X
l?$I
X
UNIVERSITEIT VAN AMSTERDAM
The influence of the Financial Crisis on the relationship between
cash retention and the dividend payout policy of US corporate
firms.
Name Student number Programme Specialization Coordinator Supervisor Date Gwen Witteman 10875093BSc Economics and Business Finance and Economics dhr. dr. D.F. Damsma E. Zhivotova
June 2018
Abstract:
This paper examines the role of cash retention in order to sustain continuous dividend policies during the Financial Crisis (2007-2009). The Economic Recession Effect is researched through analyzing data on the main factors determining the dividend payout policy. The sample size contains data of 3,649 US corporate firms active on the NASDAQ and NYSE stock exchange. The results show that US stock-listed firms did reduce
dividend payments and shifted their emphasize towards financial flexibility. Pre-saved cash functions thereby as a buffer for potential economic downturns. Finally, the Financial Crisis caused a shift in cash retention through US corporate firms. Larger cash savings are used during the crisis to finance capital investments instead of paying dividends.
Statement of Originality
This document is written by Student Gwen Witteman who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of contents
1. Introduction 4
2. Literature Review 6
2.1 The Financial Crisis 2007-2009 6
2.2 Dividend Policy 7
2.2.1 Bird-in-the-Hand Hypothesis 8
2.2.2 Taxation 9
2.2.3 Leverage and Cash Retention 10
2.2.4 Agency Costs 10
2.2.5 Asymmetric Information 11
2.3 Correlation between Economic Recession and Dividend Policy 12
2.4 Hypotheses 12 3. Methodology 14 3.1 Choice of Exchanges 14 3.2 Data 14 3.3 Research Method 15 4. Results 16 4.1 Model Modification 16
4.2 Summary and Descriptive Statistics 17
4.3 Regression Results 18
4.4 Regression Results on Cash Retention and Dividend Policy 19 5. Conclusion and Discussion
6. References 7. Appendix
7 .1 Variable Description 7.2 Hausman Test
7. 3 Cash Retention Effect
22 23 25 25 25 26
1. Introduction
Corporate firms have multiple options available to do with their ( excess) cash. One of the most commonly used options, concerns the payout policy of a firm. Either repurchasing shares or doing dividend payments to shareholders. The dividend payment consists of a fixed amount per share. This fixed amount may vary over the years of the firms lifespan (Fama and French, 1988).
The dividend payout policy of a firm is affected by various determinants. Many of these determinants are described and tested in previous research. There is however limited research to be found on the effect of the Financial Crisis on the relationship between cash retention and dividend policy of US stock-listed firms. Chen et al. (2018) were the first to describe the possible benefits of precautionary cash savings for firms who are affected by an economic recession. This paper seeks to build further on this knowledge and extend the research with new empirical evidence by analyzing the effect of cash retention on the dividend payout policy of a firm. With an emphasis on the effect of an economic
recession, in particular the Financial Crisis, on the dividend payout policy of US stock-listed firms. The Financial Crisis of 2008 may cause financial distress for corporate firms, resulting in less dividend payments and external finance becoming too costly. Expensive finance could be prevented through pre-saved cash. Cash savings with a precautionary motive helps firms providing financial flexibility (Gamba and Triantis, 2008). Moreover can pre-saved cash ensure the maintenance of dividend payout policies.
As suggested by Bliss et al. (2015), the effect of an economic recession on payout policy is unclear. As well, the impact of the Financial Crisis on the capital investments financed by cash savings is not clear either. The expectation would be that dividends are cut in order to finance future capital investments. However according to Brav et al. (2005), CFO's would rather cut their capital investments to sustain its dividend payout policy. Even firms with limited cash savings are reluctant to cut its dividends (Bliss et al., 2015). While DeAngelo and DeAngelo ( 1990) found in their paper evidence for dividend cuts caused by financial distress.
In order to research the impact of the Financial Crisis on determinants affecting the dividend payout policy, this paper seeks to answer the following research question:
Does a difference exist in terms of the Economic Recession effect on Cash Retention among US corporate firms: Is there a difference in the Economic Recession Effect on the relationship between Cash Retention and Dividend Payout Policy?
With the help of a simple OLS (Ordinary Least Squares) regression will be tested what the effect is of different variables on the dividend payout policy of US firms. This regression contains data of US common stock listed on NASDAQ and NYSE. To stress out the feasible differences in payout policies through time, an analysis would be made on three different time periods.
In the following section is the literature described that have been used for this research. Thereafter in section 3 is the methodology of the research
discussed. After which in section 4 the output of the OLS regression is presented and analyzed. Finally, in section 5 are the conclusions of the research formulated and will the research question being answered.
2. Literature review
In this section we first start with the fundamentals of dividend relevancy. After which in section 2.1 the structural causes of the Financial Crisis are
demonstrated. Followed by describing the fundamentals of dividend policy in section 2.2. Thereafter in section 2.2.1 we explain the bird-in-the-hand
hypothesis. Then in sections 2.2.2 till 2.2.5 we outline the market imperfections affecting its dividend policy. Followed by the correlation between the economic recession and dividend policy in section 2.3. Finally in section 2.4, the
formulation of the hypotheses and the expectations regarding these hypotheses. A dividend is a payment made by a firm to its shareholders. The
distribution of dividends are part of the firm's after-tax profit. The amount of a firm's dividend is determined by the firms' public board of directors (Berk and DeMarzo, 2013). Dividend payments play an important role in the payout decision of a firm. Announcements of future dividend payments can contain information about future performance. More on the importance of dividend signaling is discussed later on in section 2.2.5.
