Liu Nguyen

Open market share repurchase: A perspective from up sided stock liquidity Yifan Liu∗and Ha Diep Nguyen† January 15, 2018...

3 downloads 159 Views 309KB Size
Open market share repurchase: A perspective from up sided stock liquidity Yifan Liu∗and Ha Diep Nguyen† January 15, 2018

Abstract We propose a novel method to analyze the role of stock liquidity in share repurchase decisions. We find that higher up sided stock liquidity encourages use of repurchases over dividends, and that net insider buying prior to repurchase announcement increases announcement buy and hold abnormal return, and this value effect is stronger when stock liquidity is low. In addition, net insider buying prior to repurchase announcement predicts a decline in volatility of future operating performance, and this association is stronger when stock liquidity is low. Performance based compensation and independent board member monitoring can reduce information advantages of corporate insiders.

1

Introduction

Payout policy is important not only because of the amount of capital involved and the repeated nature of the decision, but also because payout policy closely interacts with other financing and investment decisions. In recent decades, share repurchase has become increasingly popular among U.S. firms, especially small and young firms. The proportion of firms paying cash dividends fell from 66.5% in 1978 to 20.8% in 1999 (Fama and French (2001)) and in 1999 for the first time, the aggregate amount of share repurchase exceeded that of dividends (Grullon and Michaely (2002)). Share repurchase is operationally more flexible than dividends in the sense that a firm can alter an announced share repurchase program in response to subsequent changes in the price of its shares, unexpected shocks to cash flow and/or investment, etc. Stephens and Weisbach (1998) find that, 3 years after an open market share repurchase announcement, a substantial number of firms have repurchased no shares, about 10% of firms repurchased less than 5% of the shares authorized, and just more ∗ PhD

candidate at Spears School of Buisness, Oklahoma State University. Email: [email protected] † PhD candidate at Kelley School of Buisness, Indiana University. Email: [email protected]

1

than half of firms bought back the total number of shares authorized. This noticable transformation in payout policy and the flexibility of share repurchase have attracted considerable interests among academics, regulators, and partitioners. Previous studies have documented multiple factors that affect stock buyback and have investigated the benefits and costs of the flexibility of share repurchase, in this paper we find that stock liquidity plays an important role in share repurchase decisions and it enables us to further examine the potential moral hazard, insider trading, associated with the flexibility of share repurchase. In contrast to perfect capital market, real-world managers operate in an economy with asymmetric information, agency conflicts, transaction costs, and other frictions. When stock market deepth is not unlimited, liquidity cost becomes an non-trivial transaction cost for repurchasing firms. When a firm enters secondary market to buy back its own stocks, the price impact will widen the bid-ask spread and move the lowest ask price to a higher level, making it more costly to buy back stocks. Brav et al. (2005)’s survey shows that managers take this liquidity transaction cost into consideration when repurchasing stocks. Brockman et al. (2008) use traditional liquidity measure to examine the role of stock liquidity in repurchase decisions and find that stock liquidity matters in payout policy. In this paper, we propose to use the up sided stock liquidity to reexamine this association. Up sided liquidity is the stock liquidity corresponding to transactions with positive returns and down sided liquidity is the stock liquidity corresponding to negative return transactions (Brennan et al. (2013)). Managers are more concerned with the up sided liquidity because when they buy back stocks, the ex ante expectation is that the bid orders from the repurchasing firm’s designated broker account will drive up the market price of its stocks. Consequently, the managers make repurchase decisions on basis of historical up sided stock liquidity which reflects the transation costs when there is upward price pressure. In this paper, we find that higher up sided liquidity is associated with higher amount of actual share repurchase and lower buy and hold abnormal return of repurchase announcement. Also, we find that after the 2008 financial crisis, actual share repurchase become more sensitive to the change of up sided liquidity, implying the systematic shock has changed how managers measure liquidity risk of buying back stocks. In addition, we find that the association between dividends and liquidity depends on stock exchange transaction mechanisms, and this result could partially reconcile the mixed evidences found in previous studies. More specifically, for firms listed on NYSE and AMEX, higher stock liquidity is associated with lower dividends payout, which is consistent with Banerjee et al. (2007) results; for firms listed on NASDAQ, stock liquidity has no effect on dividends payout, which is consistent with Brockman et al. (2008) results. Insider trading prior to corporate event (equity issues, dividends changes, M&A, etc.) contains value-relevant information. An interesting question is whether the information in the insider trading and the information in the corporate event are complementary, i.e. whether the interaction between the insider 2

trading and corporate event contains addtional information beyond the separate information in the insider trading and corporate event. In this paper, we focuses on the interaction between the information in the insider trading prior to repurchase announcements and information in the announcements, and we find that the value effect of repurchase announcements is larger when there is more net insider buying prior to the announcements and when stock liquidity is low. In this case, using up sided liquidity enables us to disentangle the different insider information contained in insider trading and share repurchase announcement. Bonaimé and Ryngaert (2013) find that both insider buying and selling increase significantly during the month of share repurchase announcements, which implies investors’ disagreement about the information content of repurchase announcements. Inside buyers are concerned with up sided liquidity and inside sellers are concerned with down sided liquidity, thus separating up sided liquidity and down sided liquidity is critical to investigating the interaction between stock liquidity and insider trading prior to repurchase announcements. In this paper, we find that lower up sided liquidity and higher net insider buying prior to repurchase announcements in combination enhance the value effect of share repurchase announcement, implying that insider buyers will buy more stocks prior to repurchase announcement if the insider information will generate a permium return high enough to compensate the high risk of being identified as informed trader when liquidity is low (Kyle (1985)). It is important to study the content of insider information contained in the insider trading prior to share repurchase announcements, i.e. what investors know about the future valuation of the firm. DeAngelo et al. (2009) in their review paper summarise empirical evidences and conclude that repurchase program do not contain information about future firm valuation (sum of discounted contigent cash flows). We argue that repurchase announcements do contain information about future firm valuation. More specifically, we propose that repurchase announcements contain information about volatility of future firm operating performance, instead of the level of future firm operating performance. In this sense, insiders are incentivized to trade at the same direction of the announcement because they are informed that the future firm operating performance will be more stable and investment on the firm’s stocks will be less risky and more valuable (Campbell and Shiller (1988a) and Vuolteenaho (2002)). We find that higher net insider buying and lower up sided liquidity in combination predicts lower standard deviation of future return on assets (ROA). We also inspect the role of CEO compensation and board independence in the insider trading-stock liquidity-volatility of future operating performance relationship. We follow Coles et al. (2006) and Coles et al. (2014) methodology to use the firm related wealth hold by CEO as a proxy to measure the alignment of agent and principle interests and use the ratio of independent directors who are appointed before CEO assumed office as a proxy to measure the board indpendence, and find that firms with CEO holding more firm related wealth and more indpendent directors appointed before CEO assumed office have fewer 3

