Motivation and Reason: Fama and French (2001) and DeAngelo and DeAngelo (2006) show that all advance life-cycle explanations for dividends that rely on the trade-off between the advantages and the costs of retention, but the literature offers only a rough empirical idea of the characteristics that differentiate firms that pay dividends from those that do not. Therefore, the purpose of this paper is to test the life-cycle theory by assessing whether the probability a firm pays dividends is positively related to its mix of earned and contributed capital. Moreover, this paper gives the empirical reason about decline of the number of dividends paid firms indicated by Fama and French (2001) and DeAngelo and DeAngelo (2004).
Overall Context of the Literature: this paper does not quote too many literatures. For example, authors still focus on comparing with Fama and French's (2001) findings about the propensity to pay reduction of dividends for firms. In addition, they also discuss about other theories, such as signaling and catering by Baker and Wurgler(2004), which is inconsistent with their results. Meanwhile, they use Jensen's (1986) agency cost to the life-cycle theory in order to explain the massive payouts of the largest dividends paying firms. Moreover, they consider the floating costs or asymmetric information problems as in Myers and Majluf's (1984) pecking order theory to cause firms to forego dividends entirely.
Data and Methodology: the sample of data includes the Center for Research Prices (CRSP) industrial firms with dividends and earnings on Compustat exclude nonfinancial and nonutility firms from 1973 to 2002. In addition, they use the Fama and French's (2001) and Fama and Macbeth's (1973) statistical methodology to test whether the probability that a firm pays dividends depends on the mix of internal versus external equity in the capital structure.
Results and Meaning: Consistent with a life-cycle theory of dividends, the fraction of publicly traded industrial firms that pay dividends is high when retained earnings are a large ratio of total equity (and of total assets) and falls to near zero when most equity is contributed rather than earned. They observe a highly significant relation between the decision to pay dividends and the earned/contributed capital mix, controlling for profitability, growth, firm size, total equity, cash balances, and dividend history, a relation that also holds for dividend initiations and omissions. The mix of earned/contributed capital has a quantitatively greater impact than measures of profitability and growth opportunities. They show a massive increase in firms with negative retained earnings. Controlling for the earned/contributed capital mix, firms with negative retained earnings show no change in their propensity to pay dividends from the mid-1970s to 2002, while those whose earned equity makes them reasonable candidates to pay dividends have a propensity reduction that is twice reduction in Fama and French. Finally, agency cost-inclusive life-cycle theory most powerfully explains the dividend decisions of the 25 largest longstanding dividend payers.
My impressions of the paper: they are the first to connect the life-cycle of firms with agency cost to explain the firms' dividends policy. In addition, they also discuss about other theories, such as signaling and catering, and asymmetric information. However, this paper only test the US market and interpret data first, use the theory to explain later, which cause the data mining problem.
Written Report
Introduction
This paper is the further research of Fama and French (2001) and DeAngelo and DeAngelo (2006), because when they describe life-cycle explanations for dividends that rely on the trade-off between the advantages and the costs of retention, they only offer a rough empirical idea. In addition, this paper answers the question why the number of dividends paid firms decrease showed in DeAngelo and DeAngelo (2004). Therefore, the purpose of this paper is to test the life-cycle theory by assessing whether the probability a firm pays dividends is positively related to its mix of earned and contributed capital.
Data and Methodology
Comparing with previous researches, this paper uses the similar sample of data covering from 1973 to 2002. However, authors test many other explanatory variables. For example, they estimate a firm's stage in its financial life cycle by the amount of its earned equity both relative to total common equity (RE/TE) and to total assets (RE/TA). They also measure other explanatory variables to affect the decision of paying dividends, such as profitability measured by return on assets (ROA), growth measured by sales growth rate (SGR), size measured by the asset (NYA) and the equity value (NYE).
