (-0.17) (0.12) (0.08) (0.29) Insider Net Buy [FQ1, FQ2] -0.0465** -0.0294 (-2.10) (-1.30) Insider Holding as of FQ1 0.0008 0.0014 (0.50) (0.88) Constant -0.3640*** -0.3780*** -0.2535*** -0.3601-0.6785* (-5.43) (-5.37) (-2.60) (-1.36) (-1.81) Industry Fixed Effects No No No Yes Yes Year Fixed Effects No No Yes No Yes Observations 2363 2308 2363 2363 2308 R-squared 0.017 0.020 0.032 0.051 0.069
Table 6: Profitability of Institutional Trading around Open-Market Repurchases This table reports univariate results of the profitability of institutional trading around open-market repurchase programs (OMRs). We consider the trading horizon starting from the first fiscal quarter after the OMR announcement and ending at the fourth fiscal quarter after the OMR announcement. Panel A reports the result where we split the sample by the cumulative actual repurchase by the firm during the first two fiscal quarters after the OMR announcement. Panel B reports the result where split the sample by the OMR program size, defined as the dollar amount value of the OMR program, normalized by the market capitalization of the firm as of the most recent month-end before the announcement date. Raw Profit is the total raw profit earned by institutions using actual transaction prices net of commissions, with the net position marked to market at the end of the trading horizon. Buy Principal is the sum of the actual dollar amount of all the buy transactions including commissions spent by sample institutions during the trading horizon. Maximum Investment is the maximum dollar amount committed to trading the sample firms shares during the trading horizon by the institutions. Raw Return on Buy Principal is defined as the ratio of Raw Profit to Buy Principal. Raw Return on Maximum Investment is defined as the ratio of Raw Profit to Maximum Investment. We also discount profit and investment amount back to the first day of the trading horizon using the buy-and-hold value-weighted return from the Fama and French 25 portfolios matched on size and book-to-market. For example, Risk-adjusted Profit is computed by discounting the raw profit back to the first day of the trading horizon using the benchmark return from the matched Fama- French 25 portfolios; and Risk-adjusted Return on Buy Principal equals Risk-adjusted Profit divided by Risk-adjusted Buy Principal. T-tests are performed on the profits and the returns, respectively. T-tests are also performed on the difference in the profits, the difference in the investment amount, and the difference in the returns. Statistical significance is indicated by *** for 1% level, ** for 5% level, and * for 10% level. Institutional Trading over [FQ1,FQ4] Panel A: Split by Actual Repurchase Panel B: Split by OMR Program Size Below Median Above Median Diff (B-A) Below Median Above Median Diff (B-A) Number of Observations (Events) 1,392 1,392-1,392 1,392 - Raw Profit ($ thousands) 3,590.33-2,187.52 5777.86 4,488.44-3,085.63 7,574.08 Risk-adjusted Raw Profit ($ thousands) 4,901.88-3,261.33 8163.21* 4,479.27-2,838.73 7,317.99 Buy Principal ($ millions) 674.67 1,028.45-353.78*** 837.72 865.93-28.21 Risk-adjusted Buy Principal ($ millions) 691.90 1,028.56-336.65*** 840.05 880.86-40.80 Maximum Investment ($ millions) 468.23 716.12-247.89*** 582.48 601.87-19.39 Risk-adjusted Maximum Investment ($ millions) 474.75 713.21-238.46*** 580.83 607.12-26.29 Raw Return on Buy Principal (%) 0.54* -0.21 0.75** 0.54** -0.36 0.90** Risk-adjusted Return on Buy Principal (%) 0.72*** -0.32* 1.03*** 0.54** -0.32* 0.86*** Raw Return on Maximum Investment (%) 0.77** -0.31 1.07*** 0.77*** -0.51* 1.28*** Risk-adjusted Return on Maximum Investment (%) 1.03*** -0.46** 1.49*** 0.77*** -0.47* 1.24***
