The stickiness of payout channels

(1)

The stickiness of payout channels

Author:

Wenwen Yang

Date: August 2011

Supervisor: Dr. H. Von Eije

Second evaluator: Dr. H. Gonenc

(2)

The stickiness of payout channels

Abstract

This study investigates the stickiness of payout channels. With a panel of 7113 listed firms in the US, I run multinomial regression analyses to answer the research questions: 1. Do sticky payout channels exits? 2. Which kind of firms stick to their payout channels? I show that stickiness of the payout channel does indeed exist. Furthermore, the likelihood to stick to pay dividends reduces, but sticking to repurchases rises over time. In addition, profitability and size are significantly positively and growth is negatively related to the stickiness payout decision, which indicates that larger, more profitable and less growth firms are likely to payout both in the previous and the current year. What’s more, the impact of size, profitability and growth on the decision to pay out in prior and current year is stronger than the impact on the decision to stick to paying only dividends. Also the 2003 tax cut does have an impact on the payout channel. Additionally, the influence of size, profitability and growth on the sticky firms is stronger than on the non-sticky firms. However, the result shows that investment opportunities only positively affect the decision to stick to pay only dividends, while it has no impact on the decision to stick to repurchase and stick to both channels.

JEL Classification: G35

Key words: Payout policy, Cash dividends, Repurchase, stickiness

(3)

1. Introduction

In this article, I focus on the stickiness of the payout channel. Miller and Modigliani (1961) developed the dividend irrelevance principle and founded the bedrock for the modern financial payout theories. The irrelevance principle states that in the frictionless market, the value of a firm is not affected by the firms’ payout policy. In other words, it is the investment policy alone which determines the shareholder wealth and dividend policy has no impact on the firm value. Nevertheless, much literature provides empirical descriptives and theories on payout policies. Lintner (1956) was the first to examine the payout policy by looking at the payout situation of a prior year. He interviewed managers from 28 companies and concluded that dividends are sticky. Managers decided to reduce dividend only when they have no other choice, and increase dividend only when they are confident that the future cash flows sustain the new dividend level. He concluded that firms that paid dividends last year may be much more likely to do so again this year and firms that do not pay dividends last year prefer not paying them this year. Almost fifty years later, Brav, Graham, Harvey, and Michaely (2005) survey 384 financial executives and conduct in depth interviews to find out the determinants of paying dividend and repurchasing. Like Linter, they found that maintaining a sustainable level of dividend is still a crucial consideration for managers when they make the dividend decisions, while repurchases are made out of residual cash flow after the investment spending. In addition, Mia Twu (2010) also shows that the prior status is an important determinant for the likelihood to pay cash dividends, meaning that dividend stickiness does still exist.

All of these papers find stickiness in terms of cash dividends. Nonetheless, recently published research shows that repurchases, as a more flexible and tax-favored method to distribute cash to shareholders, have been used by a growing number of firms. As repurchases are becoming more and more important when managers make their payout decision, there are four kinds of payout channels mainly used by firms (Renneboog and Trojanowski (2010)), which are (i) do not pay out anything, (ii) only pay out by cash dividends, (iii) only pay out by repurchases, (iv) pay out by both cash dividend and repurchases. I am curious to find out whether stickiness also exists among these four channels like it is the case for cash dividends.

(5)

sticky payout channels exist? And if so, 2. Which kind of firms stick to their payout channels, in other words, which characteristics determine the stickiness of the payout channel?

This paper complements the existing literature in two major ways: firstly, many prior studies examine payout policy and topics related to payout policy, but no published paper investigates the stickiness of payout policy channels (except for that of dividends). Secondly, as the investigation period of this paper is from 2000-2008, it covers the 2003 tax cut.The Jobs and Growth Tax Relief Reconciliation Act of 2003 was signed into law in late May 2003, which introduced favorable tax treatment of individual dividend income. The tax rate on dividend income decreased from 38.1% to 15%, which is also the top tax rate on long-term capital gains. Julio and Ikenberry (2004) and Chetty and Saez (2006) illustrate that although the tax cut cannot explain the whole story of payout policy after 2003, it indeed influences firms’ payouts to some extent. This paper evaluates this aspect on the stickiness of the payout channels.

(6)

2. Literature review

2.1 Signaling theory

Miller and Modigliani’s (1961) dividend irrelevance principle shows that the value of firms is not influenced by the firms’ payout policy. Miller (1986), however, questioned this theory by asking why the stock market reacts to the dividend change if the payout policy is truly irrelevant. As he described in his paper, he explained that the stock market is not reacting to the payout change but rather to the information of future earnings. When the dividends indeed covey some information about the earnings information, it may be used as a costly signal available to managers to convey the information about the firms’ future prospects (e.g., Bhattacharya (1979), Miller and Rock (1985), John and Williams (1985), Allen et al. (2000)). Due to asymmetric information, investors have only limited and incomplete information regarding the firms and this makes them looking for information as much as they can to better make their investment decisions. On the other hand, insider managers have more information than investors about the firms and this information can affect their payout polices. According to the signaling model, good firms initiate and increase dividends or repurchase in order to distinguish themselves from bad firms. They may be doing so, because it is difficult for bad firms without good projects to mimic the actions taken by the good firms.

Share repurchases, are another way of distribute free cash flow. It is also perceived by investors as a credible signal. When managers believe that the currently low stock price is the result from the incorrect perception of market about the firms’ future prospect, they would repurchase shares to signal that the stock price is below the fundamental value (Wansley, Lane, and Sarkar (1989), Brav et al. (2005)). If investors recognize this signal and decide to buy the stocks, share prices are likely to increase quickly. However, repurchase do not signal completely the same information as dividend. Dividend are distributed to investors to signal that the firms have many good project (NPV>0), while repurchase signal that firms had few good projects and are willing to pay out the cash flow to investors. This is also in line with the findings of Guay and Harford (2000); and Brav et al. (2005), who state that dividend and repurchase are not perfect substitutes and that firms use the two types of payout in different situations. Repurchases are more flexible than dividend and they are likely to be used to pay out transient cash flows.

2.2 Agency theory

(7)

conditions, management may not act for the best interest of shareholders and may choose not to pay out full free cash flows. In other words, they would retain some free cash flow and allocate resources to their own benefit. In this sense, the goal of the agent (managers) and the goal of the principal (shareholders) are not the same, which leads to the so-called “agency problem”. In practice, one ways to solve this problem is to pay out excess cash to shareholders which reduces the free cash that managers can use. The less cash managers can use, the less opportunities they have to lavish excess cash for their own interests. But why would firms pay out through dividends rather than repurchase, since repurchase is a relatively cheaper and more flexible way to distribute free cash flows. The earlier work of Shleifer and Vishny (1986) and the more recent work by Allen, Bernardo and Welch (2000) provide some explanation for this question. According to Shleifer and Vishny (1986) and Allen, Bernardo and Welch (2000), managers have to be monitored due to the agency problem and this can best be done by large shareholders. Then the presence of these large shareholders will benefit board and increase the value of firms due to their monitoring. Hence boards are willing to attract more large shareholders to buy their shares, especially when there is much excess cash flow in the firms. From the perspective of the larger shareholders, they prefer to receive cash dividends instead of repurchases given the favorable dividend tax rate for large shareholders such as institutions or corporation compared to individuals and because the dividends restrict managers more.