Moreover, they use the Fama and French's (2001) and Fama and Macbeth's (1973) statistical methodology to test whether the probability that a firm pays dividends depends on the mix of internal versus external equity in the capital structure. They use the multivariate logit model that takes the payment/nonpayment of dividends as the dependent variable, with RE/TE (or RE/TA) and profitability, growth, size, etc., as the explanatory variables. They run separate logit regressions for each of the 30 sample years to obtain a time series of fitted coefficients, which are inputs to t-statistics that gauge the statistical significance of the explanatory variables. They report t-statistics unadjusted for serial correlation and compute t-statistics adjusted for serial correlation using the Newey and West (1987) procedure. While the resultant statistics on RE/TE and RE/TA are reduced under this approach, the estimated coefficients on these variables remain highly significant in every model specification.
Empirical Results
The impact of the earned/contributed capital mix
Fully consistent with the findings of Fama and French, the probability that a firm pays dividends is significantly and positively related to profitability and size, and negatively related to growth and cash holdings. Moreover, the highly significant coefficients on lagged dividend status show that whether a firm paid dividends last year is a strong predictor of whether it will do so this year. The weak evidence of a systematic TE/TA effect indicates that the strong positive impact of RE/TE (and RE/TA) on the probability of paying dividends that they document it is not attributable to a firm's use of equity per se, but rather to its mix of internal and external capital, controlling for profitability, growth, firm size, total equity, cash balances, and dividend history, a relation that also holds for dividend initiations and omissions.
The impact of RE/TE versus profitability/growth
The very largest firms exhibit a substantially higher probability of paying dividends than do firms that rank in the middle or at the low end of the NYSE equity value spectrum, after controlling for profitability, growth, RE/TE, etc. Additional the mix of earned/contributed capital like the size has a quantitatively greater impact than measures of profitability and growth opportunities.
The upsurge in firms with negative RE/TE and the reduced propensity to pay dividends
The secular decline in firms' propensity to pay dividends could simply reflect the contemporaneous massive increase in firms with negative retained earnings. Controlling for the earned/contributed capital mix, firms with negative retained earnings show virtually no change in their propensity to pay dividends from the mid-1970s to 2002, while those whose earned equity makes them reasonable candidates to pay dividends have a propensity reduction that is twice the overall reduction in Fama and French (2001).
Retention versus distribution in the life-cycle theory
The sheer scale of 25 firms' dividends, both in the absolute and in terms of the asset and capital structure consequences of a full retention policy, and their persistence over many years are difficult to explain by non-life cycle theories such as signaling and catering, since Signaling theory predicts that a firm will pay dividends when outside investors find it especially difficult to gauge the firm's future prospects, but few such firms pay dividends at early stage of life cycle. In addition, firms have catering incentives to pay dividends when the market overvalues dividend payers, so firms pay dividends depending on stock prices and independent of their RE/TE ratios, but this is inconsistent with their empirical results. Moreover, with flotation costs and asymmetric information problems as in Myers and Majluf's (1984) pecking order theory, managers will distribute the full value of the free cash flow stream over the life of the enterprise, but will distribute nothing until the probability is zero that unanticipated attractive new investments might force them to seek outside capital. Such asymmetric information problems can cause firms to forego dividends entirely until the final period of their lives. Therefore, they conclude that the agency cost-inclusive life-cycle theory offers a more plausible explanation for the massive payouts of the 25 firms, because firms pay dividends when the agency and other costs of retaining free cash flow exceed the flotation cost and other benefits of retention to distribute substantial dividends consistently over long horizons.
Conclusion
Consistent with a life-cycle theory of dividends, the fraction of publicly traded industrial firms that pay dividends is high when retained earnings are a large portion of total equity (and of total assets) and falls to near zero when most equity is contributed rather than earned.
My impressions
After DeAngelo and DeAngelo critique on Fama and French's (2001) conclusion and describe that aggregate real dividends paid by industrial firms increased over the past two decades, even though the number of dividend payers decreased by over 50%, they propose the agency cost-inclusive life-cycle theory of dividends. Therefore, they are the first to connect the life-cycle of firms with agency cost to explain the firms' dividends policy.
Meanwhile, they have different opinions about the agency problem. Fama and French (2001) show that better corporate governance technologies will lower the benefits of dividends in controlling agency problems, when they observe the decline number of dividends paying firms. While, when DeAngelo and DeAngelo indicate dividends concentration and firms with largest dividends paying still account for a majority of the aggregate dividends, agency cost is useful to explain the life-cycle theory of dividends. Therefore, this is really strong point to attack Fama and French (2001).