informed insider tradings prior to repurchase announcements. The rest of our paper proceeds as follows. In section 2, we briefly discuss related studies and hypotheses. In section 3, we provide a description of our data and variable construction. In section 4, we present and analyze our empirical findings. In section 5, we discuss robustness check and in section 6, we summarize and conclude the paper.

2

Literature review and hypotheses development

The irrelevance propositions by Miller and Modigliani (1961) suggest that under perfect capital markets the choice between dividends and share repurchase, as a method to distribute residual earnings to shareholders, have no effect on firm value. In the real-world economy, payout policy does affect firm valuation in many ways, including capital structure, information asymmetry, capital budgeting, employee incentives, etc. Empirical evidences show that payout policy plays a critical role in corporate financial decision makings (Barclay and Smith (1988) etc.) and share repurchases have gained popularity since 1980s (Grullon and Michaely (2002)) and more U.S. firms choose not to pay any dividends since 1990s (Fama and French (2001)). The total amount of cash distribution to stockholders have been stable while the composition of payouts has grown from a repurchase-to-dividends ratio of 8% in 1972 to 113% in 2000(Grullon and Michaely (2002)). This prominent trend of transformation has attracted considerable inspection from reseachers and practioners, and many of the proposed causes for this trend are attributed to the fact that share repurchases allow firms to directly enter the secondary market and trade against other less informed investors. On one hand, share repurchases provide the firms with more flexibility because share repurchase announcements are not legal commitment and most repurchasing firms do not complete repurchasing the authorized number of shares in one quarter. Stephens and Weisbach (1998) show that, 3 years after an open market share repurchase announcement, a substantial number of firms have repurchased no shares, about 10% of firms repurchased less than 5% of the shares authorized, and just more than half of firms bought back the total number of shares authorized. With such flexibility, firms are able to time the market to lower the cost of repurchase and thus reducing the ex post cost of capital. Ikenberry and Vermaelen (1996) treat share repurchase announcements as exchange options that enable the firm to exchange the market value of the firm for its true value at management’s discretion. They suggest that the stock-price increase around the announcement comes in part from the creation of the option to buy back shares. Bagwell (1991) find that firms use share repurchases as takeover defense by making it more costly for rival management teams to acquire their shares. In addition, share repurchases can mitigate the dilution effect of option exercises by employees (Fenn and Liang (2001)). 4

On the other hand, the moral hazards associated with share repurchase should not be ignored. Almeida et al. (2016) find that managers use stock buybacks to reduce the number of shares outstanding to increase EPS. Besides corporate performance measure manipulation, insider trading is another important issue associated with share repurchase. Bonaimé and Ryngaert (2013) find that insider transactions increase significantly during the month of share repurchase announcements. Our paper mainly focuses on the insider trading activities prior to share repurchase announcement and the information content of the insider trading. Stock liquidity reflects the impact of order flow on price that results from adverse selection costs and inventory costs (Amihud and Mendelson, (1980); Glosten and Milgrom, (1985)), and high stock liquidity means that the stock price is insensitive to the impact of order flow and thus causes lower transaction costs. Previous literature mainly focuses on how share repurchases impact stock liquidity (Brockman and Chung (2001); Ginglinger and Hamon (2007)), but what role stock liquidity plays in share repurchase decision is as well important (Brockman et al. (2008) and Lipson (2003)). Brav et al. (2005)’s survey shows that managers are aware that price impact matters for investors, and that managers condition repurchase decisions on a sufficient level of liquidity. Our paper studies the relationship between liquidity and uses up sided liquidity to test this relationship. Following Brennan et al. (2013), we decompose stock liquidity into components that correspond to up (positive return) and down (negative return) transactions. Separating the up sided and downsided components of liquidity is critical in testing the association between stock market liquidity and share repurchase, because managers are more concerned about the transaction cost of price impact when there is upward price pressure. In general, for firms with very low up sided liquidity, buying back stocks is more expensive because a small amount of transaction volume can drive up the stock price. In addition, Brennan et al. (2013) find that only down sided liquidity has an effect on stock return, while up sided liquidity is not priced. This result eliminate the possibility that up sided liquidity affects stock return and stock return affects repurchase. Consequently, it is both theoretically and empirically more accurate to use up sided liquidity to test the association between stock liquidity and share repurchase. With this in mind, we propose our first hypothesis: H1: Up sided stock liquidity is positively associated with managerial share repurchase decisions. When there is more net insider buying activities prior to share repurchase announcements, i.e. investors put their money where their mouth is, the buy and hold abnormal return around announcement will be higher (complementary effects). When more insider investors trade in the opposite direction and sell 5

