Global Economic Review Vol. 37, No. 1, 107Á123, March 2008

Performance and Capital Structure of Privatized Firms in Europe
´ ´ MARIA JOSE ARCAS & PATRICIA BACHILLER
Faculty of Economics and Business Administration, Department of Accounting and Finance, University of Zaragoza, Zaragoza, Spain

ABSTRACT The objective of this paper is to analyze whether there are differences in performance between private firms and recently privatized firms in the European Union, as well as to determine whether ownership (state-owned versus private) and regulation affect capital structure. Focusing on economic reasons that justify privatizations, we analyze whether there are differences between recently privatized state-owned enterprises (SOEs) and private firms in their profitability, leverage and efficiency during the period 1999Á2002. Also, we analyze the determinants of the capital structure of these firms. Contrary to previous studies, our results show that privatized firms are not less efficient than firms with private ownership. KEY WORDS: Privatization, efficiency, capital structure, regression, Wilcoxon test

Summary In recent decades, the privatization of state-owned enterprises (SOEs) has been one of the most common policies carried out in economies worldwide and the Organization for Economic Cooperation and Development (OECD) countries have also been involved in this phenomenon. Many studies have focused on the impact of privatization on the performance of the privatized firms, but results are not conclusive. Therefore, the objective of this paper is to analyze whether there are differences in performance between private firms and recently privatized firms in the European Union (EU), as well as to determine whether ownership (state-owned versus private) and regulation affect capital structure. In consequence, we carry out a comparative study of firms in the process of privatization and private firms in the EU countries during the period 1999Á2002, and we also analyze the factors that affect their capital structure. More specifically, we analyze the economic and financial situation of recently privatized firms in the EU countries compared to other private firms by carrying out statistical tests in order to investigate whether privatization means better business performance. The work consists of analysing whether there are differences between the profitability, leverage and efficiency in recently privatized SOEs and
Correspondence Address: Faculty of Economics and Business Administration, Department of Accounting and Finance, University of Zaragoza, Gran Vıa, 2, 50005 Zaragoza, Spain. Tel.: '34 976 761000 (ext. ´ 4724); Fax: '34 976 761769; Email: pbachiller@unizar.es 1226-508X Print/1744-3873 Online/08/010107Á17 # 2008 Institute of East and West Studies, Yonsei University, Seoul DOI: 10.1080/12265080801911980

108

M. J. Arcas & P. Bachiller

private firms. Once we have seen the differences between both groups, we analyze the determinants of the capital structure of these firms. To do this, we focus on the impact of the intervention of the Administration on the leverage of the firms in our sample, distinguishing between privatized and non-privatized firms, and also between regulated and competitive companies. From the firms privatized in recent years, we selected the companies whose privatization or final phase of privatization (if there were several phases) took place after 1995. From the 1449 privatized firms in Europe, we selected those that report consolidated balance sheets. Furthermore, we excluded financial firms and those for which data could not be obtained for 2002. Consequently, the sample is comprised of 133 privatized firms and 451 firms-year observations. Having selected the privatized firms, the control group was constructed with private firms. For this, we chose a private firm for every privatized company belonging to the same geographical zone and in the same industry. Our results show that privatized firms are more profitable, less leveraged and less labour intensive than private firms. Contrary to previous studies, we have found evidence that privatized firms are not less efficient than firms with private ownership. Nevertheless, results differ for each geographical zone. This difference between zones may be due to different economic and legal environments. With respect to the capital structure, privatization, regulation and profitability determine leverage. However, the results also differ between zones. In all zones, the privatized firms are significantly less leveraged and the regulated firms are significantly more leveraged. With respect to previous research, this paper adds new evidence about the privatization process and the capital structure of firms in the EU. Previous research refers to a specific European country. By contrast, this paper provides an analysis of the effects of the privatization process on firm performance for all the EU. Furthermore, this paper analyzes the differences between privatized and nonprivatized firms using a matched-pairs sample, which allows comparisons between companies of similar characteristics. Introduction In recent decades, the privatization of SOEs has been one of the most common policies carried out in economies worldwide and OECD countries have also been involved in this phenomenon (OECD, 2003). Many studies have focused on the impact of privatization on the performance of the privatized firms (Megginson et al., 1994; Dewenter & Malatesta, 2001). Reasons for privatization can be categorized as economic, financial and political (Vickers & Yarrow, 1988). Economic aims are associated with improvements in efficiency of the privatized firm. Financial reasons are based on the positive effects on public finances, specifically the public deficit, because the State obtains an important source of income from the sale of these enterprises. Finally, political reasons refer to distributional effects on society because privatization allows increasing the number of stockholders and facilitates the access of citizens to capital markets. Our interest focuses on the study of the economic reasons that justify privatizations. The economic reasons are based on micro-economic theory. This asserts that

