Preponderant factors in the definition of dividend policies in Brazil Marcelo

The aim of this study was to verify whether company managers take into account the factors assigned by the literature as the one to influence dividend policies the most in decisions regarding profit sharing. The research was made with 106 public companies in Brazil. The results showed that managers give more attention to factors such as legislation, Articles of Incorporation, cash availability, opportunities for future investments, budget, and expectations regarding future profits. In consideration, factors such as the effect of inflation, policies regarding dividends of other companies, alternative investment sources for shareholders, and gains or losses with stock variation are taken less into account.


INTRODUCTION
In the field of corporate finances, there are three key decisions which are guidelines for the life of companies: investment decision, financing decision, and dividend decision (Damodaran, 2002, p. 31).Naturally, these are not the only issues regarding corporate finances, but they summarize the most important decisions the finance manager needs to take on a daily basis.
The guideline for investment decisions simply proposes that a single company must invest in an asset when the return offered by that asset is greater than the minimum return acceptable by the company (Damodaran, 2002, p. 31).This minimum acceptable return is partially or totally based on the cost of capital from third parties, and on the cost of the equity raised for financing the asset.Therefore, if the company systematically applies its financial resources in investment projects which offer a higher return to the equity employed in that project, it will be a wise decision.
The financing decision, in turn, is the one which relates to the source the funds were raised from.In that source, one makes an effort to determine the ideal proportion between shareholders' and creditors' equity, which a company must obtain in order to get financed.The determination of that ideal percentage between onerous liabilities and equity is known as the optimal mix of financing, and it allows the company to minimize the weighted average of capital costs, thus increasing its economic value.
For the two decisions which were previously presented, there are very objective parameters which may be summarized like this: the investment decisions must be taken whenever an asset offers a return which is higher *Corresponding author.E-mail: taba.jr@usp.br.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License than the cost of the capital which will finance it, and the financing decisions must always be taken based on the search for the company's optimal structure of capital.
However, the third great decision regarding corporate finances is related to the destination of profits in the company.This decision is not based on a single parameter.At the end of an accounting period, managers, along with Board of Director's members, with the approval from the audit committee and the shareholders, have to decide whether they will reinvest the profit in the company or give it back to shareholders, and how much of it they will return and retain.Besides that, if there is profit sharing, what is the ideal way -among all other ways to give back funds to shareholders -that will be used (e.g., dividends, Interest on Equity (IOE) or the repurchase of shares); how ofter are payments made (monthly, quarterly, biannually, annually, etc.Besides those decision factors inside the company, managers must also be aware of the several aspects related to the information content dividends sent to the market and the aspects related to shareholder's behavioral finances.That happens because several studies, among which is Graham et al. (1962, p. 518), show that, due to market imperfections, the distribution of dividends itself is only able to raise stock value, based on the fact that the price of a share is due to its capacity to pay dividends.Hence, out of two companies with the same power to generate profits and with the same placement in their sector, the one which pays greater dividend volumes will always be negotiated for a higher price.
Regarding behavioral aspects, Gordon (1963) states that dividends are desired by shareholders who feel more aversion to risks, because dividends are certain gains in the present, whereas the appreciation of stocks is an uncertain event in the future.The several studies in that research area regarding dividends gave rise to the socalled bird-in-hand theory, based on the dictum which says that one bird in the hand is worth two in the bush.
Another dividend theory related to behavioral aspects from shareholders is called clientele effect.Some studies in that area were developed, for example, by Pettit (1977), who observed that safer companies with older and less capitalized investors were inclined to pay higher volumes of dividends than companies with richer and younger investors.
Therefore, the dividend decision is observed to have a singularity in investment and financing decisions.While the investment decision has as parameter the return from investment projects which exceed the cost of capital and that the financing decision tries to reach the optimal structure of capital which is capable of minimizing the weighted average of equity and third-party capital costs, the dividend decision does not have only one parameter for its definition.
