The Determinants of Capital Structure in Central and Eastern European Firms under the Influence of the Euro

The Determinants of Capital Structure in Central

and Eastern European Firms under the Influence

of the Euro

Master Thesis

MSc International Financial Management

By KASPARS UKIS (s1845411)

Supervisors: WIM WESTERMAN and HEIN VROLIJK

ABSTRACT

This research focuses on determining the effect of country and firm specific factors and the effect of establishment of euro on leverage of listed firms in Central and Eastern Europe. First, the author accumulates literature findings on capital structure theories and compares these findings between core (West) and new (CEE) European countries. Second, the author gathers and analyses data on CEE firms. The most popular factors are accompanied by some factors that have not yet been analyzed in literature and findings are presented. Findings suggest a decreasing direct influence of country factors compared to firms specific determinants. Almost all firm specific factors studied are found significant, however the effects vary across countries, suggesting that a single uniform capital structure theory cannot be used to explain different leverage ratios. It is established that the effect of adoption of euro and overall integration of EU financial markets has positive influence on leverage.

2

Introduction

One could say that introduction of the euro was the biggest landmark event in European economic history. Establishment of European Monetary Union (EMU) created an opportunity for fully integrated financial market on a scale that can be compared to the United States. The main facilitator of integration of financial markets is the elimination of currency rate risk within the Eurozone, however it is but one of the benefits that the euro offers its members. By relaxing the regulatory, technical and psychological barriers among its members, the Eurozone market has overcome previous segmentation. On the other hand, slow recovery of Eurozone countries indicates that the euro might not be a cure for every setting. Moreover the experience we had with the common currency is too little to make conclusions for its wider applicability. Only 11 EU states have had the euro for as long as 14 years which makes it understandable that many of the new EMU candidates hesitate to jump on the band wagon. Although the decision to join the Eurozone or not does not lie in our or even our governments hands (except Denmark and the UK who opted out due to economic sovereignty), the benefits of the euro are still up for debate. Aim of this paper is to contribute to discussion about the effects of euro on corporations within the member states.

3 relationship to a strong monetary system are to benefit significantly from giving up its currency in favor of the stronger one.

The major economic downturn experienced in the euro area between 2008 and 2009, and the subsequent recovery observed in recent quarters, coincided with strong movements in the growth rate of credit to the private sector (European Central Bank, 2011). Collapse of large financial institutions has led to a somewhat skeptical attitude towards debt financing. In order to stimulate the Eurozone growth, the current monetary policy of ECB is offering unseen low interest rates. ECB has cut its benchmark interest rate to 0.15% and, for the first time in its history, introduced a negative deposit rate for banks. As a result, banks will be penalized for holding on to the money and offer cheap long-term loans to private sector. This paper will try to establish whether the change in interest rate offered by the ECB has effect on firm leverage. In particular, how does introduction of the euro in new member states affect the debt financing patterns of corporations? Moreover, it will be tested whether other factors influence firm’s debt capacity and do these factors behave differently across countries. Similar studies by Bris et al (2009, 2011) and Jóveer (2013) have looked at the financing structure and factors influencing it separately, this paper however will try to integrate the establishment of the euro in a country, the local capital structure and the variables affecting it in one study. Moreover, the countries under the question here will be those that have been under the radar of related research. Majority of academic attention so far has been devoted to the equity financing benefits of euro-area firms. Additionally, the scope of academia usually limits to the original 11 members of the EMU, if not just a hand-picked few most appealing to the author. This does not reflect the full picture. As suggested by the literature, the influence of the euro on firm’s financial behavior is by far not uniform to all situations. To gain a deeper understanding of the “euro-effects”, we must broaden the perspective and take into consideration the new Eurozone members that have enjoyed the benefits of the euro for a relatively brief time or just about to start enjoying them. This paper will establish whether there is a difference in debt financing between “new” Eurozone members (countries that joined as of 2008) and the candidates (the countries that are yet to join). The purpose of doing this is to discover whether the decision to take on debt is more correlated to Eurozone benefits of country specific factors such as GDP growth.

4 understanding of what was going on left of the Wall. Only in the end of 1980’s, when the Soviet bloc collapsed, the former communist states could get a breath of freedom. With that freedom came the free market economy and another ‘abnormality’ for the local entrepreneurs: competition. Severe economic and institutional and social reforms were needed to adjust post-communist countries and the people’s mindsets towards the real world as we know it now. Shortcomings and mistakes were made which deprived the economies of the region to grow and foster an enterprise-friendly environment. Loss-making state owned enterprises (SOEs) were not fit for the new world. Moreover, most transition economies did not have the finance needed for restructuring and growth of new formed private firms (Nivorozhkin, 2005). This might seem irrelevant, but most of these countries had only a mere 20 something years of experience with market economy and all its mechanisms. It seems only logical that many of the management norms of the West are not that rooted in the Eastern European countries or that they have rooted differently under the institutional and cultural pressures. So we could expect that findings about capital structure and its determinants will be different when compared to Western European counterparts. Some researchers (Nivorozhkin, 2005, Delcoure, 2007, Jõeveer, 2013) have studied the leverage determinants in CEE countries during transition period, however there seems to be no research of such kind for the years following the transition.

Results of this paper indicate that effects of various leverage factors are not uniform across countries and that no single best capital structure theory can explain all the leverage variations. The leverage of the firms in the current dataset seems to be almost unaffected by country specific factors such as GDP growth, inflation and country overall risk index. At the same time, firm factors still explain a great deal of the leverage variety. Engagement into Eurozone has shown to have positive impact on leverage in Slovenia and Slovakia, moreover ascension to the European Union has a positive impact on leverage in Bulgaria. The coefficient of the crisis dummy suggested a negative impact on leverage, however findings were not significant.

