MACEDONIAN TOURISM INSIGHT THROUGH THE ANALYSIS OF STOCKS RETURNS OF QUOTED TOURISM COMPANIES AT MSE

: In this paper we present the results of the statistical analysis, focused on de-termining the character of relationship between stocks of two quoted tourism companies at Macedonian Stock Exchange (MSE) using the data for the last 52-weeks series of daily stocks’ closing prices as well for the Macedonian Stock Exchange Index (MBI-10). The linear regression and correlation analysis of two securities provide evidence for statistical significance of their stocks’ daily returns at MSE. On the contrary, regression analysis did not reveal a statistically significant relationship between MTUR, MPOL and MBI-10. Through the analysis of daily stocks’ returns at MSE we could not determine a statistically significant relationship between tourism growth in the Republic of Macedonia and tourism companies’ securities prices in the last two years. Some implica-tions for tourism planning and portfolio management can be drawn.


INTRODUCTION
Macedonian Stock Exchange (MSE) was established in September 1995, but its real start was with the fi rst ring of Stock-Exchange bells on March 28, 1996. MSE started on November 01, 2001 to calculate Macedonian Stock Exchange Index (MBI), which consists of fi ve most liquid stocks at MSE. MBI was price not weighted index, and as a fi rst index, it fi nished its function as an aggregate indicator for stock exchange movement quantifi cation. On January 04, 2005, a new MSE index was introduced (MBI-10), as weighted average indicator. It enables using market capitalization more realistically, following the price movements at MSE. MBI-10 calculation is in accordance with teh methodology for MBI-10 calculation and it consists of ten quoted stocks on MSE offi cial market segment. Stock Index Committee regularly (two times per year) and adhoc (in special circumstances) makes updates of MBI-10 structure in accordance with market conditions. Th ere are three quoted companies at MSE: Makedonija turist AD Skopje (MTUR), Metropol hoteli AD Ohrid (MPOL) and International hotels AD Skopje (INHO). Th e authors shall analyses returns data of MTUR and MPOL for a period of two years:: fi rstly, to determine mutual correlation coeffi cients and statistical signifi cance of stock prices of these two tourism companies; secondly, to determine their linear and multiplied regression with MBI-10 daily closing prices. INHO security is not liquid at MSE since January 2013, and there is no trading data, so it will not be analyzed in this paper.
Th rough the analysis of stock prices of two tourism companies we are also trying to determine if the official state statistics data for the growth of Macedonian tourism can be confi rmed with the stock prices rise of quoted tourism companies at MSE. We try to identify the reliable results of regression analysis that can be used for tourism planning and portfolio management. However, the basic task of our research is examination of basic parameters and character of tourism companies stocks' returns at MSE.
Th e remainder of this paper is structured into three sections. In Section 1, we present literature overview, while in Section 2 we present the fi nancial information about the observed companies at MSE. Section 3 presents the empirical results of the regression analysis of stocks at MSE. Section 4 summarizes the main conclusions.

LITERATURE OVERVIEW
Markowitz Portfolio Th eory (MPT) models an asset's return as a normally distributed function. Th is theory calculates risk as the standard deviation of return, as well as the return of portfolio as the weighted combination of the assets' returns. By combining diff erent assets whose returns are not positively correlated, MPT suggest the possibility to reduce the total variance of the portfolio return. Th is theory also has the basic assumption that investors are rational and markets are effi cient (Markowitz, 1952).
Although MPT was developed in the 1950s, it gained popularity in the 1970s and was considered the basis in the mathematical fi nancial modeling. However, since then, many theoretical and practical criticisms have been raised on this matter. Th e theory and practice provide a lot of evidence that fi nancial returns do not follow a Gaussian distribution or any symmetric distribution. Th ere is also skepticism that the correlations between asset classes are not fi xed, but can vary depending on the external events (especially during crises).
Long ago in empirical studies, it was noticed that returns of stocks (indexes, funds) are badly fi tted by Gaussian distribution because of the heavy tails and strong asymmetry (Mandelbrot, 1960;Mandelbrot, 1963). Mandelbrot and Hudson (2006) elaborate that the random walk and Gaussian daily returns simply do not correspond to reality, and grossly underestimate the risk of huge market swings . Fama (1965) reported that daily returns of stocks on the Dow Jones Industrial Average (DJIA) display more kurtosis than permitted under the normality hypothesis. Since that early work of Fama, it has typically been found that daily returns display more kurtosis than that permitted under the assumptions of normality, while skewness has also been prevalent (Mills, 1995).
Th e expected returns and variances are almost always estimated using past returns rather than future returns. Th e assumption that all risk-returns models use when they use historical variances is that past return distributions are good indicators of future return distributions. When this assumption is violated, as is the case when the asset's characteristics have changed signifi cantly over time, the historical estimates may not be good measures of risk .
Th e bias towards positive or negative returns is represented by the skewness of the distribution. If distribution is positively skewed, there is higher probability of large positive returns than negative returns. Th e shape of the tails of the distribution is measured by the kurtosis of the distribution; fatter tails lead to higher kurtosis. In investment terms, this represent the tendency of the price of this investment to jump (up or down from current levels) in either direction (Damodaran, 2006).
In case where distribution of returns is normal, investors do not have to worry about skewness and kurtosis. Normal distributions are symmetric (no skewness) and defi ned to have a kurtosis of zero.
When return distributions take this form, the characteristics of any investment can be measured with two variables -the expected return, which represents the opportunity in the investment, and the standard deviation or variance, which represents the danger .

