3. The value of Market Capitalization in based Market Capitalization Method, which assumes that the current company's capitalization is the multiplication of the number of shares by the share price. Market Capitalization has been calculated by using the following formula:

4. The next stage is to calculate companies' Net Debt. It is worth to mention that Net Debt is one of the financial metrics for analyzing a company's liquidity, which represents company's ability to service its liabilities or the ability to repay the entire debt. Net Debt is a very useful indicator as it allows to correlate the available funds with the size of the debt. The formula for calculating Net Debt using balance sheet is the following:

5. After obtaining Net Debt and Market Capitalization indicators, it is possible to calculate Enterprise Value and, consequently, the EV/EBITDA ratio.

In order to calculate P/E multiple is was necessary to collect the following data concerning the values of this indicator:

1. Determining the price of share by using the information concerning share prices of the chosen public companies the telecommunications industry, taken from the global financial portal, providing streaming quotes - investing.com. The second way to obtain the price of share is to divide the company's market capitalization by the number of shares:

2. Lastly, these is a need to obtain Earnings per Share indicator or EPS. EPS is a financial indicator represents the ratio of the company's Net Income available for the further distribution to the average annual number of ordinary shares. Earnings per Share is one of the important financial indicators used to estimate an enterprise in the stock market, in other words, to compare the investment attractiveness of companies and their performance. EPS formula is given below:

3.4 Approaches to measuring multiples EV/EBITDA and Net Income

Another thing to describe is the possible approaches in the given research to predicting Price per Share by using EV/EBITDA or Net Income indicator.

The first one is the use of benchmark of the telecommunication industry in Russia. Benchmark it is the market standard against which the yield of a separate security can be measured. Thus, when evaluating the performance of any investment, it is important to compare it with the corresponding benchmark value. Hence, it could be an approach to compare the given EV/EBITDA and P/E ratios to the benchmark (standard) of the telecommunication industry in order to understand whether or not the company undervalued or overvalued regarding this indicator. If it is needed to predict future share price using benchmark, the following steps should be done (for EBITDA): (1) Choose a company from the sample and obtain its EBITDA, (2) Multiply EBITDA by benchmark value, (3) Obtain company's Enterprise Value and then, thanks to the formulas provided above, to extract Market Capitalization and, consequently, the price per share, (4) Compare predicted price per share based on the benchmark value with the actual price, (5) Compare price per share from EV/EBITDA and P/E with the actual price and determine which multiple predicts better. Since, during the research it was not possible to find the telecommunication benchmark values using the open sources of information, another prediction option was chosen - the regression analysis.

Based on the data collected, it became possible to predict the multiples EV/EBITDA and P/E for the 2018, in order to compare which multiple value will be closer to the actual values of this year. Since the data were by quarter (5 companies from the same industry, 3 years, 12 quarters = 60 observations), it was decided to implement the regression analysis - the Time Series model. A time series (or a series of dynamics) is a time-ordered sequence of values ??of some arbitrary variable.

In other words, it is a sequence of values that describe a process that takes place in time, measured at successive times, usually at regular intervals. Thus, the time series differs significantly from simple data sampling.

Considering the time series as a set of observations of the process under study, conducted successively over time, the main purpose of the Time Series Model is to identify and analyze the characteristic change of the parameter Y (independent variable), estimating the possible change of the parameter in the future (forecasting). There are two main objectives for Time Series analysis: (1) determining the nature of the series and (2) forecasting or predicting future values of the time series for current and past values.

Most of the regular variables of time series belong to two categories: it can be either a trend or a seasonal variable. A trend is a general systematic linear or non-linear variables that may change over time.

3.5 Regression analysis

5.4 MTS company (EV/EBITDA)

Table 9

This table represents the forecasted values EV/EBITDA regression for the VEON company:

Table 10

Figure 5

The graph repersents the gradual decrease of MTS's EV/EBITDA multiple until the beginning of 2017, then it leveled-off, as well as the predicted values.

Table 11

This table represents the forecasted values EV/EBITDA regression for the Rostelecom company:

Table 12

The graph represents the gradual decrease of MTS's EV/EBITDA multiple until the beginning of 2017, then it leveled-off, as well as the predicted values.

Figure 6

The graph represents the gradual increase of Rostelecom's EV/EBITDA multiple until the beginning of 2017, then it started decrease, as well as the predicted values. This maybe caused by the fact that net income of Rostelecom according to IFRS in 2016 fell by 15% and amounted to 12.2 billion rubles, thus the value of EV/EBITDA started to fell.

5.5 The Tattelecom company (EV/EBITDA)

Table 13

This table represents the forecasted values EV/EBITDA regression for the Tattelecom company:

Table 14

Figure 7

The graph represents the increase of Tattelecom's EV/EBITDA multiple until the beginning of 2016, then it started decrease, until the new increase in 2017. The predicted values leveled-off. This could happen be due to the acquisition of “SMARTS-Kazan”, thanks to which Tattelecom was able to enter the mobile-telecom market.

5.6 MegaFon company (EV/EBITDA)

Table 15

This table represents the forecasted values EV/EBITDA regression for the MegaFon company:

Table 16.