In addition, they also discuss about other theories, such as signaling and catering, and asymmetric information, but those theories fail to explain the behavior and motivation of firms paid dividends.
However, this paper only test the US market and interpret data first, use the theory to explain later, which cause the data mining problem. Therefore, they should test other markets, such as Japan, UK, and Canada. Moreover, this agency cost-inclusive life-cycle theory may not be useful in the small business. For example, family business has more trust, so it seems that agency cost can not explain the behavior of small business. On the other hand, low law protected shareholders take whatever dividends they can get without considering the stage of the firms, so life-cycle theory can not explain behavior of firms in those countries with poor protection of law.
Executive summary 2: Why do firms pay dividends? International evidence on the determinants of dividend policy
Motivation This paper examines the cross-sectional and time-series evidence on the propensity to pay dividends in several developed financial markets (the US, Canada, the UK, Germany, France and Japan) over the period 1989 to 2002. The motivations of this article is to see whether the characteristics of dividend payers and nonpayers are common across countries, whether these characteristics have changed over time, and whether firms in other countries exhibit a declining propensity to pay dividends in recent years.
Fits into the literature This paper finds little evidence supporting the signaling, clientele and catering theories but provides evidence for the life-cycle theory. The findings complement and extend the recent studies of payout policies outside of US. In addition, this is the first paper provides international evidence on the importance of the earned/contributed equity mix in dividend policies.
Data and methodology The data selection is similar as Fama and French (2001). The utilities, financial firms and firms with negative book equity are excluded in the sample. The data covering from 1989 to 2000 are collected via Thomson One Banker Analytics. Univariate analysis and multivariate analysis (annual logit regressions) are used to describe the determinants of the propensity to pay dividends. Baseline estimates are used to examine the changes in the propensity to pay dividends.
Results Consistent with Fama and French (2001), the findings in this paper show that dividend payers tend to be larger and more profitable firms. The likelihood of paying dividends is positively related to firm size, profitability, and the earned/contributed equity mix in all six countries. The proportion of dividend-paying firms declines over time in all six countries. Little evidence exists that the shortfall in dividend payers is due to firms unexpectedly abandoning dividends. Also the results provide evidence on the concentration of earnings and market capitalization among dividend payers.
Meaning of the results The empirical results show that the all of the six countries have the similar determinants of the propensity to pay dividends that large, profitable and those with greater earned/contributed equity are more likely to pay dividends. In addition, in the last decade, firms form a greater representation of typical characteristics of nonpayers. The level of the decline in the propensity to pay is much smaller in this paper compares with the results reported in Fama and French (2001). And the shortfall is due to the newly listed firms that fail to initiate dividend. The evidence of concentration of dividends among firms with high earnings inconsistent with clientele and signaling explanation as it is not likely for the investors to form a well-diversified portfolio when 90% of the market's capitalization is in dividend paying stocks. Finally, the findings in this paper do not support the catering incentives in explaining the changes in the propensity. In fact, there is few firms change their dividend status more than once over the sample period and even fewer firms corresponding to the changes in the dividend premium.
Impressions This paper is quite impressive as it provides the most recent international evidence on the dividend policy but also re-examine the other conventional ideas on dividends, such as clientele and signaling. The methodology in this paper is quite comprehensive and the whole article is quite easy to read. However, this paper is the first paper provides international evidence on earned/contributed equity in examining the dividend policy. This makes the current paper distinct from the similar studies on dividend policy. I found one flow of this paper is that it missed the investment opportunity and information asymmetry theories out. The traditional view on the dividend policy is that the managers have more private information than the investors and the dividend may convey information to the public. Whether the investment opportunity and information asymmetry theories relating to the propensity to pay dividend remains unsolved in this paper.
Written Report 2
Denis, D.J. & Osobov, I. (2007). Why do firms pay dividends? International evidence on the determinants of dividend policy. Journal of Financial Economics, In Press.