5

Literature review

The establishment of the EMU in 1999 has had major influence on integration of European capital markets by making them more efficient, less costly and more competitive. The idea behind the integration is that financial intermediaries such as banks have an increasing number of investment opportunities as the EMU is growing. According to the law of one price, an integrated market offers same price for securities with same cash flows. This means that a household or a firm within EMU should be able to access external finance on the same terms regardless of the country in which it is located (Jappelli and Pagano, 2008). Unfortunately this is only possible in an ideal world. Jappelli and Pagano (2008) describe hurdles that stand in way of financial integration. First of all, national differences in taxes and subsidies can create differences in after tax cost of capital. Second, countries can impose regulations that diminish the competition between financial intermediaries, for instance harder requirements for foreign banks. Fourth, inefficient judicial systems can lead to higher interest rates, because lending institution will have to insure itself for loss in case of default. Lastly, information asymmetry between domestic and potential foreign intermediaries can serve in favor of those who know the local market better. Additionally, Albulescu (2011) finds that post-crisis economic instability (underdevelopment, vulnerability and banking fragility) could be the biggest obstacle for financial integration of Central and Eastern European countries (CEECs). Overall, the impact of the EMU has been considerable, but the progress towards integration of the various national markets has been uneven across different market segments (Galati and Tsatsaronis, 2002) As countries overcome barriers for integration the costs of euro-denominated capital decline. Bris et al (2009) find evidence that the euro has decreased the market risk of firms within the EMU. Stock returns within the EMU have become more sensitive to euro-risks rather than to country-specific risks.

6 increase of external financing, except for Denmark. This makes sense considering that Denmark maintained a fixed exchange rate with the euro.

Bris et al. (2011) identify supply as a major contributing factor, explaining the increased use of external financing. Countries with previously constrained financing markets were suddenly part of a larger pan-European market that was attracting a lot of capital which helped to alleviate financial constraints they were facing prior to the euro. Nivorozhkin (2005) confirms the supply side influence on capital structure. He concludes that firms in transition economies of Eastern Europe have low levels of leverage due to the supply-side phenomenon.

As supported by Bris et al (2009, 2011), firms from countries that were exposed to currency risks (high inflation rates) before euro adoption benefit more from joining the Eurozone. This is because the euro eliminates currency risk and thus decreases the risk premium required by investors which as a consequence lowers the cost of equity. Thus they conclude that firms from countries suffering from high inflation benefit considerably once the euro is established. Hardouvelis et al (2007) argue that there is important causal relationship between joining EMU and reduction in cost of capital. They find that establishment of the euro has led to a significant reduction in cost of equity in the vast majority of sectors (Hardouvelis et al, 2007, p.326). Moreover, EU countries that did not join the euro, Denmark, Sweden and the UK, demonstrated a much smaller and statistically insignificant reduction in cost of equity. Hardouvelis et al (2007) also find that introduction of euro fosters convergence in equity costs across countries, however the differences across sectors seem to remain.

So far we identified several factors that affect the financing behavior of firms under the presence of euro. Financing decisions however are affected by many other factors. Literature offers an endless list of different measures and conditions that affect firms leverage. These factors can fundamentally be clustered in three general groups: firm-specific factors, country-specific factors (often separated into macroeconomic and institutional factors) and industry specific factors. However, this distinction tends to vary across studies. The variables within these groups and the measurements of these variables tend to differ.

7 concentration of an industry. Munificence is industry’s capacity to support a sustained growth. Such industries have abundant resources and low competition normally resulting in high profitability. Dynamism is a measure of non-predictability or instability of industry. Market concentration is measured as Herfindahl–Hirshman index- sum of squares of market shares of all firms in the given industry .All three factors show significant negative influence on leverage and account for 11,6% leverage variance in the dataset.

Another inconsistency in the literature is (non-)separation of country-specific factors into macroeconomic and institutional. Or, another case, De Jong et al. (2008) conclude that country-specific factors matter in determining and affecting the leverage choice around the world, however legal enforcement, creditor and shareholder right protection account for more variance than macroeconomic measures such as capital formation and GDP growth rate. One could say that macroeconomic and institutional factors should be separated because they follow different paths of development and are affected by different forces. These inconsistencies can affect the interpretation of findings, however as long as the definitions and measurements of factors are the same across studies, the findings should also be consistent.

Theories of Capital structure

Modigliani and Miller (1958) established one of the most cited theories in corporate finance, which suggest capital structure irrelevance. If debt policy was completely irrelevant then leverage would vary randomly across firms and industries. The drawback of MM is simple yet important- it only holds in a perfect capital market. In order to destroy the utopic image of capital market we must consider taxes, imperfect information availability and human factors, specifically agency costs. When bringing these disturbances into picture, it turns out that the financing choice is important after all. The two most influential theories of capital structure are the trade-off theory (TOT) and the pecking order theory (POT).

8 for investors, which consequently will negatively reflect on the market value of the firm. As a result, TOT assumes that for every firm there is an optimal profit maximizing leverage ratio, which managers should maintain. Large firms with safe and tangible assets and with higher profitability could enjoy larger tax shield benefits of debt and hence should have higher leverage. TOT is highly appealing for capital structure scholars because its explanations are commonsense, however it fails to explain correlation between high profitability and low debt ratio. According to TOT, when a firm is profitable it will avoid free-cash-flow agency problem by paying dividends. Therefore, in order to discipline managers, internally generated profits will be substituted with debt. Nevertheless, in reality it is often observed that most profitable firms and firms who are leaders in their industry often are reluctant to use large amounts of debt and utilize internally generated cash to finance its investments.

9 on leverage according to TOT and POT. Correlation of various factors and leverage will be discussed in detail in the next part of the paper.

Table 1 Effect of firm-specific variables on leverage according to trade-off theory and

pecking order theory Theory\Factor

Tangibility Earnings volatility

Size Profitability Growth

Trade-off + - + + -/+

Pecking order - - - - -/+

Another, less influential, capital structure theory is market timing theory (MTT). Although a relatively old idea, literature still offers some support for it. MTT suggests that managers will issue stock or debt capitalizing on price run-up. Managers will closely monitor both markets and use the one which looks more favorable. While this approach might be used by managers, it has nothing to say about most traditional factors that influence corporate structure, however it does suggest that market conditions and opportunistic behavior of managers play a role in capital structure decisions (Frank and Goyal, 2009). Behavioral explanations for financing choices of firm are documented by Bertrand and Schoar (2003). They present two important findings: first, managers of older generation seem to be financially more conservative, whilst managers with MBA degree follow more aggressive strategies; second, the behavior of managers depends on corporate performance and ownership structure.

All in all, Frank and Goyal (2009) identify the following factors as ones most considerable for studying effects on firms leverage: profitability, firms size, growth, industry conditions, nature of assets, risk, supply-side factors, stock market conditions, debt market conditions, macroeconomic conditions.

The aim of the following part is to analyze this list considering the aim of this paper and pick out the few factors that are most applicable.

