“Does stock ownership impact liquidity and dividends?”

This study investigates the interactions among stock ownership, liquidity and divi-dends in the UK stock market over the period 2002–2016. Using different liquidity measures, it is shown that stocks with higher levels of free float (institutional ownership) are associated with higher (lower) levels of liquidity. In addition, a positive and significant relation is found between institutional ownership and dividend payout policy, which, as a result, highlights the comparative tax advantages that UK institutions have for dividend income. These relations hold even after controlling for firm-specific characteristics. Finally, a negative relation is found between dividends and liquidity, implying that investors with less (more) liquid stocks are more (less) likely to receive dividend payments.


INTRODUCTION
The relation between stock ownership and liquidity has attracted a great amount of interest in the academic finance literature.Prior research has mainly focused on two channels through which liquidity can be affected by share ownership structure.First, the trading behavior of investors is regarded as an important determinant of stock market liquidity (Ding et al., 2016).The presence of concentrated ownership can reduce the number of common shares available for trading, and, therefore, can lower stock market liquidity through reducing trading activity (Demsetz, 1968;Ding et al., 2017).Second, stock market liquidity can be reduced by increasing share ownership, as larger institutional ownership can influence information asymmetries (Rubin, 2007).Specifically, large shareholders who possess private information increase the risk of adverse selection faced by market makers.This can lower stock liquidity as market makers are forced to increase bid-ask spreads in the presence of insiders (Glosten & Milgrom, 1985;Heflin & Shaw, 2000;Jacoby & Zheng, 2010).A number of studies including Ciner and Karagozoglu (2008) and Ding et al. (2016) suggest that firms with more shares in free float can alleviate information asymmetry problems.This study hypothesizes that stock liquidity rises (falls) with increased levels of free float (institutional ownership).Furthermore, corporate finance theories propose different reasons for why stock ownership structure and dividend payout policy may be associated.First, agency theory suggests that the distribution of dividends can help alleviate conflicts of interest between owners and managers (Khan, 2006).Dividend payments can serve as a monitoring mechanism substitute in the absence of adequate monitoring (Rozeff, 1982)  1 .Higher dividends can reduce firms' free cash flows and can drive managers to seek external financing, which exposes them to market scrutiny (Jensen, 1986;Khan, 2006).Presuming that institutional investors are better monitors, the agency related perspective implies a positive association between dividends and institutional ownership (Short et al., 2002;Grinstein & Michaely, 2005).Second, adverse selecation problems can induce uninformed investors to favor dividends over stock repurchases, as suggested by Barclay and Smith (1988) and Brennan and Thakor (1990).Since institutional investors are more likely to possess superior information, this theory suggests that institutions favor firms that pay out in the form of stock repurchases rather than dividends (Grinstein & Michaely, 2005).Third, institutions have a preference for dividend income as a result of the common institutional charter and prudent man rule restrictions (Del Guercio, 1996;Jun et al., 2011), and because of the comparative tax advantage that institutions have for dividend payments (Allen et al., 2000).Indeed, Short et al. (2002) claim that there are strong motivations for tax-exempt institutions to request higher dividends as a result of the bias in the UK tax system, which favors dividend income for tax-exempt investors 2 .Thus, this study hypothesizes a positive relation between institutional ownership and dividends.
The empirical findings are consistent with Ding et al. (2016), Rhee and Wang (2009) and Jiang et al. (2011) showing strong evidence that higher levels of free float (institutional ownership) are accompanied with higher (lower) levels of liquidity.Moreover, institutions are found to favor higher dividend payments, due to the bias in the UK tax system, which favors dividend income over capital gains, a finding, which is consistent with Short et al. (2002).These relations hold even after controlling for firm-specific factors, such as stock price, size, financial leverage and asset tangibility.Finally, a negative relationship is found between dividends and liquidity, suggesting that investors with less (more) liquid stocks are more (less) likely to receive dividend payments.This is consistent with Banerjee et al. (2007) who show that dividend-paying firms are positively related to investor demand for dividend payments and, thus, negatively associated with stock liquidity, suggesting that stock liquidity and dividend payments are regarded as substitutes in the view of investors.
The remainder of this study is structured as follows.Section 1 presents the literature review.Section 2 describes the sample, variable definitions and regression models.Section 3 discusses the results and findings.Finally last section concludes. 1 La Porta et al. (2000) provide strong evidence supporting the outcome agency model of dividends.Using a sample of 4,000 firms across 33 countries, they find that firms with better protection of shareholders are associated with higher payouts.In these countries, rapid growth firms are found to pay lower dividends as compared to slow growth firms, consistent with the idea that legally protected shareholders are willing to postpone their dividends when good investment opportunities exist. 2 In terms of institutional ownership, there are significant differences between the US and UK.These differences result from differences in legal restrictions and tax incentives (Short et al., 2002).The UK has a partial imputation tax system, which provides clear incentives for tax-exempt institutions to demand higher dividend payments, unlike the US, which has a classical company tax system whereby dividends are taxed twice: first, on the level of the firm company (via corporate tax on profits) and, then, on the level of the shareholder (via income tax on dividend income).Hence, the tax treatment of dividends in the UK is more favorable than the classical tax system of the US (Renneboog & Trojanowski, 2011).3 Brockman and Yan (2009) also find that stocks with higher concentrated ownership exhibit a higher probability of informed trading and higher unsystematic volatility, while Heflin and Shaw (2000) find a positive relation between concentrated ownership and quoted spread, and higher adverse selection costs.

