DETERMINANTS OF THE FINANCING STRUCTURE OF THE ACQUIRING COMPANIES IN THE PRE-MERGER PERIOD. RESULTS OF THE RESEARCH

: The article’s main objective is to identify the factors influencing the financing structure and capital changes in the pre-merger periods in the acquiring companies. The authors examined the relations between various industries and different size companies. Based on the analyses, it can be concluded that in small companies, the explanatory variables identified based on the literature review were statistically insignificant in many periods (in fact almost all). Completely different behaviour was observed in the group of large companies, where the same set of explanatory variables was statistically significant. The result of the research on the existence of non-linear relationships between company parameters is that in the case of some variables, there is no question of a linear nature of dependence. The study analysed the five years preceding the mergers of 307 business entities. The source of the survey data was the database prepared by the company InfoCredit SA for the Accountants Association in Poland. The authors used Statistica software and inductive reasoning for the study – supported by Spearman’s rank correlation analysis, linear and polynomial regression analysis, and variable scatter analysis.


Introduction
The process of capital concentration is a crucial stage in companies' operation, requiring specific preparation. Mergers of economic units are considered a form of entity development through external development, i.e. taking over other entities.
Analyses often show that acquiring companies owe their strength to the good macroeconomic situation and the booming domestic economy. The companies they take over come from countries with a slower economy and in recession. This situation is related to access to financing sources, which are the driving force behind companies' development. Many studies have confirmed this statement (Acharya, Shin, & Yorulmaz, 2010;Aguiar & Gopinath, 2005;Desai, Foley, & Forbs, 2007;Krugman, 2000). The regression analysis (Makaew, 2010) indicated that more cross-border merger transactions were carried out when companies were in good economic conditions. This research covers only domestic capital concentration transactions, and therefore examines the macroeconomic situation's impact in this preliminary study. Reddy (2015) indicates that in research on capital concentration processes (mergers of companies), including transaction failures, most often the financial aspects -the reduction of value for owners -are analysed in scientific studies. Economic and market data are examined when announcing the intention to merge and after the date of the merger (cf. De Bernardis & Giustiniano, 2015;McCann & Ackrill, 2015;Munjal & Pereira, 2015). Only a few studies indicate improperly conducted negotiations and the preparation of companies for mergers, i.e. periods preceding joining businesses , as the reason for the failure of mergers and acquisitions (cf. Ahammad, Tarba, Liu, Glaister, & Cooper, 2016;Caiazza & Volpe, 2015;Friedman, Carmeli, Tishler, & Shimizu, 2016, as well as Lee, Park & Kim, 2014). Therefore, this research focuses on the periods preceding the merger of economic entities.
An essential factor shaping capital concentration is the legal environment, which organises many activities related to mergers. Studies on the influence of the legal environment and political transactions, such as Porta, Lopez-de-Silanes, Shleifer and Vishna (1998), Klapper and Love (2004), as well as Feldman and Kumar (1995), were examined. The valuation of the companies' shares is sensitive to the financial information sent by the merging companies. Developing capital markets are characterised by the more significant asymmetry of information and unequal access to financial information. These issues were the research subject for Brunnermeier (2005) and Cornett, McNutt, Strahan and Tehranian (2011). The authors of this article focused on the Polish market to avoid the influence of multinational legal and political factors, however this is a preliminary study that will be extended to include other developing countries from Central and Eastern Europe.
In scientific research, there are many discrepancies about a company's financial condition after a merger, regardless of the motives for conducting merger transactions. The companies and the business environment's internal factors prove that mergers can positively impact on the financial results (Ghosh, 2001;Heron & Lie, 2002;Linn & Switzer, 2001;Martynova, Oosting, & Renneboog, 2006;Powell & Stark, 2005). Studies show that mergers do not significantly improve the financial situation (Martynova et al., 2006;Moeller & Schlingemann, 2004). The complexity of the conditions in which entities merge, justifies the need to examine mergers.
There are many publications concerning the field of the shaping of the capital structure. Numerous theories attempt to justify the behaviour of companies in terms of the selection of equity or liabilities. For example, based on 11,553 observations of European companies (excluding Polish ones), Castro, Fernandez, Amor-Tapia, and de Miguel (2016) noted that the financing structure (liabilities/total capital) is positively correlated with the operating profitability of assets, asset structure (fixed assets/total assets) and company size (Ln assets). They examined companies in various stages of development (start-up, growth and maturity), and analysed the financing structure determinants using GMM regression analysis and the LEV variable as a dependent variable. They established that the asset structure positively impacted on the financing structure in all company development phases. This means that the greater the share of fixed assets, the smaller the financing structure's share of liabilities. The asset structure's impact on the financing structure is the greatest (based on the variable's regression coefficient) in the group of companies starting the business activity. In the subsequent phases it gradually decreases, which confirms the occurrence of financial difficulties in the initial stages of company development. Another determinant of the financing structure is the operational profitability of assets (cf. Castro, Fernandez, Amor-Tapia and de Miguel). In all the company development phases considered by them, operational profitability had a negative impact on the structure of the capital. The higher the operating profitability, the lower the share of liabilities in the financing structure. This negative relation can be explained by the pecking order theory, according to which companies prefer to use their equity first. From the company size perspective, this theory seems more suited to companies with a stabilised operating and financial situation, i.e. mature companies. Small companies, as well as companies from various sectors of economic activity, may show other dependencies.