Another conflict that will affect the decision of payout policy is between bondholders and stockholders. Based on the result of Myers (1977), in some circumstances stockholders expropriate wealth from bondholders in the form of excessive cash dividend. They try to cut investment opportunities and use the cash which should be used in the project to pay and to amortize debt.

2.3 Clientele effect

(8)

minor influence on the payout policy. Nonetheless, tax difference among investors cannot be ignored because most clientele theory focuses on the impact of personal tax on the payout policy. Tax clientele theory ((Shefrin and Statman (1984)) show that tax minimization is an important factor for managers to consider in making payout policy. Thus different policies are adopted in order to attract different tax clienteles. If most of shareholders in a given firm face relatively high tax rates on dividend income, the firm will be inclined to repurchase stocks rather than dividends to avoid the disadvantage of dividend. When a firm is primarily held by investors who face low tax rate, it may, however, pay out through cash dividends.

On the other hand, Black and Scholes (1974) argue in their paper that firms can actually increase firm value by choosing their payout policy to meet the preference of a specific clientele. They indicate that firms are indeed able to adjust payout policy to investor’s needs and increase firm value, but this increased value will, however, disappear soon because other firms can also use the same policy. Hence, there is an equilibrium in which no firm is able to alter firm value by changing their payout policy. In addition, they figured out the fact that the clientele theories are incompatible with the observed homogeneity in real world. This is also in line with the findings of DeAngelo et al. (2004). Furthermore, much other research found no reaction in institutional and individual ownership to a firm’s payout policy change.Brav and Heaton (1998) find that the number of institutional stockholders declined by 18.1% after dividend omissions, but this rebound to a 6.5% decline after three years, and neither decline is significantly different from zero. Similarly, Hoberg and Prabhala (2008) report that there is no significant relationship between dividend increases and institutional ownership change.

Hence, I can conclude that the heterogeneous investor tax clientele demands are not a major determinant of firm payout decisions. But there may be exceptions, since some firms with a controlling shareholders or a dominant stockholder group will inevitably choose the payout policy which may not be value maximization but satisfy their own preference.

2.4 catering theory

(9)

Baker and Wurgler’s (2004) catering theory explains the observed U.S. payout patterns from the perspective of behavioral arguments. They assert that the investors’ demand influences the companies’ payout decision. Simply put, firms supply dividends to cater investors when investors assign premium to dividend-paying stocks in some period, and they do not pay when investors put a stock price discount on payers in other period. Li and Lie (2006) extend Baker and Wurgler’s (2004a) catering theory and similarly find that the change of dividends and the magnitude of the change indeed have relations with the investors’ sentiment for dividends. Baker and Wurgler’s (2004b) show the trend in propensity to pay dividends as Fama and French (2001), and also prove that the propensity to pay depends on the stock market dividend premium. In a similar vein, Ferris et al. (2006) attribute the recent changes in U.K. payout policy to the shifting in catering incentives. In contrast, Denis and Osobov (2008) exame the dividend policies in the United States, United Kingdom, Canada, France, Germany, and Japan and find little evidence support for the catering theory. Also, von Eije and Megginson (2008) report that catering theory cannot explain the European dividend policies.

2.5 Life-cycle model

Miller and Modigliani’s (1961) dividend irrelevance theorem documents that in the frictionless market all feasible capital structure are optimal if investment policy is given and 100% free cash flows are pay out. DeAngelo and DeAngelo (2006) questioned Miller and Modigliani’s model and develop their own full payout model. By relaxing the assumption of Miller and Modigliani’s model to allow retention, they concluded that investment policy is not the sole determinant of firm value but payout policy should also be taken into account in order to maximize the firm value. This full payout model lays the groundwork for the life-cycle theory.

(10)

changes over time as firms’ earnings and investment opportunities change. Young firms which have plenty of investment opportunities and less profits prefer to retain cash flows to fund growth opportunities. As firms mature, more earnings will be generated while the investment opportunities stabilize or even perhaps fade. In this case, the optimal payout policy is to retain sufficient cash flows to fund investment projects and use the excess cash flows to pay out to shareholders.

2.6 Payout and earnings

(11)

relationship between earnings and overall firm payout is now stronger than the relation between earnings and dividends. This means dividend policy becomes increasingly conservative over time to minimize the possibility of dividend cuts and omission. What’s more, they also proved for firms that exclusively use repurchases, earnings are more likely to explain the repurchase in a manner similar to the traditional relation (Linter) between dividends and earnings.

Besides, Brav et al. (2005) find the managers are more likely to repurchase instead of paying dividend if they initiate payout to shareholders. Consistent with this finding, Grullon and Michaely (2002) illustrate that the number of firms which choose share repurchase to initiate payout is more than that firms that choose dividends as the first type of payout. Bulan et al. (2007) however figure out that more firms initiated dividends in 1990s compare to 1980s, but these firms tend to be large and profitable. In a word, young firms which generally have only temporary profitability prefer repurchase initiations while mature firms with consistent profitability and low growth rates are inclined to use dividend initiations.

2.7 Are dividends disappearing?

Fama and French (2001) illustrate that the period between 1978 to 1999 witnesses a continuously decline of the proportion of firms paying cash dividend, as 66.5% firms paid cash dividends in 1987 and only 20.8% in 1999. Consistent with Fama and French, DeAngelo et al. (2004) find that the number of dividend-paying firms declines. They, however, also show that aggregate dollar dividends exhibit an increase in recent years. For the European Union, Von Eije and Megginson (2008) also show a declined propensity to pay cash dividends and a surge in the total real dividends paid and in the number of firms which repurchase.

(12)

paid dividends or repurchased stock say that their firms may never do so. What’s more, Lie (2005) find that the firms with low earning volatility are more likely to increase payout than firms with high earning volatility. This is also one of the explanations for the decreased propensity to pay dividend since high earning volatility in recent years make young firms reluctant to initiate cash dividend.

However, some evidence shows that cash dividend may be “reappearing” in the United State. According to the finding of Julio and Ikenberry(2004), the proportion of dividend payer rise from 15% in 2001 to around 20% in 2004. They argue that this rebound partly results from the 2003 Bush tax cut and partly due to remarkable number of IPO firms during 1990s becoming mature in recent years.