2.1 The Financial Crisis 2007-2009
The Financial Crisis started in 2007 was triggered by the US subprime mortgage crisis. Although the percentage of securities backed by subprime mortgages was only around 3% of all US financial assets, the impact of the subprime mortgage crisis was observable on a global scale (Eichengreen et al., 2012).
In the early 2000s the housing prices kept sharply increasing. The sharp increases of the housing prices for consecutive years, raised the concerns of multiple economists. Case and Shiller provided evidence in their paper (2003) for a possible bubble in the housing market. In 2006 the expectations of Case and Shiller, and many other economists became reality. It started with the burst of the US real estate bubble, followed by a large decline of housing prices during 2006-2007. The declining housing prices resulted in the reduction of overall wealth. The reduction in wealth means fewer available growth opportunities for firms, and therefore less demand for external funding. Fewer growth
opportunities issued greater agency costs through cash retention (Bliss et al., 2015). Because it would be beneficial to use the retained cash for the few good investment opportunities left.
Prior literature attributes to the costs and benefits associated with cash retention. The agency costs incurred by cash retention, as stated earlier, is further discussed in section 2.2.4. As for the benefits, the main reason to retain cash is to build cash reserves as a buffer for severe economic downturns.
Consequently, firms will hold greater cash reserves when external finance is too costly. Indeed, this is supported by Almeida et al. (2004). They found that financially constrained firm's save a greater proportion of their cash flows, depending on the firm's capacity for external finance.
On September 15, 2008 Lehman Brothers filed for bankruptcy. The collapse of Lehman Brothers, the largest bankruptcy in US history, was an important event during the Financial Crisis. The impact was primarily on large banks, savings and loan firms. Even the relationships between (investment) banks were suffering on a mutual basis (Fernando, 2012). Mistrust between banks resulted in reduction in line of credit to other banks. New loans to large borrowers fell by 47% relative to the peak of the credit boom (lvashina and Scharfstein, 2010). Ivashina and Scharfstein (2010) found in their paper that the reduced credit-line increased the costs of all types of external finance. Almeida et al. (2012) found evidence for firms cutting dividends and reducing cash reserves in order to fulfill their costly debt payments. More on the correlation between an economic recession and dividend policy is discussed later on in section 2.3.
2.2 Dividend Policy
Modigliani and Miller ( 1958) were the first economists to research the
fundamentals of a firms dividend payout policy; they started the debate about dividend policies when they came up with the Dividend Irrelevance Theorem. Assuming perfect capital markets, they stated that the firm's choice of dividend policy is irrelevant and does not affect the initial share price (Modigliani and Miller, 1961).
The motive why firms choose to pay or not to pay dividends, has puzzled many economists. Most common, financial economists explain dividends by the use of dividend signaling (see section 2.2.5). The firms dividend choice is based on the desire to communicate information towards its shareholders.
A different view on dividends was proposed by DeAngelo and DeAngelo (2006). They argued that relaxation of the MM assumptions of allowing retention is needed to have a payout policy that matters. Relaxation of this assumption is crucial as it would imply that cash savings determines the value of a firm. Now this adds the payout policy as determinant of value. By then it is not merely the investment policy that determines the value of a firm (DeAngelo and DeAngelo, 2006).
Another determinant influencing the dividends is investigated by Denis and Osobov (2008). They provided evidence for the relationship between retained earnings and dividends paid. Firms primarily consisting of retained earnings are more likely to pay dividends. These findings are similar of those of Fama and French (2001). They researched the effect of the firm size, growth opportunities, and probability on the likeliness to pay dividends (see section 2.2.3).
In practice, the capital markets are not likely to be efficient and perfect. The perfect capital markets assumption of Modigliani and Miller (1961) has been studied in numerous papers. The developments of market imperfections affecting the dividend payout policy are described in the remainder of this section.
2.2.1 Bird-in-the-Hand Hypothesis
"A bird in the hand is worth two in the bush." (Ray, 1670)
The bird-in-the-hand hypothesis captures the time preference of shareholders. Shareholders prefer higher current dividends over higher future dividends. These higher current dividends have a positive effect on the share price (Berk and DeMarzo, 2013). Indeed, this was pointed out by Gordon in his study (1963) on the optimal investment and financing policy. He discusses the limitations of the MM proposition, as being too simplified. Retaining dividends and investing in a later time-period, changes shareholders dividend expectations (Gordon, 1963). Using Gordon's pricing model, reducing current dividends and raising future dividends, causes an increase in the discount rate. The higher discount rate results in a lower share price. This effect could be interpreted as shareholders risk aversion and minimization of the uncertainty regarding future cash
2. 2. 2 Taxation
Dividend policies should be designed in a way that minimizes the taxation, leverage, agency costs and asymmetric information. Taxes are a major market imperfection that influences the firm's decision to pay dividends or to make share repurchases. It depends on the differences between tax rates on dividends and capital gains. To illustrate this: if dividends were to be taxed at a higher tax rate, in this situation it would be optimal for investors to invest in a non-dividend paying firm.