Performance and Capital Structure of Privatized Firms in Europe

109

with perfect competition, private firms achieve efficiency by maximizing profits. In principle, SOEs can reach the same levels of efficiency as private firms. Nevertheless, there are some problems that make it difficult to reach this objective, such as principalÁagent relationships, the non-definition of objectives and the contradiction between the objectives of the principal and the agent. Megginson et al. (1994) is the first study that analyzes the improvement of the efficiency of privatized firms from 18 countries. The objective of this paper is to analyze whether there are differences in performance between private firms and recently privatized firms in the EU, as well as to determine whether ownership (state-owned versus private) and regulation affect capital structure. In consequence, we carry out a comparative study of firms in the process of privatization and private firms in the EU countries during the period 1999Á2002, and we also analyze the factors that affect their capital structure. More specifically, we analyze the economic and financial situation of recently privatized firms in the EU countries compared to other private firms in order to investigate whether privatization means better business performance. The empirical work consists of analysing whether there are differences between the profitability, leverage and efficiency in recently privatized SOEs and private firms. Once we have seen the differences between both groups, we analyze the determinants of the capital structure of these firms. To do this, we focus on the impact of the intervention of the Administration on the leverage of the firms in our sample, distinguishing between privatized and nonprivatized firms, and also between regulated and competitive companies. With respect to previous research, this paper adds new evidence about the privatization process and the capital structure of firms in the EU. Previous research refers to a specific European country. By contrast, this paper provides an analysis of the effects of the privatization process on firm performance for all the EU. Furthermore, this paper analyzes the differences between privatized and nonprivatized firms using a matched-pairs sample, which allows comparisons between companies of similar characteristics. The paper is organized as follows. In the next section, we review the theory and previous literature. The third section details the research methodology. The fourth section describes the data employed. The fifth section presents the empirical results and, finally, the last section contains the conclusions. Literature and Theory Previous research shows the differences between the performance of private firms and SOEs. The agency theory (Jensen & Meckling, 1976) justifies these divergences as a result of the differences in the principalÁagent relationship. In SOEs, the principal’s objectives are related to the public interest, whereas in private firms, the principal’s objectives are related to maximizing the firm’s value. Also, the manager (agent) of SOEs has two principals, voters and government, whereas in private companies the principal is the shareholders. According to this theory, there are two fundamental problems: the adverse selection of the agent and the moral hazard because the agent’s behaviours can not be observed at all times (Cohen, 2001). Boycko et al. (1996) also highlight agent problems in the SOEs using a model that shows the gap between the