Based on the information exposed so far, the main question which guides this research is: 'In practice, which factors are taken into consideration by Brazilian public company managers at the moment they decide to destine the profits obtained by the company?Do Brazilian public company managers take into account the factors pointed out by scientific studies which were developed so far, ones which should be considered upon the dividend decision?' The goal of this research was to survey the main factors which are taken into account by Brazilian managers when they define how the profits gained by their companies will be destined.
To fulfill this research goal, a review of literature was initially conducted in order to survey the factors which were pointed out by the scientific studies that were developed so far, which should be taken into consideration upon the dividend decision.Following that, the singularities of Brazilian legislation on the profitdestination process were analyzed.Those surveys served as base to prepare a questionnaire which will try to verify if, in practice, Brazilian public company managers take into account the factors pointed out by the academic studies as being fundamental upon the dividend decision.The questionnaire answers served to find possible deficiencies in the decision-making process from Brazilian corporate managers to destine profits.

REVIEW OF THE LITERATURE
The seminal article which inserted dividends in academic discussions and which gave rise to the whole discussion on the information content of dividends was the one by John Lintner, called ""Distribution of Incomes of Corporations among Dividends, Retained Earnings, and Taxes", which was published in 1956.In that paper, Lintner conducted a survey with more than 600 industrial companies, and interviewed the manager of 28 American companies from 1947 to 1953.The objective of the author was not to analyze only one specific aspect of dividends, but companies' policies regarding dividends, in its broad sense.Because of that, his conclusions are broad, but they can be summarized in four great findings.
Firstly, Lintner (1956) observed that managers are reluctant to alter their company's dividend policies, for fear of having to change their decision in the future.In that regards, as observed by the researcher, what concerns managers the most is the fact they cannot maintain a high payout index, once a future cut on dividends could be misinterpreted by investors, which would case share negotiation values to drop.
Due to that fact, Lintner (1956) obtained a second important finding.The fact of managers being reluctant to alter their dividend policies makes the historical series of dividends to be softened.As dividends are calculated based on profits assessed by accountants, whenever there is a rise in profits, dividends also rise; however, not at the same proportion.The same happens when profits decline: Dividends are reduced at a lower proportion than the decrease in profits.Thus, great profit alterations are not reflected on dividends.
This behavior leads us to Lintner's (1956) third finding.Managers are more concerned with percentage variations in dividends than with their absolute value in monetary units.That is so because managers believe that the market privileges companies which can establish the series of dividend payments.Because of that, in case companies feel the need to adjust their dividend policies, they need to do it gradually, aiming to minimize possible negative reactions from the market, which tend to be partial to payout ratio stability.
Finally, the author observed that dividends had a relation with company life cycles.The researcher observed that companies in their initial stages, with a high growth phase, paid insignificant dividends.In some cases, they paid not dividends at all.In turn, companies which were consolidated in their operating sectors and which had higher profit predictability had a more elevated payout, which would rise smoothly in time, and would rarely fall.In the long term, when companies face a low growth stage or even decrease in profits, they pay high volumes of dividends.
With that study, Lintner (1956) started the academic discussions on the determining factors to define dividend policies, and provided the bases to develop that research line, which would be much explored afterwards.
It was not long after the publication of Lintner's (1956) work that an opposite school of thought was created, one which stated dividends are relevant to shareholders.The first ones to do it were Franco Modigliani and Merton Milller, in 1961.That year, the authors presented a study which defended that dividends are irrelevant to maximize shareholders' wealth.According to the researchers, stockholders would feel indifferent as to whether they received dividends or had their share of profit reinvested in the company, because future gains with the appreciation of the share would correspond to dividends received during present time.For that argument to be supported, Modigliani and Miller (1961) created a hypothetical world in which companies live under conditions expected in the best of worlds regarding the market.