In turn, Vithessonthi and Tongurai (2015), in the period 2007-2009, examined over 170 thousand companies from Thailand to find out whether a company's size has an impact on the shaping of the relationship between the financing structure and the profitability of assets. They observed that the literature's discrepancies in the positive or negative relationship between variables might result from this relationship's nonlinear nature. In the case of polynomial relations (e.g. quadratic), there may be a positive relationship in some intervals, while in others -a negative one. Vithessonthi and Tongurai also determined that in small companies, the relation between capital and profitability is positive, whereas in the group of large companies this relation is negative. This observation is in line with the agency theory proposed by Margaritis and Psillaki (2010), who assumed that the independent variable would be the capital structure and the dependent variable -the return on assets in the regression models.
The relations between the ratios representing assets and liabilities of companies are often analysed during studies of net working capital and its impact on companies' results. Thus Afrifa, Tauringana, and Tingbani (2015) examined 1,126 companies from alternative investment markets, 141 of which met the assumed criteria for the availability of financial data over eight years and the definition criteria for small and medium-sized companies. The conducted correlation analysis indicated a positive relationship between the financing structure and the structure of assets. Regression analysis showed the LEV variable's statistical irrelevance in the model describing the shaping of the assets' operating profitability. The structure of assets with a negative coefficient was an essential variable in the model.
By analysing a group of 250 companies, Afrifa and Padachi (2016) proved that while the structure of financing is positively correlated with the structure of assets, it is not significantly statistically associated with assets' profitability. The lack of impact of the financing structure on the return on assets disclosed in regression models may suggest restrictions in selecting finance opportunities in the group of small and medium-sized companies. Klapper, Sarria-Allende and Zaidi (2006) analysed the financing method of Polish companies from the small and medium-sized enterprise sectors. The study covered the period 1998-2002 and the research group comprised over 17,000 companies. The authors determined that larger companies had higher capital structure ratios, while older companies had lower financing structure ratios. The regression analysis also showed positive relationships between the structure of assets and the structure of financing. The main conclusion is that larger, younger, growing and more profitable companies with a larger share of fixed assets make greater use of liabilities as an element of financing. The authors determined the relations between selected variables (company's size, profitability, asset structure) and the shaping of the capital structure in line with leading financial theories -the trade-off and pecking order theories. The authors' predictions indicate a positive relation between the structure of assets and the financing structure in both approaches, which means that fixed assets serve as collateral for the incurred liabilities. According to the pecking order theory, more profitable companies will finance their operations more often using equity. Using the example of Polish small and medium-sized enterprises, the authors showed that the companies' profitability was negatively related to their capital structure, which is another argument for the pecking order theory. In the analysed companies, the structure of assets was positively associated with the financing structure. The authors also found that the larger the share of fixed assets in total assets, the larger the share of long-term liabilities in the financing structure. Grabiński (2016) (who audited 36,361 financial statements of companies for 2007-2014 from 27 European countries) and García-Teruela and Martínez-Solano (2007) (who audited 8,872 small and medium-sized enterprises in Europe in the period from 1996 to 2002) provided a specific view on the relation between the parameters describing the activity of economic entities. They noted that the return of assets (ROA) is positively correlated with the size of companies and negatively correlated with the debt ratio.
Studies on a group of Polish companies indicate that the financing method's choice cannot be justified based solely on one theory. According the trade-off theory, companies have an optimal financing structure that balances the benefits and costs associated with restrictions on access to capital, and the size of income tax. The existence of an optimal level of financing makes companies focus on achieving this level. Castro et al. (2016) claim that the existence of an optimal level of financing Determinants of the financing structure of the acquiring companies in the pre-merger period 21 does not contradict the pecking order theory. Companies do not aim at the optimal financing level, but maintain the proper relationship between cash flows and investment requirements. Depending on the company's life cycle phase, the benefits and costs of selecting the company's financing sources will change. Growth-phase companies have a large share of fixed assets that mainly act as collateral for foreign financing (Titman & Wessels, 1988). Additionally, companies in their growth phase are of a size that allows them to be diversified (González & González, 2008). In their maturity phase, companies have the greater trust of both owners and the market, hence those with higher profitability may be more heavily indebted, benefiting from more considerable tax savings.
The process of capital concentration as a form of further company development should occur while taking internal development opportunities. This situation can take place in the mature phase of a company's life cycle. Numerous studies on the structure of capital most frequently indicate the following set of determinants: profitability (understood as return on assets), the intensity of growth (change in sales), the durability of asset structure, and company size (measured with the total balance sheet assets) (Castro et al., 2016;Mataigne & Vermaelen, 2016;Zhou, Tan, Faff & Zhu, 2016). Studies on companies in their maturity phase showed a negative relationship between the structure of capital and companies' profitability. This confirms that companies with a stable position retain a part of the financial result, in line with the pecking order theory, and the relation between the share of fixed assets and the financing structure is positive, while the regression coefficient is smaller than in the growth (initial) phase. This situation also indicates the greater involvement of the funds achieved by companies in fixed assets.