Stock ownership and dividends
The association between share ownership structure and dividend policy continues to receive a great deal of attention from academics and practitioners.Miller and Modigliani (1961) 1996) shows that UK firms set their dividend policies to minimize (maximize) their tax liability (after-tax profit).Firms that are unable to deduct advanced corporation tax from their tax liability pay lower dividends.
dividends and institutional ownership for a panel of 211 UK firms for the period between 1988 and 1992.They claim that the effect of institutional ownership can also produce a positive earnings trend component.Renneboog and Trojanowski (2007) show that profitability drives dividend payout decisions of UK firms, and document a negative association between concentrated ownership and dividend dynamics.Grinstein and Michaely (2005) show that although institutional investors avoid investing in non-dividend paying firms, they are not more attracted to firms with higher-dividend distributions.Khan (2006)

Liquidity
For each stock, daily and monthly data are obtained on share prices, trading volume and shares outstanding to construct three measures of stock liquidity, namely the Amihud measure, liquidity ratio, and turnover 5 .The Amihud (2002) illiquidity ratio is a price impact measure of liquidity.It is defined as the ratio of absolute value of daily return and daily trading volume in GBP, averaged over the trading days of a given month: where i r is the daily stock return and i V is the re- spective daily volume in GBP.Second, the liquidity ratio (Amivest ratio) is used to measure the trading volume, which is associated with a unit change in stock prices.For each month, the liquidity ratio is calculated as the sum of daily shares traded to the sum of daily absolute stock returns: where i V denotes the daily trading volume, and i r is the daily stock return.Finally, turnover can be defined as the number of shares traded as a percentage of the total number of shares outstanding.Amihud and Mendelson (1986) find that liquidity is correlated with the trading frequency and argue that the turnover rate can be used as a proxy for liquidity: where i V is the monthly trading volume and i N is the total number of shares outstanding.

Stock ownership
Following Ding et al. (2016), this study uses the number of free float shares, which represents total ownership excluding ownership by government, corporations, key employees, and other strategic investors 6 .It is defined as the fraction of total shares available to the public for trade.Further, institutional ownership is measured by the percentage of shares held by investment banks or institutions.

Dividends
This study uses two measures of dividends.First, the dividend payout ratio (DPR) is measured as the ratio between dividends paid out by the firm and earnings, following La

Control variables
This study also includes a number of control variables that are considered to be significant factors in explaining the cross-sectional variation in stock liquidity and dividends.Based on prior literature (Rubin, 2007

Methodology
The following regression models are constructed to investigate the interactions among stock ownership, liquidity and dividends in the UK stock market over the period 2002-2016.The analysis is conducted using the Fama-MacBeth (1973) method, where the reported t-statistics are based on Newey-West (1987) heteroskedasticity and autocorrelation consistent standard errors: , , , , , where i Liq includes the Amihud ratio, liquid- ity ratio (Amivest), and turnover, i Finally, i Control represents firm-specific control vari- ables, which include stock price, size, financial leverage and asset tangibility.The Figure 1 presents the annual average percentage of shares held by free float across the sample period.Further, stocks are divided into three groups based on the fraction of total shares available to the public for trade: low (bottom three deciles), medium (middle four deciles) and high (top three deciles).

Summary statistics
The Figure 2 presents the annual average percentage of shares held by institutions across the sample period.Further, stocks are divided into three groups based on institutional holdings: low (bottom three deciles), medium (middle four deciles) and high (top three deciles).
Moreover, Panel B shows the correlation among liquidity, ownership and dividends.It can be seen that free float (institutional ownership) is significantly positively (negatively) correlated with liquidity, implying that higher free float (institutional holdings) is accompanied with greater (lower) levels of liquidity.In addition, institution- al ownership is positively and significantly correlated with dividends, but the free float-dividends correlation is insignificant.The correlation results, however, should be observed with caution, as they do not control for different factors that influence liquidity and dividends, such as stock price, firm size, financial leverage and asset tangibility.