In the United States were until 2003 double taxation on dividends
received (Morck and Yeung, 2005). The maximum individual tax rate is now set at 15%. The effective individual tax rate is related to an investors income. Investors with a lower income fall into a lower tax bracket and are able to make use of deductible tax rates. Furthermore does the investment horizon affect the taxation of capital gains. Longer-term investors can defer tax payments and don't need to pay higher taxes on short-term investments. At last does the tax jurisdiction determine the taxation of an investor. In the United States taxes differ by state.
Chetty and Saez (2005) have studied the effect of an tax cut on the dividend payout of a firm. The main result of the tax cut in 2003 was the shift from share repurchases to dividend payouts. The size of dividend payments and number of firms paying dividends did significantly rose after the implementation of the new dividend taxation policy. While the previously two decades, indicated a trend in firms' making share repurchases (Chetty and Saez, 2005). Share repurchases reducing the existing amount of shares, could generate capital gains for the investor. Taxes on capital gains need to be paid by shareholders when selling their shares. These capital gains can be seen as a homemade dividend. Creating homemade dividend is used by Modigliani and Miller as response to the bird-in-the-hand hypothesis (Modigliani and Miller, 1961). They argue that a firms dividend policy choice doesn't matter while investors can create their own dividend.
In general investors prefer share repurchases over dividend payments. Equal tax rates gives still a tax advantage to share repurchases over dividend payouts. Capital gain taxes are deferred until the share is sold. Therefore there
exists a tax timing advantage for share repurchases compared to dividend payments, for the long-term investor.
2.2.3 Leverage and Cash Retention
Leverage allows a firm to acquire higher returns on their investments. External finance through issuing more debt, exaggerates the leverage ratio. During an economic recession it would be difficult for firms to finance its investment opportunities by external finance (Chen et al., 2018). However smaller firms with high leverage ratios and liquidity problems in the past, are less likely to default during future economic downturns. Chen et al. (2018) argue that firms should learn from previous events and need to adapt by pre-saving more aggressively to mitigate the consequences of a new exogenous shock.
The available investment opportunity set of a firm is another aspect that defines its external financing and dividends. Firms with high growth options tend to have lower leverage and do less dividend payments (Smith and Watts, 1992). Fama and French (2001) argued this as smaller firms with high growth
opportunities, having a lower propensity to pay (e.g. information technology sector). These kind of firms have numerous of profitable investment
opportunities available, any excess cash would be retained to undertake future investment opportunities. Reinvesting profits enhances value maximization, they do less dividend payments as this would dilute value of the firm. Partly
reinvestment of these firms occurs by using the cash for research and
development (R&D) purposes. The first to study the effect of R&D on dividend payout policy was done by Jensen et al. (1992). They came up with a negative effect of R&D expenditures on the dividend payout policy. Smith and Watts (1992) also found that regulated firms have higher leverage and do pay higher dividends. The same holds for the larger firms, higher leverage and higher dividends. Therefore this implies that there is a positive correlation between leverage and dividends.
2.2.4 Agency Costs
Agency costs arise when agents pursue their own interests instead of the principals (Jensen and Meckling, 1976). To prevent this behavior, investors and managers find it advantageous to set up devices. Devices that enable principals to better monitor their agents. Setting up these devices is costly and harmful for
the shareholders gains (Easterbrook, 1984).
Easterbrook ( 1984) found in his study that dividends are a mechanism to reduce the agency costs of a firm. Declaring dividends and issuing new debt helps the firm to rebalance the debt-to-equity ratio. The risk adjusted debt-to- equity ratio helps managers to monitor the capital market and reduce the agency costs incurred. However the downside from dividends functioning as monitor device, is the increase of transaction costs associated with the issuance of external finance (Rozeff, 1982).
Agency costs play also a role in the cash savings of a firm (Jensen, 1986). Managers retain more cash than needed to reduce the pressure of operating efficiently. The cash savings of a firm are misused by managers to the aim of empire building (Jensen, 1986). The conflict of interest between managers and shareholders results in a manager willingness to engage even in negative NPV investments so, there doesn't need to be paid any dividends to its shareholders.
2.2.5 Asymmetric Information
The last market imperfection not covered yet is the asymmetric information. Asymmetric information occurs when managers have better information than investors regarding the future prospects of the firm. Payout decisions, including dividend policies, may signal this information.
Lintner ( 1956) suggested investors preferring stable dividends with
sustained growth. When firms use dividend smoothing, the firm's dividend choice will contain information about future prospects. The increase in dividends of a firm sends a positive signal to the investors. Investors believe that the
management expects to be able to afford these higher dividends. On the other hand, a firm cutting its dividends will send a negative signal to the investors. The belief of dividends functioning as a signaling mechanism is called the dividend signaling hypothesis. Dividend smoothing is the key insight for this signaling hypothesis (AI-Malkawi et al, 2014). Results of AI-Malkawi et al. (2014) are consistent with the Lintner dividend smoothing model ( 1956 ), suggesting current earnings and past dividends are important for a firms dividend decision.