Research methodology
In the study the authors used a database prepared by InfoCredit SA for the Accountants Association in Poland. The database includes all mergers entered into the National Court Register that were available when creating the database (307 transactions).
The database was adapted to this article's needs, so that all the analysed merger cases covered exactly the five years preceding the mergers. According to the size criterion, the research sample was divided into small and large companies ( Table 1). As the companies included in the study are subject to the Polish Accounting Act and the Code of Commercial Companies, this division was based on the Accounting Act guidelines, according to which small companies should not exceed at least two of the following three amounts: a) PLN 17,000,000 -in the case of total balance sheet assets at the end of the financial year, b) PLN 34,000,000 -in the case of net revenues from the sale of goods and products for the financial year, c) 50 people -in the case of average annual employment calculated as full-time positions. 5 Due to the lack of information on average annual employment, the total balance sheet criterion and the value of revenues were taken into account in company classification. Companies that did not exceed these two parameters were considered to be small entities, while all the others were included in the group of large companies.
The transition from being small companies to becoming large entities indicates their development in the studied period. Between the fifth and second year before the merger, the size of the large companies' group increased by 28 entities, i.e. by nearly 15%.
All the companies included in the study were also assigned to a relevant sector, based on the first entry in the business entity classification.
The variables used in the study include quantitative and qualitative variables (transformed into binary variables) defining the economic activity sector and the reporting year.
In the group of scientific publications in finance, as many as 74% of authors, based on research conducted by Berent (2013), measured financial leverage using the D/ (D+E) measure; D/E (14%) as well as D (10%) and (E+D)/E (2%) were also used. Similar results of the study were obtained by Berent (2013) for financial leverage measures used in accounting journals. Interesting results can be obtained by introducing into the analysis a variable determining the number of years of economic activity of the surveyed economic unit, which allows to examine differences in the behaviour of young and mature companies. In the research conducted by Afrifa and Padachi (2016), a positive relation (correlation matrix) between the operational profitability of assets (ROA) and the variable determining the company's age can be observed.
The literature review shows a positive correlation between the asset structure (often referred to interchangeably as asset sustainability in the literature) and the funding structure (Frank & Goyal, 2009;Jõeveer, 2013;López-Iturriaga & Rodriguez-Sanz, 2008). However, the assets structure -such as maintenance costs or depreciation costs -affects companies' business performance. Numerous studies have shown that the relationship between the asset structure and the return on assets is negative (Afrifa & Padachi 2016;Matias & Serrasqueiro, 2017). The smaller the share of fixed assets, the higher the companies' profitability from their assets. Based on a literature review covering the period 2004-2016 (Matias & Serrasqueiro, 2017), it can be concluded that in the different types of companies, the relationship between the asset structure and the financing structure as well as the profitability of the companies, is not unambiguous (i.e. only positive or only negative).
Quantitative variables are: • ROA ROA = Net profit or loss , Total assets

Total assets
The verification of the research hypotheses was carried out using Statistica software.
The study used the following statistical tools: Spearman's rank correlation analysis, the Mann Whitney equality median test, the Kołmogorow Smirnow equality mean test, and regression analysis.