Ownership and liquidity regression results
Table 2 shows the regression results of the influence of stock ownership on stock liquidity.Models 1 and 2 report the regression results using the Amihud measure as the dependent variable.The free float coefficient is negative and significant, which is consistent with the hypothesis that increased free float is accompanied with higher liquidity.This result is in line with When calculating the economic significance of the free float (institutional ownership) effect in model 1, it is observed that a 1 standard deviation increase in free float (institutional ownership) enhances (reduces) stock liquidity by lowering (increasing) the Amihud illiquidity measure by approximately 21.8% (5.3%) 8 .In addition, models 3 and 4 (models 5 and 6) report the regression results using the liquidity ratio (turnover ratio) as the dependent variable.In each of these models, the coefficients of free float (institutional own-  2016), stock liquidity is greater for larger stocks.This is because larger firms experience smaller adverse selection risks.Indeed, the effect of informed trading within larger firms can be reduced, since these firms tend to have more shareholders, and more information is available about them due to greater media and analyst coverage (Stoll & Whaley, 1983).Asset tangibility is positively linked with liquidity.This finding is in line with Gopalan et al. (2012) who demonstrate that asset liquidity increases stock liquidity, particularly for firms with low growth opportunities and that are less likely to reinvest their more liquid assets 10 .Finally, the leverage coefficient is positive, 9 Bekaert et al. (2007) find that illiquid assets and assets with high transaction costs trade at low prices relative to their expected cash flows.Brennan and Tamarowski (2005) point out that an increase in stock liquidity reduces a firm's cost of capital and increases its stock price.They suggest that a firm can reduce (increase) its cost of capital (stock price) through more effective investor relations activities, which reduce the cost of information to the market.10 According to Prommin et al. (2014), "tangible assets' payoffs are easier to observe and, therefore, will result in lower information asymmetry" (p.136).
implying that firms with higher levels of liquidity are more leveraged, as these firms view debt financing to be more attractive.

Ownership and dividends regression results
The results shown in Table 3 provide strong evidence of a positive association between institutional ownership and dividend payout policy measured by DPR and DPS.This finding is consistent with Allen et al. (2000), Khan (2006), Al-Najjar and Belghitar (2014) and Firth et al. (2016).The positive relation highlights the tax preferences of institutional ownership in favor of dividend income (Short et al., 2002), and that institutional investors can detect high firm quality.It is observed that a 1 standard deviation increase in institutional ownership increases dividend payout policy by increasing DPR (DPS) by approximately 2% (2.5%) in models 1 and 3, respectively.The negative but insignificant association between free float and dividends can possibly be explained by the fact that highly taxed individuals prefer holding stocks with low-or zero-dividends and trade out of firms that rise dividend payouts, as documented by Lee et al. (2006).Furthermore, the firm-specific control variables are found to be significant in explaining dividend payout policy.The stock price coefficient is positive and significant, suggesting that higher stock prices are associated with higher dividend payouts.This finding is inconsistent with Allen et al. (2000) who find that dividend-paying firms tend to have lower stock prices compared to non-dividend paying firms.Firm size coefficient is positive and significant, and is in line with the transaction cost theory of dividends.Jain (2007) and Al-Najjar and Belghitar (2014) find that larger firms are more likely to be well-established and have more ability to pay dividends due to their lower earnings volatility.A positive relation exists between financial leverage and dividend payout policy, and is in line with the signalling theory (Chang & Rhee, 1990).This finding suggests that firms that are highly levered are also are more inclined to pay out a higher portion of earnings as dividends.Finally, asset tangibility is found to have a positive impact on dividend payouts.This finding is inconsistent with the agency theory of dividends, which proposes that firms with greater tangible assets have less current assets that can be used to borrow against (Al-Najjar & Belghitar, 2014).Indeed, if a large percentage of a firm's assets are tangible, then these assets should serve as collateral, thereby reducing the risk of lenders suffering the agency costs of debt (Rajan & Zingales, 1995).
11 According to Brockman et al. (2008), "when stock liquidity is low, managers are reluctant to share repurchases because their market transactions could increase the price impact of trading" (p.446).