prior to share repurchase announcements, the buy and hold abnormal return around announcement will be lower (substitute effects). Kyle (1985) suggests that insider traders can better hide themselves when market liquidity is high, and are more likely to be identified when market liquidity is low. When stock liquidity is low, non-insider traders can judge if they are trading against insider traders at lower cost. When insiders choose to trade under low stock liquidity condition, the risk of exposing themselves is high. To justify this risk, the insiders need to be certain about the value of the insider information. A higher buy and hold abnormal return is necessary to incentivize insider buying when stock impact and transaction cost is high. Our second hypothesis is: H2: Net insider buying prior to share repurchase announcement increases buy and hold abnormal return around the announcement, and this association is stronger for firms with low up sided stock liquidity. Lie (2005) finds that operating performance improvement is limited to those firms that actually repurchase shares during the same fiscal quarter, suggesting share repurchase announcement by itself does not contain information about the level of future performance. We propose that instead of signaling better operating performance, share announcements signal safer operating performance. The information content in announcement and net insider buying prior to the announcement are complementary, and the information is about the decrease of volatility of future operating performance. Hence, we propose: H3: Net insider buying prior to repurchase announcement is negatively related to the volatility of future operating performance, and this association is stronger for firms with low up sided stock liquidity. We further inspect if corporate governance is able to dampen the insider trading and mitigate the relationship between net insider buying prior to repurchase announcement and volatility of future operating performance. In firms with better corporate governance, insider trading is more risky and costly. More specifically, we look at how well the interests of top managers and stock holders are aligned and how many independent board members are really independent (appointed before CEO assumed office and thus not affect by CEO preference). Theorectically, these two measures approximate the two main aspects of corporate governance, i.e. management incentives and board monitoring. Hence, we propose: H4: The association in H3 is weaker for firms with more real independent board member and firms using more performance based compensations.

6

3

Sample and variable construction

We obtain our sample data from Compustat North American Annual file, Standard & Poor’s Execucomp Annual file, CRSP-Compustat merged annual file, Thomas Reuters Insider file (Table 1 stock transaction report), and SDC Platium M&A stock buyback file. We exclude all firms in the financial and utilities sectors (SIC codes between 6000-6999 and 4900-4999). Our sample period begins in 1983 and ends in 2016. Prior to the SEC’s amendment of safe harbor Rule 10b-18 in 1982, uncertainty about potential charges of price manipulation discouraged managerial preference of share repurchase. Our payout (dependent) variables include REP (repurchases/assets), DIV (dividends/assets), REPR (repurchases/total payouts), and AREP (announced repurchases/assets). Following Grullon and Michaely (2002), we compute repurchases using the purchases of common and preferred stock (Compustat item #115) and then subtract any reductions in the redemption value of preferred stock (Compustat item #56). Dividends correspond to Compustat item #21, and total assets to Compustat item #6. Total payout is the sum of repurchases and dividends. Following Brennan et al. (2013), we use Amihud (2002) illiquidity measure and its sided components as liquidity (independent) variables, and high illiquidity measure means low stock liquidity. we denote the Amihud (2002) measure, which is constructed using the daily return and the dollar volume of trading, by ILLIQ. It is defined as the annual average of the daily values of |rit | DV OLit where r is a daily stock return and DVOL is daily dollar volume of transactions. It measures the sensitivity of stock return to the market value of transaction. UILLIQ corresponds to the annual average of this sensitivity when r is positive, and DILLIQ corresponds to the annual average of this sensitivity when r is negative. We use the Lakonishok and Lee (2001) measure of net buy amount, denoted by INSBUY, to proxy for insider trading during the 100 days period prior to repurchase announcement. It is calculated as: P −S P +S where P is the number of shares purchased and S is the number of shares sold by insiders. To measure the performance-based proportion of CEO compensation, we use FWEALTH (firm related wealth or inside equity) using the method of Coles et 7

al. (2006). To measure the independent board member monitoring power, we use IND (proportion of independent directors who are appointed before CEO assumed office) following the method of Coles et al. (2014). Following Brockman et al. (2008), we use measures of cash flow permanence, cash flow volatility, firm size, leverage, and investment opportunities as control variables. Following Jagannathan et al. (2000), we use operating cash flows to proxy for permanent cash flows (Compustat item # 13), denoted by PCASH, and non-operating cash flows to proxy for temporary cash flows (Compustat item # 61), denoted by TCASH. Both of these measures are scaled by the firm’s book value of assets. We define cash flow volatility as the standard deviation of the firm’s operating cash flows to total assets during the previous five-year period (i.e., from year −4 through year 0), denoted by CASHVOL. We define firm size as the logarithm of book value of assets, denoted by SIZE, and leverage as long term debt (Compustat item # 9) scaled by the book value of assets, denoted by LEV. Following Dittmar (2000), we use the market-to-book (M/B) ratio as a proxy for investment opportunities, denoted by MB. The market value of equity is the closing fiscal year-end price (Compustat item # 199) multiplied by the total number of shares outstanding (Compustat item # 25). The book value of equity corresponds to Compustat item #60. We exclude firms that have prices below $5 per share or total assets less than $1 million, and we trim the most extreme values from the top and bottom of each variable using a 0.5% cutoff. ILLIQ100 is the 100-day average of ILLIQ prior to repurchase announcement, UILLIQ100 is the 100-day average of UILLIQ prior to repurchase announcement, and DILLIQ100 is the 100-day average of DILLIQ prior to repurchase announcement. ROAVOL is the 3-year leading volatility of return on assets, i.e. standard deviation of ROA from year 1 through year 3 where the announcement occurs in year 0. BHAR is the 3-day buy and hold abnormal return of share repurchase announcement, and the benchmark is Fama-French 3 factor model and the estimation window is 100 days prior to the announcement date.

4

Empirical results

Dividends signal information in an one-time discrete fashion, while repurchases signal information in a more complicated way that the repurchase announcement signals a fraction of information in an one-time discrete way and the subsequent actual repurchasing transactions signal the rest of information in a continuous way. Consequently, we divide our empirical analysis into two parts and test the association between stock liquidity and repurchase on both actual share repurchase data and announced share repurchase data. For actual share repurchase, we use the annual actual amount of share repurchased as dependent variable; for announced share repurchase, we use the buy and hold abnormal return around the announcement date as dependent variable. 8

NYSE and NASDAQ are very different in their treatment of limit orders and commissions, resulting different execution efficiency and liquidity costs (Huang and Stoll (1996)). Consequently, following Brennan et al. (2013) we create subsample NYAM (corresponding to stocks listed on NYSE and AMEX) and subsample NASDAQ (corresponding to stocks listed on NASDAQ) to account for the difference in exchange transaction mechanism.

