The Characteristics of Takeover Targets : Evidence from Food Industry in Serbia

This study analyses the characteristics of target companies from food industry that were taken over by foreign investors in the Serbian market. The aim of the study is to determine: Can future target companies be predicted based on the characteristics of target companies that have already been taken over in the Serbian market? The study compares the indicators of profitability, liquidity, leverage and efficiency of 30 target companies from the food industry with the industry' average. The study includes multivariate analysis of variance (Manova) and Discriminant analysis in order to identify the difference between observed indicators. The liquidity and leverage indicator showed a statistically significant difference compared to average values. Group classification is 95% and sample homogeneity is 90% which leads to a conclusion that target companies can be predicted with extremely high reliability.


Introduction
In the initial phase of transition, takeover processes were done in a form of acquisition of assetsprivatization of the existing state-owned companies.The largest number of foreign investments came to Serbia through the privatization of state-owned companies operating in telecommunications, banking and food industry.According to D 'Souza et al (2017), privatization processes involve a whole series of reforms in the economic, political, legal and financial system, and the privatization of state-owned enterprises has a significant role in transition countries.Foreign direct investments grew from 2001, reaching their maximum in 2006 when they amounted to 14.3% of the GDP (gross domestic product).Onset of the world economic crisis in the summer of 2008 led to a rapid decline in investment activity worldwide and in Serbia as well.The gross domestic product grew at an average annual rate of 3.5% from 2001 to 2011, despite the fact that the first moderate decline of 3.5% was observed in 2009 due to negative effects of the global crisis.In the period from 2001-2011, growth rate of production industry was 0.7%, where processing industry dictates growth pace to a significant extent 0.4% (food production amounts to 1.6%, while beverage production amounts to 0.5%).Food industry accounts for 18.6% of the total Industry, which is the highest percentage share according to the Chamber of commerce and industry of Serbia (2017).Contribution of the food industry to GDP is 3.3%.Analysing the period of transition in Serbia, the authors Nikolić and Zubović (2013) conclude that there have been significant structural changes in the manufacturing industry, but that did not result in a sufficient improvement in comparison with other transition countries and the average EU countries.
Since 2001 the new beginning of transition in Serbia has been aimed at reaching the macroeconomic stability.There was an increase in growth rate of the gross domestic product, the decline in inflation and market liberalization which gave a positive signal to investors to come to the Serbian market.Since Serbia in the second round of transitional reforms did not even achieve 60% of the GDP it had had in 1989 when the transition first had begun, the Serbian market was appealing to foreign investors due to its great growth potential.Provided that there is no economic growth of the country without foreign investments, many authors in their research dealt with studying determinants which influence the choice of foreign investors to invest their capital in a certain country (Savoiu et al., 2013;Brouthers & Brouthers, 2000;Mudambi & Mudambi, 2002;Bevan & Estrin 2004;Mahmoodi & Mahmoodi, 2016;Andrašić et al., 2017).
According to Alhenawi andStilwell (2017, p.1042) "Intuitively, an M&A transaction creates value when the target's potential value as a division of a competent acquirer is greater than the target's pre-acquisition value as a stand-alone entity."Through a discriminatory analysis, the authors have tested the hypothesis, that acquisitor's competencies are important for the success of acquisition in addition to the pre-acquisition value of the goals.Campa and Kedia (2002) also support the view that firm-specific characteristics should be considered when evaluating corporate acquisition events.
In recent years, there is increasing evidence that in the takeover process some other factors get importance over the commonly used data from financial statements by analysing and selecting the target companies.In their researches, authors Alimov and Officer (2017), Ahern et al (2015), Fresard et al (2017), Contractor et al (2014) have shown that these factors include cultural distance between acquirers and target companies, the difference in the protection of the intellectual property rights and sectoral affiliation between the target company and the acquirer.Also, a numerous studies researched predictive models for determining future targets based on the characteristics of companies which were already taken over (Graham et al., 2001;Baker & Kennedy, 2002;Branch & Yang, 2003;Officer, 2003;Schwert, 2000).
The aim of the study is to determine: 1.Who were targets, i.e. target companies in the Serbian market?2. Do the characteristics of target companies in the Serbian market coincide with the characteristics of those in other empirical studies?
3. Can future targets, i.e. target companies in the food industry be predicted based on the characteristics of target companies that have already been taken over in the Serbian market?
However, in transition countries, such as Serbia, there are problems with the availability of any data.In this paper, the authors analyse target companies in the food industry using data from financial statements.In Serbia, there is no database or agency that provides data on mergers and acquisitions, so it is very difficult to get information on the impact of the takeover process on the target company.Bearing that in mind, this authors' research significantly contributes to the economic policymakers, giving them information about the characteristics of the companies in the food industry which might become targets in the takeover process.