2.8 Pattern in payout policy and payout channel choice

(13)

3 Methodology

Renneboog and Trojanowski (2010) cover four groups of firms in their paper, which are B1, B2, B3 and B4 respectively shown in table 1. They investigate the payout channel in a particular year Regardless of the situation in the prior year.

Table 1 comparison of Renneboog and Trojanowski’s (2010) groups and sticky payout channel groups in a matrix that covers the possible choices in two consecutive periods. (t and t-1)

t t-1

pay nothing Only pay dividend Only repurchase Both pay dividend and repurchase Pay nothing A1

Only pay dividend A2

Only repurchase A3

Both pay dividend and repurchase

A4

Renneboog and Trojanowski

B1 B2 B3 B4

Unlike Renneboog and Trojanowski (2010), I focus on the payout channel on both this year and last year, which means that the payout channel of this year is the same with that of last year. I call this stickiness of the payout channel and there are A1, A2, A3 and A4 respectively in table 1. Multinomial regression (Brooks, 2008) is used to explain the likelihood that a firm prefers a particular payout channel (group B1, B2, B3 and B4) and whether a firm sticks to a particular payout channel (group A1, A2, A3 and A4).

Model 1

(14)

Payout channel it=β0+ β1(Size)it-1+ β2(Profit)it-1+ β3(Leverage)it-1+ β4(Mbf)it-1+ β5(Growth)it-1

+ β6 (Tax)it + β7(Year)it+ εit (1)

In equation 1 the following meaning is attached to the variables:

Payout channel it :the dependent variable of regression, taking 0 if a firm does not pay out

anything, 1 if a firm pays out only dividends, 2 if a firm only repurchases and 3 if a firm uses both payout channels(all in year t).

(Size) it-1: the natural logarithm of the total assets of firm i in year t-1.

(Profit) it-1: the earnings ratio of the firms defined as the earnings before interest but after tax

divided by the total assets in year t-1.

(Leverage) it-1: the ratio of total debt divided by total assets for firm i in year t-1.

(Mbf) it-1: the natural logarithm of the ratio of market capitalization divided by common equity

for firm i in year t-1.

(Growth) it-1: the relative change of total assets for firm i in year t-1.

(Tax) it:dividend tax equals to 0.38 before year 2003 and 0.15 after 2002.

(Year)it: year equals to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 for year 1999, 2000, 2001, 2002, 2003, 2004,

2005, 2006, 2007, 2008 respectively.

Model 2

Secondly, to test the likelihood that a firm sticks to a particular payout channel (group A1, A2, A3 and A4), the multinomial regression includes five possible value of the dependent variable: (i) not sticking to the payout channel of the prior year, (ii) not paying out anything in the current and prior year, (iii) paying out only dividends in the current and prior year, (iv) only repurchase in the current and prior year, (v) use both payout channels in the current and prior year.

Payout channel it=β0+ β1(Size)it-1+ β2(Profit)it-1+ β3(Leverage)it-1+ β4(Mbf)it-1+ β5(Growth)it-1

+ β6(Tax)it + β7(Year)it +εit (2)

In equation 2 the following meaning is attached to the variables:

Payout channel it:the dependent variable of regression, taking 0 if the firm does not stick to the

(15)

the firm do not pay out anything in the current and prior year, 2 if the firm pays out only dividends in the current and prior year, 3 if the firm only repurchase in the current and prior year and 4 if the firm use both payout channels in the current and prior year.

(Size) it-1: the natural logarithm of the total asset of a firm i in year t-1.

(Profit) it-1: the earnings ratio of the firms defined as the earnings before interest but after tax

divided by the total assets in year t-1.

(Leverage) it-1: the ratio of total debt divided by total assets for firm i in year t-1.

(Mbf) it-1: the natural logarithm of the ratio of market capitalization divided by common equity

for firm i in year t-1.

(Growth) it-1: the relative change of total assets for firm i in year t-1.

(Tax) it : dividend tax equals to 0.38 before year 2003 and 0.15 after 2002.

(Year)it : year equals to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 for year 1999, 2000, 2001, 2002, 2003, 2004,

(16)

4 Data

4.1 Sample selection

Our sample covers all United State firms listed on AMEX, NYSE and NASDAQ. The investigation period is from 2000 to 2008. All the independent variables are lagged one year to reduce the endogeneity problem, so they are collected from 1999 to 2007. I exclude banks, insurance companies, financial firms and utilities following Trojanowski and Renneboog (2005), which result in a sample of 10978 firms. Banks, insurance and financial firms are excluded because their financial reporting standards are different. Utilities are also excluded due to the regulation of their payout policy and the access to external financing. What’s more, the sample excludes firms whose market-to-book values is less than 0, whose leverage is less than 0, whose leverage is larger than 1 and whose size is equal to 0. All of these decisions result in a final sample of 7112 firms and a total of 64008 observations. The variables are found in the DataStream database, except the variable share repurchase which is derived from Worldscope. An overview of all variables and their corresponding code are shown in Appendix A.

4.2 Dependent variables

Following Fama and French (2001), I measure the repurchase as net repurchase. If the firms use the treasury stock method for repurchase, I measure the repurchase as the increase in common treasury stock. If the firms’ treasury stock is zero in the current and prior year, which represents the “retirement” method, I measure the repurchase as the difference between stock purchases and stock issuances. If either the increase in common stock or the difference is negative, I set repurchase to zero.

To find out the likelihood that a firm opts form a particular payout channel in a particular year (group B1, B2, B3 and B4), the multinomial regression consists of four dependent variables, which stand for the four mutually exclusive choice that a firm make: (i) not pay out anything, (ii) pay out only dividends, (iii) pay out only through repurchases, (iv) using both payout channels. I code these outcomes as 0, 1, 2, and 3, respectively.

(17)

paying only dividends in the current and prior year, (iv) only repurchase in the current and prior year, (v) use both payout channels in the current and prior year.

4.3 Independent variables

In this section, I will present the independent variables, namely size, profitability, leverage, investment opportunity, growth, tax and year. I will illustrate a summary of hypothesis in table 2. For descriptive statistics of the variables, I will refer to table 3. Also, I will indicate that the independent variables will be used to explain both the channel and the stickiness of the channel and the also give the expected relationships.

Size

Based on the life-cycle model, the trade-off between retention and payout changes over time as firms’ earnings and investment opportunities change. Large firms which have profits prefer to payout to shareholders. Fama and French (2001) and Von Eije and Megginson (2008) show that size plays an important role in firms’ payout policy. Similarly, Renneboog and Trojanowski (2010) illustrate that size are positively related to likelihood to pay out to shareholders. This is similar in both UK market and US market. Hence, size is proxy by total asset.