The effect of the dividend signaling hypothesis during the Financial Crisis is rather ambiguous. AI-Malkawi et al. (2014) argues that there is no significant effect of the Financial Crisis on the dividend policy. While Reddemann et al. (2010) findings support that dividend signaling is not relevant during the crisis.
The ambiguous effect could be explained by the setting of their sample.
Reddeman et al.'s sample concerned a developed country whereas AI-Malkawi et al.'s sample contained data from a developing country. The two samples concern differences between the establishment of these markets.
2.3 Correlation between Economic Recession and Dividend Policy
The importance of the correlation between an economic recession and the
dividend policy of a firm follows from the dividend signaling hypothesis. Although firms are in times of an economic recession, they still reserve some cash to smooth dividends across the years. Meanwhile the economic recession causes firm to reduce their expenditures. Reducing expenditures can be achieved
through retaining cash instead of cash distributions through dividends otherwise. As described in the previous section, dividend reductions or slowdowns can imply a firm bothered by financial distress. This would send a negative signal to their shareholders and negatively affects its share price. The conflicting interest of dividend policy during an economic recession is investigated in several papers. DeAngelo and DeAngelo (1990) provided evidence in their paper of the effect of an economic crisis on the dividend policy of 80 NYSE firms. They concluded in their paper that most firms significantly reduced its dividends, strategically seen as the most attractive option. Managers perception concerning the firms'
dividend policy, was to maintain long dividend policies history and refuse to suspend dividends.
Chen et al. (2018) confirmed the findings of DeAngelo and DeAngelo ( 1990). They came up with the same results on dividend cuts during the
Financial Crisis. However, they used a different perspective to get to this results. Chen et al. argue (2018) that firms have a learning curve and that most firms saved more money after 2000 (dot-com crisis). Pre-saved cash helps them to survive an economic recession, when external finance is too costly.
2.4 Hypotheses
Based on past literature on this subject, the effect of an economic recession on the dividend policy is an ongoing debate. As discussed in the introduction, according to Bliss et al. (2015), the effect of an economic recession on the dividend policy is unclear. However most recent studies (Hauser, 2013; Chen et
al., 2018) suggest an effect of the economic recession on the dividend policy of firms. Because the presence of an economic recession increases the probability of firms ending up with higher financial distress costs, whereby cost of external finance rises. Precautionary cash would help firms future access to external finance. Therefore this paper seeks to combine previous literature and
investigate the effect of an economic recession on the relationship between cash retention and dividend payout policy. A different data set is used in order to conclude the effect of the Financial Crisis. The purpose is to study the market imperfections determining the dividend policy. Thereby to study thoroughly, the effect of the Financial Crisis on the relationship between cash retention and dividend policy. In this paper is therefore expected that, US stock-listed firms active on the NASDAQ and NYSE Stock Exchange:
1) An Economic Recession Effect is to be found in the dividend policy of US stock-listed firms.
2) An Economic Recession Effect is to be found on the cash retention of US stock-listed firms.
These hypotheses enable this paper to draw a conclusion on the behavior of firms' public board of directors concerning their dividend policy, in times of an economic recession. The conclusion would be formulated and discussed in
section 5. This conclusion helps to formulate an answer on the research question asked in the introduction.
3. Methodology
In this section the methodology applied is described. First, in section 3.1 the choice of the particular stock exchanges is explained. Then in section 3.2 the data of the sample is covered. Conclusively in section 3.3 the research method is clarified and also the model being used.
3.1 Choice of Stock Exchanges
In this empirical research, the effect of an economic regression on the dividend policy is tested on US common stocks listed at the NASDAQ and NYSE exchange. These stock exchanges are the two largest in the world. Differences between the two stock exchanges in volatility, market capitalization and market type
encourages the soundness of the data. Another reason to justify the choice for these stock exchanges is the availability of data and the origins of the Financial Crisis in the United States.
3.2 Data
The sample of individual stocks is created from panel data originating from both the NASDAQ and NYSE stock exchange. The time interval of this panel data ranges from 2000-2015. Monthly data is preferred due to its availability and accuracy. Data on a monthly basis is more informative as multiple economic events during a fiscal year, have a short-term effect on some of the variables. The monthly data is obtained from the database Wharton WRDS on the following variables: Dividend Payout Ratio, Debt-to-Equity Ratio, Book-to-Market Ratio, Return on Assets, Cash Ratio, Effective Tax Rate, R&D- and Labor Expenditures.
The total dataset contains 1,920 observations for each variable over 3,649 firms. The firms are categorized into ten industry sectors in alignment with the Global Industry Classification Standard (GICS). The categorization of firms improves the ability to test the model on random- and fixed effects.
3.3 Research Method
To test the effect of an economic recession on the relationship between cash retention and dividend policy, a similar approach to Hauser (2013) and Chen et al. (2018) is used. These methods are used since the papers cover both aspects of the regression on dividend policy. The recession effect is investigated by
Hauser (2013) while the cash retention relationship is investigated by Chen et al. (2018). Also, these are the two most recent published papers on the related subjects. The combination of the papers would be the framework to run the OLS regression. The main variables of the regression are kept the same and
additional variables are attached, as contributed by the most relevant existing theories affecting the dividend policy of a firm. In addition to prior literature, fixed year effects of industry sectors are included. More on the fixed effects is discussed in section 4.1.