Those ideas were soon disputed.The first researcher to oppose Modigliani and Miller's (1961) conclusions was Myron Gordon, in an article called "Optimal Investment and Financing Policy", published in 1963.In that paper, Gordon defends the thesis that, in real life, shareholders are not individuals who take perfectly rational measures, and, thanks to the uncertainties linked to the future of any company, prefer to receive dividends today rather than wishing to wait for a certain appreciation in that company's shares.Those ideas from Gordon (1963) were called "bird-in-hand theory", in regards to the popular adage which says that having a bird -dividends -in the hand today is more worth than having two birds in the Ambrozini and Pimenta 533 bush -shares appreciating for the future.The lack of taxes, a condition which is seen in the perfect market envisioned by Modigliani and Miller (1961) was disputed by Elton and Gruber (1970).The authors showed that shareholders from companies located in countries which charge heavier taxes on dividends than on incomes which arise from stock sales -such as observed in the USA from 1966 to 1969 -tend to prefer to reinvest the profits rather than receiving the dividends.In turn, the premise from Modigliani and Miller (1961) concerning shareholder's rationality was disputed by Gordon (1963), who showed shareholders not only act rationally, and, due to future economic uncertainties, prefer guaranteed dividends in the present to the futureand uncertain -appreciation of their shares.
In Brazil, generally speaking, the surveys conducted on dividends aimed at testing the empiric findings verified in the American market.The clientele effect pointed out by Elton and Gruber (1970) was initially rejected in the Brazilian market by Brito and Rietti (1981), and later on, in many other studies, among which are Procianoy and Verdi (2003) and Santos et al. (2004).
Studies aiming to verify the signaling power of dividends in Brazil -many of which using similar methods to the ones used by the American authors -were conducted by Martinez (2005), Nossa and Nossa ( 2007), and Decourt, Procianoy and Pietro Neto (2007).Nonetheless, the investigation regarding dividend/IOE changes, and changes in future profits in Brazilian companies did not find unanimous results.There is no consensus as to the ability dividends have to influence the Brazilian companies' stock market price.
Brazilian legislation interferes, in countless aspects, in the decision from Brazilian companies to destine profits.Among them, some stand out: (i) the mandatory payment of a minimum 50% dividend of net reserve-adjusted profits, in case of companies with Articles of Incorporation which do not mention that; (ii) the mandatory payout of 25% of profits in case the Articles of Incorporation are altered; (iii) the existence of a type of stocks with priority over voting stocks -that priority also applies to dividends; (iv) an authorization for the company to pay its share of profits to managers only in case it has Articles of Incorporation which provided a distribution of 25% of dividends; (v) the assignment of voting rights to preferred stocks if the company fails to pay fixed or minimum dividends which are provided in the Articles of Incorporation for more than three accounting periods in a row; and (vi) the existence of a legal provision which replaces part of the dividends.That is called Interest on Equity.
Those Brazilian tax singularities based on the policy governing dividends were studied by Procianoy and Poli (1993), Procianoy (1996), and Procianoy and Poli (2004).More recently, the introduction of IOEs in Brazil through Law 9,249/95 got many researchers interested, among whom are Zani and Ness Júnior (2000); Paiva and Lima Besides tax and legal aspects, the economic reality Brazilian companies are inserted, mainly in inflation, justifies some studies on dividend policies from Brazilian companies.The distortion disregarding inflation causes in profit calculation, and consequently in the calculation of dividends was studied by Jaloreto (1994); Padoveze et al. (1994);and Santos (1997).
Important studies conducted in Brazil were aware of the determining factors in dividend policies in a broader sense.One of those first studies was developed by Heineberg and Procianoy (2003).The theoretical aspects studied by the authors were divided into profits or losses; taxing on dividends; inflation; indebtedness from companies; investment/growth opportunities; companies' size; and operating sector.More recently, Mota and Júnior (2007) conducted a study which aimed to evaluate the motivations which lead Brazilian companies to choose one out of three instruments to share profits: dividends, Interest on Equity, and repurchase of stocks.The authors verified that the factors which determine the choice of dividends and IOE were the existence of cash flow, the stability thereof, the constraining aspect of indebtedness, the concern with corporate management, and the existence of only few investment opportunities.This paper's literature review intended, besides offering understanding of how studies in this area were developed, to provide the bases for preparing the questionnaire which aimed to identify whether financial managers from Brazilian public companies follow, in practice, what is established by the dividend theory.The research method used in this study and the instrument for collecting data are introduced.