I expect that size is positively associated with the likelihood to payout to shareholders (B1-B4), but this relationship is less strong than for sticky firms (A1-A4). I refer to the first part of table 2 for the signs and magnitudes of channel choice. For the sticky firms, size is expected to positively associated with the likelihood to stick to only pay cash dividends (A2) and stick to use both channels (A4) and the association with the latter (A4) is stronger. For the sticky channels I refer to the second part of table 2 for the expected signs and magnitude.

Profitability

As also a traditional life-cycle variable, profitability is positively related to the likelihood to pay according to the life-cycle theory. Von Eije and Megginson (2008) show significantly positive relation between profitability and likelihood to distribute to shareholders by cash dividends or repurchase. Skinner (2008) also indicates that profitability is positively associated with the likelihood to repurchase or total payout mix. So I define profitability as earnings ratio of divided by the total asset.

(18)

(A1-A4) than for non-sticky firms (B1-B4). For the sticky firms, I expect the profitability is positively related to stick to only pay cash dividends (A2) and use both channels (A4) and the association with the latter (A4) is stronger. I refer hypothesis of sticky firms to the second part of table 2.

Table 2 summary of hypotheses

“Non-paying” for the non-sticky groups means firms not payout anything in the current year, “only dividend” for the non-sticky groups means firms pays out only dividends in the current year, “only repurchases” for the non-sticky firms means firms only repurchase in the current year, “both dividend and repurchases” for the non-sticky firms means firms use both payout channel in the current year. “Non-paying” for the sticky groups means firms not payout anything in the current and prior year, “only dividend” for the stick firms means firms pays out only dividends in the current and prior year, “only repurchases” for the sticky firms means firms only repurchase in the current and prior year, “both dividend and repurchases” for the sticky firms means firms use both payout channel in the current and prior year. Firm size is defined as the natural logarithm of the total asset of the firm. Profitability is defined as the earnings before interest but after tax divided by the total asset. Leverage is defined as the natural logarithm of the ratio of total debt divided by total asset for the firm. MBF is defined as the natural logarithm of the ratio of market capitalization divided by Common equity. Growth is defined as the relative change of total assets.

Non-sticky groups Sticky groups

Non-paying (B1) Only dividend (B2) Only repurchases (B3) Both dividend and repurchases (B4) Non-paying (A1) Only dividend (A2) Only repurchases (A3) Both dividend and repurchases (A4) Size - + + + -- ++ ? +++ Profitability - + + + -- ++ ? +++ Leverage - + + + -- ++ ? +++ Investment Opportunity + - - - ++ -- ? --- Growth + - - - ++ -- ? --- Leverage

High leverage may mitigate the agency problem, which reduces the need for managers to payout to shareholders through cash dividends and repurchases. Furthermore, firms with high leverage may be the ones that are larger, more mature and more profitable that they are able to payout to shareholders. Fama and French (2001) find that higher payout ratios of dividend payers are related to higher book leverage in the US market. However, Von Eije and Megginson (2008) find a negative relationship between these two variables. Leverage is measure as the total debt divided by total asset.

(19)

sticky firms (A1-A4 in the second part of table 2). For the sticky firms, leverage is expected to positively associate with the likelihood to stick to only pay cash dividends (A2) and stick to use both channels (A4) and the association with the latter (A4) is expected to be stronger.

Investment opportunity

Life-cycle theory document that firms with abundant investment opportunities may not payout to shareholders but to retain cash flows to fund investment. In other words, there is negatively relationship between investment opportunities and likelihood to payout. Renneboog and Trojanowski(2010) use Tobin’s Q to represent investment opportunities and finding that firms are more likely to pay out with less investment opportunities. Von Eije and Megginson (2008) also verify that market-to-book ratio is significantly and negatively related to likelihood to pay out. Accordingly, Investment opportunity is measured as market-to-book ratio of the firm, which is de fined as market capitalization divided by common equity.

I predict that investment opportunity is negatively associated with the likelihood to payout to shareholders (B1-B4 in the first part of table 2) but this association is not as strong as for the sticky firms (A1-A4 in the second part of table 2). For the sticky firms, investment opportunity is expected to negatively associate with the likelihood to stick to only pay cash dividends (A2) and stick to use both channels (A4) and the association with the latter (A4) is stronger.

Growth

As life-cycle theory predict, retention dominates distribution if firms are young and grow rapidly. Fama and French (2001) find asset growth is negatively associated with high likelihood to pay cash dividends. Von Eije and Megginson (2008) indicate firms with less growth rate tend to pay out cash dividends or repurchase to shareholders. Coulton, Jeff J. and Ruddock, Caitlin M. S. (2010) prove the negative relationship between asset growth and probability to payout. Therefore, I use asset growth rate to measure growth.

(20)

Tax

According to the clientele theory, tax difference among investors influence the payout policy to some extent. Our investigation period covers 2003, in which a more favorable tax treatment of individual dividend income was introduced. Julio and Ikenberry (2004), Chetty and Saez (2006) illustrate that although the tax cut cannot explain the whole story of payout policy after 2003, it indeed influences firms’ payout channel to some extent. Hence, we add the tax variable (tax=0.38 before 2003 and tax=0.15 after 2002) to find out whether it indeed has impact on the payout policy.

Year

Besides the variables mentioned above, I also include the year variables to find out whether the payout channel changes over time. I set year equals to 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 for year 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, and 2008 respectively.

Table 3 shows the descriptive statistics variables for the whole sample. The difference between cash dividends and repurchases are quite large, which proves that although repurchases are becoming important in recent years, cash dividends still account for a large part in the total payout. The average level of growth rate is larger than 1, meaning that generally firms are growing. However, the profitability is negative on average, which suggests that firms performed poorly over the period 2000 to 2008.

Table 3 descriptive statistics of variables

Descriptive statistics variables for the whole sample. “CD” is the cash dividend. “Repurchases” is the amount of repurchases. “Size” is the natural logarithm of total assets. “Profitability” is the earnings before interest but after tax divided by the total assets. “Leverage” is the ratio of total debt divided by the total assets. “Investment opportunity” is the natural logarithm of market capitalization divided by common equity. “Growth” is the asset growth rate.

Variable Mean Median Max Min Std.Dev

(21)

5 Results

5.1 Descriptive statistics

Before proceeding to a more detailed analysis of the payout policy of US firms, I briefly analyze some general patterns and trends in dividends and share repurchases between 2000 and 2008.

Figure 1. Proportions of firms in different groups between 2000 and 2008.

Figure2. Proportion of firms initiating cash dividends and repurchases.