4.1

Actual share repurchase

We present summary statistics of NYAM and NASDAQ firms on Table 1 and Table 2. Table 1 shows that annually NYAM firms on average distribute cash that weighs 1.73% of total assets via share repurchase and 4.37% via dividends, and that 38.15% of the total cash distributions are made by repurchase. Table 2 shows that annually NASDAQ firms on average distribute cash that weighs 2.01% of total assets via share repurchase and 0.88% via dividends, and that 61.18% of the total cash distributions are made by repurchase. In addition, NYAM firms on average are larger in terms of size, uses more leverage, and are more mature and have fewer future growth opportunities. These results are consistent with the findings of Fama and French (2001). In addtion, for NYSE stocks, repurchase is more stable than dividends (standard deviation of 0.07 compared to 1.694), which contradicts the idea that repurchase has more flexibility over dividends. For NASDAQ stocks, repurchase is more volatile than dividends. Table 3 and Table 4 show the Pearson correlation between key variables for NYAM and NASDAQ stocks. On average, the correlation between liquidity and repurchase is smaller in magnitude for NASDAQ stocks, and large and mature firms tend to have higher liquidity measures (lower ILLIQ). We test our first hypothesis with the regression model as below:

P AY OU Ti,t = α+β1 ∗LIQU IDIT Yi,t−1 +β2 ∗P CASHi,t−1 +β3 ∗T CASHi,t−1 + β4 ∗CASHV OLi,t−1 +β5 ∗SIZEi,t−1 +β6 ∗LEVi,t−1 +β7 ∗M Bi,t−1 +δi +ηt +ei,t where PAYOUT represents 3 dependent variables, i.e. REP, DIV, and REPR, and LIQUIDITY represents 3 independent variables, i.e. ILLIQ, UILLIQ, and DILLIQ. This gives us 9 combinations of regression models, and their results are listed on Table 5 and Table 6. For NYAM firms, 1 unit increase in ILLIQ will cause 21 basis points of decrease in the annual amount of actual share repurchase (higher ILLIQ means lower stock market liquidity), which is consistent with the findings of Brockman et al. (2008). 1 unit increase in ILLIQ will lead to 149 basis points of increase in dividends payout, which contradicts the findings of Brockman et al. (2008) but is consistent with that of Banerjee et al. (2007). This result offers support to the dividends and share repurchase substitution hypothesis of Grullon and Michaely (2002). When stock liquidity is low, firms are more willing to distribute cash via dividends to avoid higher repurchase cost 9

caused by price impact, and when stock liquidity is high, the direction of substitution will be reverse. More importantly, we find that UILLIQ negatively affects actual share repurchase with a coefficient of -0.0019, which supports our first hypothesis that up sided stock liquidity is positively associated with managerial share repurchase decisions. In this setting, up sided liquidity is more appropriate in studying the stock liquidity-repurchase relationship in the sense that managers have an ex ante expectation that stock price will go up after share repurchase and thus they are more concerned about the liquidity cost and price impact in the state when price goes up. In addtition, Brennan et al. (2013) find that up sided liquidity has no effect on stock return and only down sided liquidity is priced, implying investors are more concerned with liquidity risk on down side. The possible expanation for Brennan et al. (2013)’s findings could be the behavior factors or constrain on short sale, etc. Their empirical results eliminate the possibility that up sided liquidity affects stock return first and then stock return affects repurchase, which makes the channel through which liquidity impact repurchase uninterrupted. Consequently, up sided liquidity is more appropriate for testing the relationship between stock liquidity and repurchase decisions. For NASDAQ firms, liquidity and up sided liquidity show a similar pattern of effects on repurchase, while dividends are negatively related to ILLIQ and UILLIQ. This result is consistent with the findings in Jiang et al. (2017) which use Chinese stock market as test sample, implying in terms of response to stock liquidity change, dividends and repurchase are not substitute of each other. In addition, we find that on average larger firms prefer to pay dividends over repurchase, increase in temporary cash flow only increases repurchase but not dividends payout, and that volatility of cash flow affects both dividends and repurchase. Bliss et al. (2015) document a sharpe decline in both dividends and repurchases in the 2007-2008 financial crisis. To inspect if the relationship between stock liquidity and repurchase changes after this systematic shock, we further look into the post-crisis period subsamples and their results are shown on Table 7 and Table 8. We find in the post-crisis NYAM sample firms, liquidity has no effect on dividends, which is consistent with Brockman et al. (2008). However, repurchase is still positively affected by stock liquidity. For post-crisis NASDAQ sample firms, we find that the ratio of repurchase to dividends is not associated with stock liquidity though the amount of repurchase and dividends are affected by liquidity, and this implys that repurchase and dividends have very similar sensitivity to stock liquidity. Compared to dividends, repurchases have a stable and clear response mechanism to stock liquidity change, which enables us to better understand how market microstructure interacts with corporate events (Lipson (2003)).