Figure 8

The graph represents the increase of MegaFon's EV/EBITDA multiple until the end of 2015, then it started decrease, until it leveled-off in the end 2016.

The second table represents the P/E input data for the regression analysis (Appendix 2). Some of the P/E are zero, it means that the value of P/E for that period was negative (less than zero). EPS can be zero, negative, or insignificantly small relative to price, but in this case P/E does not make economic sense with a zero, negative, or insignificantly small denominator. Moreover, it should be noted that some data in the Dividend Yield column are also zero, it means that in that period the companies didn't pay dividends, thus the value is zero.

5.7 VEON company (P/E)

Table 17

This table represents the forecasted values P/E regression for the VEON company:

Table 18

Figure 9

The graph represents the increase of VEON's P/E multiple from the end of 2015 till the end of 2016, then it started decrease, until it leveled-off in the beginning 2017. The value zero (1Q of 2015 - 4Q of 2015, as well as 1Q 2017 till the 4Q of 2018) means that the P/E ratio was negative, thus it makes no sense, because the data is not applicable (N/A).

5.8 MegaFon company (P/E)

Table 19

This table represents the forecasted values P/E regression for the MegaFon company:

Table 20

Figure 10

The graph represents the positive values of MegaFon's P/E multiple from the end of 2015 till the end of 2017. The predicted values according to the model will be negative, but it can be positive due to predicted upper confidence interval (UCL). Since, negative P/E does not make sense, it was decided to consider the most positive prediction (UCL).

5.9 MTS company (P/E)

Table 21

This table represents the forecasted values P/E regression for the MTS company:

Table 22

Figure 11

The graph represents the sharp increase in values of MTS's P/E multiple from the beginning of 2015 till the end of 2016. Then, the gradual decrease is observed, but the predicted values should grow.

5.10 Rostelecom company (P/E)

Table 23

This table represents the forecasted values P/E regression for the Rostelecom company:

Table 24

Figure 12

The graph represents the gradual increase in values of Rostelecom's P/E multiple from the beginning of 2015 till the end of 2016. Then, the gradual decrease is observed, as well as the predicted values will decrease.

5.11 Tattelecom company (P/E)

Table 25

This table represents the forecasted values P/E regression for the Tattelecom company:

Table 26

Figure 13

The graph represents the overall increase of Tattelecom's EV/EBITDA multiple. In the beginning of the period the P/E values were negative (thus, value = zero, N/A). This could happen be due to the acquisition of “SMARTS-Kazan”, thanks to which Tattelecom was able to enter the mobile-telecom market. Then, all the P/Es are growing, as well as the predicted values.

6.Prediction of share prices

This subsection of the results description presents a comparison of the predicted prices of shares of telecommunication companies quarterly for 2018. After comparing the results of the predictions with the real values of that period, it will be revealed which regression model - EV/EBITDA or P/E is better able to predict the future price of a share based on historical data collected and calculated.

First of all, it is worth to compare the R-squared of the models created. R-square is the coefficient of linear determination. The coefficient is one of the most effective indicators to estimate the adequacy of the regression model, thus, it is a measure of the quality of the regression equation. If the R-square is > 0,95, it is believed that the model represents a high accuracy of approximation, in other words, the model describes the data well. If the R-square lies in the range from 0,8 to 0,95, it is believed that the model represents a satisfactory approximation, or the model as a whole is adequate to the described data. If the R-square is < 0,6, it is considered that the accuracy of the approximation is insufficient and the model needs to be improved through the introducing new independent variables, considering non-linearities, etc. However, the R-square statistic has one serious drawback: with an increase in the number of predictors, this statistic can only increase. Therefore, it may seem that a model with more predictors is better than.

Table 27

As it can be seen from the table, the average R-squared of both models are approximately the same, but in average P/E model a little bit accurate in comparison the EV/EBITDA model. This might could happen due to the drivers of the models (X dependent variables), in P/E model drivers better influence the independent variable, than in EV/EBITDA model.

Nevertheless, the detailed comparison between the predicted values and the actual ones shows that EV/EBITDA model predicts its values for the following periods better.

Table 28

The calculation for the determining which model better predicts values for the Russian telecommunication is the following: (1) Firstly, to calculate the deviation between the predicted value and the actual; (2) take the resulting values by module (ABS function in Excel); (3) calculate the median (Median function in Excel) in order to obtain the further difference between models. The results (0,21 and 6,275) indicate that on average, the difference between the predicted and real values of P/E is greater than the difference between the predicted and real values of EV/EBITDA. Thus, EV/EBITDA accurately predicts its values.

Lastly, the predicted values extracted from EV/EBITDA and P/E models should be converted to the price of share values. In order to obtain the price from P/E model, it is needed to multiple P/E predicted to the Earning per Share actual, thus, the predicted price is extracted. In order to obtain the price from EV/EBITDA model, the following steps should be done: (1) to have the actual EBITDA; (2) to multiple EV/EBITDA predicted and EBITDA, thus, the Enterprise Value is calculated; (3) to calculate Market capitalization by subtracting Net Debt from the EV; (4) divide the resulting market capitalization by the number of shares, thus, obtaining the predicted price per share. For the MegaFon company the UCL was chosen as the predicted values, since the negative value of P/E does not make sense.