As can be seen from figure 1, the proportion of firms that pay dividends grows gradually between 2000 and 2008, which is consistent with the “reappearing” evidenc (Julio and Ikenberry (2004), Chetty and Saez (2006)). The start of the 21st century was a highly unusual time in U.S. economic history. The escalation of stock prices in the late 1990s and the NASDAQ plunge that began in March 2000 made investors subject to a crisis of confidence. In this case, paying dividends, a way signal well-managed companies and management’s optimism about future performance, and this can attract more investors. On the other hand, the maturity hypothesis is also one of the reasons why firms are more likely to pay dividend. The 1990s saw large waves of

0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 2000 2001 2002 2003 2004 2005 2006 2007 2008

(22)

IPOs of technology and internet companies. These firms, though small, have larger investment opportunities. Much research shows that during 1990s these firms retain their funds to improve growth opportunities instead of paying out. However, after 10 years, they becoming mature and the value of their growth opportunities shrink. It is time for them to rethink their payout policy and many firms decide to initiate dividend or increase the amount of dividend. Thirdly, the Bush tax cut of 2003 is critical to explain the increased trend of paying dividends. It is well known that the cash dividends were originally higher taxed than share repurchases, which make managers prefer repurchases. The 2003 tax cut lowered the marginal rate on dividends to that of capital gains, thus reducing the dividends tax disadvantage.

Unlike firms which only pay cash dividends, the proportion of firms that only repurchase decreases to around 30% at 2004, after which it rises continuously until 2008. Although repurchase is also a way to signal the firms’ future prospects, dividends may be viewed by investors as a more reliable signal. Moreover, many firms that announce buyback programs end up buying back no shares at all (Julio and Ikenberry(2004)). For this reason, while a repurchase may be a credible indicator of management’s optimism about near-term performance, dividends may be interpreted by investors as a sign that management expects stability over the long run. Furthermore, due to the crisis of confidence at the beginning of the 21st century, investors may prefer receiving cash dividends also because dividends are cash in hand and make investors feel safer about the program. So Firms are more willing to pay cash dividend and the number of firms repurchasing declined. Repurchase initiation can also partly explain the trend of firms to only repurchase. From figure 2, before 2004 the number of firms initiating payout through repurchases decreased, but after 2004 it increases again.

Figure 3. Proportions of firms in different groups between 2000 and 2008.

0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 2000 2001 2002 2003 2004 2005 2006 2007 2008

only cash dividend only repurchase

(23)

According to figure 3, the proportion of firms that do not pay out decreases continuously, from 48.33% in 2000 to 35.65% in 2008, which indicate that there is growing number of firms choose to pay out to shareholders. As for firms that only pay dividends (figure 3), the trend is similar with that of the firms which pay dividends (figure 1), except after 2004 the proportion of firms only paying dividends reduces to around 20%. This is because firms that only paid dividends in the prior year choose to pay both instead of only cash dividends after 2004, which may prove that firms are more willing to repurchase to some extent. Furthermore, the declined proportion of firms that initiate dividends after 2004 can also partly explain the story. On the other hand, with regard to the firms that only repurchase the change for the whole period is generally the same with the firms that repurchase in table 1. It is worth noticing that the proportion of firms that repurchase is larger than that of firms paying dividends after 2007, implying that repurchases are used by more and more firms.

5.2 The Existence of payout channel stickiness

In this section, I will show the number of firms choosing certain payout channel in a particular year (group B1, B2, B3 and B4 of table 1) and the number and proportion of firms in different groups (group A1, A2, A3 and A4 of table 1).

Table 4 the Number of firms which choose a particular payout channel in each year.

(24)

Table 5 Number of firms sticking to their prior payout channel in each year.

“No payout” means that firms payout nothing in the current and prior year. “Dividend only” means that firms only pay dividends in the current and prior year. “Repurchase only” means firms only payout through repurchases in the current and prior year. “Both dividend and repurchase” means that firms use both channels in the current and prior year. “Change channels” means the channel which firms use in the current and prior year is different. “Total sticky firms” is the sum of firms which stick to payout channel of prior year. “Firms absent” means firms were absent in the current year. “New firms” are the firms which is available in the current year but were absent in the prior year. “Sum” is the total number of firms in the current year. The percentages corresponding to “no payout”, “dividend only”, “repurchase only” and “both dividend and repurchase” are the ratio of the number of firms which stick to the payout channel divided by the total number of firms which use the same channel last year. The percentage of total sticky firms is the ratio of the total sticky firms divided by the total number of firms in the previous year. The percentage of the change channel is the ratio of the number of firms which change payout channel this year divided by the total number of firms in the previous year. The percentage of the firms absent is the ratio of number of firms which were absent in the current year divided by the total number of firms in the previous year. No payout Dividend only Repurchase only Both dividend and repurchase Total sticky firms Change channels Firms absent Sum 2000 1409(68%) 422(55%) 397(51%) 417(63%) 2645(62%) 984(23%) 629(15%) 3996 2001 1329(68%) 439(66%) 353(50%) 354(54%) 2475(62%) 1001(25%) 520(13%) 3795 2002 1299(70%) 476(66%) 342(49%) 292(57%) 2409(63%) 1000(26%) 386(10%) 3723 2003 1327(72%) 515(69%) 331(49%) 304(67%) 2477(67%) 864(23%) 382(10%) 3584 2004 1259(72%) 519(71%) 230(41%) 358(67%) 2366(66%) 843(24%) 375(10%) 3505 2005 1218(71%) 515(65%) 240(54%) 355(66%) 2328(66%) 754(22%) 423(12%) 3362 2006 1090(69%) 447(60%) 266(58%) 415(72%) 2218(66%) 795(24%) 349(10%) 3324 2007 1028(69%) 392(59%) 337(64%) 488(75%) 2245(68%) 776(23%) 303(9%) 3279 2008 859(63%) 353(58%) 397(67%) 493(69%) 2102(64%) 853(26%) 324(10%) 3144

(25)

353 out of 709 firms only repurchase as they did last year(50%), 439 out of 668 still paid cash dividends(66%), 354 out of 661 still used both dividend and repurchase(54%) (table 5). The total number of firms sticking to payout channel is 2457(62%). These firms all stick to their payout channel. Based on table 5, the average proportion of four groups of firms sticking to their prior payout channel is above 50%, which means that the payout stickiness does exist.

5.3 Characteristics of firms which stick to the payout channel

Some firms which do not stick to their payout channel may change to other channel based on table 5. Here in table 6, I investigate which kind of payout channel firms will choose if they change their payout policy.

Table 6 Payout matrix: average number of firms choosing different payout channel between 2000 and 2008.

T

T-1 No payout Repurchase only

Cash dividend only

Both cash dividend and repurchase

No payout 1202 228 44 14

Repurchase only 192 321 10 20

Cash dividend

only 39 8 453 151

Both cash dividend

and repurchase 11 10 147 386

(26)

(2008) finds that newer firms without a dividend history are unlikely to initiate dividends but repurchase, especially the firms whose earnings are relatively volatile. Finally, firms repurchasing stocks last year are more likely to change to paying out nothing if they do not want to stick to their payout channel. This is also plausible since repurchases are used as a vehicle to distribute transitory cash flows.