MATERIALS AND METHODS
The goal of this research was to survey the main factors which are taken into account in the decision regarding what to do with companies' profits.In order to do that, a questionnaire was prepared and sent to all public companies in the country.The questions which make up the questionnaire were prepared based on the review of literature on dividends, and also based on the review of Brazilian legislation, which pointed out legal and tax singularities Brazilian companies are subjected to.Such singularities interfere in their decision to destine profits.
The questionnaire comprises 20 closed-ended questions, regarding the possible factors taken into account in dividend decisions.So the factor taken into account upon the dividend decision could be quantitatively measured; in each of the questions in the questionnaire the so-called dummy variables were used . Hair Jr. et al. (2005, p. 176) define dummy variables as variables which represent a single component out of a broader structure.As per the authors, when they are analyzed together, those dummy variables measure a concept.The identification of those variables is important because they allow assigning numeric scores to the measurement of concepts in quantitative terms.
Each of the questions in the instrument to collect data aimed at measuring the degree of importance placed by managers in Brazilian public companies to factors deemed relevant when policies regarding the dividends brought up by the review of literature are defined.The list of options of answers comprised five statements which described the degree of importance given by companies to each of the factors in the policy of dividends which were brought up by each question.The statements are in increasing order of importance, and range from "disregarded factor" to "very important", and the caption of each answer was based on one example of metric scale which was presented by Hair Jr. et al. (2005, p. 187).
Each of the answers for dividend-policy variables found in each question received a score.The assignment of scores to each of the variables on companies' dividend policies intended to globally measure the degree of importance assigned by each company to each of the factors deemed fundamental to the dividend decision, as mentioned in the review of literature and in the legal review.Table 1 lists the scores from each answer according to the degree of importance for each factor, as per the methodology in Likert scale.
The assignment of numeric scores for the answers in the questionnaire is justified in order to transform qualitative into quantitative data.Flick (2009, p. 45), during the discussion on the transformation of qualitative into quantitative data, states qualitative observations may also be analyzed as per their frequency.Those frequencies in each category may be specified and compared.There are several statistical methods available for the calculation of those data.
The questionnaires were sent by e-mail to the 516 Brazilian public companies, which make up this research population.Table 2 presents a summary of questionnaires sent, with the success index of each attempt.
After the answers were collected and the data were organized, the statistic techniques regarding description, exploration, and comparison of data sets known as descriptive statistics were applied.The first measure calculated based on the results obtained through the data-collection instrument was frequency distribution.A frequency distribution lists data values individually or per interval groups, along with their frequencies; that is, their corresponding counts.Each frequency represents the number of times that answer was chosen for each given question (variable).
After frequency distributions are presented, the central tendency and dispersion measures are calculated and analyzed for each of the variables.Once the research method in this study is presented, the following section will present the results obtained with the application of the questionnaire, and will also analyze the results, comparing them with the literature regarding dividends.

RESULTS AND DISCUSSION
The list of variables used in the data-collection instrument is described in Table 3.Each of those variables gave rise to one question in the questionnaire.