ANALYSIS OF FINANCIAL STATEMENTS OF MTUR, MPOL AND INHO
Makedonija turist AD Skopje is a company that works in the hotel, tourism & catering industry. Th e company manages seven hotels: Holiday Inn, Best Western Hotel Turist, Hotel Karpos, Hotel Jadran, Hotel Bristol, Best Western Hotel Bellevue, Hotel Vodno, all located in Skopje. Th e company also runs two restaurants in Skopje: Pivnica and Ogniste -Makedonija and the London Pub.
We present the vertical and horizontal analysis as well as the main fi nancial ratios of MTUR (See Table 1 and Table 2).
Internesenel Hotels AD Skopje is a company engaged in the tourism industry. It operates the Hotel Continental in Skopje. We present the vertical and horizontal analysis as well as the main fi nancial ratios of MTUR (See Table  3 and Table 4).
Hotels-Metropol AD Ohrid is a tourism company. Th e company manages three hotels located at Lake Ohrid: Hotel Metropol, Hotel Tourist and Hotel Bellevue. Th e hotels are equipped with conference facilities, one casino, restaurants and bars, sports and recreation center, including   basketball, tennis, volleyball and soccer courts, swimming pools, as well as the spa center. Th e main shareholder of the hotel is Metropol AD Ohridis Fersped AD Skopje. We present the vertical and horizontal analysis as well as the main fi nancial ratios of MTUR in Table 5 and Table 6.
Th e fi nancial analysis gives information about the fi nancial results of those three companies. It is obvious that only MPOL revenues raise in 2015 and 2014 compared to 2013, while other ratios do not indicate increased profi tability. We present the current three securities' price multiplies in Table 7.