10

4.2

Share repurchase announcement

Unlike dividends announcement, share repurchase announcement is not required to be fulfilled. Stephens and Weisbach (1998) find that, 3 years after an open market share repurchase announcement, a substantial number of firms have repurchased no shares, about 10% of firms repurchased less than 5% of the shares authorized. However, stock market normally considers share repurchase announcement as good news, implying the announcement itself contains positive information about firm fundamentals. To test the role stock liquidity plays in the settings of share repurchase announcement, we construct an event study frame and use the 3-day buy and hold abnormal return to measure the value effect of the announcements. Insider trading prior to corporate events contains insider information that is important for understanding corporate finance issues like corporate governance, management compensation, etc (Marin and Olivier (2008)). We argue that stock liquidity plays an important role in analyzing insider trading prior to share repurchase announcements. To test our second hypothesis, we use the this regression model: BHARi,t = α+β1 ∗LIQU IDIT Y 100i,t +β2 ∗IN SBU Yi,t +β3 ∗LIQU IDIT Yi,t−1 + β4 ∗ AREPi,t + β5 ∗ P CASHi,t−1 + β6 ∗ T CASHi,t−1 + β7 ∗ CASHV OLi,t−1 + β8 ∗ SIZEi,t−1 + β9 ∗ LEVi,t−1 + β10 ∗ M Bi,t−1 +β11 ∗LIQIDU IT Y 100i,t ×IN SBU Yi,t +β12 ∗IN SBU Yi,t ×AREPi,t +δi +ηt +ei,t where LIQUIDITY100 represents 3 independent variables, i.e. ILLIQ100, UILLIQ100, and DILLIQ100, and this gives us 3 regression models and the results are in Table 9 and Table 10. For NYAM stocks, 1 unit increase in the authorized amount of share repurchase at announcement date will improve buy and hold abnormal return by 0.0865, and this implies that higher announced amount of shares repurchases signal more positive information. 1 unit increse in net insider buying prior to share repurchase announcement will increase 0.1604 in the buy and hold abnormal retrun, suggesting when insiders trade in the same direction as the repurchasing firm the value effect of the announcement will be larger. The interaction between net insider buying and announced amount of repurchase is positive, implying complementarity between the information in announcement and the insider trading. The interaction between net insider buying and liquidity is positive, implying that when liquidity cost is high and insiders still buy stocks at the high risk of being identified as informed trader, the value effect of repurchase announcement will be larger. For NASDAQ stocks, we observe similar patterns of associations. For the post-crisis subperiod, the effects of liquidity become stronger for NASDAQ firms but weaker for NYAM stocks. The reason could be that there are more young and small firms listed on NASDAQ and these firms have lower liquidity which makes liquidity an important factor when repurchasing stocks. Also, NASDAQ firms rely more 11

on stock buyback to distribute residual revenues to stockholders, thus making it difficult to use dividend as an alternative payout method when liquidity is low. Bonaimé and Ryngaert (2013) find that both insider buying and insider selling increase signicantly during the month of share repurchase announcements, which implies investors’ disagreement about the information content of repurchase announcements. To disentagle the investors that hold different opinions about the information content, we need to use the up sided stock liquidity to inspect the net insider buying activity (proxy of insider trading in the same direction of repurchasing firm). Insider buyers are concerned with the upward price risk prior to announcement. If we use the tradtional non-sided liquidity, then the interaction between liquidity and insider trading will get opposite effects mixed and cancelled off, which makes it difficult to isolate the directional insider trading activities. We use the up sided stock liquidity and find that net insider buying and stock liquidity have complementary effects, suggesting that insider buyers are willing to take the risk of low stock liquidity (being identified as informed traders and other uninformed traders could infer information from these transactions and thus reduce the information advantage of the insiders) when they can be compensated by higher buy and hold abnormal return around the announcement date. From the previous test, we see that when insiders trade in the same direction as the repuchasing firm, the buy and hold abnormal return will be higher. We further argue in hypothsesis 3 that insider trading prior to repurchase announcement and the announcement contain complementary information about the fundamentals of the firm. Unlike early studies that examine if payout announcements signal improvement of future operating performance, we inspect the risk of future operating performance. To test our 3rd hypothesis, we use the regression model: ROAV OLi,t+3 = α+β1 ∗LIQU IDIT Y 100i,t +β2 ∗IN SBU Yi,t +β3 ∗LIQU IDIT Yi,t−1 + β8 ∗ SIZEi,t−1 + β4 ∗ LEVi,t−1 + β5 ∗ M Bi,t−1 + β6 ∗ LIQIDU IT Y 100i,t × IN SBU Yi,t + δi + ηt + ei,t The regression results are in Table 13 and Table 14. For NYAM firms, 1 unit increase in net insider buying prior to repurchase announcement reduces 23 basis points in the standard deviation of ROA in the next 3 years and the interaction between stock illiquidity and net insider buying is negative. This implies that net insider buying prior to repurchase announcement contains information about firm fundamental, i.e. the decrease in volatility of operating performance for the 3 years following the announcement. The level of ROA has not changed significantly, but has become more stable. When stock liquidity is low, net insider buying predicts more reduction in the volatility of future operating performance, implying insider trading predicts higher fall in the risk of 12

future operating performance when stock liquidity is low. Our results are consistent with the implications of Campbell and Shiller (1988a) and Vuolteenaho (2002). Solid corporate governance can reduce the amount of insider trading activities and the abnormal return of insider trading (Gompers et al. (2003)). In this paper, we look at 2 aspects of corporate governance, i.e. the alignment of management and firm interests and the monitoring power of independent board directors. To test hypothesis 4, we use this regression model:

ROAV OLi,t+3 = α+β1 ∗LIQU IDIT Y 100i,t +β2 ∗IN SBU Yi,t +β3 ∗LIQU IDIT Yi,t−1 + β8 ∗ SIZEi,t−1 + β4 ∗ LEVi,t−1 + β5 ∗ M Bi,t−1 +β6 ∗LIQIDU IT Y 100i,t ×IN SBU Yi,t +β7 ∗LIQIDU IT Y 100i,t ×IN SBU Yi,t ×F W EALT Hi,t + β7 ∗ LIQIDU IT Y 100i,t × IN SBU Yi,t × IN Di,t + δi + ηt + ei,t In the results of Table 15 and Table 16, we find that when there are more independent directors who are appointed before CEO assumed office (not influenced by CEO preference) or higher firm related wealth hold by CEO, the association between net insider buying, stock liquidity and voaltility of future operating performance is lower, implying better corporate governance mitigate the information advantage of insiders. These results suggest that performance based compensations alleviate agency conflicts in the aspect of insider trading and when there are more independent board directors appointed without influence of CEO, insiders are less likely to profit on insider information in stock market.