The dependent variable to be explained is the Dividend Payout Ratio. In order to help explaining the dependent variable, nine explanatory variables are added to the model. Including one dummy variable to test the effect of an economic recession. This dummy variable is indicated by 1 during an economic recession and if there is no recession this is equal to 0.The Financial Crisis in the dataset is represented during the years 2007, 2008 and 2009. The particular years chosen are roughly a representation of the period concerning the Financial Crisis. The model is as follows:
DPR
=
/3
0 +/3
1(Recession\t +/3
2(Cash)it +/3
3(ROA\t +/3
4(DE)it +/3
5(BM)it +/3
6(R&D\t+
/3
7(Tax)it +/3
8 (SNWC\t + f39(Labor)it +ai+ uitSetting up these variables is done with taken into account the size, profitability and investment opportunities of a firm (Fama and French, 2001). Furthermore is tested with the Hausman test if including random- or fixed effects is desirable. The Hausman test concluded that there is enough evidence to reject the H0
hypothesis with alpha is 5 percent (see appendix table 2). Adding fixed effects would be desirable due to efficiency reasons.
In order to test whether the crisis did affected the firms' dividend policy, the panel data sample is divided into three time periods. Namely the pre-crisis period (2000-2006), the Financial Crisis (2007-2009) and respectively the post- crisis period (2010-2015).
4. Results
In this section the results of the regression are described. First in section 4.1 the modification of the model due to multicollinearity issues. Then in section 4.2 summary and descriptive statistics. Afterwards in section 4.3 the results of the economic recession effect. Finally, in section 4.4 the regression on the
relationship between cash retention and dividend policy.
4.1 Model Modification
The first regression tested on the dividend payout model, resulted in
multicollinearity of two variables. Modification of the model through omitting variables was necessary to improve the empirical results.
DPR Cash DE RD Labor DPR 1 Cash -0.3279 1 DE 0.3617 -0.0023 1 RD -0.3583 0.8223 -0.3010 1 Labor 0.2089 0.1760 0.9250 -0.1626 1
Table 1 Pair wise correlation matrix
The first variables where possibly a multicollinearity relationship exists, is between the R&D Expenditures Ratio and the Cash Ratio (0.8223). Fama and French (2001) discussed that firms with high growth opportunities use cash to finance new investments. Firms with high growth opportunities are mostly active in the information technology sector, taken this environment into consideration, explains the effect of cash savings usage for their R&D department. However the Cash Ratio variable has a lower correlation (-0.3279) with the Dividend Payout Ratio than the R&D Expenditures Ratio (-0.3583). Nevertheless the Cash Ratio variable would have a higher explanatory effect in accordance with previous literature, describing the factors determining the Payout Ratio. Hence omitting the R&D variable was necessary to address this multicollinearity issue.
The second variables that suffer from multicollinearity are the Labor Expenditures with respect to the Debt-to-Equity ratio. These variables are very strong correlated (0.9250) with each other. Previous literature doesn't explain this strong correlation. To cancel out this effect, the Labor Expenditures variable
needs to be omitted to prevent this factor influencing the errors and some other variables. The choice of dropping the Labor Expenditures instead of the Debt-to- Equity Ratio is based on its explaining power (0.3617>0.2089).
The new model is now as follows:
DPR
=
/3
0+
/3
1 tRecessiorû.,+
/3
2(Cash)it+
/3
3(ROA)it+
/3
4(DE)it+
/3
5(BM)it+
f36(Tax)it+
/3
7 (SNWC)it +ai+ uit4.2 Summary and Descriptive Statistics
Table 2 provides an overview of the descriptive statistics of all the variables which are included in the revised model. The variables exists of 1,920
observations for each variable over the period 2000-2015. Thereby should be noted that the dividend payout consists of only positive amounts. Even if income in this period is negative.
Variable Mean Std. Dev. Min Max
DPR 0.1406 0.1940 0 0. 7319 Recession 0.1875 0.3904 0 1 Cash 0.6323 0.5593 0.0518 3.0889 ROA 0.0919 0.0570 -0.2150 0.2196 DE 1.8695 2.2837 0.3550 9.5321
BM
0.5999 0.2095 0.1714 1.4713 Tax 0.3390 0.0290 0.2480 0.4052 SNWC 5.6620 3.7448 1.0733 25.9797Table 2 Descriptive Statistics
Table 3 provides an overview of the correlations between each variable. The overview consists of the revised model (excluding the R&D- and Labor
Expenditures). The excluded variables exceeded the 0.8 correlation, the rule of thumb value, which determines if there exists a possible multicollinearity relation between the two variables.
DPR Recession Cash ROA DE BM Tax SNWC DPR 1 Recession 0.0438 1 Cash -0.3279 -0.0117 1 ROA 0.0229 0.0378 -0. 7627 1 DE 0.3617 -0.0034 -0.0023 -0.3241 1 BM 0.4760 0.0414 -0.4349 0.0346 0.4861 1 Tax 0.0147 -0.0017 -0.4490 0.2678 -0.1892 0.0340 1 SNWC 0.6187 0.0270 -0.6326 0.4032 0.0232 0.3813 0.3624 1
Table 3 Correlation Matrix
As can seen from table 3, the two variables with the highest correlation are ROA and Cash. The high correlation between the two variables is described by
Kallapur ( 1994). In this paper he tests the earlier work of Jensen ( 1986). Jensen argued that managers overinvest retained earnings. Managers seek to engage in empire building (section 2.2.4 Agency Costs) to prevent paying dividends to outsiders (i.e. shareholders). Kallapur (1994) findings support this by stating that retained earnings financed investments show significantly lower returns than a proxy of the market's required rate of return.