The results are presented in the table:

Table 29

As it can be seen from the table, the results (42,25 and 27,48) indicate that on average, the difference between the predicted spare price and actual share price of P/E is greater than the difference between the predicted and real values of EV/EBITDA. Thus, the share prices predicted with the use of EV/EBITDA model more accurate in comparison with the actual prices, whereas P/E model showed the overall tendency to predict share prices worse. The results show that highest difference in absolute values for all companies occurs in the 4^th quartile.

Conclusion

telecommunication market strategy

This bachelor thesis was aimed at analyzing telecommunications industry in Russia from the perspective of EBITDA and Net Income indicators. The following 5 public companies were analyzed - VEON Russia (Vimpel-Communications), MegaFon, MTS, Rostelecom and Tattelecom. These public companies consist 80% of the industry's capitalization, therefore, it can be said that these companies represent the whole industry. All the companies are very capital-intensive, which can be observed from the average share of fixed-assets in companies, thus the first hypotheses was approved, as well as all the tasks of the bachelor thesis were completed and achieved. As it was mentioned before, the major assumption of the given bachelor thesis is to compare EBITDA and Net Income indicator by using market-based method of analyzing multiples, which were described above - EV/EBITDA and P/E (EPS) multiples.

The major hypothesis concerning that EBITDA may accurate reflect the company's share price rather than Net Income was approved. In order to test it, the financial data for the last 3 years by quarters (from 2015 to 2018) for 5 companies operating in the telecommunications industry were taken as primary data, on the basis of which EV/EBITDA and P/E multiples were calculated manually, as well as their intermediate indicators necessary for the calculation of these two multiples. Further, the regression analysis has been implemented in a form of a Time Series model, in order predict the values of each multiple for their historical values. According to the data obtained, it became evident that the regression of the EV/EBITDA multiple more accurately reflects the dependence of the previous year, followed by, while the P/E repression gives a less accurate forecast. Based on the obtained forecast values of the EV/EBITDA and P/E multiples, calculations were carried out, which allowed both multiples to lead to the share price. Further, the share price derived from the predicted EV/EBITDA multiple more accurately reflected the market price of the stock of those periods compared to the share price derived from the predicted values of the P/E multiple. Thus, it was found out that EBITDA more accurately reflects the companies' value than Net Income, and consequently, the companies' share prices.

The possible limitation of the given research it is short-term orientation. The recent data for the past three years was chosen as it better reflects the economic aspects and conditions of the current market state. Since 2013 there were the currency crisis in Russia, which was marked with great weakness of the Russian ruble against foreign currencies caused by the rapid decline in world oil prices, the export of which largely determines the revenue side of the Russian budget, as well as the introduction of economic sanctions against Russia due to events in Ukraine. These factors caused a significant depreciation of the ruble against foreign currencies, and then led to an increase in inflation, a decrease in consumer demand and finally - economic decline. The unstable economic situation in Russia has had a negative impact on the state of the economy as well as it resulted almost in all spheres of the market, also affecting telecommunication industry. Thus, when the situation became rather stable (since 2015) it became possible to use financial data without the need to make additional adjustments. Otherwise, the predictions might be not so accurate. The short-term orientation of the given research is one of the objective ways to predict a market situation is extrapolation, the spread of trends in the past for the future. At the same time, the forecast for the long-term period should take into account as much as possible the probability of a change in the conditions in which the market will function.

Taking everything into consideration is can be concluded that nowadays EBITDA, despite its possible criticism, is one the most commonly-used financial metrics for evaluation of the company or business segment. Since, EBITDA is near to the operating cash flow and operating earnings, it provides an opportunity to use this indicator in assessing the return on investment and self-financing reserve. Thus, more and more companies tend to disclosure this indicator. Moreover, EBITDA is often used when it is necessary to evaluate the current market position of an enterprise in its industry and to compare the company's current activity with the similar activities of its competitors. In this case, EBITDA is the most appropriate indicator showing the exact financial situation in the organization (Ivshin, 2019). It resulted, that EBITDA reflects the share price better that Net Income for the chosen telecommunication companies, thus, EBITDA better reflects the value of businesses in this market segment.

The analysis of financial results of the business segment can bring a benefit both for the investors wants to be profitable thanks to the company, and for the management of the company itself as well. The main goal of any enterprise is to make a profit. Securities provide investors the opportunity to become owners of the business. The value of the shares depends on the profits received by the company, it is the most important factor influencing the price of shares. Thus, is the share price can be predicted by using appropriate models for prediction share price for the concrete industry, it would become possible to earn. It depends on the industry, which financial indicator reflects better the value of business, in the given case it is the EBITDA indicator. Finalizing, it can be concluded that the right managerial decisions can be made upon the use of EBITDA, but at the same time other financial performance indicator such as Net Income indicator should be considered in complex.

References

Размещено на Allbest.ru