Introduction
Why do some firms pay dividends while others do not? The financial researchers have tried to answer this question by using the signaling and clientele theories. More recently, DeAngelo and DeAngelo (2006) combine the agency with investment opportunities to form a life-cycle theory. While Fama and French (2001) report a substantial decline in the proportion of firms paying dividends. The purpose of this paper is to extend this literature by conducting both cross-sectional and time-series tests on the propensity to pay dividend in six developed financial markets, including the US, Canada, the UK, Germany, France, and Japan. The authors are motivated to find out whether the characteristics of dividend payers and nonpayers are common across countries, whether these characteristics have changed over time, and whether firms in other countries out of US exhibit a declining propensity to pay dividends in recent years.
Contribution
One of the contributions is that it is the first study to provide international evidence to show the importance of the earned/contributed equity mix in dividend policies. Another contribution is that it provided international evidence on signaling, clientele and catering explanations through the analysis of the concentrations of dividend payment and also the association between the dividend premium and the propensity to pay dividend. Last important contribution is that it narrows down the scope of the disappearing dividends is due to the decisions of newly listed firms.
Sample selection
The data are collected via Thomson One Banker Analytics. The firms need to have the information on total assets, common equity, net income, interest expense, and market capitalization. Following Fama and French (2001), utility, financial and firms with negative book equity are excluded in the sample. Due to the database has several limitation, the available study period is from 1989 to 2002.
Determinants of propensity to pay dividends
Univariate analysis
This paper examines the profitability, book equity, growth opportunities and firm size as the possible determinants of the propensity to pay dividends. It is found that the typical characteristics are that the firms pay dividends are more likely to be large and profitable companies, which consistent with Fama and French (2001). Also, it seems like the dividend paying firms tend to have more valuable growth opportunities. These finding supports LaPorta, Lopes-de-Silanes, Shleifer, and Vishny (LLSV, 2000). By measuring the earned/contributed equity mix, the authors find that the dividend payers have higher ratios of retained earnings to total equity than the nonpayers. The results are strong in all of the six countries which indicate the importance of earned/contributed equity mix relating to the proportion of firms that pay dividends. The test of the evolution of firm characteristics through time showing that it is more likely the typical characteristics of non payers are represented in each country.
Multivariate analysis
Logit regressions are run to measure the marginal effects of profitability, growth, opportunities, size, and the earned/contributed equity mix on the probability of dividend payments. The regressions give the similar results as in univariate analysis that the probability of paying dividends is positively related to firm size, profitability, and the earned/contributed equity mix for all six countries. Furthermore, the results are generally robust by using the Newey and West (1987) and Peterson (2007) procedure. The authors also estimate the logit model separately for those firms that paid a dividend in the previous year (payers) and those did not (nonpayers). The results show that the firms stick to their dividend policy. That is, firms pay dividends currently is more likely to continue to pay while the firms do not pay dividend currently are reluctant to initiate.
International evidence on changes in the propensity to pay dividends
As reported by Fama and French (2001), there is substantial decline in the proportion of dividend payers among public traded companies in the US in the last two decades. A similar declining is found in this paper in all six countries and the authors believe that the changes in the characteristics of publicly traded companies lead to a large proportion of the decline in dividend payers in the sample countries.
On the other hand, the authors argue that the magnitude of the decline found in the article is economically small. This modest decline may be caused by the short coverage of the study period in this research or may be due to the firms included in the database overtime. It is likely that the database includes the large firms first, especially during the period from 1981-1985 and then includes more and more small firms from that point onwards. Thus, the inclusion of the small firms drags down the proportion of the dividend payers. Two additional tests were taken to see whether it is the case. Not surprisingly, for the specific period from 1989 to 1993, smaller reduction are found which means the firms enter the sample after 1978 drives the reduction in the propensity to pay dividends. The observations show that the firms enter the sample in 1993 failing to initiate the dividends cause the shortfall in the proportion of paying dividend. The reductions occur in this period account for 80% of the shortfall. Little evidence is found that the reduction is due to the dividend abandonment. The only exception is Japan as it experienced special financial problems (distress) during the sample period. The findings on the little evidence of the reduction in the propensity to pay dividend consistent with Benito and Young (2001) and Rennneboog and Trojanowski (2005) basing on the UK data. The results on changes of propensity changes over time contradict to the conclusion from Ferris, Sen, and Yui (2006) and von Eijie and Megginson (2006) while the authors argue that those two studies have defects in their data sets and methodology.