Liquidity and dividends regression results
The final section examines the relation between liquidity and dividends.Banerjee et al. (2007) argue that in markets with trading friction, dividend-paying firms enable investors to satisfy their liquidity needs with minimum-or no-trading and, thus, allow them to avoid trading friction.Consequently, investors with future liquidity needs might favor dividend-paying firms.This preference will be positively associated with the level of trading friction, so that higher (lower) trading friction will result in higher (lower) demand for dividends.  1.They claim that value-maximizing managers will seek the payout method that transaction and informational costs.

CONCLUSION
A great deal of attention has been paid to free float in recent years for its implications on market liquidity and corporate governance.However, little attention has been devoted to examining the free float-liquidity relation.(2016), this study provides strong evidence suggesting that higher levels of free float improves liquidity.This relationship is significant regardless of the liquidity measure employed, and holds after controlling for different firm-specific factors acknowledged in the finance literature to be related to stock liquidity, such as stock price, size, financial leverage and asset tangibility.Further, it is observed that a rise in institutional ownership is associated with poorer liquidity.Institutional ownership reduces liquidity of domestic firms through reduced trading activity.As pointed out by Ng et al. (2015), foreign institutional investors increase the degree of asymmetric information between firms and outsiders, and that the effects of asymmetric information influence stock liquidity.Moreover, Short et al. (2002) argue that there are strong incentives for tax-exempt institutions to demand higher dividends due to the bias in the UK tax system, which favors dividend income over capital gains.This study investigates the ownership-dividends relationship, and finds that institutional ownership is positively associated with dividend payout policy, and, hence, provides strong evidence of the comparative tax advantages that UK institutions have for dividend income.
payout ratio (DPR) and dividend per share (DPS) for firm , i i FF represents the percentage of shares held by free float in firm , i and i INS is the percentage of shares held by in- vestment banks or institutions in firm .
Ding et al. (2016)8To measure the economic magnitude, the coefficient of free-float in Model 1 is multiplied by the standard deviation of free float (-1.186 x 0.211 = -0.25).Thus, an increase in free float by 1 standard deviation reduces theAmihud (2002) illiquidity by 0.25.Since the average illiquidity is 1.146, a change by 0.25 represents 21.8% of the average.The same procedure is used to calculate the economic magnitude of institutional ownership.who also document a positive free float-liquidity relation.Further, an increase in institutional ownership is accompanied with lower levels of liquidity.This is consistent with Rhee and Wang (2009) and Jiang et al. (2011), among others, who document an inverse relation between institutional investors and liquidity.

Table 1
displays the summary statistics.As shown in Panel A, all three liquidity measures have a positive mean, which is significant and different from zero.The sample's average percentage free float is 7For comparison, Figure2presents time-series plots of the average percentage of shares held by institutions.It is observed that the percentage of institutional holdings has decreased between 2004 (34%) and 2005 (11%) and has gradually decreased thereafter.

Table 1 .
Summary statistics Note: This table presents the summary statistics for liquidity measures, ownership, dividends and control variables.The sample consists of the FTSE 100 Index constituents over the period from May 2002 to December 2016.Panel (A) displays summary statistics using monthly observations.Panel (B) reports the correlation table.* Denotes significance at 1% level.

Table 2 .
Impact of ownership on liquidityThis table shows the estimates of cross sectional regressions of liquidity measured by the Amihud illiquidity ratio, liquidity ratio, and turnover on share ownership (free float and institutions) and on other control variables.The sample includes the FTSE 100 Index constituents over the period from May 2002 to December 2016.* (**) represent significance at the 1% and 5% levels.

Table 4
This table displays the estimates of cross sectional regressions of dividends measured by DPR and DPS on free float, institutional ownership and on other control variables.The sample includes the FTSE 100 Index constituents over the period from May 2002 to December 2016.* (**) Represent significance at the 1% and 5% levels.
Ding et al. (2016)glu (2008)andDing et al. (2016)argue that firms with more shares in free float can alleviate information asymmetry problems, which, in turn, can enhance liquidity.This study examines the relationship between liquidity and free float using a sample of 15,650 firm-level observations in the UK stock market (FTSE 100 Index constituents) over the period from May 2002 to December 2016.The findings are important, because they contribute to the existing literature in several areas, mainly corporate finance and market microstructure.Consistent with the findings of Ding et al.

Table 4 .
Dividends and liquidityThis table shows the estimates of cross sectional regressions of dividends measured by DPR and DPS on liquiditymeasured by the Amihud illiquidity ratio, liquidity ratio, and turnover.The sample includes the FTSE 100 Index constituents over the period from May 2002 to December 2016.* (**) represent significance at the 1% and 5% levels.