Literature Review
A large number of empirical studies researched the characteristics of target companies.The most commonly used criterion for the comparison was the average of industry in which target companies operated.Most studies analysed the following determinants of target companies: sale, HHI (Herfindahl-Hirschman Index), participation of the board of directors in the equity capital, ROA, ROE, research and development to assets ratio, operating cash flow to sales ratioliquidity ratio, Tobin Q ratio, debt to assets ratio -leverage, growth rate, share of fixed assets in total assets, etc.
One of the first research studies done on the market for corporate control in Great Britain in the 1960s concluded that target companies were less profitable and had a lower growth rate compared to the industry average (Singh, 1997).According to Baker and Kennedy (2002), target companies are more often less profitable measured by indicators ROA and ROE compared to the industry average.This corresponds to the research of Palepu (1986) according to which companies with lower return rate are targets of hostile attacks.There is a high level of agreement in the literature on measuring the profitability of target companies.Most studies confirm that target companies show a decline in profitability and performance indicators compared to the industry average.First empirical studies to confirm this were the aforementioned studies of the authors Singh (1997) and Meeks (1977).These authors also studied takeover likelihood and came to a conclusion that companies whose profitability was below the industry average were more often takeover targets.
Company's liquidity level and leverage are considered as key characteristics in choosing target companies.Target companies are more often a takeover target, if they have liquidity level above the industry average and degree of leverage below the industry average, as reported by Palepu (1986).Ambrose and Megginson (1992) also concluded that low leverage and a big share of fixed assets in total assets increased takeover likelihood.According to Jensen (1986), low leverage is an especially appealing characteristic, because it leaves the acquirer with the possibility to use available borrowing capacity after the takeover of the target company.Harrison et al (2014) conclude that a low leverage increases the likelihood of a firm becoming an acquisition target, which is confirmed by previous studies (Wruck, 1990;Clark & Ofek, 1994).Eichholtz and Kok (2008) analyse 122 mergers and acquisitions over the 1999 -2004 period.They compare targets, acquirers and a control sample using a pre-acquisition performance study, a two-sample t-test and consequently a multinomial logistic approach.Their study confirmed that targets have lower operational performance compared to companies that represented the control group.Song and Walking (1993), North (2001) and Graham et al (2001) also point out that the lower performance of the company increases the likelihood of these companies to be targeted for acquisition.
The acquisition processes of companies are characterized by cyclical activity.The first wave of the takeover process acquires on the US market was in the late 19th and early 20th centuries.A significant takeover activity in Serbia started at the beginning of the 21st century.Considering that fact, it is to notice that there is a time gap between the analysis of the target companies in the US from the literature review and the analysis of the target companies in Serbia in this research.This research gives its contribution defining the attractiveness determinants of the Serbian market for the future investors.It is especially valuable for the investors in the food industry because this market has not been explored yet, by analysing the companies as potential targets.
In accordance with the above listed empirical studies that dealt with the characteristics of target companies, we come to a conclusion that there is a great deal of agreement in the literature that target companies have the following characteristics: lower profitability, lower leverage, higher liquidity and higher efficiency compared to the industry average.
In this part of the study, the following hypotheses will be tested on the example of the Serbian market: H1: Potential takeover candidates are companies with lower leverage compared to the industry average.H2: Potential takeover candidates are companies with higher liquidity compared to the industry average.H3: Potential takeover candidates are companies with lower profitability compared to the industry average.H4: Potential takeover candidates are companies with higher efficiency compared to the industry average.

Materials and Method
The sample includes 30 companies from the food industry in chosen sectors.Considering that there is a total of 263 companies in the chosen sectors founded by legal entities, 167 thereof are in liquidation, and of remaining 96 companies, 30 companies selected for the analysis represent 33% of the total number which is considered to be a very representative sample for drawing valid conclusions.The sample consists of companies in the following sectors: 1032 -Juice production from fruit and vegetables; 1051 -Milk and cheese production; 1072 -Production of toast, crackers, biscuits and cakes; 1082-Production of coco, chocolate and confectionary products; 1089-Production of other food products; 1105 -Beer production; 1107 -Production of refreshing drinks and mineral water.
The study compares profitability, liquidity, leverage and efficiency determinants of the target companies with the industry average a year before the takeover.The Business Registers Agency (financial reports of companies and macroeconomic communications relative to the whole economy of the country) is used as a database in this study for chosen companies' determinants and for determinants of the industry average.Most often used ratios for testing company's profitability are ROA (return on assets ratio) and ROE (return on equity ratio).According to Jakšić et al (2015), ROA ratio should be higher than 0.1 and ROE ratio should be higher than 0.15.Liquidity is the ability of a company to pay off its immediate liabilities.Ratio of current liquidity should be greater than one.Leverage gauges the share of borrowed funds in total company assets, i.e. share of long-term debt in total capital.This ratio should be as low as possible and in accordance with the efficiency ratio, which shows the amount of fixed assets in total company assets, i.e. it shows the amount of fixed assets financed with owners' equity (Gogan, 2004).