4.3 Regression Results
The results of the regression are described below. Table 4 provides a summary of the regression model, whilst table 5 provides the data of the t-test on the coefficients and corresponding p-values.
Fixed- Group Number of Number Observations R- F Prob.
effects GLS variable observations of per group squared >F
regression groups (Within,
between, overall)
gicdesc 1,920 10 192 0.1320 41.34 0.000
0.4010 0.1982 Table 4 Summary Regression Model
Table 4 provides information on the R-squared of the regression. The table shows a low value of R-squared. The small proportion of variances of DPR explained by the explanatory variables could be due to the infrequency of the
data points. The monthly data reveals that firms do not pay dividends every month. However omitting all not dividend paying data points would negatively affect the validity of the variables. The issue of data points without any
dividends is been addressed by categorizing and pooling firms in the same industry sector.
DPR Coefficient Standard t-statistic p-value Confidence
Error Interval 95% (Low/High) Recession 0.0215 0.0048 4.49 0.000*** 0.01 0.03 Cash 0.1112 0.0146 7.62 0.000*** 0.08 0.14 ROA 0.5486 0.0817 6.71 0.000*** 0.39 0.71 DE 0.0039 0.0069 0.56 0.573 -0.01 0.02
BM
0.0935 0.0138 6.77 0.000*** 0.07 0.12 Tax -0.3414 0.1313 -2.60 0.009*** -0.60 -0.08 SNWC -0.0115 0.0012 -9.59 0.000*** -0.01 -0.01 - cons 0.1335 0.0563 2.37 0.018** 0.02 0.24Table 5 Regression Results
Significance at: 1 percent(***), 5 percent(**) and 10 percent(*)
From table 5 it can be concluded that there is a significant effect of the economic recession on the dividend payout policy. The time dummy variable is significant at an alpha of 1 percent. DeAngelo and DeAngelo (1990) concluded also that there is an economic recession effect on the dividend payout policy of a firm. Hauser (2013) confirms this using logistic panel models on the effect of the Financial Crisis on the dividend payout policy. He provided evidence for dividends declining during the Financial Crisis.
4.4 Regression Results on Cash Retention and Dividend Policy
In order to investigate the effect of the Financial Crisis on the relationship
between Cash Retention and DPR, three regressions would be run to test if there changed anything during the overall period. The three time periods are as
follows: pre-crisis (2000-2006), Financial Crisis (2007-2009) and post-crisis (2010-2015).
DPR 1. Pre-crisis 2. Financial Crisis 3. Post-crisis Recession 0.0215*** (0.0048) Cash 0.0825*** -0.1469*** -0.0116 (0.0172) (0.0145) (0.0197) ROA -0.1874 -0.5964*** -0.1693 (0.1182) (0.1026) (0.1629) DE 0.0130*** 0.2755*** 0.3970*** (0.0017) (0.0228) (0.0174) BM 0.0345* -0.0972*** -0.0793** (0.0194) (0.0151) (0.0321) Tax -3.4774*** -1.5551 *** -1.3648*** (0.2311) (0.1437) (0.1400) SNWC 0.0468*** -0.057*** -0.0072*** (0.0032) (0.0054) (0.0016) _cons 0.9631 *** 0.7609*** 0.2481 *** (0.1020) (0.0632) (0.0634) Fixed effects No No No
Random effects Yes Yes Yes
R"'2 0.5855 0.5861 0.7950
Observations 920 280 720
Table 6 Regression Results per Time Period
Significance at: 1 percent(***), 5 percent(**) and 10 percent(***)
Table 6 provides an overview of the regressions over the separable time periods. Besides looking into the economic recession effect, we would also like to
investigate the influence of cash retention on the dividend policy of a firm. The first regression in table 6 indicates a positive significant influence (0.0825 at significance level 1 percent) of the cash ratio on dividend policy. When the cash ratio is 1 unit higher, the dividend payout ratio increases with 0.0825.
The second and third regression in table 6 are specified on the Financial Crisis and respectively the post-crisis period. The second regression indicates an negative significant influence (-0.1469 at significance level 1 percent) of the cash ratio on dividend policy. The influence of the cash ratio becoming negative is possibly due to the economic recession. Retaining cash before the crisis would create value for the firm if used according to the precautionary motive from agency theory. The marginal value of a dollar cash increases when it is saved on a precautionary basis (Chen et al., 2018). Larger cash savings helps firms to maintain continuous dividend payout policies, commitment to its shareholders and thus, has a positive effect on the dividend payout ratio. Meanwhile, during
the Financial Crisis US corporate firms used larger cash savings for capital investments rather than paying dividends. This is to prevent costly external finance.
The third regression in table 6 indicates a shift of cash retention determining the dividend payout ratio. After the crisis, the effect of the cash ratio (-0.0116) on dividend payout ratio is not found to be significant. There could however not be concluded with certainty that there was no effect of the cash retention during this period. Moreover, a different dataset or smaller sample size may yield a significantly result.