Next, I will show the characteristics of firms choosing a particular payout channel (B1-B4) in Table 7 and the characteristics of firms sticking to their payout channels (A1-A4) in table 8. Also comparison between these two groups would be analyzed in this part.

Table 7 Averages for firms choosing a particular payout channel (B1-B4)

Firm size is defined as the natural logarithm of the total assets of the firm. Profitability is defined as the earnings before interest but after tax divided by total asset. Leverage is defined as the ratio of total debt divided by total assets for the firm. MBF is defined as the ratio of market capitalization divided by Common equity. Growth is defined as the relative change of total assets.

No payout Repurchases only Dividend only Both dividend and repurchases Size 11.317 12.328 13.773 14.361 Profitability -0.140 0.0054 0.057 0.077 Leverage 0.408 0.419 0.512 0.519 Investment opportunity 7.087 2.905 3.186 3.156 Growth 0.335 0.131 0.165 0.103

(27)

Now I will show the averages of the characteristics for sticky firms (A1-A4) in table 8.

Table 8 Averages for firms sticking to their payout policy (A1-A4)

Firm size is defined as the natural logarithm of the total assets of the firm. Profitability is defined as the earnings before interest but after tax divided by total asset. Leverage is defined as the ratio of total debt divided by total assets for the firm. MBF is defined as the ratio of market capitalization divided by Common equity. Growth is defined as the relative change of total assets.

No payout Repurchases only Dividend

only Both dividend and repurchases Size 11.313 12.596 13.855 14.592 Profitability -0.116 0.024 0.063 0.081 Leverage 0.404 0.428 0.501 0.524 Investment opportunity 5.182 2.861 2.999 3.263 Growth 0.241 0.092 0.142 0.096

(28)

5.4 Multinomial regression

In this section, I will present the multinomial probit regression for the non-sticky channel choice (B1-B4) and for the sticky group (A1-A4) and compare these two groups.

Firstly, Table 9 shows the multinomial regression model explaining the type of payout policy adapted by US companies.

Table 9 Multinomial probit regression of the choice of payout channel (B1-B4).

The dependent variable take a value of 0 for firms not payout anything in a particular year, 1 for firms pays out only dividends in a particular year, 2 for firms only repurchase in a particular year, and 3 for firms use both payout channel in a particular year. Firm size is defined as the natural logarithm of the total asset of the firm. Profitability is defined as the earnings before interest but after tax divided by the total asset. Leverage is defined as the natural logarithm of the ratio of total debt divided by total asset for the firm. MBF is defined as the natural logarithm of the ratio of market capitalization divided by Common equity. Growth is defined as the relative change of total assets. All the independent variables are lagged by one year.

Base-case outcome is nonpaying (B1)

Only dividend (B2)

Only repurchase (B3)

Both dividend and repurchase (B4)

coefficient z-statistic coefficient z-statistic coefficient z-statistic

Size 0.351 42.9*** 0.183 25.97*** 0.476 51.96*** Profitability 2.503 11.2*** 1.429 13.72*** 5.069 15.53*** Leverage 0.326 4.58*** -0.634 -9.27*** -0.131 -1.65 Investment Opportunity -0.081 -3.72*** -0.020 -1.32 -0.018 -0.84 Growth -0.233 -7.4*** -0.147 -6.72*** -0.457 -6.56*** year -0.017 -1.75*** 0.082 8.66*** 0.035 3.41*** tax -0.264 -1.17 1.915 8.63*** 0.502 2.09 intercept -4.983 -34.78*** -3.474 -26.77*** -7.175 -45.84*** observation 24677 Wald chi2(28) 4701.14 Prob> chi2 0

* Significance at the 10% level. ** Significance at the 5% level. *** Significance at the 1% level.

(29)

investment opportunity is only positively related to the likelihood to only pay cash dividend (B2).

Table 10 Multinomial probit regression to explain which firms stick to payout channel (A1-A4)

The multinomial regressions are estimated separately for each year t for the 2000-2008 periods. The dependent variable take a value of 0 for firms do not stick to payout channel of prior year, 1 for firms not payout anything in current and prior year, 2 for firms pays out only dividends in current and prior year, 3 for firms only repurchase in current and prior year, and 4 for firms use both payout channel in current and prior year. Firm size is defined as the natural logarithm of the total asset of the firm. Profitability is defined as the earnings before interest but after tax divided by the total asset. Leverage is defined as the natural logarithm of the ratio of total debt divided by total asset for the firm. MBF is defined as the natural logarithm of the ratio of market capitalization divided by Common equity. Growth is defined as the relative change of total assets. All the independent variables are lagged by one year.

Base-case outcome is shifting group Non-paying (A1) Only dividend (A2) Only repurchase (A3)

Both dividend and repurchase(A4) coefficient z-stat coefficient z-stat coefficient z-stat coefficient z-stat Size -0.246 -35.38*** 0.103 12.65*** -0.030 -3.9*** 0.244 27.03*** Profitability -0.992 -11.44*** 2.005 9.68*** 0.938 6.63*** 5.132 20.86*** Leverage 0.060 0.93 0.400 5.26*** -0.467 -6.03*** -0.03 -0.37 Investment Opportunity 0.073 4.81*** -0.063 -3.25*** 0.014 0.79 0.027 1.37 Growth 0.151 7.15*** -0.101 -3.11*** -0.183 -4.66*** -0.619 -6.42*** year -0.028 -3.1*** -0.054 -5.23*** 0.041 3.79*** 0.030 2.76*** tax -1.145 -5.49*** -1.456 -6.07*** 0.718 2.87*** 0.293 1.14 intercept 3.513 28.07*** -1.437 -9.97*** -0.462 -3.14*** -4.332 -26.19*** observation 24677 Waldchi2(28) 5033.11 Prob> chi2 0

(30)

Furthermore, coefficients of size, profitability and leverage are all positive and significant, and investment opportunity and growth are both negatively significantly related to the likelihood to stick to pay cash dividends. This is in consistent with my expectations (A2 in the second part of table 2) that firms which are larger, more profitable, more levered and more mature are more likely to pay continuously dividends. Also, fewer firms are willing to payout only dividends over time based on the negative sign of the year variable. Since a growing number of firms use repurchases in recent years. In addition, the lower tax rates after 2003 lead to an increase in the likelihood to pay cash dividends, which prove that 2003 tax cut indeed has influence on the decision to pay.