As a way to summarize the results from frequency The first question in the questionnaire tried to find, among the companies in the sample, the degree of importance assigned to the mandatory minimum dividend provided in corporate legislation when the Articles of Incorporation omit the minimum payout for profit distribution.As observed in Table 4, most Brazilian companies in this research sample assign a high degree of importance to the legal aspects of dividend sharing.However, the dividend decision is understood to be, ultimately, a managerial decision; the legal interference in regards to that, may, in some cases, compromise the company's financial status.Table 4 shows central tendency and dispersion of variable 1.As observed in Table 5, the average is located between relatively important answers, which are assigned a score of 3, and very important answers, assigned 4, with a standard deviation of less than 1 score point, which, as compared to the remaining variables, is one of the smallest in this study.The second question in the questionnaire aimed to find out the degree of importance assigned by the company in regards to its Articles of Incorporation.Table 4 presents the frequency distribution of the second variable in the questionnaire.As observed in Table 4, the number of companies which answered that the verification of what is provided in the Articles of Incorporation influences their definitions of dividend policies is very important (76 respondents) is even higher than the companies which assigned that maximum degree of importance to legal aspects.This result was expected, once the Articles of Incorporation supersedes the legislation in dividend definition.Out of the 104 companies in the sample, only 4 (2.9%) responded the Articles of Incorporation are not a relevant factor on the decision to share profits.That can be explained by the fact that those companies have incomplete Articles of Incorporation as regards dividends.
The third question in the questionnaire aimed at verifying the degree of importance assigned by company managers to the cash availability for profit sharing.The obtained data are summarized in Table 4. Taking into account what was observed in the first two variables, cash availability when paying dividends is also a very important factor for Brazilian companies.Out of the 20 questions asked in the Likert scale for the data-collection instrument in this study, this third factor received the highest number of very important answers (76% of the total).
The fourth variable in this study is also related to a financial aspect of companies.However, unlike the immediate cash availability, the fourth dividend decision factor to be analyzed in this study is related to the company's budget for the following accounting period.As seen in Table 4, most companies in this study sample classify the budget for next year as very important in its decision to share profits.This variable received the third highest number of very important answers in the whole study, and only stayed behind variable 3 (cash availability), which received 79 very important answers, and variable 2 (Articles of Incorporation) which received 76 very important answers.
The fifth question in the questionnaire relates more specifically to new future investment opportunities, not only in order to replace assets for companies to keep operating, but also in order to extend the operational capacity, and thus increase the ability to generate cash and profits.As it can be seen in Table 4, the number of companies which assigned the maximum level of importance to "new opportunities for future investments" (61) was smaller than the one in the previous variable.However, if "relatively important" category (35 observations) is added to the 61 observations of "very important" category, a total of 95 observations are obtained.It exceeds the 93 observations from variable 4, regarding the budget for the next accounting period.That shows the relative importance degree of that factor when defining the dividend policies in Brazilian public companies.
The relationship between the expectation for future profits (variable 6) and the previous one (new future investments) lies in the fact that financing sources for Ambrozini and Pimenta 537 new investments are basically three: internally-generated and retained incomes, external indebtedness, and the further payment of capital from shareholders.When a company envisions new opportunities for future investments, one of the ways to finance that investment is through profit retention.If companies decide to share current profits as dividends, there is an expectation regardiing new profits, so new investments can be financed (variable 6), or else the company will be able to raise funds externally (variable 11).Once the insertion of variables 6 and 11 is justified, Table 4 presents the descriptive statistics explaining the importance assigned to future profit expectations in the company's decision to either share dividends or not in the current accounting period.
According to what is observed in Table 4 and having as a parameter the results observed in the other five previous variables, one can notice a significant drop in the number of answers within the "very important" category.So far, it is the variable of least expression.Nonetheless, by verifying the relative accumulated frequency, only 11.5% of companies are observed to assign low importance (7.7%) or disregard (3.8%) that factor in their dividend decisions.
Variable 7 sought to verify how important managers from Brazilian companies think the ability to maintain dividends in the future is.According to data from Table 4, 24 companies, or a little more than 23% of the sample, consider dividend stability in future periods as very important.If categories of scores ranging from 0 to 3 are summed, 36% of the companies in the sample are verified to assign "low importance" to the ability to maintain dividends in the future.The category with the highest number of observations was "relatively important".That shows that 40.4% of Brazilian companies assign a relative importance to stable dividends.