Table 1 Chosen company determinants
Provided that the variables are parametric, the analysis will be done using parametric methods.Methods used in the study are multivariate statistical methods MANOVA, discriminant analysis of univariate methods ANOVA and t-test.Discriminant coefficient serves to distinguish which variables determine the subsample specificity and which variables are to be excluded from further analysis.Purpose of using mathematical-statistical analysis is to determine the characteristics of both subsamples -chosen variables and the industry average, homogeneity and distance between them, in order to perform precise prediction and forecast with certain reliability.The study tests whether there is a difference between chosen variables of the food industry and the industry average, in order to prove the stated hypotheses.

Results and Discussion
Based on descriptive statistics, results show: mean value, standard deviation (Sd), minimums and maximums of all values, the coefficient of variation (C.var.), confidence interval, a measure of asymmetry -Skewness, a measure of flatness -Kurtosis and Kolmogorov-Smirnov test values.After the layout of the results of descriptive statistics, this part of the study will show the difference between chosen company determinants, i.e. the stated hypotheses will be accepted or rejected, in order to assess the results and usefulness of further consideration, determine the direction and methodological priorities for their processing.Then, if the conditions are met, the characteristics and homogeneity of each chosen company determinant will be defined, as well as the distance between them.Based on the above shown table 4, it can be concluded that the result of multivariate statistical method MANOVA (p = .000),is below the significance threshold (p<0.05).That further implies that there is the statistically significant difference between analysed determinants of studied companies in the food industry and industry average.Based on the above shown table 5, it can be concluded that the result of the discriminant analysis method MANOVA (p = .000),is below the significance threshold (p<0.05).That further implies that there is the statistically significant difference between analysed determinants of studied companies in the food industry and industry average.

Source. Authors
Canonical correlation coefficient of .805implies a very strong model and important significance and correlation of discriminant variables in the formation of differences.Both canonical coefficient and the Wilks' lambda result (sig=.000)confirm a good choice of company determinants in the formation of differences.

Source. Authors
Calculating the discrimination coefficient isolates the variables that define the specificity of subsamples (subsample: companies and subsample: industry average) and variables which are to be excluded from further processing.The coefficient of discrimination indicates that the biggest contribution to discrimination between chosen company determinants and their average values was with (i.e. the biggest difference was observed with) Leverage (1.045), ROE (.827), ROA (.393), Efficiency (.273), Debt/Capital (.213) and Liquidity (.035).Since p < 0.01 is below the significance threshold with Leverage (.000) and Liquidity (.001), it can be concluded that these two determinants differ substantially between chosen company sample and the industry average.Since p > 0.01 is above the significance threshold with ROE (.167), ROA (.791), Debt/Capital (.758) and Efficiency (.476), it can be concluded that there isn't a substantial difference of these determinants between company sample and the industry average.However, discriminant analysis, which is more complex than MANOVA, included these determinants into analysis as latent variables and showed a significant difference between chosen company determinants and average values.Based on the above considerations and analysis of the sample of 30 companies and in accordance with the used methodology, logical sequence of the study is determining the characteristics and homogeneity of all chosen company determinants.Since the discriminant analysis showed (p= .000)that there is a significant difference, which further implies that there is a clearly defined border between chosen company determinants, i.e. it is possible to determine the characteristics of all chosen determinants a year before the takeover..580*p<0.05,note.chosen company determinants-1, average values for chosen determinants -2; note.↑(higher); ↓(lower).

Source. Authors
The above presented table 8, based on the contribution of the variable to the characteristics (% contribution) leads to a conclusion that the biggest contribution in creating the difference between determinants of chosen companies and industry average is made by determinants in the following order Leverage (37.51%),ROE (29.68%),ROA (14.11%),Efficiency (9.80%), Debt/Capital (7.65%) and Liquidity (1.26%).Based on the obtained results from the above presented table, it can be concluded that the target companies from the food industry in the Republic of Serbia have the following characteristics: lower leverage (Leverage*, Debt/Capital) than the industry average, lower profitability (ROE, ROA) than the industry average, higher efficiency than the industry average and higher liquidity* than the industry average.In order to make a prognosis with certain reliability, it is necessary to examine the sample homogeneity and whether the groups were classified in a manner to ensure model validity.