Thirdly, for the group of firms that only repurchased last year and also stick to it this year, the relationship between profitability and likelihood to stick to repurchasing is positive, which is in line with the Skinner (2008). Skinner (2008) proves that for firms which only repurchase, there is a strong relationship between repurchases and earnings. Newer firms without a dividend history which have comparable characteristics of a traditional dividend payer would nowadays choose to initiate repurchases, such as some mature technology firms. They are more likely to initiate repurchases instead of dividends. These firms may not only repurchase but may also repurchase regularly. This is consistent with table 10 since the profitability is significantly positive and growth is significantly negative, which may to some extent stand for young and profitable technology firms. Year is positively significantly related to the propensity to repurchase, which also proves that more and more firms are willing repurchases and also stick to it over time. Also the relationship between tax and likelihood to repurchase is positive, which verifies that 2003 tax cut did affect the decision to repurchases negatively.

(31)

to table 10, firms sticking to a payout mix are larger, more profitable and have less growth. In addition, the absolute value of the coefficients for the size, profitability and growth are 0.244, 5.132 and 0.619 respectively, which are all larger than the absolute value of coefficient of the firms that only sticking to pay dividends (0.103, 2.005 and 0.101). All of these are in line with my expectation for group A4 in the second part of table 2.

Table 11 Multinomial probit regression to explain which firms stick to payout channel using Fama and Macbeth method (A1-A4)

The dependent variable take a value of 0 for firms do not stick to payout channel of prior year, 1 for firms not payout anything in current and prior year, 2 for firms pays out only dividends in current and prior year, 3 for firms only repurchase in current and prior year, and 4 for firms use both payout channel in current and prior year. Firm size is defined as the natural logarithm of the total asset of the firm. Profitability is defined as the earnings before interest but after tax divided by the total asset. Leverage is defined as the natural logarithm of the ratio of total debt divided by total asset for the firm. MBF is defined as the natural logarithm of the ratio of market capitalization divided by Common equity. Growth is defined as the relative change of total assets. All the independent variables are lagged by one year. The table shows means of the regression intercepts and slope and the t-statistics which defined as the mean divided by its standard error (the time-series standard deviation of the regression coefficient divided by the square root of the number of years minus 1)

Base-case outcome is shifting group

Non-paying Only dividend Only repurchase Both dividend and repurchase Average coefficient t-stat Average coefficient t-stat Average coefficient t-stat Average coefficient t-stat Size -0.249 -24.97*** 0.098 4.17*** -0.027 -1.80 0.246 19.03*** Profitability -1.159 -6.48*** 2.062 17.87*** 1.014 5.52*** 5.058 18.34*** Leverage 0.098 1.06 0.407 5.05*** -0.509 -9.83*** -0.039 -0.43 Investment Opportunity 0.053 2.30*** -0.085 -4.01*** 0.072 1.71 0.056 1.97 Growth 0.143 4.16*** -0.082 -2.55*** -0.194 -3.41*** -0.721 -7.33*** intercept 3.136 22.29*** -2.017 -6.46*** -0.176 -0.86*** -4.137 -22.18***

(32)

In order to compare the sticky groups with non-sticky groups, I calculate the difference of the coefficient of the “only dividend”, “only repurchases” and “both dividend and repurchases” (table 10) and the coefficient of the “non-paying” (table 10). In table 12, the coefficient for “only dividend” stands for the likelihood of firms sticking to pay only cash dividend (A2) compared to firms sticking to pay nothing (A1). The coefficient for “only repurchase” represents the likelihood of firms sticking to only repurchase (A3) comparing to firms sticking to pay nothing (A1). The coefficient for “both dividend and repurchase” stands for the likelihood of firms sticking to use both channel (A4) comparing to firms sticking to pay nothing (A1). In this sense, the non-paying firms (A1) are reference choice and I can compare the effect of the characteristics on the decision to payout through only dividend (B2), only repurchases (B3) and use both channels (B4) (from table 9) with the decision to stick to only pay dividend (A2), stick to only repurchases (A3) and stick to use both channels (A4).

Table 12 comparison the coefficient of characteristics for non-sticky firms and sticky firms

“Non-paying” (B1) means firms not payout anything in the current and prior year, “only dividend” (B2) means firms pays out only dividends in the current and prior year, “only repurchases” (B3) means firms only repurchase in the current and prior year, “both dividend and repurchases” (B4) means firms use both payout channel in the current and prior year. “Non-paying” (A1) means firms not payout anything in the current and prior year, “only dividend” (A2) means firms pays out only dividends in the current and prior year, “only repurchases” (A3) means firms only repurchase in the current and prior year, “both dividend and repurchases” (A4) means firms use both payout channel in the current and prior year. Firm size is defined as the natural logarithm of the total asset of the firm. Profitability is defined as the earnings before interest but after tax divided by the total asset. Leverage is defined as the natural logarithm of the ratio of total debt divided by total asset for the firm. MBF is defined as the natural logarithm of the ratio of market capitalization divided by Common equity. Growth is defined as the relative change of total assets. The coefficient for B1, B2, B3 and B4 are from the coefficient of each independent variable for each group. The coefficient for “only dividend” (A2), “only repurchases” (A3) and “both dividend and repurchases” (A4) in table 12 is the difference of the coefficient of the “only dividend” (A2), “only repurchases” (A3) and “both dividend and repurchases” (A4) in table 10 and the coefficient of the “non-paying” (A1) in table 10.

base-case outcome is non-paying(B1) base-case outcome is non-paying (A1) Only dividend (B2) Only repurchase (B3) Both dividend and repurchase (B4) Only dividend (A2) Only repurchase (A3)

(33)

(34)

6 Conclusion and limitation

6.1 conclusions

With panel data of 7113 listed firms in the US, this study investigates the stickiness of payout channels. I run multinomial regression analyses to answer the research questions: 1. Do sticky payout channels exits? 2. Which kind of firms stick to their payout policy? The analysis is based on the paper of Renneboog and Trojanowski (2010). They investigate the payout channel in a particular year regardless of the situation in the prior year. I extend their analysis by focusing on the payout channel in both current year and prior year, which I call the stickiness of payout channel.

In order to answer the first question, I calculate the proportion of four groups of firms sticking to their prior payout channel and I find the average of these proportions is above 50%, which means that the payout stickiness does exist.

(35)

6.2 limitations

There are some limitations in the data selection in this paper. Firstly, the investigation period is from 2000 to 2008. However, after 2007, US market was affected to a large extent by the global financial crisis. It is worthwhile to find out whether the stickiness still exists under financial crisis. If it is not the case, how does the financial crisis influence the firms’ payout policy?

(36)

Reference

Allen, F., A. E. Bernardo, and I. Welch, 2000. Theory of Dividends Based on Tax Clienteles. Journal of Finance 55, 2499-2536.

Baker, M. and J. Wurgler (2004a), ‘A catering theory of dividends’. Journal of Finance 59, 1125–1165.