Variable 8 was related to the impact of inflation effects in the formation of corporate profits.The reason this variable was included in the study was to verify whether Brazilian public company managers take into account the loss of purchasing power by currencies, which is caused by inflation, in the process to attain profits, and their related destination.As one can observe in the data from Table 4, although academic studies show that inflation, the moderate one included, has the power to dent accounting profits which are built up throughout a certain period, only 12.5% of public companies in this study sample consider that factor very important in their dividend decisions.Another important finding is that, among all variables included in the questionnaire, the effect from inflation in building profits was the factor in dividend policies which received the highest number of answers in "neither important, nor unimportant" category, with 28 observations.Those results supply evidence to support the fact that the distortion power of inflation in building profits is underrated by Brazilian public companies.
Question 9 aimed to verify the degree of importance assigned to the ability for managers to raise external funds to finance investments at dividend decision time.As one can observe in Table 4, the majority of answers are concentrated on "relatively important" and "very important" categories.When they are added up, the number of observations in those two categories (72 companies) makes up for almost 70% of all answers (72/100=69.2%).
Also based on the relationship between financing and dividend decisions, variable 10 tried to verify the degree of importance assigned to managers in capital structure changes which originate from companies' decisions to share profits.The decision to retain incomes is known to increase the ratio between equity and liabilities in a company.On the contrary, returning funds to shareholders reduces equity and increases the share of thirdparty capital in an organization's financing structure.As presented in Table 4, a few more than 30% of this study's respondents stated that the shift in the company's capital structure is a very important factor to define how much profit is going to be returned to shareholders or reinvested in the company.Most answers are concentrated in "relatively important" category.To 24% of Brazilian companies, changes in the percentage of onerous liabilities and equity in companies' financing sources represent a less important factor.Only 2.9% of companies in the sample said they disregarded changes in capital structures when deciding how to manage dividends.
One of the findings observed by Lintner (1956) in his seminal study on dividends in the USA was the one which postulated that dividends were related to companies' life cycles.This finding motivated the inclusion, in this study, of a variable which aimed at discovering how important the stage of a company's life cycle was in the decisions from managers to share their profits.From Table 4, ten respondent managers stated they share dividends without regard to the current stage of their companies through their life cycle.The ones which assigned low importance to that factor were 4 in total, and the number of respondents who stated they were indifferent to their companies' life cycles reached 14 observations.Added up, those three categories made up for 26.9% of the companies in the sample.The remaining respondents were equally divided in "relatively important" and "very important" categories.
Complementing the analyses related to the stage the company is in its life cycle, question 12 aimed to find out the importance degree assigned to companies' degree of consolidation within their economic activity sectors.That is so because companies which are more consolidated within their sectors are expected to have a higher ability to share profits than companies that still need to invest funds to increase their market share, such as when getting new clients.According to the data from Table 4, the company's degree of consolidation within its operating sector is still regarded as a factor of less importance in dividend decision than it is in the life cycle of a company.The addition of companies which assigned low importance (16) to companies which disregard that factor when deciding how share profits ( 16) is higher than the 24 companies which stated that factor is very important.
Variable 13, in turn, is concerned about how important the company's profit stability is in dividend decision.The analysis of that variable must be conducted along variable 7, which relates to the importance assigned by managers on the ability to maintain dividends in the future, in order to avoid oscillations.That is so because variable 13 is related to the concern from managers as to future profit stability in the company, whereas variable 7 relates to the concern with dividend stability.
In order to measure how important the operational risk from the economic activity sector the company operates in is on dividend decision, this research included a question which relates to that.This question, measured by variable 14, may be related to variables 11 -current company stage within its life cycle -and 12 -degree of consolidation of the company in its sector -because it refers to a factor which is related to the company's operating sector.According to the data presented in Table 4, the relatively important category was the answer given by 44 companies when they were asked about the importance assigned to the operational risk in the sector the company operates in.In the remaining categories, the answers were distributed in a very similar way.According to Table 4, variable 11 is observed to present, at the same time, the highest total of points, and the smallest standard deviation as compared to variables 12 and 14.In turn, variable 12 -the degree to which the company is consolidated in its sector -presents the lowest total of points and the smallest standard deviation.Variable 14, the operational risk within the sector the company operates in, is presented at an intermediate position.
