نقدینگی بازار سهام و تصمیم به خرید مجدد

We examine the impact of stock market liquidity on managerial payout decisions. We argue that stock market liquidity influences payout policy through a first-order effect on the share repurchase decision, and a second-order or residual effect on the dividend decision. Managers compare the tax and flexibility advantages of a repurchase against its liquidity cost disadvantage. All else equal, higher market liquidity encourages the use of repurchases over dividends. Our empirical results confirm that stock market liquidity plays a significant role in repurchase and dividend initiations, as well as in recurring payout decisions. Unlike previous studies that measure liquidity changes following the repurchase decision, we examine liquidity levels prior to the payout decision. We show that managers condition their repurchase decision on a sufficient level of market liquidity, consistent with Barclay and Smith's [Barclay, M.J., Smith, C.W. Jr., 1988. Corporate payout policy: cash dividends versus open-market repurchases. Journal of Financial Economics 22, 61–82.] theoretical analysis and Brav et al.'s [Brav, A., Graham, J.R., Campbell, R.H., Michaely, R., 2005. Payout policy in the 21st century. Journal of Financial Economics 77, 483–528.] CFO survey results. Repurchases have recently become the payout decision of choice in part because of rising stock market liquidity.

Managers establish payout policy by selecting the level, timing, and composition of cash remittances. Recent research has shown that while the level and timing of aggregate cash flows have changed relatively little since the 1970s, the composition of these payouts has changed significantly. The level of aggregate cash payouts has remained constant at roughly 3–5% of equity value (Allen and Michaely, 2003), and managers have persistently paid dividends at regular quarterly intervals. In contrast, the composition of corporate payouts has grown from a repurchase-to-dividend ratio of 8.44% in 1972 to 113.11% in 2000 (Grullon and Michaely, 2002). The shift towards repurchases is not only meaningful in percentage terms. During the five-year period ending in 2000, managers repurchased over $846 billion of their companies' equity (Grullon and Michaely, 2002). This striking transformation in payout policy has attracted considerable interest among academics, regulators, and practitioners. Although it is unlikely that any single variable can fully account for all of these empirical regularities, we argue that stock market liquidity plays a significant role in explaining changes in the composition of corporate payouts. In contrast to payout policy irrelevancy under perfect capital markets, real-world managers operate in a business environment characterized by asymmetric information, incentive problems, and transaction costs. Under these conditions, alternative payout policies have a direct impact on the firm's cost of capital and market value.1 Value-maximizing managers will search for the payout mechanism that minimizes the sum of transaction, incentive, and information costs. When market liquidity is low, managers are more reluctant to repurchase shares and reduce float because their market transactions could increase the price impact of trading — and survey evidence shows that managers are aware that price impact matters to investors (Brav et al., 2005). Managers are also reluctant to repurchase shares when liquidity is low because their trading activity could impact transaction costs by widening bid–ask spreads (Barclay and Smith, 1988). Thus, we hypothesize that managers will prefer repurchases to dividends when stock market liquidity is relatively high. We further posit that stock market liquidity will have a direct impact on repurchases and a residual impact on dividends through the substitution effect described in Grullon and Michaely (2002). This analysis suggests one important channel through which a firm's market microstructure can influence its corporate decisions (Lipson, 2003). Barclay and Smith (1988) treat the level and timing of payouts as predetermined and examine why managers prefer one payout mechanism over another. Managers attempt to maximize firm value by minimizing the total cost of cash distributions. Although tax advantages appear to favor share repurchases over dividends, Barclay and Smith (1988) show that repurchases also induce higher asymmetric information costs. When managers announce repurchase programs, uninformed investors realize that they are exposed to a higher probability of trading against informed insiders. This realization impairs the firm's information environment and results in higher liquidity costs. Dividend payments, on the other hand, do not increase the probability of trading against informed managers and therefore do not increase liquidity costs. A direct consequence of this analysis is that managers consider the liquidity of the stock in making the decision about the form of the payout. We refer to this hypothesis as the liquidity hypothesis of repurchases. One testable implication of this hypothesis is that the firm's current liquidity level will significantly influence subsequent payout choices. Previous studies have examined the effect of the current payout decision on subsequent changes in liquidity (e.g., Miller and McConnell, 1995, Brockman and Chung, 2001 and Cook et al., 2004). While these studies provide useful results about the consequences of repurchases, the purpose of this study is to examine the determinants of repurchases. A second testable implication of our liquidity hypothesis is that market liquidity will have a stronger impact on repurchase decisions than on dividend decisions. In a related study, Banerjee et al. (2007) examine the impact of stock market liquidity on dividend policy. Although their dividend hypothesis yields some observationally-equivalent predictions, our results suggest that stock market liquidity has a first-order effect on the repurchase decision and a residual effect on the dividend decision. High levels of liquidity allow managers to benefit from the tax and flexibility advantages of repurchase programs — and dividends decline as a consequence. Consistent with our central claim, Brav et al. (2005) provide evidence based on financial executive surveys and interviews that supports our liquidity hypothesis of repurchases. Managers express a keen awareness that their “stock price would decrease if the overall liquidity of the