Baker, M. and J. Wurgler (2004b), ‘Appearing and disappearing dividends: The link to catering incentives’. Journal of Financial Economics 73, 271–288.

Benartzi, S., R. Michaely, and R. Thaler (1997), ‘Do changes in dividends signal the future or the past?’. Journal of Finance 52, 1007– 1043.

Bhattacharya, S., 1979. Imperfect information, dividend policy, and ‘The Bird-in-the-Hand’ fallacy, Bell Journal of Economics 10, 259–270.

Brav, A., Graham, J., Harvey, C., and R. Michaely (2005). Payout policy in the 21st century. Journal of Financial Economics 77, pp. 483-527.

Brav, A. and J. B. Heaton (1998). ‘Did ERISA’s prudent man rule change the pricing of dividend omitting firms?’ Working paper, Duke University and University of Chicago.

Brooks, C., 2008. Introductory Econometrics For Finance, 2nd Edition, Cambridge University Press, Cambridge UK.

Bulan, L. T. and N. Subramanian (2008), ‘The firm life cycle theory of dividends’. In: H. Kent Baker (ed.): The Blackwell Companion to Dividends and Dividend Policy. Malden, MA: Blackwell Publishing.

Chetty, Raj, and Emmanuel Saez. “The Effects of the 2003 Dividend Tax Cut on Corporate Behavior: Interpreting the Evidence.” American Economic Review Papers Review and Proceedings 96 No. 2 (May, 2006): 124–9.

Choe, H. (1990), ‘Intertemporal and cross-sectional variation of corporate dividend policy’. Ph.D. dissertation, University of Chicago.

Coulton, Jeff J. and Ruddock, Caitlin M. S., Corporate Payout Policy in Australia and a Test of the Life Cycle Theory (March 9, 2010). Australian School of Business Research Paper No. 2010 ACCT 03.

(37)

DeAngelo, H., L. DeAngelo, and D. J. Skinner (2004), ‘Are dividends disappearing? Dividend concentration and the consolidation of earnings’. Journal of Financial Economics 72, 425–456.

DeAngelo, H., L. DeAngelo, and R. Stulz (2006), ‘Dividend policy and the earned/contributed capital mix: A test of the life-cycle theory’. Journal of Financial Economics 81, 227–254.

Denis, D. J. and I. Osobov (2008), ‘Disappearing dividends, the earned/contributed capital mix, and catering incentives: International evidence on the determinants of dividend policy’. Journal of Financial Economics 89, 62–82.

Fama, E. F. and H. Babiak (1968), ‘Dividend policy: An empirical analysis’. Journal of the American Statistical Association 63, 1132–1161.

Fama, E. F. and K. R. French (2001). ‘Disappearing dividends: Changing firm characteristics or lower propensity to pay?’ Journal of Financial Economics 60, 3–40.

Grullon, G. and R. Michaely (2002), ‘Dividends, share repurchases and the substitution hypothesis’. Journal of Finance 57, 1649–1684.

Guay W., and J. Harford (2000). The cash-flow permanence and information content of dividend increases versus repurchases. Journal of Financial Economics 57, pp. 385-415.

Hoberg, G. and N. R. Prabhala (2008), ‘Disappearing dividends, catering, and risk’. Review of Financial Studies 22, 79–116.

John, K. and J. Williams, (1985). Dividends, dilutions, and taxes: A signaling equilibrium, Journal of Finance 40, 1053–1070.

Julio, B. and D. Ikenberry (2004), ‘Reappearing dividends’. Journal of Applied Corporate Finance 16, 89–100.

Lease, R. C., K. John, A. Kalay, U. Loewenstein, and O. H. Sarig (2000), Dividend Policy: Its Impact on Firm Value. Boston, MA: Harvard Business School Press.

Lie, E. (2005), ‘Financial flexibility, performance, and the corporate payout choice’. Journal of Business

78, 1–23.

(38)

Li, W. and E. Lie (2006), ‘Dividend changes and catering incentives’. Journal of Financial Economics 80, 293–308.

Mia Twu (2010), “Prior Payment Status and the Likelihood to Pay Dividends: International Evidence” the Financial Review Volume 45, Issue 3, pages 785–802.

Miller, M. (1986), ‘The informational content of dividends’. In: J. Bossons, R. Dornbusch, and S. Fischer (eds.): Macroeconomics: Essays in Honor of Franco Modigliani. Boston, MA: MIT Press.

Modigliani, F. and M. Miller (1959), ‘The cost of capital, corporation finance, and the theory of investment: Reply’. American Economic Review 49, 655–669.

Modigliani, F. and M. Miller, (1961). Dividend policy, growth, and the valuation of shares. Journal of Business 34, 411–433.

Miller, M.H. and K. Rock, (1985). Dividend policy under asymmetric information, Journal of Finance 40, 1031–1051.

Mueller, D. C. (1972), ‘A life cycle theory of the firm’. Journal of Industrial Economics 20, 199–219.

Senchack, Jr., A. J. and W. Y. Lee (1980), ‘Comparative dynamics in a life cycle theory of the firm’. Journal of Business Research 8, 159–185.

Shefrin, H. and M. Statman (1984), ‘Explaining investor preference for cash dividends’. Journal of Financial Economics 13, 253–282.

Shleifer .A.and Robert .V. (1986), ‘Large shareholders and corporate control’. journal of Political Economy 94 (3), 461-488.

Sinn, H.-W. (1991), ‘The vanishing Harberger triangle’. Journal of Public Economics 45, 271–300.

Skinner, D. J. (2008), ‘The evolving relation between earnings, dividends, and stock repurchases’. Journal of Financial Economics 87, 582–609.

(39)

APPENDIX A:Description of the variables and their corresponding codes

This table reports the variable codes used in the dataset. The data is gathered from DataStream and Worldscope.

Description Code

Market Capitalization WC08001

Total Asset WC02999

Common Equity WC03501

Earnings before Interest and Taxes (EBIT)

WC18191

Income Taxes WC01451

Treasury stock ws.TreasuryStock

Purchase of common and preferred stock ws.PurchOfComAndPfdStkCFStmt Sale of common and preferred stock ws.SaleOfComAndPfdStkCFStmt

APPENDIX B: Correlation matrix for independent variables

“Size” is the natural logarithm of total assets. “Profitability” is the earnings before interest but after tax divided by the total assets. “Leverage” is the ratio of total debt divided by the total assets. “Investment opportunity” is the natural logarithm of market capitalization divided by common equity. “Growth” is the asset growth rate.

Size Profitability Leverage

The stickiness of payout channels

The stickiness of payout channels

Author:

Wenwen Yang

Date: August 2011

Supervisor: Dr. H. Von Eije

Second evaluator: Dr. H. Gonenc

The stickiness of payout channels

Table of Contents

1. Introduction

2. Literature review

3 Methodology

4 Data

5 Results