Still analyzing variables in a comparative way, the 15th question asked managers to assign the importance degree given to dividend policies of other companies in the same sector, in regards to their own dividend decisions.Among all twenty variables included in this research, variable 15 was the one with the higher number of observations in the category with the smallest importance score.No less than 29 companies (27.9% of the sample) stated they disregard the policies from their competitors in their dividend decisions.More than 70% of the observations in that variable relate to companies in the sample which said their competitors' payoff indices are irrelevant (25.0%) when they define their dividend policies or to companies which assigned low (17.3%) or no importance (27.9%) to that factor.As it can be observed in the answers to that question, there is a predominance of observations in the four least important categories, even though no scale stands out individually.
As previously presented, variable 16 aimed to measure a factor which is similar to the one in variable 15.The 16th question in the data-collection instrument asked managers to assign an importance degree to dividend policies of public companies in general.Table 4 summarizes the answers to that variable.The number of respondents who stated not to consider their competitors' payoff indices (29) was higher than the ones who said they disregard dividend policies of the other companies in the stock market as a whole (23).Besides that, the number of observations in categories of score 3 (relatively important) and 4 (very important) of variable 16 is higher than the one from variable 15.
Question 17, in turn, focused again on the organization's internal aspects which influence dividend decision.Variable 17 intended to measure the importance degree assigned by managers to the company's longterm objective to the profit sharing index.The inclusion of that variable intended to verify whether the companies' long-term financial planning and the determination of a future target-dividend affect the definition of current profit payouts.That variable 6 also relates to variable 6 (future profit expectation), because it relates to the company's prediction horizon and to aspects from the organization's capital budget.
The three last questions in the questionnaire relate to variables regarding the concern from managers with the shareholder.Thus, the 18th question intended to measure the degree of importance assigned by dividend policy defining items in terms of profits or losses obtained by shareholders from shares in the recent past.This variable, therefore, intends to verify whether the current payoff definition of dividends is influenced or not by the appreciation of shares.The answers collected for question 18 seem to indicate most Brazilian public companies in the sample do not try to set off losses from the depreciation of their shares with increased dividend yield, and neither do they try to reduce the yield from dividends when their shares are recently appreciated in value.In a nutshell, the managers who responded to the data-collection instrument of this research seem not to consider dividends as a way to counterbalance the shareholders' profits from the appreciation or depreciation of shares in a short-term period.
Question 19 is also related to shareholders in a company.In it, managers were asked whether they considered alternative investment sources available to shareholders when defining their companies' dividend policies.The results obtained with answers to variable 19 seem to provide clues that managers are not strongly influenced by alternative investments available to their companies' shareholders when they need to decide whether they will return the profits generated in their operations or reinvest them in the company.What seems to justify that behavior is the fact that, in Brazil, there have always been investment sources which were an alternative to fixed-income investments.That is so, because fixed-income investments have always offered increased returns when one takes the low risks into account, due to the high basic interest rates which have been present in Brazilian economy as compared to developed countries, for a long time.
Finally, question 20 tried to find out the importance degree assigned by managers, in their decisions regarding dividends, to the returns required by shareholders in returns offered by the company.If financial managers base their actions on adding value to shares, as dictated by the financial literature as the company's objective, the return the company offers to its shareholders should be one of the main decision factors in dividend policy.
According to the data from Table 4, out of the three variables in this study which relate to factors regarding shareholders (questions 18, 19, and 20), the return required by company shareholders in relation to which is effectively obtained by the company is the factor which influences payout index decision the most.The results from variable 20 show that some corporate managers seem to be aware of the role dividends play in the generation of wealth to shareholders, and they take that factor into account when they decide whether to retain or return profits.The dividends and the gains obtained with share appreciation compose the total returns received by shareholders.