Table 7: Institutional Trading after OMR Announcements and Subsequent Actual Repurchase by the Firm This table presents OLS regression analysis of the effects that Net Buy from institutional investors after an OMR announcement has on the subsequent actual repurchase by the firm. Net Buy i is the aggregate Net Buy from institutional investors during the first fiscal quarter after the OMR announcement. Subsequent Actual Repurchase, the dependent variable, is the actual repurchase by the firm (normalized by the number of shares outstanding) during the subsequent two fiscal quarters. OMR Program Size is the value of the OMR program (in dollar amount), normalized by the market capitalization of the firm as of the most recent month-end before the announcement date. Log Total Assets is the natural logarithm of total asset of firm i at the most recent fiscal year end before the subsequent actual repurchase measurement period; similarly, Log B/M is the natural logarithm of firm i's Book-to-Market ratio; Cash Holdings is firm i's cash holdings normalized by total assets; Capital Expenditure is firm i's capital expenditure normalized by total assets; R&D is firm i's R&D expenditure normalized by total assets; Dividend Paying Dummy is a dummy variable which equals one if firm i pays out dividends and zero otherwise. Prior Stock Performance is the market adjusted stock return during the oneyear period before the subsequent actual repurchase measurement period. Actual Repurchase FQ1 is the actual repurchase by the firm during the first fiscal quarter after the OMR announcement. Insider Net Buy is the aggregate Net Buy from top level insiders of firm i during the first fiscal quarter after the OMR announcement. Insider Holding is the aggregate stock holding of top level insiders of firm i. Industry and Year fixed effects are included. T-statistics are in parentheses. ***, **, and * represent statistical significance at the 1, 5, and 10 percent levels. Dependent Variable: Subsequent Actual Repurchase (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Net Buy FQ 1 0.0291** 0.0230* 0.0336** 0.0269** 0.0271* 0.0290** 0.0314** 0.0331** 0.0249* 0.0267* (2.14) (1.70) (2.48) (1.99) (1.96) (2.03) (2.28) (2.31) (1.82) (1.87) Log Total Assets 0.1845*** 0.1595*** 0.1879*** 0.1649*** 0.1818*** 0.1870*** 0.1805*** 0.1888*** 0.1576*** 0.1672*** (5.53) (4.79) (5.27) (4.63) (5.26) (4.79) (4.91) (4.54) (4.29) (4.02) Log B/M -0.1050-0.0305-0.1098-0.0254-0.0866-0.0941-0.0904-0.1136-0.0151-0.0412 (-1.30) (-0.37) (-1.28) (-0.29) (-1.04) (-1.08) (-1.02) (-1.21) (-0.17) (-0.43) Past Return -0.5084*** -0.3968** -0.5117*** -0.4178** -0.6066*** -0.6256*** -0.5934*** -0.6117*** -0.5456*** -0.5467*** (-3.25) (-2.30) (-3.25) (-2.41) (-3.46) (-3.39) (-3.36) (-3.25) (-2.77) (-2.61) Cash Holdings 0.9570** 0.9887** 0.8058* 0.8044* 0.7642* 0.5896 0.5517 0.4077 0.5533 0.4096 (2.42) (2.51) (1.93) (1.93) (1.86) (1.36) (1.27) (0.89) (1.28) (0.89) CapEx -0.9711-0.6241-3.3248** -2.9439* -0.7749-1.2220-3.2946** -3.5405** -2.9128* -3.2166* (-0.83) (-0.54) (-2.09) (-1.86) (-0.65) (-0.98) (-2.03) (-2.07) (-1.80) (-1.88) R&D -3.3914*** -2.8293** -2.7749* -2.2479-2.9771** -2.7955* -2.2026-2.4484-1.7081-1.9021 (-2.65) (-2.22) (-1.91) (-1.55) (-2.18) (-1.93) (-1.41) (-1.48) (-1.10) (-1.15)
Div. Paying Dummy -0.4960*** -0.4949*** -0.3502** -0.3428** -0.4852*** -0.5814*** -0.3285** -0.3659** -0.3307** -0.3623** (-3.75) (-3.76) (-2.49) (-2.45) (-3.53) (-3.93) (-2.24) (-2.29) (-2.27) (-2.27) Insider Net Buy FQ 1-0.1808* -0.1847* -0.1802* -0.2012** -0.1369-0.1661* (-1.91) (-1.91) (-1.91) (-2.06) (-1.45) (-1.70) Insider Holding -0.0005 0.0021 0.0029 (-0.05) (0.23) (0.32) OMR Program Size 0.0752*** 0.0757*** 0.0695*** 0.0696*** 0.0764*** 0.0780*** 0.0701*** 0.0718*** 0.0699*** 0.0716*** (8.25) (8.37) (7.49) (7.55) (8.08) (7.73) (7.27) (6.92) (7.29) (6.93) Actual Repur. FQ 1 0.2045*** 0.2024*** 0.1882*** 0.1848*** 0.2496*** 0.2329*** 0.2318*** 0.2180*** 0.2280*** 0.2133*** (7.48) (7.41) (6.86) (6.73) (8.19) (7.32) (7.58) (6.77) (7.44) (6.59) Constant 0.0777 0.3633-0.9234-0.5001 0.0454 0.0804-0.2144 1.4623 0.4348 2.1629 (0.28) (1.09) (-0.60) (-0.33) (0.16) (0.24) (-0.12) (0.48) (0.25) (0.71) Industry Fixed Effects No No Yes Yes No No Yes Yes Yes Yes Year Fixed Effects No Yes No Yes No No No No Yes Yes Observations 2843 2843 2843 2843 2676 2411 2676 2411 2676 2411 R-squared 0.073 0.091 0.111 0.129 0.080 0.079 0.121 0.109 0.137 0.123