Firm-specific factors

10 therefore less prone to bankruptcy, which increases their capacity to tolerate debt, favoring the TOT. However, POT advocates claim that larger firms avoid information asymmetry typical to smaller firms, thus in case of necessity they can issue equity more easily. Third, firms with tangible assets are less exposed to financial distress and thus can access debt with less cost. Fourth, market-to-book ratio conveys similar results to both TOT and POT. TOT suggests that market-to-book implies higher financial distress costs, hence a lower debt ratio. POT interprets market-to-book ratio as another profitability indicator, also suggesting lower leverage.

Nivorozhkin (2005) argues that the only variables with a uniform effect on leverage across all CEE countries are firm’s profitability and age. He also suggests that ownership by managers and ownership concentration (large amount of shares in hands of one shareholder) seem to have a significant influence on leverage in several CEE countries, however due to inability of research to separate types of owners, the results are mixed. Theoretically, large ownership by managers decreases agency costs, which reduces the need for debt as disciplinary measure. Moreover, managers owning shares are expected to be more risk averse and reduce the exposure to risk of bankruptcy, hence a decreasing leverage. On the other hand, if ownership is concentrated, the major shareholder has incentives to closely monitor and control the management, possibly using increased levels of debt as a control mechanism.

De Jong et al (2008) suggest that even the core firm factors such as tangibility, firm size, risk, growth and profitability may be insignificant in some countries and in rare cases even inconsistent with theories (having opposite significant signs). These findings imply that it is not valid to construct a model with a single pool of all companies in the world and test the impact of country-specific factors assuming that cross-country firm-specific determinants are equal (De Jong, 2008). The findings of Nivorozhkin (2005) also imply that effect of some factors tends to vary across EU countries. In order to assess the national differences, country dummies are suggested to be used in further research.

11 leverage, while the remaining factors only add 2%. Four of these factors are firm specific. Negative influence on the level of leverage is attributed to firms with high market-to-book ratio and firms with high profits. More leverage is observed in firms that have more tangible assets and large firms (high book assets).

The main findings of most relevant capital structure researches are presented in Table 2.

Table 2 Summary of recent related capital structure research (continued in page 12)

Titman and Wessels (1988)

Profitability (-) Size (+) Tangibility (+)

Uniqueness of a firm contributes to less leverage. Transaction costs significantly affect capital structure choice

Rajan and Zingales (1995)

Tangibility (+) Profitability (-)

Market-to-book ratio (-) Size (+)

Data from G7 firms suggests that leverage and its determinants are similar across countries. In order to establish fundamental factors of leverage, deeper understanding of institutional differences is needed.

Nivorozhkin (2005) Size (+) Profitability (-) Age (-)

Tangibility (-  +)

Leverage in transition economies is lower than rest of EU but tends to converge. Dynamic adjustment model illustrates potentially costly nature of adjustment to target leverage. Large firms adjust their leverage faster and less costly.

Delcoure (2006) Size (+) Tangibility (+) Profitability (-)

Non-debt tax shield (+) Corporate Tax rate (+)

Trade-off theory and pecking order theory when modified partially explain corporate capital structure choices in the CEE countries. CEE firms have more factors to consider. Firms from transition economies rely more on short term debt.

Influence of institutional structure, ownership and dividend policy should be studied. De Jong et al (2008) Profitability (-) Size (+) Tangibility (+) Growth opportunities (-) Risk (-) GDP Growth (+) Development of financial market (+)

Capital structure theories work well in developed legal and economic environments. Firm-specific determinants vary across countries.

Direct and indirect influence of country factors is observed. Indirect influence changes the role/importance of firm factors across countries. Indirect impact can differ from direct impact in sign.

Frank and Goyal (2009) Tangibility (+) Size (+) Profitability (-) Market-to-book assets(-) Expected inflation (+) Industry median leverage (+) Dividend payments (+)

TOT accounts for many of the factors. POT seems intuitively pleasing, however theoretical upgrades are necessary to consider it reliable.

12 Črnigoj and Mramor

(2009) Tangibility (-) Earnings volatility (-) Profitability (-) Employee- governance (-) Size (+) Growth (+)

Shareholder wealth maximization assumption does not hold in Slovenia. Slovenian firms after transition are characterized by high ownership by employees, who try to hold out the “outsiders”.

Kayo and Kimura (2011) Profitability (-) Tangibility (+) Size (+) Growth opportunities (-) Munificence (- and -) Stock market development (-and-)

Firm characteristics account for 78% of leverage variance. Country and industry variables must not be ignored due to hidden indirect effects on firm characteristics. Some country factors show in developing economies opposite effect on leverage compared to developed countries.

Bris et al (2011) Euro (+)

Inflation (+ and -) Development of

financial market (+ and -)

Size (+)

Industry dependence on debt (+)

Introduction of euro significantly increases external financing of firms. Firms previously experiencing currency fluctuations and poor financial market conditions increase both debt and equity once euro is introduced. Large firms and ones depending on external finance increase their leverage due to the euro. Jõeveer(2013) Profitability (-)

Size (+) Tangibility (-) GDP Growth (-) Inflation (-)

Country credit rating (-)

Country-specific factors most important for small and unlisted firms, firm-specific factors more important for large and listed firms. Unmeasurable institutional factors more important than thought (half of country-specific variance)

Country-specific factors

13 country-level influence on capital structure is very low, only 3.3% of total variance, however few variables (stock market development, bond market development and GDP growth) show significant negative influence on leverage indirectly. Kayo and Kimura (2011) and De Jong et al (2008) construct “country factors x firm factor” interaction variables in order to measure whether country-specific factor increases the impact of firm-specific factor or makes it less important. In other words, country characteristics may not have a direct strong influence on leverage, but they can determine the importance of a firm-specific factor on capital structure. For example, the more developed is the stock market is, the more negatively growth opportunities of a firm will affect the leverage. Gungoraydinoglu and Öztekin (2011) state that country’s legal, political and financial institutions greatly affect cross-country capital structure choices through the influence of bankruptcy costs, agency costs and information asymmetry costs levied on firms. Thus, the choices of capital structure are made in accordance with firms own characteristics and also its external environment and traditions. Kayo and Kimura (2011) also add that despite major institutional and macroeconomic differences among countries, the variables that explain capital structures in the US, Europe and emerging countries, are the same.