110

M. J. Arcas & P. Bachiller

objectives of politicians and managers. They also show the relative inefficiency of SOEs and the improvements of efficiency after privatization. Most empirical studies conclude that SOEs significantly improve their efficiency after privatization. In the international area, Megginson et al. (1994) compare the pre- and post-privatization financial and operating performance of companies from 18 countries during the period 1961Á1990, and find an increase of sales and capital investment spending and a decrease of debt levels after privatization. Dewenter and Malatesta (2001) carry out a cross-sectional and longitudinal analysis. The comparison between SOEs and private companies shows that SOEs are significantly less profitable than privately owned firms and use more leverage than private firms do. In the European countries, the results are less conclusive. In the UK, Boussofiane et al. (1997) conclude that the improvement in the efficiency of privatized firms does not take place after the change in ownership, but 2 or 3 years before the privatization. In Spain, Villalonga (2000) shows that the effect of privatizations between 1985 and 1995 has not always led to increases in efficiency and is influenced by political and organizational factors. In Germany, Knieps (2004) analyzes the role of liberalization and privatization in German network industries and concludes that, after deregulation, service quality increased and industry efficiency improved. In Italy, Goldstein (2003) finds a few significant improvements in efficiency but significant increases in investments. In France, Berne and Pogorel (2004) conclude that the privatization process has forced the government to make privatized firms profitable. In Sweden, Springdal and Mador (2004) assert that privatization increased competition in the information technology sector and this led to an improvement in service quality and a drop in price. Finally, in Eastern Europe, Hyclak and Kina (1994) demonstrate that the privatization experience in this zone is different to any other region, because of the lack of developed capital markets. As regards non-European countries, Tian (2001) finds that Chinese firms without state shareholding perform significantly better than SOEs. Also, Omran (2004) documents significant increases in the financial and operating performance of privatized Egyptian SOEs. The interventionism of government is also reflected in the financial structure of the firm. Titman and Wessels (1988) document the theoretical determinants of the capital structure. These factors are: competition, regulation, profitability, asset tangibility, size, growth opportunities and institutional environment. According to the pecking order theory, companies that do not operate in competitive industries achieve higher profit levels and lower debt levels. According to Kole and Mulherin (1997), the existence of competitive markets and control of the managers of privatized firms are essential requirements to improve the performance. Likewise, Hodges (1997) considers that privatization without increasing market competition produces ambiguous welfare results as public monopolies are replaced by private monopolies with a minimum of competition. About regulation, Parker (1998) asserts that an optimal regulatory system incentives for management to manage business assets efficiently and Newbery (1997) argues that regulation affects capital structure. The greater the fluctuations in a company’s cash flow, the greater the chance of being unable to meet its obligations. In consequence, firms with rate regulation have lower cash flow risk and can support higher debt levels.

Performance and Capital Structure of Privatized Firms in Europe

111

Profitability is positively correlated with leverage, because highly profitable companies tend to use more debt because of tax incentives. Another factor that may affect the firm’s debtÁequity choice is the tangibility of assets. According to bankruptcy costs theory, tangible assets keep their value even in bankruptcy, so firms with more tangible assets can support higher levels of debt at lower costs because of the ability of firms to collateralize their debt (Jensen et al., 1992). By contrast, Titman and Wessels (1988) assert that firms with fewer intangibles may choose higher debt levels to limit their managers consumption of perquisites because bondholders or bankers will closely monitor such firms. Furthermore, according to Rajan and Zingales (1995), size is positively correlated with leverage; nevertheless, very large firms pay less than small firms to issue equity, so they prefer to issue equity in the capital markets (Ryen et al., 1997). Leverage is also related to growth opportunities (Bevan & Danbolt, 2002). Finally, institutional and legal characteristics determine capital structure because the legal environment has a significant effect on the ability of firms to raise external finance (La Porta et al., 1997). Rajan and Zingales (1995) carry out a comparative study of the capital structure of the firms from the G7 countries and show that, at aggregate level, differences between the leverage of firms are not explained by institutional differences. Also at the international level, Wald (1999) points out that variables associated with risk, growth, firm size and inventories show different effects across countries. In Europe, Hall et al. (1999) carry out an analysis of small and medium-sized enterprises. The authors conclude that variations in the determinants of capital structure between countries are due to the financing requirements of firms, their relationship with banks, taxation and other national economic, social and cultural differences. In the UK, Ozkan (2001) concludes that profitability, liquidity and growth opportunities have a negative effect on the leverage ratios. In Spain, Arrondo and Gomez-Anso (2003) find that the agency model is determinant in the firm’s ´ ´n security issue choice. Likewise, De Miguel and Pindado (2001) compare the determining factors of capital structure in Spanish and US firms, finding that transaction costs in Spain are lower than those for US firms due to the higher percentage of private debt in Spain. Methodology Firstly, we compare the economic-financial characteristics of private firms and those that are privatized or in the process of privatization by carrying out the Wilcoxon two-sample paired signed rank test, which does not require normality. In accordance with the previous section, the factors to be compared are profitability, leverage and efficiency. As in Dewenter and Malatesta (2001), profitability is measured using the return on sales (ROS), return on assets (ROA) and return on equity (ROE). With respect to leverage, there are several measures in the literature. Following Bevan and Danbolt (2002), we have chosen the following ratios: Total Debt Total Assets

LEV1 0

112

M. J. Arcas & P. Bachiller LEV20 LEV30 Long Term Debt Total Assets Total Assets

Total Debt ( Trade Credit and Equivalent

These authors define two variants of measures using the book value or the market value; we only use book value measures because not all firms are quoted on the Stock Exchange. Finally, we measure efficiency as labour intensity. Following Dewenter and Malatesta (2001), we use two different measures of labour intensity: LAB10 LAB20 Employees Sales (in million of dollars) Employees Assets (in million of dollars)