stock were to fall” (pp. 515–516). In addition, One-half of firms feel that the liquidity of their stock is an important or very important factor affecting their repurchase decisions (Table 8, row 4). Interview discussion clarifies that the executives think that reduced liquidity can hurt their stock price because demand for a stock falls if investors think that their trades would move the stock price. Therefore, a company would restrict repurchases if it feels that doing so would reduce liquidity below some critical level. Managers are clearly concerned that repurchase decisions can impair their firms' market liquidity, and that any such impairment would reduce market values. In contrast, managers do not appear to condition their dividend decisions on stock market liquidity. They describe the role of liquidity in dividend decisions as “not important.” Our empirical findings fit these survey and interview results very closely. Stock market liquidity influences payout policy primarily through the repurchase decision. We conduct empirical tests of the liquidity hypothesis of repurchases using company payout data from 1983 to 2006. We divide our empirical analysis into two main sections corresponding to payout initiation decisions and ongoing payout decisions. Similar to Skinner's (2008) results, we find that there are two main groups of corporate payers: firms that make repurchases only, and firms that make both repurchases and dividend payments. The latter group consists of large, mature firms that have a history of paying dividends. There is also a much smaller group of firms that make dividend payments only. We provide separate analyses for each of these groups. In our payout initiation analysis, we show that repurchase-initiating firms are significantly more liquid than non-initiating firms. We find that dividend-initiating firms are generally less liquid than non-initiating firms, although the dividend-initiating results are not as robust or economically significant as their repurchase-initiating counterparts. These differential results corroborate the claim that the repurchase decision is more sensitive to stock market liquidity than the dividend decision. Our findings also verify that, given a payout initiation, the probability of a repurchase increases with the liquidity of the initiating firm. We use alternative measures of liquidity, including multiple control variables, and mitigate endogeneity concerns by using the current period's liquidity measure to explain the subsequent period's payout decision. Overall, our payout initiation results provide support for the liquidity hypothesis of repurchases. After confirming that liquidity plays a significant role in payout initiations, we test its explanatory power in ongoing payout decisions. We find that the size of the repurchase increases significantly with the stock market liquidity of the repurchasing firm. Parallel to the initiation results, we find that the size of the dividend generally decreases with the liquidity of the dividend-paying firm, although these results are again weaker than their repurchase counterparts. We also show that the repurchase portion of a firm's total payout is an increasing function of the firm's market liquidity. That is, stock market liquidity helps to explain the substitution between repurchases and dividends.2 Taken together, our empirical findings strongly support the hypothesis that stock market liquidity affects payout policy primarily through its impact on the repurchase decision. Our study contributes to a growing literature that connects market microstructure to corporate finance. Previous research has linked market microstructure to security offerings (Ellul and Pagano, 2006, Karolyi, 2003 and Butler et al., 2005), asset pricing and cost of capital (Amihud and Mendelson, 1986 and Amihud et al., 1997), mergers and acquisitions (Lipson and Mortal, 2007), and announcement effects related to dividend signaling (Fuller, 2003) and analyst recommendations (Kim et al., 1997 and Irvine, 2003). Other related studies have linked market microstructure to specific corporate decisions, including capital structure decisions (Lipson and Mortal, 2008) and investment decisions (Becker-Blease and Paul, 2006). Our study contributes to this line of research by establishing a significant role for market liquidity in corporate payout decisions. Most related studies examine the impact of the repurchase decision on the firm's subsequent market liquidity (Barclay and Smith, 1988, Wiggins, 1994, Singh et al., 1994, Miller and McConnell, 1995, Brockman and Chung, 2001, Cook et al., 2004 and Ginglinger and Hamon, 2005). These studies have produced decidedly mixed results.3 In contrast, we begin with the firm's market liquidity and then examine its influence on the repurchase decision. If the decision to repurchase is conditioned on its expected liquidity impact, then measuring the liquidity impact for self-selected repurchasing firms will tend to understate the significance of liquidity as a managerial decision variable. The (ex-post) repurchase sample consists predominantly of liquid firms whose managers chose to repurchase shares only after considering their firms' ability to absorb the liquidity costs of this payout mechanism. Using our approach, we are able to show that stock market liquidity is a significant determinant of the repurchase decision. Another contribution of this study is that we identify the precise mechanism through which stock market liquidity affects payout policy. As stated by Skinner (2008, p.1), “the question of why firms are less likely to pay dividends, along with the relation of repurchases to this phenomenon, remains unresolved in corporate finance.” Our liquidity hypothesis of repurchases asserts that stock market liquidity encourages managers to substitute repurchases for dividends. Our empirical results confirm that higher levels of stock market liquidity enable managers to take advantage of the tax and flexibility advantages of repurchases. When liquidity is relatively high, non-payout firms will initiate with repurchases instead of dividends, and positive payout firms will increase repurchases relative to dividends. It is the repurchase dog that wags the dividend tail. The rest of our paper proceeds as follows. In Section 2, we provide a description of our data and sample selection. In Section 3, we present and analyze our empirical findings, and in Section 4, we summarize and conclude the paper.