Jõeveer (2013) argues that country characteristics are significant factors in CEE countries, especially in case of unlisted firms. Unmeasurable country institutional differences account up to 25% of leverage variation for unlisted firms. Jõeveer (2013) addresses the importance of a country’s policy maker’s influence on capital structure. For example, policy makers may want to lower leverage of local firms with high leverage because those may be prone to face financial distress. In such a case, policy makers will decrease banking market concentration and/or cut corporate tax rates, which will consequently affect the capital structure of firms (Jõeveer, 2013).

14 leverage increased (Nivorozhkin, 2005, p. 159). Firm’s age and profitability- only factors that have uniform effect on target capital across CEE.

Transition toward market economy has paid an important role in corporate governance of CEE firms. State owned enterprises were mainly privatized through free distribution of vouchers; it recognized the special role of employees, giving them priority in investing the vouchers and additional funds in the firm where they were employed and regarding them with a major stake in the transformed firm (Črnigoj and Mramor, 2009). As a result, ownership was often characterized by inside owners (employees) that prevented external shareholders from taking over. Such firms typically faced limited supply of debt resulting from information asymmetry.

Country profiles* Estonia

Estonia, a member of the European Union since 2004 that adopted the euro on 1 January 2011, has a modern market-based economy and one of the higher per capita income levels in Central Europe and the Baltic region. Estonia's successive governments have pursued a free market, business economic agenda and have wavered little in their commitment to pro-market reforms. Similar to Latvia, Estonia's fast growing economy fell into aching recession in mid-2008, experiencing high inflation and high negative GDP growth (see Table 3 for GDP growth rate and Table 4 for and inflation). Estonia escaped crisis more successfully than other Baltic states by rapidly restoring GDP growth, decreasing budgetary deficit under 3% of GDP and public debt less than 10% of GDP. Estonia is one of the least corrupt of the new EU entrants. The legal environment has significantly improved and included measures to strengthen corporate governance and promote competition.

*

15

Table 3 Real GDP growth rate – volume

Percentage change on previous year

geo\time 2004 2005 2006 2007 2008 2009 2010 2011 2012 European Union 2,6 2,2 3,4 3,2 0,4 -4,5 2 1,7 -0,4 Bulgaria 6,7 6,4 6,5 6,4 6,2 -5,5 0,4 1,8 0,6 Czech Republic 4,7 6,8 7 5,7 3,1 -4,5 2,5 1,8 -1 Estonia 6,3 8,9 10,1 7,5 -4,2 -14,1 2,6 9,6 3,9 Latvia 8,8 10,1 11 10 -2,8 -17,7 -1,3 5,3 5,2 Lithuania 7,4 7,8 7,8 9,8 2,9 -14,8 1,6 6 3,7 Hungary 4,8 4 3,9 0,1 0,9 -6,8 1,1 1,6 -1,7 Poland 5,3 3,6 6,2 6,8 5,1 1,6 3,9 4,5 1,9 Romania 8,5 4,2 7,9 6,3 7,3 -6,6 -1,1 2,3 0,6 Slovenia 4,4 4 5,8 7 3,4 -7,9 1,3 0,7 -2,5 Slovakia 5,1 6,7 8,3 10,5 5,8 -4,9 4,4 3 1,8 Source: Eurostat

Table 4HICP - inflation rate Annual average rate of change (%)

geo\time 2004 2005 2006 2007 2008 2009 2010 2011 2012 European Union 2 2,2 2,2 2,3 3,7 1 2,1 3,1 2,6 Bulgaria 6,1 6 7,4 7,6 12 2,5 3 3,4 2,4 Czech Republic 2,6 1,6 2,1 3 6,3 0,6 1,2 2,1 3,5 Estonia 3 4,1 4,4 6,7 10,6 0,2 2,7 5,1 4,2 Latvia 6,2 6,9 6,6 10,1 15,3 3,3 -1,2 4,2 2,3 Lithuania 1,2 2,7 3,8 5,8 11,1 4,2 1,2 4,1 3,2 Hungary 6,8 3,5 4 7,9 6 4 4,7 3,9 5,7 Poland 3,6 2,2 1,3 2,6 4,2 4 2,7 3,9 3,7 Romania 11,9 9,1 6,6 4,9 7,9 5,6 6,1 5,8 3,4 Slovenia 3,7 2,5 2,5 3,8 5,5 0,9 2,1 2,1 2,8 Slovakia 7,5 2,8 4,3 1,9 3,9 0,9 0,7 4,1 3,7 Source: Eurostat Latvia

16 in 2004 and EMU as of 1 January 2014. Latvia’s transition from lat (LVL) to euro has been noted as fast and well arbitrated.

Lithuania

Foreign investment and business support have helped in the transition from the old command economy to a market economy. Same as the neighboring Baltic republics Lithuania was among the hardest hit by the 2008-2009 financial crisis. The government's struggles to invite foreign investment in order to develop export markets, and to pursue broad economic reforms has been of crucial importance to Lithuanias rapid recovery from a deep recession, making Lithuania one of the fastest growing economies in the EU. Lithuania is committed to meeting the Maastricht criteria to join the euro zone, and to be the next Eurozone member as of 2015.

Poland

Poland has pursued a policy of economic liberalization since 1990 and Poland's economy was the only one in the EU to avoid a recession through the 2008-2009 economic downturn. Although EU membership and access to EU structural funds have provided a major boost to the economy since 2004, GDP per capita remains significantly below the EU average. The Polish economy went through an economic downturn by skillfully managing public finances and adopting controversial pension and tax reforms to further shore up public finances. While the Polish economy has performed well over the past five years, growth slowed in 2012 and 2013, in part due to the ongoing economic difficulties in the euro zone. In the short-term, the key policy challenge will be to consolidate debt and spending without stifling economic growth. Poland is still characterized by a poor business environment, a commercial court system, much government red tape, and a burdensome tax system. There is no date set for Poland to adopt the euro.

Slovakia

state-17 owned banking, energy, and telecommunications sectors; and amending the constitution to reform the judiciary system.

Slovenia

Slovenia became the first of CEE countries to adopt the euro (on 1 January 2007) and has experienced one of the most stable political transitions in CEE. In March 2004, Slovenia became the first transition country to progress from borrower status to donor status at the World Bank. Since 2012, long-delayed privatizations, particularly within Slovenia’s largely state-owned and increasingly indebted banking sector, have fueled investor concerns that the country would need ECB and/or IMF financial assistance. The government has embarked on a program of state asset sales intended to bolster investor confidence in the economy, which is poised to contract 1% in 2014, its third-year of recession.