In the second place, we carry out a multivariate regression to analyze leverage determinants. As we have mentioned, the uniqueness of privatized firms means, according to pecking order theory, less leverage. Therefore, we introduce the variable PRIV that indicates whether a company has been involved in a privatization process. It is a dummy variable that has value one in the affirmative case and zero otherwise. Another factor is regulation, so the REG variable shows whether the firm operates in a regulated industry; this is also a dummy variable with value one for the energy, tobacco, telecommunications and transport sectors. To study the effect of size on leverage, we use the SIZE variable, measured by the total assets in million of dollars.1 Following Titman and Wessels (1988), we include the tangibility (TAN) and the profitability measured as ROA in the model. TAN is defined as the proportion of tangible assets over total assets. Finally, to analyze variations in geographical zones, we introduce the dummies variables ZFRAN, ZGERM, ZESC and ZBRIT in the regression. These indicate whether the company belongs to the French, German, Scandinavian or British zone respectively; the value is equal to one in the affirmative case and zero otherwise. Growth opportunities have not been included because not all the firms have market value. Accordingly, we estimate the following model: Leveragei 0a'b1 PRIVi 'b2 REGi 'b3 SIZEi 'b4 TANi 'b5 ROAi 'b6 ZFRANi 'b7 ZGERMi 'b8 ZESCi 'b9 ZBRITi 'oi (1) where Leverage is the LEV3 ratio aforementioned. We have repeated this analysis with other leverage measures and the results obtained have been similar but with less statistical significance. Data The list of privatized firms in Europe has been obtained from the Internet website http://www.privatizationbarometer.net. The economic-financial data was derived from Amadeus Bureau van Dijk Electronic Publishing.

Performance and Capital Structure of Privatized Firms in Europe

113

From the firms privatized in recent years, we selected the companies whose privatization or final phase of privatization (if there were several phases) took place after 1995, because the 2001 OECD report establishes this year as the most significant in Europe, except for the UK. From the 1449 privatized firms in Europe, we selected those that report consolidated balance sheets. Furthermore, we excluded financial firms and those for which data could not be obtained for 2002. Consequently, the sample is comprised of 133 privatized firms and 451 firms-year observations. Having selected the privatized firms, the control group was constructed with private firms. For this, we chose a private firm for every privatized company belonging to the same geographical zone, following La Porta et al. (1997),2 and in the same industry. In this way, we control for differences in accounting rules and firms are paired up with other firms which operate in the same sector. Tables 1 and 2 present the distribution of the companies in the sample by industry and geographical zone. It can be seen that the number of French firms is greater than those in the other zones studied. In the British zone the number of pairs is lower because the privatization process began before 1995.3 Tables 3 and 4 provide summary statistics (mean, minimum, maximum and number of observations) of the variables used in the study and the size of the firms by zones. The mean size of the privatized firms (Table 3) is greater than that of the private firms. This is due to the fact that most SOEs have been in a monopolistic situation which caused their expansion. As Table 4 shows, British firms are the smallest in the sample because, as we have said, this zone began its privatizations before the other zones, which means that the most important companies were privatized previously and are not included in the sample. The profitability of privatized firms is greater than that of private firms, as ROS denotes. On average, privatized companies use less leverage than private firms do

Table 1. Sample distribution of privatized firms by industry (NACE code) Sector Construction Energy Holding Manufacture of food and tobacco products Manufacture of basic metals Transport Computer and related activities Manufacture of chemical products Manufacture of petroleum Telecommunications Wholesale trade Other sectors Total Number of companies % Companies 6 26 18 6 7 12 4 6 5 14 6 23 133 4.51% 19.55% 13.53% 4.51% 5.26% 9.02% 3.01% 4.51% 3.76% 10.53% 4.51% 17.29% 100%