DESCRIPTIVE STATISTICS
We analyze two years (2014-2015) stocks daily returns data at MSE in order to determine stock correlation and comprehensive regression analysis. We believe that the results provided will be useful for stock valuation. Th e basic task of our research is determination of returns character at MSE and identifi cation of mutual dependence and correlation of stocks returns. We argue that our fi ndings have practical application for stock value forecast and tourism planning.
We use the sample of two quoted tourism companies'stocks from the offi cial market segment of MSE: MTUR and MPOL. Th e basic criterion for the analysis was to analyze stocks of tourism companies. MSE was the main source of data through the offi cial stock newsletters and annual reports. Th e two-year period allowed us to make the appropriate conclusion. Th e analysis was performed using the daily closing prices of traded stocks as well for MBI-10 for the period from January 01, 2014 to December 31, 2015.  By means of the regression analysis, we have determined a strong positive correlation between two stock prices at MSE (value oscillate around 0.83), and a weak correlation between stocks and MBI-10, as shown in the following table:  Table 8 provides correlations among two stocks and MBI-10 at MSE. We can see lower but still positive correlation among stocks and MBI-10. Th e diff erence of correlation among stocks and MBI-10 compared with only mutual stocks correlation coeffi cients suggest that MBI-10 changes are not immediately followed by the other stocks on MSE. Th e conclusion about lower statistical signifi cance between stocks' price movements and MBI-10 daily values can aff ect using MBI-10 for predicting tourist companies stocks' market prices at MSE. Th is fi nding will be tested by means of regression analysis.
We explore the correlation of MSE stocks' daily returns in order to determine mutual dependence and correlation of stocks returns as tools for stock value forecasting. Using regression analysis we are trying to determine if there is a statistically signifi cant relationship between the variables (two stock prices or daily index values and stock price). We fi rst analyze the Multiple R (coeffi cient of correlation) and R Square (R 2 ). Th e R 2 is the coeffi cient of determination and tells us the proportion of the total variation in the dependentvariable explained by the independent variable. If there is a stronger relationship (higher coeffi cient of determination), it indicates that this relationship is statistically signifi cant and prediction of dependent variable will be accurate if we have a good forecast of independent variable. By using variance statistics, we determine f-test that confi rms if regression analysis is statistically signifi cant. A very low level of signifi cance F value confi rms statistical signifi cance of the analyzed relationship. Next, we look at the t-statistics for our regression coeffi cients. We analyze whether a t-statistic coeffi cient is statistically distinguishable from zero (i.e. statistically signifi cant). Th e magnitude of the coeffi cient is not the issue of our interest. If the coeffi cient for one stock price is signifi cantly diff erent from zero, then we know that independent value (stock price) is useful in predicting other company's stock price. Th e t-statistic indicates the number of standard deviations from zero the coeffi cient. Obviously, the higher this number, the more confi dence we have that the coeffi cient is diff erent from zero. Generally for large samples, a t-statistics greater than 2.00 is signifi cant at the 95% confi dence level or more (Neter et al., 2004). We also use the p-value to determine the exact confi dence level. We calculate p-value by subtracting the p-value from 1 to fi nd the confi dence level. Th is number is simply the best point estimate given to our set of sample data. We also present the result: Lower 95%. Th is gives us a range of values between which we can be 95% sure that the true value of this coeffi cient lies. Since we are merely using this forecasting model, the significance of the intercept is not important. In our regression statistics, we asked for 95% level of confi dence.
Th e results of descriptive statistics and regression analysis of daily stock prices at MSE are given in the tables below, as follows: It is obvious from the results of descriptive statistics that stocks at MSE have high volatility, negative skewness and low kurtosis. Distributions with zero kurtosis are called mesokurtic. Normal distribution has zero kurtosis. Distributions with high kurtosis distribution are called leptokurtic, and tend to have a distinct peak near the mean, decline rather rapidly, and have heavy SITCON 2016 QUALITY AS A BASIS FOR TOURISM DESTINATION COMPETITIVENESS tails. Distributions with negative kurtosis (platykurtic) have a fl at top near the mean and shorter, thinner tails .
Th e daily return series for MPOL stocks are leptokurtic, while for MTUR is platykurtic. Th is means that signifi cant variations in the daily prices are very common. Both MSE stocks have large kurtosis values.
In Table 10, we report the results of linear regression statistics for MTUR stock as dependent variable (Multiple R, R Square, Adjusted R Square, Standard Error, Number of Observations, df, SS, MS, Signi cance F, t Stat, P-Value) where MPOL stock is independent variable. Table 10 reports the results of the stocks analysis for two companies (MPOL and MTUR) Skopje) from tourism sector in the Republic of Macedonia. Values for Multiple R (coeffi cient of correlation) and R Square (coeffi cient of determination, variance) for MTUR daily stocks returns as dependent and MPOL as independent value are around 0,70, which leads to the conclusion that there is a statistical signifi cant relationship between these two variables. In fact, there is almost 70% significant relationship between the outcomes and predicted value. Th e R 2 tells us that the proportion of the total variation in the dependent variable (MTUR stock market price) can be explained by the independent variable (MPOL stock price). Using the variance statistics, we determine f-test that confi rms the signifi cance of regression analysis. A very low level of signifi cance F confi rms statistical signifi cance of the analyzed relationship. A tstatistics is high and confi rms signifi cance. We can also see that p-value (probability value -that explains that results occur randomly) is zero, which means that we are 100% confi dent that our coeffi cient (MTUR) is significant for predicting MPOL stock price changes. Table 11 reports the results of regression analysis between MTUR and MBI-10, where MTUR stock price as a dependent variable was tested using MBI-10 as independent variables. We did not fi nd a statistically signifi cant relationship between MTUR and MBI-10 index. Our fi ndings are supported by low values for R Square (13%) and appropriate values of t-statistics and p-values. Table 12 provides multiple regression statistics results for MBI-10 index. Multiple regression analysis is not signifi cant (Adjusted R 2 is 12%), which indicates a low level of relationship between MBI-10 and MPOL and MTUR stocks prices. Regression statistics confi rms our fi ndings (with 95% level of confi dence) that the proportion of the total correlation in the dependent variable (MBI-10 index value) cannot be explained by the independent variables (stock prices of tourism companies-MTUR and MPOL). We try to identify if the offi cial statistical data for tourism rise in the Republic of Macedonia over the last two years (2014 and 2015) can be confi rmed by stock returns of tourism companies. In particular, we fi rst identify the correlation between tourism companies' stocks at MSE. By means of the regression analysis, we have determined a strong positive correlation between stock prices at MSE (most of the values oscillate around 0.70). We determine lower but still positive correlation among the analyzed stocks and MBI-10. Th e correlation diff erence among stocks and MBI-10 compared Note: Signi cance at the 95% con dence level Note: Signifi cance at the 95% confi dence level with mutual stocks correlation coeffi cients suggests that MBI-10 changes are not immediately followed by other stocks on MSE.
Th e results of linear and multiple regression analysis lead us to the conclusion that there is a statistical signifi cance between stock prices at MSE, as well as that regression analysis is a useful tool for stocks market prices forecasting at MSE. Th e research results identify that stocks from the same industry (tourism) have statistically signifi cant relations. Th e R 2 values confi rmed that the proportion of the total correlation in the dependent variable (one tourist company stock price) can be explained by the independent variable (other stock price) as well as that accurate forecasting of one stock price movements will lead us to reliable valuation and prediction of other stocks' future price.
Multiplie regression analysis is not signifi cant (Adjusted R 2 is 12%), which indicates a low-level relationship between MBI-10 and MPOL and MTUR stocks prices.
We did not fi nd evidence that the tourism growth of Macedonia identifi ed by state statistics data have direct infl uence on the stock prices rise of quoted tourism companies at MSE We determine the statistical signifi cance among tourism companies' stocks, which leads us to the conclusion that we can use one stock price for other stock price forecasting at MSE. Th is fi nding can be used for portfolio management at MSE as well as tourism planning in the Republic of Macedonia.