Zaltsman A.A.

National Research University Higher School of Economics, Russia

Determinants of dividend policy of Russian public companies

Many companies all over the world pay dividends to their shareholders. These payments attract the attention of rating agencies (for example, S&P Dividend aristocrats index series), there are mutual funds that are focused on investing in dividend shares (ETF SPDR S&P Dividend, ETF iShares Dow Jones Select Dividend Index Fund, T. Rowe Price Equity Income and many others). A variety of researches were devoted to dividend payments on the developed markets. Nevertheless several new approaches to identifying the determinants of dividend policy were proposed in the last decade. Most large Russian companies pay dividends. For example, dividends were paid per 22 out of 30 shares which were constituents of MICEX index for the fiscal 2009. Since the dividend policy of Russian companies is studied to a much lesser extent than that of companies on developed markets, this study attempts to reveal factors that determine it.

Search for possible determinants lie in three areas, respectively the study consists of three parts. Namely, it checks the applicability of the catering theory and Lintner model to the Russian market. Also it tests the dependence of the quantitative indicators of dividend policy on the broadest possible set of explanatory variables, including financial and macroeconomic factors and the ownership structure. According to catering theory (Baker, Wurgler, 2004) [1] some investors have irrational, changing in time sentiments about dividends and managers of firms adapt dividend policy to them. If investors prefer dividend-paying stocks then these stocks have relatively higher prices than stocks of companies which dont pay dividends. Managers perceive this dividend premium and have additional incentives to initiate dividend payments when it is high. Lintner model (Lintner, 1956) [4] concludes that firms have a target payout ratio and smooth dividends. Managers increase dividends with an increase in earnings per share, but they use some adjustment coefficient because they are very reluctant to reduce dividends. And according to several recent studies [2, 3, 5, 6] dividends are somehow related to key financial indicators of a company, which will be listed further.

The following model is tested using OLS to check the applicability of catering theory to the Russian market: Initiatet = α0 + α1 PD-NDt+ εt. Initiate is the percentage of companies which start to pay dividends in a given period. Too check results of empirical testing other dependent variables are used too: the percentage of companies continuing to pay dividends and the percentage of companies paying dividends which are new in the sample in a given period. PD-ND is the dividend premium on the market in a given period. It is calculated as the difference between the logarithm of the average market-to-book of payers and this of nonpayers. Averaging is conducted both using equal weights and weighting by the book value.

Lintner model is tested by the means of the following regression: Divit = ai + biEPSit + diDivi(t-1) + uit, where Div is a dividend per share, EPS is earnings per share. This regression is applied to each firm separately. Also it is tested on the whole sample using panel data analysis (this method is used in some recent studies due to short time series).

In the third part of the study a variety of models with different dependent and explanatory variables are tested. Each possible determinant of dividend policy is calculated in several ways (Table 1). Close attention is paid to excluding of unusual observations. Dependent variables are: dividend payout ratio, ratio of dividends to retained earnings, ratio of dividends to free cash flow.

Table 1. List of explanatory variables used in the third part of the study.

The observation period is 1999-2009 fiscal years. Sample consists only of Russian public corporations. Only common shares are taken into account. 594 companies are observed in order to test catering theory. Information about whether companies paid dividends or not in given year was collected manually. Other financial data was obtained from Bloomberg. Initial sample for Lintner model and testing the dependence of the quantitative indicators of dividend policy on the broadest possible set of explanatory variables consists of 56 largest companies, which usually paid dividends in the observation period. Because some figures are missing and some observations have unusual values of variables final sample is reduced to 300 observation for Lintner model and from 62 to 107 observations for the third part of the study. Financial data was obtained from Bloomberg, macroeconomic data was obtained from IFS.

Catering theory find no empirical evidence on the Russian market for any method of calculation of variables. So, managers of Russian companies dont try to maximize firms value through responding to irrational investors sentiments about dividends. Probably it can be explained by the existing civil law legal system (as proposed in Ferris, Jayaraman, Sabherwal, 2009 [3]), which provides fewer investor protections for minority shareholders than common law system does.

Hypotheses of Lintner model are generally rejected too, because dividends of only 4 companies can be clearly explained by this model. These companies are: MMC Norilsk Nickel, Mobile TeleSystems, Tatneft, North-West Telecom (from 01.04.2011 part of Rostelecom).

When testing dependence of quantitative variables of dividend policy on the broad list of factors negative influence of the share of the state in a company is revealed at 1% significance level using all dependent variables. Other explanatory variables seem to be either not significant at all, or weakly significant only in some models.

Thereby, it can be argued, that dividend policy of most Russian public companies is quite unstable and doesnt depend on factors which influence dividend policy on the developed markets. The single factor that has significant influence on dividends is the share of the state in a company.

References.

1. Baker, M., J. Wurgler, 2004. A Catering Theory of Dividends. The Journal of Finance, Vol. 59, No. 3 (Jun., 2004), pp. 1125-1165 2.

2. DeAngelo, H., L. DeAngelo, R. Stultz, 2006. Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory. Journal of Financial Economics 81 (2006), pp. 227254.

3. Ferris, S., N. Jayaraman, S. Sabherwal, 2009. Catering effects in corporate dividend policy: The international evidence. Journal of Banking & Finance 33 (2009) 1730-1738.

4. Lintner, 1956. J. Distribution of Income of Corporations. American Economic Review, May 1956, p. 97-113.

5. Neves, E., J. Pindado, C. Torre, 2006. Dividends: New evidence on the catering theory. Universidad de Salamanca.

http://www3.uva.es/empresa/uploads/dt_14_06.pdf

6. Teplova, T., G. Shagaleeva, 2010, Dividend policy of companies on emerging markets, New Bulgarian University, Academia, Sofia.