Source. Authors
The table shows that 27 of 30 companies have the characteristics of chosen determinants (n/m) and thus homogeneity of the sample is 90%, leaving 3 companies with other characteristics and not those of the chosen sample.Since sample homogeneity is 90%, the forecast can be made with certain reliability.In other words, it can be concluded with the certainty of 95% that companies in the food industry whose characteristics are similar to those of the chosen company determinants in the sample can become targets of hostile attack in the Serbian market.Reliability of 95% represents a very good indicator of the original classification of groups and classification of groups through the coefficient of determination.
Results confirm the hypothesis H1 that potential takeover candidates are companies in the food industry with lower leverage compared to the industry average, where that difference was also significant.This research is consistent with the previously undertaken studies by Palepu (1986), Ambrose and Megginson (1992), Wruck (1990), Jensen (1986), Clark and Ofek (1994).Hypothesis H2 that potential takeover candidates are companies in the food industry with higher liquidity compared to the industry average is also accepted which is in line with previous studies of Palepu (1986) and Harrison et al (2014).As previously shown, two determinants leverage and liquidity differ substantially, i.e. those two determinants of chosen companies differ significantly from the results of those two determinants for the whole branch of industry.This leads to a conclusion that in the process of takeover, foreign investors were particularly interested in companies with lower leverage because, as we have seen, the contribution to creating a difference for that variable was the biggest (37.51%).Low leverage enabled foreign investors to significantly increase the leverage in first years after the takeover, in accordance with the fact that most takeovers are financed through bank loans or another form of leverage.Unutilized borrowing capacity along with high liquidity (higher than the industry average), indicates that companies which were targets of the takeovers, were capable of covering their immediate liabilities and that leverage over a certain point should not endanger their liquidity.Although profitability and efficiency determinants did not differ significantly, they were included in the analysis by the discriminant analysis approach based on the coefficient of discrimination and contribution of variables to characteristics.
Classification of groups of 95%, as well as high canonical correlation coefficient, confirm the model's strength, thus proving hypotheses H3 and H4 that target companies in the food industry in the Republic of Serbia had lower profitability and higher efficiency compared to the industry average.Low profitability in comparison to the industry average indicates that companies are not run in the optimal manner and that after the takeover with the increase in leverage and realizing the effects of financial lever there can be a significant boost in profitability and therefore positive postoperative performances for the acquirer.The low profitability of the target in their research was confirmed by Singh (1997), Palepu (1986), Meeks (1977), Baker and Kennedy (2002).

Conclusion
According to the performed study, it can be concluded that the following study aims have been met.
(1) In the Republic of Serbia, takeover targets were companies in the food industry with the following characteristics: low leverage (leverage*, debt/capital), low profitability (ROA, ROE), higher liquidity* and higher efficiency compared to the industry average, where the determinants leverage and liquidity differed significantly from the average.
(2) Characteristics of target companies in the Republic of Serbia coincide with those of the target companies in other empirical studies shown in this paper.
(3) Since the homogeneity of the sample of 90% is considered very high it is possible to predict takeover targets.That is, with a certainty of 90% it can be concluded that companies operating in the Serbian market which have not yet been taken over but have low leverage, low profitability, higher liquidity and higher efficiency compared to the industry average, may become targets of hostile attacks.
Based on the given analysis of differences of chosen determinants between companies in the sample and the industry average, it can be concluded that takeover targets in the food industry did not use the effects of financial leverage in an appropriate way, or an increase in the share of borrowed funds to total equity, which reflected negatively in their profitability.Companies in the Serbian market which are not optimally run but have high efficiency and liquidity are especially attractive to takeovers, since foreign investors can significantly improve financial performances of the company by simply replacing bad management and financial politics.The market for corporate control forces managers to optimally run companies since missed financial opportunities can encourage foreign investors to make an offer for a takeover and achieve better financial performances.

Table 2
Measures of central tendency and dispersion parameters and measures of asymmetry and flatness of company's business performance ayear before the takeover for chosen company determinants

Table 4
Significance of the difference between chosen company determinants and average values for chosen determinants a year before the takeover

Table 7
Analysis of differences between chosen company determinants and average values for chosen determinants a year before the takeover

Table 8
Characteristics and homogeneity of chosen company determinants and average values a year before the takeover

Table 9
Classification Resultsa,c Cross validation is done only for those cases in the analysis.In cross validation, each case is classified by the functions derived from all cases other than that case.c.95.0% of cross-validated grouped cases correctly classified.