We hypothesize that stock market liquidity affects corporate payout policy primarily through its in fl uence on the decision to repurchase. Our liquidity hypothesis of repurchases is consistent with Barclay and Smith's (1988) theoretical framework as well as with Brav et al.'s (2005) survey and interview results. Managers prefer repurchases over dividends because of tax and fl exibility advantages, although their ability to conduct repurchases is subject to various constraints. Prior to the enactment of safe harbor Rule 10b-18, managers were constrained by uncertainty about charges of price manipulation. Repurchase activity increased signi fi cantly after this regulatory constraint was lifted by the SEC's ruling in 1982. We argue that stock market liquidity directly impacts the repurchase decision and, through the substitution effect, indirectly impacts the dividend decision. Our empirical results con fi rm that managers use stock market liquidity as a decision variable in setting payout policies. We divide our results into two sections corresponding to payout initiation decisions and ongoing payout decisions. We fi nd that repurchase-initiating fi rms are signi fi cantly more liquid than non-initiating fi rms, and that dividend-initiating fi rms are generally less liquid than non-initiating fi rms. The repurchase-initiating results are stronger than their dividend-initiation counterparts, as hypothesized. We con fi rm parallel patterns for ongoing payouts. We show that the size of the repurchase increases signi fi cantly with the market liquidity of the repurchasing fi rm. These repurchase results are signi fi cant for every measure of market liquidity. We fi nd that the size of the dividend generally decreases with the liquidity of the dividend-paying fi rm, although these dividend results are considerably weaker than their repurchase counterparts, both in terms of statistical and economic signi fi cance. Similar to our initiation fi ndings, stock market liquidity has a fi rst-order effect on the repurchase decision and a weaker, residual effect on the dividend decision. Overall, our payout initiation and ongoing payout results provide considerable support for the liquidity hypothesis of repurchases. In addition to connecting market microstructure to corporate fi nance, our study contributes to a coherent picture of corporate payout policy that is currently emerging from the literature. Fama and French (2001) show that dividend policy has changed substantially since the 1970s, and that these changes cannot be fully attributed to fi rm characteristics. DeAngelo et al. (2004) con fi rm that there has been a downward shift in the percentage of fi rms that pay dividends, in spite of the fact that total dividend payouts have continued to increase since the 1970s. Grullon and Michaely (2002) , among others, document the secular rise in share repurchases over this same time period. Given the evidence that managers have been substituting repurchases fordividends, the next step is to examine the underlying causes of this substitution effect. Skinner (2008) shows that two groups of payers have emerged since the 1980s: fi rms that use repurchases exclusively, and fi rms that use repurchases and dividends combined. The latter group consists of well-established, mature fi rms that have been paying dividends for a number of years. Managers of these fi rms are reluctant to cut dividends because of perceived “ negative consequences ” ( Brav et al., 2005 ), but they use repurchases for payout increases if there is suf fi cient liquidity. Stock market liquidity determines the extent of the substitution effect for mature, dividend-paying fi rms. For the former group of repurchase- only fi rms, stock market liquidity plays an even more pervasive role since their entire payout depends on liquidity thresholds. When payout opportunities arise, these managers will choose repurchases as long as there is suf fi cient liquidity.