S. Conclusion and Discussion
This paper seeks to answer the following research question: 'Is there a
difference in the Economic Recession Effect on the relationship between Cash Retention and Dividend Payout Policy?' For this purpose, an empirical analysis is carried out on the Economic Recession Effect and factors determining the
dividend payout policy.
The available data indicate that there is an Economic Recession Effect to be found in the dividend policy of US stock-listed firms (table 5). This evidence supports the first hypothesis, in section 2.4. The results of the first regression are also in line with previous literature (DeAngelo and DeAngelo, 2006; Hauser, 2013; Chen et al., 2018). The dividend policy did shift during the Financial Crisis. Hauser (2013) allocates the shift of dividend policy to firms emphasizing the financial flexibility instead of its dividend policy.
The Economic Recession Effect on the relationship between cash retention and dividend payout is less clear. The data indicate that a shift in cash retention occurred, as consequence of the Financial Crisis (table 6). The influence of the economic recession may justify the shift in the relationship of the cash ratio with respect to the dividend payout ratio. However the insignificant effect of cash retention after the crisis is not mentioned by existing literature. Therefore no definitive conclusion can be drawn. For now it can be concluded that there is enough evidence to conclude that the Financial Crisis most likely negatively affected the cash retention of firms.
Another limitation of this paper is the omission of a industry sector specific variable. Including such a variable may give a better insight in the management of dividend policies. It may also explain the cash retention of firms among firms located in different industry sectors.
The lack of explaining the Economic Recession Effect on cash retention and the omission of a industry sector variable, provides an opportunity for further research. Perhaps another, more balanced sample size and, additional variables would improve the trade-off between potential bias and efficiency issues.
6. References
AI-Malkawi, H. N., Bhatti, M. I., & Magableh, S. I. (2014). On the dividend smoothing, signaling and the global financial crisis, Economic Modelling, 42, 159-165.
Almeida, H., Campello, M., & Weisbenner, S. (2004). The cash flow sensitivity of cash, Journal of Finance, 59, 1777-1804.
Berk, J., & DeMarzo P. (2013). Corporate Finance (third edition). Esssex: Pearson Education Limited, ISBN: 9780273792086.
Bliss, B. A., Cheng, Y., & Denis, D. J. (2015). Corporate payout, cash retention, and the supply of credit: Evidence from the 2008-2009 credit crisis, Journal
of Financial Economics, 115, 521-540
Brav, A., Graham, J. R., Harvey, C. R., & Michaely, R. (2005). Payout policy in the 21st century, Journal of Financial Economics, 77, 483-527.
Case, K. E., & Shiller, R. J. (2003). Is There a Bubble in the Housing Market?,
Brooking Papers on Economic Activity, 2, 299-342.
Chen, H., Chou, R., & Lu, C. (2018). "Saving for a rainy day: Evidence from the
2000 dot-com crash and the 2008 credit crisis", Journal of Corporate Finance
48, 680-699
Chetty, R., & Saez, E. (2005). Dividend Taxes and Corporate Behavior: Evidence from the 2003 Dividend Tax Cut, The Quarterly Journal of Economics,
120(3), 791-833.
DeAngelo, H., & DeAngelo, L. (1990). An Empirical Investigation of Troubled NYSE Firms, Journal of Finance, 45(5), 1415-1431.
DeAngelo, H., & DeAngelo, L. (2006). The irrelevance of the MM dividend irrelevance theorem, The Journal of Financial Economics, 79, 293-315.
Denis, D., & Osobov, I. (2008). Why do firms pay dividend: international
evidence on determinants of dividend policy, Journal of Financial Economics,
89, 62-82
Easterbrook, F. H. (1984). Two Agency-Cost Explanations of Dividends, The
American Economic Review, 74( 4 ), 650-659.
Eichengreen, B., Mody, A., Nedeljkovic, M., & Sarno, L. (2012). How the Subprime Crisis went global: Evidence from bank credit default swap spreads, Journal of International Money and Finance, 31(5), 1299-1318.
Fama, E., & French, K. (1988). Dividend yields and expected stock returns,
Fama, E., & French, K. (2001). Disappearing dividends: changing firm
characteristics or lower propensity to pay?, Journal of Financial Economics,
60, 3-43.
Fernando, C. S., May, A. D., & Megginson W. L. (2012). The Value of
Investment Banking Relationships: Evidence from the Collapse of Lehman Brothers, Journal of Finance, 67(1), 235-270.
Gamba, A., & Triantis, A. (2008). The Value of Financial Flexibility, Journal of
Finance, 63(5), 2263-2296.
Gordon, M. J. (1963). Optimal Investment and Financing Policy, Journal of
Finance, 18(2), 264-272.
Hauser, R. (2013). Did dividend policy change during the financial crisis?,
Managerial Finance, 39(6), 584-606.
Ivashina, V., & Scharfstein, D. (2010). Bank lending during the financial crisis of 2008, Journal of Financial Economics, 97(3), 319-338.
Jensen, G. R., Solberg, D. P., & Zorn, T. S. (1992). Simultaneous Determination of Insider Ownership, Debt, and Dividend Policies, The Journal of Financial
and Quantitative Analysis, 27(2), 247-263.