Research results
During the research the authors used commonly accepted determinants of companies' financing structure. First of all the study examined the correlation between variables, emphasising the correlation between the LEV variable and other variables, and Spearman's rank correlation analysis due to being less sensitive to deviations from the linearity of the scatter of variables.
The authors conducted the dependency in the periods preceding mergers, starting five years before the merger and ending two years before it. The groups of small companies and those of large companies were analysed separately. Table 2 shows Spearman's rank correlation for small companies five years before the merger.

Five years before the merger -small companies
Based on the table, it was concluded that the LEV variable was significantly correlated with the ROA OP variable only. This correlation is negative, which may mean a decrease in the profitability of assets and debt increase, or vice versa -a lower debt coefficient for more profitable companies. It is worth adding a graphical representation of the relations between variables, which makes it easier to assess the nature of this dependency -linear or non-linear ( Figure 1).
An assessment of the nature of the dependency, based on the scatterplot, may not be unequivocal. The authors extended the analysis with a linear and polynomial regression analysis, and based on the results selected the relation, characterised by the greater suitability expressed by the R2 determination coefficient. An evaluation of the parameters is presented in Tables 3 and 4. Based on Table 4, the authors concluded that there is a linear relationship between the ROA OP variable (explanatory variable) and the LEV variable (explained variable). Table 5 shows Spearman's rank correlation for large companies five years before the merger.

Five years before the merger -large companies
The LEV variable is correlated with the TANG quantitative and qualitative sector 1 and sector 2 variables (Table 2). Companies belonging to the production sector had a lower share of foreign capital than companies belonging to the commercial sector. Determinants of the financing structure of the acquiring companies in the pre-merger period

25
In addition, an increase in the share of fixed assets in the assets structure impacted on the decrease in the share of liabilities in equity and liabilities. This fact can also be interpreted as the fact that the companies getting into debt allocated funds to current assets, which reduced the share of fixed assets. The LEV and TANG variables scatter is shown in Figure 2. Figure 2 shows the polynomial relationship between the LEV and TANG variables. To verify the non-linear nature of the relationship, the authors carried out a linear and polynomial regression analysis: TANG variable (explanatory variable) and LEV variable (explained variable). Table 6 shows the polynomial regression analysis, and Table 8 the linear regression analysis.
The results of the polynomial model regression analysis are presented in Table 7.
The results of the linear model analysis are presented in Table 9. Based on the regression analysis, the authors concluded that the relationship between the TANG and LEV variables is non-linear with a critical (minimum) point. Table 10 shows Spearman's rank correlation for small companies four years before the merger.

Four years before the merger -small companies
There was no statistically significant correlation between the LEV variable and other variables in the group of small companies. Table 11 shows Spearman's rank correlation for large companies four years before the merger.

Four years before the merger of large companies
Based on Table 11, it was observed -similarly as in five years before the merger in large companies -that the financing structure was correlated with the economic activity sector (sector 1 -negative correlation, sector 2 -positive correlation) and the structure of assets (negative correlation coefficient). Figure 3 shows the scatter of LEV and TANG variables. Based on Figure 3, the authors concluded that there is a polynomial relationship between these variables: TANG and LEV. Table 12 shows the polynomial regression model, and Table 13 summarises the polynomial model. Table 14 provides a summary of the linear model.
Based on the regression analysis, the authors concluded that the determination coefficient is higher in the polynomial model, which means the better suitability of this model to variables.
Three years before the merger -small companies Table 15 shows Spearman's rank correlation for small companies three years before the merger.
The correlation analysis in Table 15 indicates that the LEV variable was negatively correlated with the asset structure and positively correlated with the sales volume, which means that companies with higher sales were more indebted. This would be in line with the trade-off theory, in which increased sales and thus tax income would be offset by the higher costs of interest on liabilities. This situation coincides with the initial phase of company life. Figure 4 shows the scatter of LEV and TANG variables. Based on Figure 4, the authors concluded that the dependence of the TANG variable (explanatory variable) and the LEV variable (explained variable) is polynomial. An analysis of the summary of regression models (polynomial -Tables  16 and 17, linear -Table 18) indicates a better fit in the polynomial model.
There is no linear relationship between the TANG variable and the LEV variable. Table 19 shows Spearman's rank correlation for large companies three years before the merger. Based on the correlation analysis in Table 19, the authors concluded that the LEV variable was negatively correlated with the TANG variable and the sector 2 variable. This means that commercial companies had a larger share of liabilities in the financing structure than other companies.

