06 января 2016г.
Introduction.
Dividend policy is one of the most controversial aspects of the company, which has always caused a lot of controversy among financiers and scientists. Do firms pay dividends, and if so, what share of their profits? What will be there action of investors to the announcement of the dividend? Dividend policy is an important attribute of the company, as well as an important element of the financial strategy of the company.
In this case, the aim of the work is to study the impact of dividend payments on the value of Russian companies by using one of econometric methods - the method of least squares. If any dependence is to take place, the second target will be the analysis of the factors that can influence the value of dividend payments. As an example, I chose major Russian company in the real sector of the economy, whose shares are traded on the Russian stock exchange - LUKOIL. In the first part of my work, I will give a detailed overview of trends and patterns of dividend policy, which exist at the moment. In the second part of the paper will be presented to the statistical and econometric study of dividend payments of the Russian companies for 2002 - 2013, which will allow to draw some conclusions about which theories and models confirmed for Russian business.
Description of the model.
In the previous paper described the main theories in the field of dividend policy. They were all designed by foreign experts and tested on the American and British markets, which casts doubt on the applicability of the obtained results for Russian companies and Russian financial markets.
In this paper I am going to conduct a study of the influence of ads on payment of dividends on the value of shares of Russian companies, as well as to test the significance of some factors that can potentially influence the magnitude of the dividend. This study will allow us to draw some conclusions about what the classical theory of dividend payments is confirmed in the Russian data, and which are not.
The sample design.
To build the model and its further validation was taken major Russian public company whose shares are traded on the RTS stock exchange: LUKOIL.
These companies were selected based on the following criteria:
· At the annual meeting shall be announced on payment of dividends per one ordinary share);
· There is evidence of the financial performance of the companies for 2002-2013 year (net profit, capitalization, and so on);
· There is evidence of price fluctuations of stocks for 15 days before and 15 days after the announcement of the payment of the dividend;
· The company has received a credit rating according to the national scale rating Agency "Expert RA". Data collection and calculation variables
Data collection occurred for the following sources: financial statements of companies for 2002-2014 year, the website of the rating Agency "Expert RA", the website of the financial company FINAM, which publishes the exact date of the shareholders ' meetings, as well as announcements of dividend payments. Data on the price fluctuations of the shares of the companies were obtained by download from the Internet site of the company FINAM.
Variables used in the study:
Credit rating - a rating that expresses the opinion of the rating Agency on the company's ability to timely and fully meet its financial obligations. This rating is assigned to companies based on a comprehensive analysis of the risks inherent in their activities, i.e. assessment of the company's position in the market and the dynamics of the markets in which the company operates, production capacities and potential." For the analyzed models were found credit ratings companies from 2002 to 2013. To use the rating values in the regression equations were used two dummy variables (dummies) which takes the value «1» if the rating reflects the highest level of creditworthiness – «AA+» and «0» if all other ratings.
The capitalization of the company - this indicator shows the amount and size of the company and is calculated as the market price of the share multiplied by the number of shares of the company. Capitalization of the studied companies was obtained on the basis of financial reporting (number of shares) and the market price of the shares.
The carrying amount of company's nominal value of a share multiplied by the number of shares. The carrying value was obtained based on data from the financial statements of the respective companies.
The total dividend is calculated as dividend per ordinary share multiplied by the number of ordinary shares. The data mentioned above, allowed us to calculate the necessary parameters:
· The ratio of market capitalization to its carrying value. This ratio serves as an indicator of investment opportunities of the company. The higher this index, the more the firm invests in fixed capital.
· The ratio of total dividend to net profit (dividend payout). This indicator characterizes the portion of the profits that the company aims to pay dividends to its shareholders
The impact of the announcement of the dividend on the share price of the company
To find the influence of ads on payment of dividend on the share price of the company was selected period of time, which should be a similar relationship. A gap was identified in 15 days before the announcement of the dividend and 15 days (excluding days when trading in the shares was not conducted). This length will allow you to leave to exclude from the study of third-party effects that are unrelated to the information effects of dividends.
Further, there were determined the share price movements and calculated the average share price during the 15 days prior to the announcement of the dividend and 15 days after that allows you to calculate the change (absolute and percentage) price of shares of the company for the entire time period.