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure, Journal of Financial
Economics, 3(4), 305-360.
Kallapur, S. (1994). Dividend payout ratios as determinants of earnings response coefficients, Journal of Accounting and Economics, 17, 359-375.
Lintner, J. (1956). Distribution of Incomes of Corporations Among Dividends, Retained Earnings, and Taxes, The American Economic Review, 46(2), 97- 113.
Miller, M. H., & Modigliani, F. (1961). Dividend Policy, Growth, and the Valuation of Shares. The Journal of Business, 34(4), 411-433.
Modiligiani, F., & Miller, M.H. (1958). The cost of capital, corporate finance, and the theory of investment, The American Economic Review, 48, 261-297.
Morck, R., & Yeung, B. (2005). Dividend Taxation and Corporate Governance, Journal of Economic Perspectives, 19(3), 163-180.
Rozeff, M. S. (1982). Growth, beta and agency costs as determinants of payout ratios, Journal of Financial Research, 5(3), 249-260.
Smith, C. W., & Watts, R. L. (1992). The investment opportunity set and
corporate financing, dividend, and compensation prices, Journal of Financial
7. Appendix
7.lVariable Description
Variable Calculation Source
DPR Dividends as fraction of Jensen et al. ( 1976) Income Before Extra Items
Recession No Recession: 0 Hauser (2013) Recession: 1
Cash Cash+ Short - term Investments Chen et al. (2018) Current Liabilities
ROA EBITDA Hauser(2013)
Total Assets
DE Debt DeAngelo & DeAngelo
Equity
(2006)
BM Book Equity Fama & French (2001)
Market Capitalization
RD R&D Expenses Jensen et al. (1992) Sales
Tax Income Tax Chetty & Saez (2005)
Pretax Income
SNWC Sales per dollar of Working Efficiency Measure Capital Wharton WRDS
Labor Labor Expenditures Fama & French (2001)
7.2 The Hausman Test . hausman fe re
,t-3, -=-:::::c, :!.:.a-;;-
,·:_t-·:_s,,
~.:...
. 0880639 .0856848 .002379 .0010143
t = ==~-=.:..=te~t ~~:!e::: ~= a~:! ~a; :bca.:.~ej !:::~~ xc:::eç S .:.~:::~.=.:..=te~t ~~je::: ~a, e!!.:.c.:.e~t ~~je:::~-- :bca.:.~e:! !::::~ xc:::ec
Test: :;:" • .:.: I~) F::: :t · :::- . .:.: tb-SI': ,·:_t-':_3) ',-:.!: it-3, 5.50 0.0190
7.3 Cash Retention Effect
Stata output on the cash retention effect for each seperable time period. Pre-crisis (2000-2006)
. xtreg DPR Cash ROA DE BM Tax SNWC 1n 1/920, re
:::~r. ~a::at:~: g1cctescl :;~:,-J:~:: :I:.;, ;:: :·;r..: ... -
....
- :·:":.":-ei.:..:. :::::::.: 1·.: .:., X1 0.0909 0.9160 0.5855 ;,·a:.:: ::::..:..: 16) F - ->· -·-. a::ç rr.ax 920 5 152 184.C 192 1289.76 C.0000 :-:':'::. ;;:·l-
• c..•. ... .0824936 -.187362 .013007 .0344841 -3.477359 .0467756 .963139 =-- -· E::::.=
F=
: !":¾ - - - ç ::--.:.':':.:·:a:.. .0171925 4.80 o.cco .C487969 .11619C3 .11822C6 -1.58 D .113 -.4190701 .C44346 .0016758 7.76 0.000 . 009722 6 .0162915 .0194439 1. 77 0.076 -.0036253 . 0725934 . 231105 -15.05 0.000 -3.930316 -3.024402 . 003211 14.57 0.000 . 0404821 .053069 .1019742 9.44 0.000 .7632733 1.163005 =.:.~.a_·..:. 0 .05901416 0 f :::a=~.:.::-. Financial Crisis (2007-2009). xtreg DPR Cash ROA DE BM Tax SNWC in 921/1200, re
:l:.:
~==~r.
~a::at:~: gicctescl 280 3 :t.: r-~= ç:::·;r.: · .. ;.:.t::..:.:-. 0.5327 0.8820 0.5861 ~\al:i. :::..:..: 161 F:: :l: .... .., ... a·:ç r.".GX 40 93.3 192 386.52 0.0000 :a.;,:: s:-: Îc?.>: -.1468646 -.5963789 .2754804 -.0972147 -1.555118 -.0574355 .7608805 .3t.:!.-=---
z F : :-: ½ • - _ç ::-:!:-=::·:a:. . .0144873 -10.14 0.000 -.1752592 -.11847 .1025907 -5.81 0.000 -.797453 -.3953047 .0228434 12.06 0.000 .2307081 .3202526 .0151133 -6.43 0.000 -.1268363 -.0675931 .1436513 -10.83 C.000 -1.83667 -1.273567 .0053921 -10.65 0.000 -.0680037 -.0468672 .0631989 12.04 0.000 .6370129 .8847482 ::.·.:.Ç'":':.a_·.: 0 .02484435 0 ·; :)Post-crisis (2010-2015)
. xtreg DPR Cash ROA DE BM Tax SNWC 1n 1201/192C, re