Czech Republic

Since EU accession in 2004, the Czech Republic is a stable and prosperous market economy closely integrated with the EU. While the conservative, inward-looking Czech financial system has remained relatively healthy, the small, open, export-driven Czech economy remains sensitive to changes in the economic performance of its main export markets, mainly Germany. When Western Europe fell into recession in late 2008, demand for Czech goods plunged, resulting in heavy descents in industrial production and exports. As a result, real GDP fell harshly in 2009. The economy slowly recovered in the second half of 2009 and showed weak growth in the next two years. In 2012, however, the economy fell into a recession again, due both to a slump in external demand and to the government’s austerity measures. Foreign and domestic businesses are under pressure of corruption, especially in public procurement.

Hungary

18 corporate and personal taxes. Hungary has not yet set a date for adoption of euro due to its excessively high budget deficit. Only in end of 2013, the deficit has reached acceptable norm of below 3% of GDP.

Bulgaria

Being one of the newest members of EU (as of 1 January 2007), Bulgaria is also one of the economically weakest. The transition period of late 1990-ies has been hard, privatization of state owned enterprises was done in two rounds, followed by low foreign investor interest and high inflation rates. After the crisis of 2008, the growth of GDP shrank significantly, the budget deficit is currently too high to set a date when Bulgaria could be joining Eurozone. Many of the current Bulgarian firms are characterized by increasingly high ownership concentration. Many of the majority shareholders are holding companies which were encouraged during the privatization process and that are currently in great power. Moreover Bulgaria is characterized by weak securities law, confusing accounting standards and poor corporate governance (Petranov and Miller, 2000).

Romania

The developments in Romania are quite similar to those of her southern neighbor. The privatization of state owned enterprises took off slow at the end of 1990-ies; the second wave of privatization is still in progress. The majority of banks were still owned by the state up until end of 20th century. Domestic debt grew as large amounts of finance were devoted to restructuring. Due to high tax rates, political instability and corruption foreign direct investments in Romania remained low compared to rest of CEE. Romania joined EU in 1 January 2007, however it is yet to conform to the requirements for entering Eurozone and no specific date for this is set at this moment.

Leverage measures

19 was also an important criterion for several reasons. First post-Soviet countries are the ones that experience the most rapid growth and institutional changes, thus research that uses the data of the 1990-ies does not offer findings that could be currently applicable. Second, the enlargement of EMU and continuous financial integration of the EU now affects many more countries (including the CEE).

The existing literature offers various measurements of leverage, depending on the aim of the research. Almost all studies reviewed for this paper assume the relationship between leverage and its determinants to be static. This approach is questioned by Nivorozhkin (2005), who adopts dynamic modeling framework to study the relationship between capital structure and the factors influencing it. He identifies that speed and cost of adjustment to the target leverage differs depending on size of the firm and size of adjustment. Nivorozhkin (2005) also observes difference higher leverage level in core EU countries than in transition economies of CEE. Nevertheless when transition economies achieve macroeconomic stability and undergo broad institutional reforms, the debt levels of firms tend to rise. Delcoure (2007) finds that CEE country firms rely more heavily on short-term than long-term debt. The differences in leverage between core EU economy and transition economy firms strive from differences and financial constraints of banking systems, disparity in legal systems governing firms' operations, shareholders and bondholders rights protection, sophistication of equity and bond markets, and corporate governance structure, Delcoure (p.415) concludes. For example, due to weaker shareholder right protection managers in transition economies see equity as “free” money and prioritize it over debt. Jõeveer (2013) finds significant differences in leverage across transition countries which cannot be explained by neither industry-specific nor firm-specific differences. She suggests importance of unmeasurable institutional variables in affecting leverage in transition countries. Additionally, leverage seems to be affected by transition period and instability.

20 two leverage measures: market leverage, which is calculated as the ratio of long-term debt to total firm value (where total firm value is the sum of debt and market value of firm equity), and book leverage- ratio of long-term debt to total firm book value and total firm book value is the sum of debt and book value of firm equity. De Jong et al (2008) argue against use of short term debt as it consists of trade credit which is affected by completely different determinants.

Bris et al (2011) use two measures: Net Debt Issues (NDI)-defined as the net change in the book value of total liabilities over a given year divided by the lagged book value of assets, Net Equity Issues (NEI) is defined as the net change in external equity divided by the lagged book value of assets. Alternatives are proposed by Frank and Goyal (2009, p.12): 1) the ratio of total debt to market value of assets (TDM), 2) the ratio of total debt to book value of assets (TDA), 3) the ratio of long term debt to market value of assets (LDM), and 4) the ratio of long-term debt to book value of assets (LDA).

Measures of leverage determinants

The literature on capital structure tends to use diverse sets of independent variables. Authors are rarely in consensus on which factors are the most important in explaining firms leverage. Moreover, same as in case of the leverage measures, scholars often employ different proxies to measure the same country- of firm-specific variable. The following part of the paper will present overview of the most common variables used in related literature.

Size

De Jong et al. (2008), Delcoure (2007), Kayo and Kimura (2011) uses the natural logarithm of total sales. Jõeveer (2013) uses total assets as measure of size and groups firms in five classes (class 1: firms with total assets <$1mln, class 5: $50+mln). Nivorozhkin (2005) uses the logarithm of total assets. Logarithm of sales is proposed by Titman and Wessels (1988).

Profitability

There seems to be general consensus on the profitability measure. Noivorozhkin (2005), Kayo and Kimura (2011), De Jong et al. (2008), Jõeveer (2013) use operating income over book value of total assets as a proxy for company profitability, however whether EBIT or EBITDA is used is not always clear. Delcoure (2007) uses return on assets (ROA).

Growth

21 2) change in log assets and 3) capital expenditure to total assets. Titman and Wessels (1988) additionally to previous also use ratio of R&D costs to sales.

Tangibility

The main rationale of this factor is that tangible assets reduce agency cost of debt due to ease of use as collateral. Tangibility measures are quite similar across reviewed studies. Delcoure (2007) uses net plant, property and inventory over total assets. De Joung (2008) and Kayo and Kimura (2011) employ the measure net fixed assets to total assets. Titman and Wessels characterize tangibility as “uniqueness” of assets and recommend measuring it as selling expenses over sales.