Table 8: Degree of Asymmetric Information around OMRs and Institutional Trading This table presents OLS regression analysis of the effects that Net Buy from institutional investors after an OMR announcement has on the change in the firm s information asymmetry from before the OMR announcement to after. Net Buy i is the aggregate Net Buy from institutional investors during the two fiscal quarters after the OMR announcement. Change in Information Asymmetry, the dependent variable, is the difference of the firm s information asymmetry from before the OMR announcement to after. For each firm, we retrieve analyst earnings forecasts for the previous and the next fiscal year-end. If the next annual earnings announcement is within 6 months of the OMR announcement date, we jump ahead to the following fiscal year. If the previous annual earnings announcement is within 6 months of the OMR announcement date, we jump backward to the preceding fiscal year. We employ two measures for analyst forecasts. The first measure is the number of analysts following the firm (Panel A). The second measure is the analyst forecasts dispersion (Panel B), defined as the standard deviation of analyst forecasts normalized by the stock price. We calculate the change in each measure by taking log of the ratio of the post-omr value to the pre-omr value of the measure. Log Total Assets is the natural logarithm of total asset of firm i at the most recent fiscal year end before the OMR announcement. Actual Repurchase Dummy is an indicator variable that equals one if the firm has actual repurchase during the first two fiscal quarters after the OMR announcement and zero otherwise. S&P 500 Dummy is an indicator variable that equals one for firms in the S&P 500 Index and zero otherwise. Announcement Returns is the 5-day abnormal returns of the OMR announcements, based on a market model. Insider Net Buy is the aggregate Net Buy from top level insiders of firm i during the two fiscal quarters after the OMR announcement. Insider Holding is the aggregate stock holding of top level insiders of firm i. Industry and Year fixed effects are included. T-statistics are in parentheses. ***, **, and * represent statistical significance at the 1, 5, and 10 percent levels. (1) (2) (3) (4) (5) (6) Net Buy [FQ1,FQ2] 0.0046** 0.0040** 0.0048** 0.0045** 0.0046** 0.0042** (2.39) (2.10) (2.46) (2.26) (2.27) (2.15) Log Total Assets -0.0101-0.0052-0.0082-0.0096-0.0081-0.0052 (-1.15) (-0.59) (-0.78) (-1.05) (-0.75) (-0.57) Actual Repurchase Dummy -0.0030 0.0058-0.0030-0.0027-0.0007 0.0077 (-0.10) (0.19) (-0.09) (-0.08) (-0.02) (0.25) S&P 500 Dummy -0.0237-0.0444-0.0377-0.0190-0.0315-0.0388 (-0.64) (-1.21) (-0.92) (-0.51) (-0.76) (-1.05) Announcement Returns -0.0579-0.0288-0.0331-0.0873-0.0662-0.0596 (-0.35) (-0.18) (-0.20) (-0.53) (-0.39) (-0.36) Insider Net Buy [FQ1, FQ2] -0.1260*** -0.1220*** -0.0945*** (-3.74) (-3.53) (-2.80) Insider Holding -0.0013-0.0009-0.0009 (-0.65) (-0.42) (-0.44) Constant 0.1463** 0.2188*** 0.0470 0.1269* 0.0574 0.1992*** (2.33) (3.26) (0.14) (1.88) (0.13) (2.76) Industry Fixed Effects No No Yes No Yes No Year Fixed Effects No Yes No No No Yes
Observations 1424 1424 1424 1395 1395 1395 R-squared 0.007 0.046 0.059 0.019 0.069 0.054 (1) (2) (3) (4) (5) (6) Net Buy [FQ1,FQ2] -0.0106** -0.0073* -0.0102** -0.0095** -0.0090* -0.0074* (-2.24) (-1.68) (-2.10) (-1.97) (-1.83) (-1.66) Log Total Assets 0.0476** 0.0469** 0.0041 0.0430* 0.0051 0.0464** (2.07) (2.20) (0.15) (1.82) (0.18) (2.11) Actual Repurchase Dummy -0.2787*** -0.2905*** -0.2541*** -0.2661*** -0.2426*** -0.2808*** (-3.43) (-3.90) (-3.00) (-3.29) (-2.87) (-3.75) S&P 500 Dummy -0.2178** -0.1561* -0.0532-0.1996** -0.0619-0.1453* (-2.38) (-1.86) (-0.52) (-2.20) (-0.61) (-1.73) Announcement Returns -0.0984-0.3534-0.1460 0.0091-0.0437-0.2731 (-0.24) (-0.92) (-0.34) (0.02) (-0.10) (-0.71) Insider Net Buy [FQ1, FQ2] 0.4351*** 0.4414*** 0.2726*** (5.27) (5.22) (3.51) Insider Holding as of FQ1 0.0138** 0.0120** 0.0096* (2.50) (2.09) (1.88) Constant 0.3646** 0.0783 0.7606 0.4100** 0.5299 0.0959 (2.22) (0.48) (0.95) (2.34) (0.49) (0.55) Industry Fixed Effects No No Yes No Yes No Year Fixed Effects No Yes No No No Yes Observations 1291 1291 1291 1264 1264 1264 R-squared 0.017 0.179 0.074 0.043 0.099 0.191