Three years before the merger of large companies
The relation between the TANG and LEV variables is presented in the scatter graph in Figure 5. Based on Figure 5, the authors observed a polynomial relation between the TANG (explanatory) and LEV (explanatory) variables. Further regression analysis indicates that the only statistically significant model describing these variables is the polynomial model. Table 21 shows the results of the polynomial regression analysis. Table 22 shows the results of the linear regression analysis.
Two years before the merger -small companies Table 23 shows Spearman's rank correlation for small companies two years before the merger.
A linear regression analysis is presented in Tables 26 and 27.  Based on Tables 24, 25, 26 and 27, the authors concluded that the relation between the LEV and ROA OP variables is linear. Table 28 shows Spearman's rank correlation for large companies two years before the merger.

Two years before the merger -large companies
As in previous periods, two years before the merger in large companies, the financing structure is negatively correlated with the TANG variable and the variable representing the production sector, and positively correlated with the variable representing the commercial sector.
The relation between the LEV and TANG variables is presented in the scatter chart in Figure 6. Based on Figure 6, one can conclude that the relation is nonlinear, however the summaries of the models (polynomial -Tables 29 and 30) do not confirm that.
A linear regression analysis revealed that the linear model is statistically significant (Tables 31 and 32). Table 33 presents the correlations of the LEV variable (financing structure) in the group of large companies.
The analysis of the data in Table 33 allows to observe certain regularities in the financing structure. In all the analysed periods there was a negative correlation between the structure of assets and financing structure. This suggests that large companies could finance their investments in fixed assets using the generated financial results, as well as their capital. In addition, a positive correlation between the financing structure and the variable representing the commercial sector is visible in all the years. This correlation means that companies from the commercial sector had a larger share of liabilities in the financing structure than other companies. In almost all the analysed periods there was also a negative correlation between the LEV variable and the variable representing the production sector. In this case, production companies preferred higher financing with equity.
In small companies no regularity in the correlation between the LEV variable and other variables can be observed. Table 34 presents the summary of correlation analysis in the group of small companies.
The lack of significant dependence of the financing structure and other parameters of the company's capital may result from the weaknesses and difficulties in conducting business by these entities.

Conclusion
The article's main objective was to identify the determinants of the financing structure in Polish companies in the periods preceding the mergers of business units. The research sample included 307 Polish acquiring companies was divided into two groups: small companies and large companies.
Based on the presented analyses, the authors concluded that in small companies the explanatory variables identified based on the literature review were statistically insignificant in most periods (almost all). The authors observed completely different behaviour in the group of large companies, where the same set of explanatory variables was statistically significant.
Additionally, during the research the authors discovered that there is no question of a linear nature of dependence in the case of some variables. This situation creates the need to thoroughly analyse the nature of the relation because there may be extremes in the examined parameters. Similarly, the strength of the relation may vary depending on the distance from the extremes of the function.   Source: authors' own work.   Source: authors' own work.  Source: authors' own work.

Tables
Determinants of the financing structure of the acquiring companies in the pre-merger period 33 Table 11. Spearman's rank correlation for large companies four years before the merger Source: authors' own work.  Source: authors' own work.  Source: authors' own work. Table 15. Spearman's rank correlation for small companies three years before the merger Source: authors' own work. Source: authors' own work. Table 19. Spearman's rank correlation for large companies three years before the merger  -0,136422 -0,180233 -0,197110 -0,183651 -0,255576 1,000000 Source: authors' own work.  Source: author's own work. Source: authors' own work.
Determinants of the financing structure of the acquiring companies in the pre-merger period  Source: authors' own work.  Source: authors' own work. Source: authors' own work.  Source: authors' own work.
Determinants of the financing structure of the acquiring companies in the pre-merger period 39 Source: authors' own work.  Source: authors' own work.