The study analyzed the influence of such factors as company size (market capitalization), investment opportunities (the ratio of market value to its book value), the rating companies (dummy variables), the amount paid dividend (the ratio of the amount of the dividend to the profits of the company) to change the prices of the stocks before and after the announcement of dividend payment.
So I will be estimating the model, where Yt is the price of shares change in absolute mining. X1t – dividend payout in million dollars, X2t – market-to-book value in million dollars, X3t – market capitalization in million dollars and X4t – credit rating - dummy. a0, a1, a2, a3 and a4 are parameters.
The project is multivariate i.e. it involves more than two variables. Regression analysis is used to study the relationship between the price of shares change, dividend payout, market-to-book value, market capitalization and credit rating. The dependent variable (Y) is the price of shares changeand dividend payout, market-to-book value, market capitalization and credit rating are the independent variables (X1, X2, X3 and X4).
PCt –current value of price change,
DPt –dividend payout,
INVESTt –market-to-book value,
SIZEt – market capitalization,
CRt – credit rating;
𝜀𝑡 – the disturbance term.
Required data for estimation
For analyzing and testing the model, it needs to found out some specific data. I took annual data from 2002 to 2013 about price of stock changes (PC),dividend payout (DP), market-to-book value (INVEST), capitalization (SIZE) and credit rating (CR) (See Appendix 1).
Moreover, it is important to mention that the information for the last period of 2013 would not be include in our analyzing right now, we will need it later for model forecasting.
In order to evaluate the model we used data that was collected by financial reports of companies and thus is in public domain.
Matrix of correlation
Correlation is used to describe the relationship between two continuous variables. In general, correlation tends to be used when there is no identified response variable. It measures the strength (qualitatively) and direction of the linear relationship between two or more variables.
In order to find correlation between our variables we construct matrix of correlation.
It could be seen that there is no strong correlation (equals 1) between independent variables. So we should not exclude some of these variables.
Scatter diagram
A scatter diagram uses a graph comprising of a horizontal axis containing the measured values of one variable and a vertical axis representing the measurements of the other variable in order to study correlation between the two variables. Scatter diagrams do not necessarily indicate or establish a cause effect owing to one variable with respect to the other, but do reflect existence (as well as type/strength) of a relationship, which may be of type such as strong linear (positive or negative correlation), quadratic or exponential relationship, outliner, damped (Sinusoidal relationship), etc.
A scatter diagram can suggest various kinds of correlations between variables with a certain degree of confidence level. The correlations can be positive (rising), negative (falling), or null (uncorrelated).
Positive Correlation – If the pattern of dots slopes from lower left to upper right, it suggests a positive correlation between the variables being studied.
Negative Correlation – If the pattern of dots slopes from upper left to lower right, it suggests a negative correlation. A line of best fit can be drawn in order to study the correlation between the variables. An equation for the correlation between the variables can be determined by the best-fit procedures. For a linear correlation, the best-fit procedure is known as linear regression and is guaranteed to generate a correct solution in a finite time.
So our next step is to draw scatter diagrams using Excel and find out whether there is correlation between price changes and independent variables.
Dependence of price changes on dividend payout shows that there is negative correlation between these two variables. The equation for the correlation is Y=-10,161x + 2,6804.
Dependence of price changes on market-to-book value shows that there is positive correlation between these two variables. The equation for the correlation between the variables is Y=2E-0,5x –0,5517.
The correlation between unemployment rate and inflation is positive.The equation for the correlation between the variables is Y = 6E-0,6x + 0,5704.
Credit rating is a dummy, so we don‘t use scatter diagram to show the dependence of price changes on credit rating.
These three scatter diagrams reflets existence of linear relationship between the variables. Now we can pass to the next stage of our project – econometric model testing.
Specification of an econometrics model
Mathematical interpretation of the model is presented below:
Where Yt is the price change in absolute mining. X1t – dividend payout
in million dollars, X2t – market-to-book value in
million dollars,
X3t – market capitalization in million
dollars and X4t – credit rating
- dummy. a0,1,2,3,4are parameters (sensitivity of the explained variable
to changes of the explainable variable),
ɛt – the disturbance term.For
the estimated model we take data from appendix
1 and we get the following results:
Table2 Regression statistics
Регрессионная статистика
|
Множественный R
|
0,92
|
R-квадрат
|
0,85
|
Нормированный R-квадрат
|
0,75
|
Стандартная ошибка
|
2,79
|
Наблюдения
|
11
|
Specification of estimated econometric model
The specification of our model with calculated parameters are presented
below:
Where a0 = 14,87 with standard
error of 5,63, a1 = -95,35 with standard error of 27,92,
a2 = 0,0008 with standard error of 0,0002, a3 = -0,0009 with standard error of 0,0002,
a4 = -9,51 with standard error of 2,08 the standard error of disturbance term is 2,79.