5

Robustness check

For all panel regressions, we use firm and year fixed effects and clustering (Petersen (2009)) and use Newey-West covariance to handle autocorrelation issue. We also have tried other measures of stock liquidity, including stock turnover, turnover version of Amihud measure (Brennan et al. (2013)), efficient bid-ask spread, Lou and Shu (2017) modified Amihud measure, and the results are similar. We used lagged independent variables to avoid endogeneity concern, and use falsification test for causality check. Our results are robust to alternative measures and fasification test, and most results remain unchanged in direction and significance of effects.

6

Summary and conclusion

This paper studies the role of stock liquidity in share repurchase decisions, and our results are consistent with the Brav et al. (2005) survey that managers condition repurchase decisions on stock market liquidity. We propose a novel 13

method to use up sided stock liquidity to analyze the relationsip between stock liquidity and repurchase, because managers have ex ante expectation that stock buyback will drive up stock price and thus are concerned with the transaction cost of price impact when stock price goes up. We find that actual share repurchases are positively associated with stock liquidity, and the mixed results of how dividends respond to stock liquidity change found by previous research could be caused by the different market mechanism adopted by NYSE, AMEX, and NASDAQ. We extend our anlysis to insider trading and corporate governance within the stock liquidity-repurchase frame. We find that firms are concerned with up sided liquidity when they buy back stocks, and that net insider buying prior to share repurchase announcement increases the buy and hold abnormal return of the announcement. We also find that the buy and hold abnormal return of repurchase announcement is higher when stock liquidity is low. When insiders trade in the same direction as the repurchasing firm, the announcement abnormal return will be higher when sock liquidity is low. Net insider buying prior to repurchase announcement and the announcement contain complementary information about decline in the volatility of future operating performance. Use of performance based compensation and better independent board member monitoring can also alleviate the the information advantage of corporate insiders. Brennan et al. (2013) find that down sided stock liquidity is a risk factor that investor requires return premium on. It would be interesting to investigate the interaction between insider trading in the oppsite direction of repurchasing firm and down sided liquidity.

7

References

Almeida, H., Fos, V. and Kronlund, M., 2016. The real effects of share repurchases. Journal of Financial Economics, 119(1), pp.168-185. Amihud, Y., 2002. Illiquidity and stock returns: cross-section and timeseries effects. Journal of financial markets, 5(1), pp.31-56. Bagwell, L., 1991, Share repurchase and takeover deterrence. Rand Journal of Economics 22, 72-88. Banerjee, S., Gatchev, V.A. and Spindt, P.A., 2007. Stock market liquidity and firm dividend policy. Journal of Financial and Quantitative Analysis, 42(2), pp.369-397. Barclay, M., Smith, C., 1988. Corporate payout policy: cash dividends versus open-market repurchases. Journal of Financial Economics 22, 61–82. Bernanke, B.S., 1983. The determinants of investment: Another look. The American Economic Review, 73(2), pp.71-75. Bliss, B.A., Cheng, Y. and Denis, D.J., 2015. Corporate payout, cash retention, and the supply of credit: Evidence from the 2008–2009 credit crisis. 14

Journal of Financial Economics, 115(3), pp.521-540. Bonaime, A.A. and Ryngaert, M.D., 2013. Insider trading and share repurchases: Do insiders and firms trade in the same direction?. Journal of Corporate Finance, 22, pp.35-53. Brav, A., J. Graham, C. Harvey, and R. Michaely, 2005, Payout policy in the 21st century. Journal of Financial Economics 77, 483-527. Brennan, M., and A. Thakor, 1990, Shareholder preferences and dividend policy. Journal of Finance 45, 992-1017. Brennan, M., Huh, S.W. and Subrahmanyam, A., 2013. An analysis of the Amihud illiquidity premium. Review of Asset Pricing Studies, 3(1), pp.133-176. Brockman, P., and D. Y. Chung, 2001, Managerial timing and corporate liquidity: evidence from actual share repurchases. Journal of Financial Economics 61, 417-448. Brockman, P., Howe, J.S. and Mortal, S., 2008. Stock market liquidity and the decision to repurchase. Journal of Corporate Finance, 14(4), pp.446-459. Chan, K., Ikenberry, D.L. and Lee, I., 2007. Do managers time the market? Evidence from open-market share repurchases. Journal of Banking & Finance, 31(9), pp.2673-2694. Campbell, John Y., and Robert J. Shiller. "Stock prices, earnings, and expected dividends." The Journal of Finance 43, no. 3 (1988): 661-676. Coles, J.L., Daniel, N.D. and Naveen, L., 2006. Managerial incentives and risk-taking. Journal of financial Economics, 79(2), pp.431-468. Coles, J.L., Daniel, N.D. and Naveen, L., 2014. Co-opted boards. The Review of Financial Studies, 27(6), pp.1751-1796. DeAngelo, H., DeAngelo, L. and Stulz, R.M., 2006. Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory. Journal of Financial economics, 81(2), pp.227-254. Dittmar, A.K., 2000. Why do firms repurchase stock. The Journal of Business, 73(3), pp.331-355. Dittmar, A. and Field, L.C., 2015. Can managers time the market? Evidence using repurchase price data. Journal of Financial Economics, 115(2), pp.261282. Fama, E., 1998, Market efficiency, long-run returns, and behavioral finance. Journal of Financial Economics 49, 283-306. Fama, E., and K. French, 1993, Common risk factors in the returns on stocks and bonds. Journal of 22 Financial Economics 33, 3-56. Fama, E., and K. French, 2001, Disappearing dividends: changing firm characteristics or lower propensity to pay? Journal of Financial Economics 60, 3-43. Fenn, G. W., and N. Liang, 2001, Corporate payout policy and managerial stock incentives. Journal of Financial Economics 60, 45-72. Ginglinger, E., and J. Hamon, 2007, Actual share repurchases, timing and liquidity. Journal of Banking & Finance 31, 915-938. Grullon, G., and D. Ikenberry, 2000, What do we know about stock repurchases? Journal of Applied Corporate Finance 13, 31-51.