Thus, cross-analyses of variables 18 and 20 provide evidence indicating managers assign a relatively higher degree of importance to the returns required by shareholders in relation to what the company offers them, than the gains or losses obtained by shareholders with company's shares recently.That seems to show managers can somehow calculate returns required by shareholders, or they can somehow estimate the cost of their equity; and they stick more to it than to recent performance of shares from their companies.

CONCLUSIONS AND FINAL CONSIDERATIONS
The review of literature in this study showed that several factors should be taken into account when one decides to share their dividends.Among them is the stage of the company in its life cycle, clientele effect, signaling power of dividends, distortion power of inflation in calculation of corporate profit, among others.Besides those, due to legal Brazilian singularities, factors such as minimum mandatory dividend, Interest on Equity, the enforcement of a minimum payout for the possibility of paying shares to managers, the difference in taxes between dividends and capital gains obtained from share appreciation, etc., make dividend decision even more complex for managers in national companies.Those variables are not included in the Residual Dividend Model.
However, in practice, are those factors taken into account by managers in Brazilian public companies?To answer that question, a questionnaire which contained twenty variables related to factors pointed out by the financial academic literature as important to dividend decision was developed.The questionnaire was sent to all 516 Brazilian public companies with active registrations in Brazilian Securities and Exchange Commission in April 2011.Out of the 516 open companies, 104 of them returned the questionnaire fully answered, causing the sample to reach a percentage of 20.2% of the population.
The results obtained showed that, in dividend decisions, managers assign a higher degree of importance to factor such as Brazilian legislation, the company's Articles of Incorporation, cash availability for the company to share profits, the new opportunities for future investments, the company's equity budget for the next accounting period, and the expectation of future profits.Those factors received the highest scores among the variables analyzed, and presented the smallest standard deviations in the average, showing an increased homogeneity of answers.
On the other hand, factors such as inflation effect on the formation of accounting profits, dividend policies of public companies in general and of other companies in the same sector, the alternative sources of investment in the economy which are available to the shareholders, and the gains or losses obtained by shareholders with the depreciation or appreciation in value of shares in their company were the variables which received the smallest scores.These factors presented the highest standard deviations as compared to the factors which received higher scores; this gives a better notion of the relatively high dispersion of answers from those variables in relation to their averages.That was due to the fact there was a higher distribution of answers in the five categories of the scale used in this research.
The main limitations of the model proposed here are believed to be related to factors from dividend policies which were not mentioned in the questionnaire, because they refer to strictly particular aspects of each company.Among those factors, one could mention the gains obtained by shareholders with the appreciation of company shares, and with dividends received recently, the degree of consolidation of the company in its sector, the company's operational risk in its economic activity, the company's long-term objective towards the profitdistribution index, among others.Further studies could be developed to find out whether those variables influence decision-making in dividends, which is important for the investment and financing decisions of an organization.

Table 1 .
Scores assigned for each variable in the questionnaire.

Table 2 .
Summary of questionnaire answers.

Table 1 .
List of variables used by the data-collection instrument.Dividends maintainability in the future, avoiding swings.8 Inflation effect on the formation of the accounting profit.9 Gathering ability of new external funding for investments. 10 Maintenance or displacement of the optimal capital structure.11 Current stage of the company in its life cycle (growth, maturity and decline).12 Company consolidation degree in your industry.13 Degree of stability of the company's profits.14 Industry operational risk in which the company operates (high or low activity risk).15 Dividend policy of other companies in the same sector.16 Dividend policy of listed companies in general.17 Long-term goal of the company for profit sharing ratio.18 Gain or loss obtained by the shareholders recently.19 Alternative sources of investment in the economy to the shareholder.20 Return required by the shareholder in relation to the return generated by the company.21 Others factors taken into consideration in the decision of the allocation of company's profit.22 Planning or simultaneity of factors considered in profits allocation decision.23 Preferred form of profit distribution.24 Frequency of profit distribution normally used by the company.25 Position of who answered the questionnaire.

Table 4 .
Summary of frequency distributions of questionnaire variables.

Table 5 .
Summary of central tendency and dispersion measures of variables in the questionnaire.