R2 is high (85%). It means that varies in X explains
85% of varies in Y. Fcrit.is less than F, therefore R2 is not random and quality
of specification of econometric model.
E (ɛt) = 0 is the first Gauss-Markov assumption that the error ɛ has an expected
value of zero given value of the explanatory variable. This means that on average the errors balance out. This is not a restrictive assumption since we can always use a0 so that this equation holds.
The second condition
is that the error term has a constant
variance. This is the assumption of homoscedasticity.
In order to estimate
the econometric model we should use software
(Excel). We input the values of endogenous and exogenous variables
from 2002 to 2012 into corresponding rows in the Regression, Analysis Toolpak.
We should mention that we do not use data of 2013 year, because we are going to use it later to check model adequacy. Level of significance is 92%.
When checking the significance of the coefficients with the help of
t-test, I found out
that all coefficients passed it, because |t| ≥tcrit. In this case we shouldn‘t
exclude coefficients from the model.
According to this model, increase in market-to book value by one
million dollars, leads to increase
in current value of price change by 0,0008 dollars.
And increase in market capitalization of the company by one million dollars, leads to decrease in current
value of price change by 0,0009
dollars.
Regression analysis allowed us to draw the following
conclusions: on the current value of price change influence the investment opportunities of the company,
capitalization, and credit rating. These results were confirmed signaling and Agency motives dividend
policy. In addition, some specific resulthas led to claims that the
theory satisfies the preferences of the investor also has the right to life in the Russian context.
Model testing
Now we
should test our model. The calculated regression coefficients а0,а1, а2, а3,а4 allow us to construct
the equation Yt = 14,87 – 95,35X1t +0,0008X2t-0,0009X3t – 9,59X4t + ɛt, where ɛtis random
value.
Value of multiple
coefficient of determination R2 equals to 0,85. It shows that 85% of total
deviation of Yt is explained by the variation of the factors X1t, X2t.X3tand X4t. Such value of R2is good, as it is close to 1. It means that the selected
factors do not effect significantly our model, which confirms the correctness of the inclusion in the estimated model.
Significance F
The calculated level of significance 0,01<0,05 (table 3) confirms
the R2 significance. F-test
This is another way of checking
R2. It is based on comparing
F (table 3) with Fcrit. F should be more than Fcrit. In our case Fcrit= FРАСПОБР(0,05;4;6) = 4,53, where 4 is the number of degrees of freedom, it equals to the number of the equation regresses
m=4, and 6 is the number of degrees of freedom,
it equals to n-(m+1).
As our F >Fcrit., the H0 hypothesis that R2=0 is rejected. That means that R2 is not random and quality of specification of
our econometrics model is high.
Standard error
Now we should test the importance of regression coefficients a0,a1, a2,a3, a4. Comparing the elements of the columns Coefficients and Standard Error
(table 4), we can say that absolute values of standard
errors is less than the corresponding values of coefficients, so, at the first stage of analysis, all the variables
should remain in the model.
t-test
Then we should
check the significance of the coefficients with the help of t-test.
That is to test the inequality |t|≥tcrit., where t is the value of t-statistics (table 4). If the inequality is right, the coefficient and the regressor are considered to be significant and vice versa.
In our case tcrit. = СТЬЮДРАСПОБР(0,05; 6) = 2,45, where 0,05 is the level of significance, 6 is the number of degrees of freedom,
it equals to n-(m+1).
After
estimation of the initial model all
absolute values of t-statistics in table 4 are more than
tcrit., therefore, all the regression coefficients are significant.
Goldfield-Quandt test
In our case after all calculations we get the following
information:
Table5 Goldfield-Quant test
RSS1 | 1,60 |
RSS2
|
1,64
|
GQ
|
0,98
|
1/GQ
|
1,03
|
FcritGQ
|
161,45
|
In our example
both inequalities are valid, so the assumption about homoscedasticity of random disturbance is adequate.