15

Grullon, G., and R. Michaely, 2002, Dividends, share repurchases and the substitution hypothesis. Journal of Finance 57, 1649-1684. Grullon, G., and R. Michaely, 2004, The information content of share repurchase programs. Journal of Finance 59, 651-680. Huang, R.D. and Stoll, H.R., 1996. Dealer versus auction markets: A paired comparison of execution costs on NASDAQ and the NYSE. Journal of Financial economics, 41(3), pp.313-357. Vancouver Ikenberry, D., J. Lakonishok, and T. Vermaelen, 1995, The underreaction to open market share repurchases. Journal of Financial Economics 39, 181-208. Ikenberry, D.L. and Vermaelen, T., 1996. The option to repurchase stock. Financial Management, pp.9-24. Jagannathan, M., Stephens, C.P. and Weisbach, M.S., 2000. Financial flexibility and the choice between dividends and stock repurchases. Journal of financial Economics, 57(3), pp.355-384. Jagannathan, M., Stephens, C.P. and Weisbach, M.S., 2000. Financial flexibility and the choice between dividends and stock repurchases. Journal of financial Economics, 57(3), pp.355-384. Jiang, F., Ma, Y. and Shi, B., 2017. Stock liquidity and dividend payouts. Journal of Corporate Finance, 42, pp.295-314. Kahle, K., 2002, When a buyback isn’t a buyback: open market repurchases and employee stock options. Journal of Financial Economics 63, 235-261. Kyle, A.S., 1985. Continuous auctions and insider trading. Econometrica: Journal of the Econometric Society, pp.1315-1335. Lakonishok, J. and Lee, I., 2001. Are insider trades informative?. The Review of Financial Studies, 14(1), pp.79-111. Lie, E., 2005. Operating performance following open market share repurchase announcements. Journal of Accounting and Economics, 39(3), pp.411-436. Lipson, M.L., 2003. Market microstructure and corporate finance. Journal of Corporate Finance, 9(4), pp.377-384. Lou, X. and Shu, T., 2017. Price Impact or Trading Volume: Why Is the Amihud (2002) Measure Priced?. The Review of Financial Studies, 30(12), pp.4481-4520. Miller, M.H. and Modigliani, F., 1961. Dividend policy, growth, and the valuation of shares. The Journal of Business, 34(4), pp.411-433. Marin, J.M. and Olivier, J.P., 2008. The dog that did not bark: Insider trading and crashes. The Journal of Finance, 63(5), pp.2429-2476. Petersen, M. 2009, Estimating standard errors in finance panel data sets: comparing approaches. Review of Financial Studies 22, 435-480. Stephens, C.P. and Weisbach, M.S., 1998. Actual share reacquisitions in open-market repurchase programs. The Journal of finance, 53(1), pp.313-333. Vuolteenaho, T., 2002. What Drives Firm-Level Stock Returns?. The Journal of Finance, 57(1), pp.233-264.

8

Appendix

Table 1: Descriptive statistics Mean St. Deviation REP 0.0173 0.0700 DIV 0.0437 1.6940 REPR 0.3815 0.4109 ILLIQ -17.8606 3.1184 UILLIQ -17.1857 2.8865 DILLIQ -17.1114 2.9527 PCASH 0.1254 1.7002 TCASH 0.0056 0.5297 CASHVOL 0.0618 0.8111 SIZE 6.7414 2.3398 LEV 0.2227 0.2096 MB 1.3215 18.5547

of key variables for firms listed on NYSE and AMEX during 1983-2016 Minimum 25th Percentile Median 75th Percentile Maximum Skew 0.0000 0.0000 0.0000 0.0102 9.5853 45.8996 -0.3737 0.0000 0.0050 0.0219 226.3104 114.1030 0.0000 0.0000 0.1935 0.8171 1.0125 0.4611 -26.5256 -20.1176 -17.9706 -15.7288 -4.7852 0.1766 -25.2727 -19.2364 -17.2437 -15.2750 -4.3103 0.1750 -25.4410 -19.2190 -17.1720 -15.1538 -3.3325 0.1755 -21.4375 0.0575 0.1136 0.1662 226.3104 108.6097 -152.3750 0.0000 0.0034 0.0120 3.0895 -284.3077 0.0000 0.0138 0.0267 0.0503 92.6040 95.3110 -1.5750 5.1401 6.7739 8.2969 15.1429 0.0718 0.0000 0.0527 0.1900 0.3337 6.9307 2.7391 0.0000 0.3430 0.6582 1.2081 2705.2507 118.6334

Kurtosis 5186.7958 13708.7878 -1.4906 -0.3692 -0.2205 -0.2259 12667.5368 81762.1091 9866.5698 -0.0511 33.5706 15388.2660

REP DIV REPR ILLIQ UILLIQ DILLIQ PCASH TCASH CASHVOL SIZE LEV MB

Table 2: Descriptive statistics of key variables for firms Mean St. Deviation Minimum 25th Percentile 0.0201 0.1223 0.0000 0.0000 0.0088 0.0986 -0.0073 0.0000 0.6118 0.4449 0.0000 0.0085 -15.7781 3.1364 -27.8892 -17.9289 -15.1327 3.0204 -26.7986 -17.2171 -14.9395 3.1207 -27.0082 -17.0786 0.0078 0.6903 -134.2381 -0.0026 0.0085 0.1318 -13.5601 0.0000 0.0959 0.3944 0.0000 0.0202 4.7345 2.0745 -4.3428 3.2379 0.1345 0.2340 0.0000 0.0000 1.6374 3.4719 0.0000 0.3356

listed on NASDAQ during 1983-2016 Median 75th Percentile Maximum 0.0000 0.0048 19.7374 0.0000 0.0016 14.1648 0.9923 1.0000 1.0000 -15.5071 -13.4935 -4.1282 -14.9734 -12.9011 -2.9110 -14.7500 -12.6731 -1.8225 0.0625 0.1422 4.5310 0.0039 0.0159 30.4302 0.0468 0.0986 58.9831 4.6367 6.1403 12.7550 0.0509 0.2037 39.5929 0.8505 1.8454 263.3318

Skew 70.1898 69.3279 -0.4476 -0.2798 -0.2266 -0.2204 -120.4477 110.0799 78.1497 0.2545 52.7583 23.4607