Liquidity

De Jong et al. (2008), Gungoraydinoglu and Öztekin (2011) use total current assets over total current liabilities.

GDP growth

GDP growth is an important factor especially in emerging economies. Almost all reviewed papers have found it as a significant variable. Eurostat data can be used to measure GDP growth.

Inflation

Although not many of capital structure scholar assume inflation as significant factor, Bris et al (2011) suggests that firm experiencing hard inflation rates before euro will benefit more once it is introduced in their country. This research will use the Harmonized Index of Consumer Prices provided by World Bank database.

Methodology

22 data often show “NA” value. To increase the number of year observations, also other sources were used such as annual reports of individual firms and websites providing missing data. Unfortunately, in many cases the secondary data sources provided inconsistent data with Datastream. Unavailability of observations consequently restricted choice of leverage measures. Many of the previously discussed leverage measures were impossible to calculate due to unavailable data for some companies. For example Datastream items such as market value of assets and debt when split into long-/short-term (WC03251/WC03051) had limited observations in many firms across countries. According to Rajan and Zingales (1995) accounting differences across countries and leverage ratios need to be accounted for. Financial firms were excluded from the sample.

The time span of this research covers the years 2004-2012. The time span of the research is affected by several factors. First, the accessibility of data becomes smaller prior to 2004 (more “NA” values). Second, most of prior research focuses on the time span before 2004, when transition economies were turbulent and had did not have current growth. Third, 2004 was the year when all subject countries became members of the EU (except Bulgaria and Romania that joined in 2007).

The sample consists only of firms that have data available from 2008 or earlier. The final sample consists of 656 firms. The sample is unbalanced and representativeness of firms across countries significantly differs- most firms are from Poland (241) and least from Czech Republic and Slovakia (14 firms each). Total amount of listed firms per country per year are specified in Table 5.

Table 5 Total number of listed firms per country. Data from World Band Database.

23 The dependent variable in this study is leverage. Due to inconsistency and limitations of data availability, measures that employ market values will not be used. Three leverage measures are used for this study. First two measures were originally recommended by Titman and Wessels (1988) and are most commonly used by scholars, namely TOTAL leverage defined as total debt to total assets and LONG term leverage- long term debt to total assets. Third measure recommended by Titman and Wessels (1988) is short-term leverage, however as suggested by De Jong et al (2008) this will be not used since short-term debt is mainly payments to suppliers of goods. Third measure of leverage (LEV) is long term debt over total capital (Datastream item WC0821). This measurement was chosen because the aim of this paper is to assess the part of debt in the capital of the firm rather than debt to assets as commonly measured in related studies. Moreover, LEV has not been employed by any previous research.

We chose natural logarithm of total sales as a measure of firm’s size (Datastream item E037). In few cases negative, value of sales resulted in errors in Datastream and these were deleted. Profitability (PROFIT) of firms is measured as operating income over book value of assets (WC01250 divided by WC02999). Both profitability and size measures are most commonly used in relevant literature. Tangibility (TANG) is measured as net fixed assets divided by total assets (WC02501/ WC02999). Liquidity is measured as total current assets divided by total current liabilities (WC02201/WC03101). Growth is measured as one year’s sales growth to revenues growth (WC08631).

The country specific factors were obtained from Datastream, Eurostat and World Bank databases. Inflation rate and GDP growth rates are extracted from Eurostat. When regression on all firms together is performed, average of inflation and GDP are weighted against the number of firms from each country.

For individual countries, the influence of risk is measured as overall country risk index (from Datastream). For tax rate total tax rate (% of commercial profit) is used (from World Bank). To measure the influence of establishment of euro in Slovenia and Slovakia, a dummy variable is introduced. This dummy takes value 0 prior to 1st January 2007 and value 1 during 2007 and after. To check for the post-crisis effect on leverage, a dummy for the year 2008 was used for individual countries and total dataset.

24 Although time span of this research does not capture current ECB low interest rate policy, the findings will suggest whether borrowing rates affect the capital structure of firms.

This study does not assess institutional effects on leverage, because the data available from Datastream, World Bank or other web-sites is questionable. The measures provided by the World Bank rates countries on scale a 1 to 10, often not changing the rating throughout all 10 years (i.e. Strenght of legal rights index). Index used by De Jong et al (2008), defined by La Porta et al. (1998) is not relevant for the time span addressed in this research, because CEE economies have undergone significant changes and progress since the early 1990-ies. This suggests that an update on institutional development measures is needed in order to assess how leverage is affected by institutional factors.

The estimation method employed for this study is OLS, performed on the statistical program Eviews 8 developed by Quantitative Micro Software. Panel data is suitable for such regressions because it provides large amount of data points, offering additional degrees of freedom. Two models are constructed: a model for individual countries and a pooled model. These models first assess the effect of all variables (firm and country specific and year dummies) on individual countries. Factors that do not show any significant results for individual countries will be removed and not addressed in pooled data sets.

The model can be written as follows:

L

=α+βF

+ βC

+D

+ D

+ ε

,

α is the intercept, L is one of three leverage measures for firm i at year t, F is one of the firm specific variables for firm i at year t, C is one of the country specific variables for country k at year t, D are year dummies, εit is the disturbance term. In case of pooled model, weighted

average for inflation and GDP is calculated.

There are no correlations among the variables that indicate causality (see Table 6)

Table 6 Correlation Matrix of leverage measures and firm-specific variables

SIZE PROFIT LONG LIQUID GROW TOTAL TANG LEV

25

Findings and discussion

The leverage measures employed here did not bring consistent results. LEV was omitted from the rest of the study due to its inconsistency across countries and firms: several extremes (outliers) and negative values, which is impossible for leverage measure. For some countries (Poland, the Baltics) it offered acceptable results, however in other cases the data were incompatible with the rest of dataset. This could be explained by different accounting standards across CEE countries and the fact that some countries had to undergo a stock market reform somewhere during 2004-2012.

Albeit that LONG had sufficient number of observations and did not have inconsistencies across countries and firms, the explanatory power of model that employed this leverage measure was much lower that of TOTAL. As a result, TOTAL was chosen as a main dependent variable for this study. See Table 7 for descriptive statistics of leverage measures and Table 8 for firm-specific determinants.