Durbin-Watson test
This test is designed to check a particular case of third assumption of the Gauss-Markov theorem
about the absence of autocorrelation between
adjacent random residuals
in the model.
Using values
of the residuals ɛtfrom appendix
3, we can compute
Durbin-Watson statistics.
Then we should find Durbin-Watson statistics critical values dL and dU with
the help of special statistical table, where n=11 – total number
of observations, k=4 – total number of factors,
α = 0,05.
There are three possible outcomes of the test:
|
|
|
Positive/negative autocorrelation of the model‘s residuals exists
|
Autocorrelation of the model‘s residuals is ambiguous
|
Positive/negative autocorrelation of the model‘s residuals does not exist
|
In our model dL = 0,59 and dU = 1,93, so there is no information about autocorrelation of the
model‘s residuals, so we
can use least square technique to estimate
the model.
Confidence interval
The purpose of
confidence of intervals
is to
determine a series
of values from recurring samples of data so that the series of values of the specific
parameter is more likely to happen within the specified probability.
We also should estimate
the lower and upper boundaries for year. We will use the following formula: 99,5% boundary = Yˆ ±
t crit* st.error , where tcrit.is calculated as it has been shown in part ―t-test‖ and standard error = 2,79(table 2), Yˆ – predicted value of Yt. Then we should compare the empirical data for each data with the resulted interval
boundaries.
Low level = -3,42 Upper level = 10,24 Adequacy checking
Now we should check whether predicted by our model Yˆ
is truly describes the empirical data correctly and test the forecasting capabilities of the model. Yt should lie within
confidence interval, predicted
by our model.
If we look at our case, we will get the
following information:
Table 6 Adequacy checking
Lower 95%
|
Upper 95%
|
Empirical
|
Empirical>Lower
|
Empirical
|
-3,42
|
10,24
|
3,41
|
True
|
True
|
So our empirical
for 3,41 of 2013 data lies between upper and lower boundaries predicted by our model. Conclusion.
Each of the above-mentioned theoretical studies have confirmed or deniedthe
viability of one of the models for those in the financial and corporate systems, the study of which it was based. Such countries, as a rule, were the United States or Britain.
But those conclusions, which were made for them in the advancedeconomies, could
not find his evidence to the emerging market of the Russian Federation. Here was the main goal of
my work: on the basis
of empirical studiesto
test the validity
of existing theories
for Russian companies.
Regression analysis allowed us to draw the following
conclusions: on the current value of price change influence the investment opportunities of the company,
capitalization, and credit rating. These results were
confirmed signaling and Agency motives
dividend policy. In addition, some specific
resulthas led to claims
that the theory satisfies the preferences of the investor
also has the right to life in the Russian context.
List of references
1.
Backer M. and Wurgler J. A Catering
Theory of Dividends. Journal of Finance.
Vol. 59. P. 1125–1166. 2004.
2.
Bhattacharya S. Imperfect Information, Dividend Policy, and ―The Bird in the Hand‖ Fallacy. The Bell Journal of Economics.
Vol. 10. P. 259–270. 1979.
3.
Fama E. and French
K. Disappearing Dividends: Changing
Firm Characteristics or Lower Propensity to Pay? Journal of Financial Economics. Vol. 60. P. 3–43. 2001.
4.
Gomez A. Going Public with Asymmetric Information, Agency Costs, and Dynamic
Trading. Journal
of Finance. 2000.
5.
La Porta R., Lopez-de-Silanes F., Shleifer
A. and Vishny R. Agency Problems and Dividend Policies around the World. Journal of Finance.
Vol. 55. P. 1–33. 2000.
6.
Shleifer
A. Inefficient Markets:
An Introduction to Behavioral Finance. Oxford.
Oxford University Press. 2000.
7.
TregubIlona V. Simulation, tutorial, M.: FA, 2000.
8.
TregubIlona V. Investment Project Risk Analysis
in the Environment of Russian Economy
/ Foreign Investment, Ljubljana Empirical Trade Conference 2012
9.
www.finam.ru;
10. www.lukoil.ru;
11.
www.raexpert.ru;
12. www.rts.ru.