Kurtosis 9598.9380 7865.1083 -1.6462 -0.2759 -0.2762 -0.2323 20445.6849 31533.6054 9284.4923 -0.1447 8541.7129 1177.9244

REP DIV REPR ILLIQ UILLIQ DILLIQ PCASH TCASH CASHVOL SIZE LEV MB

Table 3: REP 1.00 -0.01 0.31 -0.05 -0.05 -0.05 -0.02 0.02 0.15 -0.07 0.02 0.16

Correlation between key variables for firms listed on NYSE and AMEX during 1983-2016 DIV REPR ILLIQ UILLIQ DILLIQ PCASH TCASH CASHVOL SIZE LEV

MB

1.00 -0.10 0.02 0.03 0.02 0.65 0.03 0.26 -0.11 -0.05 0.65

1.00

1.00 0.00 0.00 0.01 -0.05 -0.01 0.08 -0.10 0.01 0.01

1.00 0.99 0.99 -0.10 0.07 0.13 -0.80 -0.02 -0.08

1.00 1.00 -0.10 0.07 0.12 -0.79 -0.02 -0.08

1.00 -0.10 0.07 0.13 -0.79 -0.02 -0.09

1.00 -0.03 -0.02 -0.06 -0.07 0.63

1.00 0.02 -0.09 -0.06 0.06

1.00 -0.21 0.00 0.35

1.00 0.09 -0.15

1.00 -0.13

REP DIV REPR ILLIQ UILLIQ DILLIQ PCASH TCASH CASHVOL SIZE LEV MB

Table 4: Correlation between key variables for firms listed on NASDAQ during 1983-2016 REP DIV REPR ILLIQ UILLIQ DILLIQ PCASH TCASH CASHVOL SIZE 1.00 -0.01 1.00 0.18 -0.16 1.00 -0.03 0.00 -0.02 1.00 -0.04 0.00 -0.03 0.99 1.00 -0.03 0.00 -0.02 0.99 0.99 1.00 -0.66 0.09 -0.07 -0.11 -0.11 -0.11 1.00 0.58 0.06 0.01 0.01 0.01 0.01 -0.76 1.00 0.64 0.06 0.11 0.08 0.07 0.07 -0.81 0.70 1.00 -0.09 -0.07 -0.19 -0.73 -0.70 -0.72 0.09 -0.04 -0.20 1.00 0.02 -0.03 0.02 -0.02 -0.02 -0.02 -0.01 0.02 0.01 0.14 0.16 0.15 0.10 -0.26 -0.27 -0.27 0.00 0.03 0.17 -0.14

LEV

MB

1.00 -0.15

1.00

Table 5: Actual share repurchase and stock liquidity for NYAM firms during 1983-2016 Dependent variable:

ILLIQ

REP

DIV

REPR

REP

DIV

REPR

REP

DIV

REPR

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

−0.0021∗∗∗ (0.0007)

0.0149∗∗ (0.0061)

0.0075∗∗ (0.0037) −0.0019∗∗∗ (0.0007)

0.0136∗ (0.0071)

0.0057 (0.0035) −0.0019∗∗∗ (0.0006)

0.0138∗∗ (0.0063)

0.0049 (0.0035)

UILLIQ DILLIQ PCASH

0.0080 (0.0076)

0.1120∗∗ (0.0526)

0.0839∗ (0.0436)

0.0082 (0.0076)

0.1120∗∗ (0.0527)

0.0808∗ (0.0432)

0.0082 (0.0077)

0.1120∗∗ (0.0527)

0.0800∗ (0.0429)

TCASH

0.0135∗ (0.0076)

0.3712∗∗ (0.1591)

−0.0010 (0.1452)

0.0141∗ (0.0080)

0.3763∗∗ (0.1592)

−0.0043 (0.1440)

0.0134∗ (0.0075)

0.3715∗∗ (0.1596)

−0.0050 (0.1440)

CASHVOL

0.0149∗∗ (0.0071)

1.4167∗∗∗ (0.2368)

0.1070 (0.0765)

0.0149∗∗ (0.0071)

1.4169∗∗∗ (0.2366)

0.1070 (0.0760)

0.0150∗∗ (0.0071)

1.4167∗∗∗ (0.2368)

0.1065 (0.0756)

SIZE

−0.0025∗ (0.0015)

0.0499∗∗∗ (0.0074)

−0.0024 (0.0088)

−0.0021 (0.0015)

0.0473∗∗∗ (0.0076)

−0.0049 (0.0084)

−0.0021 (0.0014)

0.0479∗∗∗ (0.0072)

−0.0057 (0.0087)

LEV

−0.0264∗∗∗ (0.0055)

−0.0489∗ (0.0249)

−0.0692∗∗ (0.0345)

−0.0269∗∗∗ (0.0055)

−0.0462∗ (0.0245)

−0.0664∗ (0.0346)

−0.0267∗∗∗ (0.0055)

−0.0474∗ (0.0249)

−0.0660∗ (0.0346)

MB

0.0025∗∗ (0.0011)

0.0241∗∗∗ (0.0047)

0.0042 (0.0027)

0.0026∗∗ (0.0011)

0.0241∗∗∗ (0.0047)

0.0038 (0.0026)

0.0026∗∗ (0.0011)

0.0241∗∗∗ (0.0047)

0.0037 (0.0026)

50,007 0.2710 0.1951 0.0651 (df = 45289)

58,330 0.8222 0.8052 0.8938 (df = 53242)

38,508 0.6185 0.5725 0.2674 (df = 34366)

50,000 0.2709 0.1950 0.0651 (df = 45282)

58,323 0.8222 0.8052 0.8939 (df = 53235)

38,507 0.6184 0.5724 0.2674 (df = 34365)

49,995 0.2709 0.1950 0.0651 (df = 45277)

58,317 0.8222 0.8052 0.8939 (df = 53229)

38,503 0.6183 0.5723 0.2674 (df = 34361)

Observations R2 Adjusted R2 Residual Std. Error Note:



p