The international evidence reinforces the importance of firm size, profitability, and the earned/contributed equity mix in dividend decisions. Since they are the primary determinants in the dividend policy, the signaling theory needs to be doubted. Dividend payers are normally large and profitable companies, unlike newly listed firms they do not have the need to signal their future to the investors. Therefore, it seems like it is not appropriate to use the signaling to explain why firms pay dividends.
The Concentration of dividends and earnings
In order to examine the concentration of dividend, the authors report the percentage of all dividends paid by the top 20% of the dividend payers in each country for three subperiods: 1989-1993, 1994-1998, and 1999-2002. In each country, top 20% of the payers contribute 73.3% to 90% out of the total aggregate dividends. Also, the authors find that US, UK, and Germany, the top payers account for 90% of the aggregate earnings, which proves the concentration of dividends are highly correlated with the concentration of earnings.
The evidence of high concentration of dividend and earning is inconsistent with clientele theory. Equilibrium clientele assumes that the investor can construct sufficiently well-diversify portfolios with the desired dividend and risk level. However, with such highly concentrated market capitalization, it seems like difficult for investors building up well-diversified portfolios. At least, the findings on the concentration of dividend show that clientele and signaling considerations on propensity to pay dividends are not the first-order determinants of dividend policy. It will be more sensible if the concern of the dividend policy is the payout of free cash flows which relate to the life-cycle explanation of dividends.
Catering theory
Finally, this paper re-examines the catering hypothesis of the propensity to pay dividend. Baker and Wurgler (2004a, 2004b) predict that managers cater to investors by paying dividend when they put a stock price premium and do not pay when the investors prefer nonpayers.
There are two ways to test the catering view on dividend. Similar as Baker and Wurgler (2004a, 2004b), the first measure is the dividend premium which is defined as the difference between the log of the weighted-average market-to-book ratio of payers and that of nonpayers, where the weight is the book value of total assets. US and Canada have consistent negative dividend premium over the testing period but positive values for Germany, France, and Japan. UK has mixed results for different subperiods. In contrast, for all those six countries, the unexpected percent of dividend payer which is differences between the expected and the actual percent of dividend payers are generally positive. In the sense of catering theory, if the managers cater the investor demand on dividend, the unexpected percent should vary in different countries and different periods but the results do not fit into this pattern.
The second test is to find how and how often the firms change their dividend statuses. In this article, the authors define those companies change dividend more than once a year as "switchers". According to the catering hypothesis, there should be more initiations (omissions) than omissions (initiations) when there are positive (negative) changes around the dividend premium. If we take the difference between the number of the initiations and omissions and then run correlations on this set of figures with the change in the dividend premium in that year, the results show that few firms switch their dividend status more than once a year. Furthermore, the direction of the changes are not closely attach to the direction of the changes in dividend premium. Thus, the bottom line here is that the catering theory cannot be used as the first-order determinant of the dividend policy in the sample countries.
Conclusion
This article finds homogeneous characteristics for dividend payers across countries. Firm size, profitability, growth opportunities and the earned/contributed equity mix tend to be the key factors affecting the dividend. There is a declining trend of the propensity to pay while the shortfall is economically small. In addition, the newly listed firms fail to initiate dividend (they are expected to do so) accounting for the declining trend. Actually, the aggregate amount of the dividends did not drop during the sample period. The concentration of the dividends and earnings reject the clientele and signaling considerations for the propensity to pay dividend. At last, the catering theory is tested again and there is no supportive results found for catering view of dividend.
My impressions
This paper is quite impressive as it not only provides the current international evidence on the dividend policy but also reexamine the other traditional dividend theories like clientele, signaling and life cycle theory at the same time. The methodology in this paper is quite comprehensive and the whole article is quite easy to read. However, this paper is the first paper includes earned/contributed equity in examining the dividend policy. This makes the current paper different from the similar studies on dividend policy. One problem of this paper is that it left out the investment opportunity and information asymmetry out. The traditional view on the dividend policy is that the managers have more private information than the investors and the dividend may convey information to the public. Whether the investment opportunity and information asymmetry theories relating to the propensity to pay dividend remains unknown.