Table 7 Descriptive statistics of leverage measures

TOTAL LONG LEV

26

Table 8 Descriptive statistics of firm-specific determinants

TANG SIZE PROFIT LIQUID GROW

Mean 401.1333 11318.82 30.57214 4620.790 169.2219 Median 392.0000 11407.00 34.00000 1442.000 6.455000 Maximum 998.0000 22437.00 4993.000 968143.0 503380.0 Minimum 0.000000 0.000000 -39562.00 4.000000 -100.0000 Std. Dev. 255.5754 2877.474 610.8908 26877.91 7433.717 Skewness 0.239968 0.071683 -52.61123 20.61527 66.78234 Kurtosis 2.207799 4.450322 3405.145 566.7103 4517.856 Jarque-Bera 186.6337 425.7709 2.51E+09 68472903 3.95E+09 Probability 0.000000 0.000000 0.000000 0.000000 0.000000 Sum 2094317. 54454865 158700.0 23769342 786881.8 Sum Sq. Dev. 3.41E+08 3.98E+10 1.94E+09 3.72E+12 2.57E+11 Observations 5221 4811 5191 5144 4650

When performing pooled regressions with only county specific variables, all but one (GROW) showed a significant probability. Growth in our data set had the highest amount of outliers, which introduced a problem: how high growth can be considered an outlier? Deleting too many data-points would corrupt the data. Moreover, in growing economies of CEE, high growth in firms should be expected. Pooled regression indicated an increase in R-squared of less than 1%, and none of the variables p-values were significant under any acceptable threshold. In many cases R-squared suffered a decline when some country factors were added to regression. Apart from a few individual country cases where GDP growth was significant beyond a 10% threshold, other country specific (Slovenia) variables do not seem to have any significant effect on any measures of leverage used for this study. Only in a few cases GDP growth was found significant to total leverage at 10% level. Surprisingly TAX rate, ECB lending rate and country risk rate did not have any significant effect on leverage whatsoever. Tax rate could be insignificant for two reasons. First, the tax rates in CEE countries are lower compared to Western Europe and the US, thus the tax shield is smaller. Second, only corporate tax rate was considered for this research, disregarding personal taxes. Rajan and Zingales (1995) argue that explanatory power of taxes increases if personal taxes of investors are also considered. When personal taxes are introduced, the firm’s objective is no longer minimization of corporate tax paid, but rather minimizing the taxes paid on all corporate income, thus also taxes for bondholders and shareholders (Brealey et al. 2011)

27 second, the ECB low interest rate policy has not yet taken place. More recent data might show different results.

Irrelevance of country factors found in this study, contradicts many of the previous studies, however they are strongly supported by Kayo and Kimura (2011). Kayo and Kimura find relatively low importance of country level determinants of leverage, however they urge not to disregard country effects, as the indirect impact on firm and industry characteristics still remains important. Indirect impact is also stressed as important by De Jong et al. (2008). It is worth mentioning that Kayo and Kimura’s paper delivers the most recent findings, covering data up until 2007. The rest of the relevant studies are performed on data on 2002 or earlier. This difference is significant in context of EU and CEE countries especially. Due to expansion of EU and integration and homogenization of its financial markets the relevance of national characteristics seems to be fading.

Table 9 Regression model for individual countries(continued in page 28)

Poland Latvia Lithuania Slovakia Slovenia

Coef. p. Coef. p. Coef. p. Coef. p. Coef. p.

28 *,**,*** donate statistical significance at 1%, 5% and 10%, respectively.

Hungary Romania Czech Republic Estonia Bulgaria

Coef. p. Coef. p. Coef. p. Coef. p. Coef. p.

29

Table 10 Pooled regression model

ALL EURO NO-EURO

TANG 0.068 0.000* -0.026 0.660 0.117 0.000* SIZE 0.006 0.000* 0.020 0.002* -0.001 0.497 PROFIT -0.228 0.000* -0.131 0.182 -0.254 0.000* LIQUID -0.003 0.000* -0.065 0.000* -0.015 0.000 GROW 0.018 0.226 0.073 0.881 0.000 0.521 C 97.627 0.000 65.344 0.474 191.272 0.000 @YEAR>2007 30.541 0.000* 102.075 0.000* 14.532 0.078*** @YEAR=2008 -5.644 0.622 -28.327 0.379 2.766 0.817 R-squared _{0.080 0.304 0.137} Adjusted R-squared _{0.078 0.283 0.135} S.E. of regression _{212.659 160.527 176.997} Sum squared

resid _{1.45E+08 6107255 6.75E+07} Log likelihood

-21714.72 -1587.797 -14254.54

F-statistic _{39.9372 14.76428 49.03594}

prob F _{0 0 0}

The effects on TOTAL leverage for individual countries and pooled regressions are summarized in Table 9 and Table 10 respectively. The explanatory power of the model varies greatly across countries, ranging from 0.10 in Bulgaria to 0.64 in Slovenia. Interpretation of these differences can be twofold. First, this can suggest that the chosen independent variables are not equally important for leverage across countries. Second, the explanatory power correlates with the quality of the data extracted from Datastream, meaning that cross-country studies are highly dependent on differences in data quality provided by each country.

30 is often too dependent on them and, in worst case scenario, governments can provide them a bail-out (Črnigoj and Mramor, 2009).

Tangibility demonstrates significant positive effect on leverage for the total dataset, as suggested by TOT and all but one of the reviewed studies. However, in Lithuania and Bulgaria the coefficient has negative sign, confirming POT. With De Jong (2008) the findings strongly support positive effect of tangibility, however Jõeveer (2013) and Črnigoj and Mramor (2009) show significant negative effects of tangibility on leverage. This shows an obvious conflict between TOT and POT. On one hand, tangible assets offer lower costs of distress and fewer debt related agency problems. On the other hand, tangible assets have lower information asymmetry, which makes equity issuance less costly. Perhaps a better explanation is suggested by Nivorozhkin (2005), who claims that although tangible assets remain a poor source of collateral in less advanced transition economies, the effect of tangibility on target leverage is moving towards the positive relationship observed in the West.

Profitability has negative significant effect on leverage for the pooled data set, as suggested by POT and the literature. Due to asymmetric information, firms will more likely use retained earnings in financing their operations and only next seek external finance. Another explanation comes from the dynamic trade-off model, which argues that profitability can be negatively related to leverage because firms tend to passively accumulate profits (Frank and Goyal, 2009), especially in case of newly privatized firms form transition economies (Nivorozhkin, 2005). Only in Romania, the effect is significantly positive. This can be explained by the fact that the Romanian bond market is still one of the least developed in EU and shareholder rights are weakly protected, therefore managers prefer equity over retained earnings and debt.

Liquidity has a significant negative effect on leverage across all countries in our dataset. Liquidity means that firms can easily trade their assets or securities in the market without affecting the assets price. It is likely that such firms would generate necessary cash from the marketability of their assets and not turn to external financing, overall supporting POT.

31 with high growth prospects will not be willing to transfer future profits from shareholders to creditors.

Lending rate in almost all cases had negative coefficient implying that decrease in interest increases leverage of firms, however values were hardly below 0.15. Albeit insignificant p-values, lending rate was the only non-firm specific factor that added explanatory power to model. From this we can conclude that although cost of debt is to be accounted for, it is not a factor that dictates whether firm will take on more debt or not.

When adding dummies for euro, explanatory power of model increases, especially in the pooled Slovenia and Slovakia model, suggesting that both events affect leverage. Nevertheless, the euro dummy is statistically significant only for Slovenia and Bulgaria. Obviously in case of Slovenia, this positive impact on leverage can be explained by introduction of euro. Slovakian sample does not show significant influence, but this could be explained by its very limited size. In case of Bulgaria, results could be influenced by its engagement into the EU, which was at the same date as the enlargement of Eurozone. A pooled model shows significant positive effects on the all three pools, although for non-euro countries it is significant only at 10% threshold. One explanation could be that CEE firms are gradually becoming more debt tolerant and are approaching the leverage levels of firms in the West (Nivorozhkin, 2005). In case of Slovenia and Slovakia, an interaction dummy of profit and euro showed positive significant (at 5%) effect. This could mean that heavy decrease of profits during the crisis forced firms to borrow.

If running a horse race between POT and TOT was the aim of this paper, the outcome would be that in the end POT was slightly ahead. However, more practical conclusion would be that both theories have flaws and are in need of adjustments. Perhaps in CEE countries there are too many forces that are affecting capital decisions and no single theory can be applied.

Conclusions

32 Factors affecting leverage become more/less important and often change the direction of their influence in time and across countries. Many of the differences in findings across studies can be accounted to inconsistent measures or proxies for the variables they are trying to observe. Moreover, the financial strategy of firms can be too complex to fit in one theory. Importance of information and agency costs are clearly major determinants of the financial behavior of a firm.

All in all, it can be said that countries in the current dataset can no longer be studied as one the Soviet Block. Differences across CEE states are more and more obvious once the market economy has “kicked in”. Moreover, previously assumed gaps between transition economy firms and the Western companies are vanishing. Trade unions open various doors, globalization forces push firms to adapt to environment and not be dependent on external factors as much as was a decade ago. However, one must not overlook the ability of the environment and the traditions to influence the firm characteristics that in turn determine the financing decisions.

33 This study has only given initiative suggestions about changes in leverage and its determinants. A next step would be to discover the underlying drivers of these changes. Interactions between various firm- and country-specific variables, could explain how macroeconomic processes affect leverage through various proxies. One could go deeper into firm environment and the behavior of managers to understand financing tactics of firms and bring to light the reasons why orthodox capital structure theories are no longer reliable without mutations and exemptions. Perhaps the search for one universally applicable theory is not the way to go, but rather understanding the context where the theory is to be applied.

REFERENCES

 Albescu, C., 2011. Economic and financial integration of CEECs: the impact of financial instability. Czech Economic Review 1, 27-45.

 Alesina A., Barro R.J., 2002. Currency unions. Working paper No. 7924, National Bureau of Economic Research.

 Bertrand, M., Schoar A., 2003. Managing with style: the effect of managers on firm policies,” Quarterly. Journal of Economics, 1169-1208.

 Brealey, R.A., Myers, S.C., Franklin, A., 2011. Principles of Corporate Finance, 10th

edition., The McGraw-Hill/Irwin, New York.

 Bris, A., Yrjo, K., Nilsson M., 2011. The euro and corporate financing. Bank of Finland Discussion paper.

 Bris, A., Koskinen, Y., Nilsson, N.,2009. The euro and corporate valuations. Review of Financial Studies 22, 3171-3209.

 CIA Factbook. Available at: https://www.cia.gov/library/publications/the-world-factbook.

 Črnigoj, M., Mramor, D., 2009. Determinants of capital structure in emerging European economies: evidence from Slovenian firms. Emerging Markets Finance and Trade 45, 72-89

 Decoure, N., 2007. The determinants of capital structure in transition economies. International Review of Economics and Finance 16, 400-415.

34

 European Commission web-site, Why the euro? Available at:

http://ec.europa.eu/economy_finance/euro/why/single_market/index_en.htm.

 Fama, E.F., French, K.R., 2002. Testing trade-off and pecking order predictions about dividends and debt. Review of Financial Studies 15, 1 -33.

 Frank, M.Z., Goyal, V.K., 2009. Capital structure decisions: Which factors are reliably important? Financial Management 38, 1-37.

 Gungoraydinoglu, A., Öztekin, Ö., 2011. Firm- and country- level determinants of corporate leverage: some new international evidence. Journal of Corporate Finance 17, 1457-1474.

 Hardouvelis, G.A., Malliaropulos, D., Priestley, R., 2006. EMU and European stock market integration. Journal of Business 79, 365–392.

 Japelli, T., Pagano, M., 2008. Financial market integration under EMU. CEPR discussion paper No. 791. Centre for Economic Policy Research.

 Jõeveer, K., 2013. Firm and coountry macroeconomic determinants of capital structure. Journal of Comparative Economics 41, 294-308.

 Kayo, E.K., Kimura, H., 2011. Hierarchical determinants of capital structure. Journal of Banking and Finance 35, 358-371.

 Miller, M.H., Modigliani, F., 1958. The cost of capital corporation finance and the theory of investment. The American Economic Review 48, 261-297.

 Nivorozhkin, E., 2005. Financial choices of firms in EU accession countries. Emerging markets review 6, 138-169.

 Petranov, S., Miller, J., 2000. Bulgaria’s capital markets in the context of EU accession: A status report. Center for the Study of Democracy.

 Rajan, R., Zingales, L., 1995. What do we know about capital structure? Some evidence from international data. Journal of Finance 505, 1421-1460.