114

M. J. Arcas & P. Bachiller
Table 2. Sample distribution of privatized firms by geographical zone

Zone French German Scandinavian British Eastern Total

Number of companies 58 13 9 13 40 133

% Companies 43.61% 9.77% 6.77% 9.77% 30.08% 100%

because the firms of the control group have made a greater investment effort to compete with firms already set up in the market. Results In this section, the results obtained in the study are presented. Firstly, we carry out a comparative analysis between the cited variables to study whether there are differences in the economicÁfinancial characteristics of privatized and private firms and, secondly we study the capital structure of the firms in the sample. Univariate Analysis We examine the differences in the profitability, leverage and labour intensity of the firms using matched pairs as was explained earlier.
Table 3. Summary statistics Mean Statistics of privatized firms Total assets (million dollars) ROS ROA ROE LEV1 LEV2 LEV3 LAB1 LAB2 Statistics of control pairs Total assets (million dollars) ROS ROA ROE LEV1 LEV2 LEV3 LAB1 LAB2 Minimum Maximum N

10,337.43 10.99 3.62 8.63 0.60 0.22 0.25 10,028.00 6,689.00 4,350.21 1.69 3.19 8.36 0.65 0.26 0.65 8,646.00 8,190.00

7.61 Á170.39 Á64.71 Á228.50 0.06 0.00 0.00 0.19 0.25 2.01 Á364.40 Á90.70 Á395.39 0.00 0.00 0.00 7.97 120.42

170,568.19 862.71 39.16 135.79 2.55 0.92 3.33 261,245 100,000 257,934.06 223.41 47.25 151.27 2.09 2.11 2.09 129,781 83,333

451 402 450 439 447 399 448 359 404 426 374 428 413 426 394 429 332 391

Performance and Capital Structure of Privatized Firms in Europe
Table 4. Size of privatized firms Zone French German Scandinavian British Eastern Total Mean total assets (million dollars) 10,351.18 27,661.62 17,074.24 2,810.80 5,692.07 12,717.98 N 198 41 34 41 137 451

115

Table 5 presents the KolmogorovÁSmirnov normality test for the data in both groups, as we have already said. We can see that the variables do not follow a normal distribution, which justifies the Wilcoxon test. The first line in Table 6 shows the differences between the profitability, leverage and labour intensity of the privatized firms and the control group. The return on sales and return on assets are significantly larger in firms recently privatized; although there are no significant differences in the return on sales. As for leverage, using both total leverage (LEV1) and long-term leverage (LEV2), we observe that the control group firms use more leverage than the privatized firms. The differences between both groups are not significant for ratio LEV3. Private firms have a significantly higher employment to assets ratio (LAB2) than privatized firms but there are no significant differences in the employees to sales ratio (LAB1). Because the privatization process has not been homogeneous throughout the countries analyzed, we divide them into several geographical zones. Analysing the results by zones, in the French zone we find that the profitability, measured by ROS, ROA or ROE, is significantly greater in firms in the process of privatization than in private firms. This result is repeated in the Scandinavian zone with the ROS and ROA ratios and in the Eastern zone where only the ROS ratio is significant. In the British zone, the results are the opposite, that is, private firms have a significantly greater ROA than privatized firms. The ROE ratio does not show significant differences between the two groups and the lack of observations does not allow the estimation of the ROS variable or any conclusions. In the German countries, profitability measured by the three ratios does not show significant differences between the two groups. As for leverage, in both the French and Scandinavian zones the three ratios show that
Table 5. KolmogorovÁSmirnov test Privatized firms Z KolmogorovÁSmirnov ROE ROA ROE LEV1 LEV2 LEV3 LAB1 LAB2 7.50 3.77 4.92 10.90 10.34 3.71 6.36 5.55 Sig. 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Non-Privatized firms Z KolmogorovÁSmirnov 6.14 3.84 5.14 8.07 4.50 2.27 10.38 10.76 Sig. 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

116

Table 6. Wilcoxon test results
Wilcoxon test Total sample Z p-value Positive ranks (Varc Varp) Negative ranks (Varc BVarp) Z p-value Positive ranks (Varc Varp) Negative ranks (Varc BVarp) Z p-value Positive ranks (Varc Varp) Negative ranks (Varc BVarp) Z p-value Positive ranks (Varc Varp) Negative ranks (Varc BVarp) Z p-value Positive ranks (Varc Varp) Negative ranks (Varc BVarp) Z p-value Positive ranks (Varc Varp) Negative ranks (Varc BVarp) ROS (6.048*** 0.000 114 209 (5.567*** 0.000 51 116 (1.366 0.172 13 20 (1.754* 0.079 6 14 * * * * (2.059** 0.039 44 59 ROA (2.851*** 0.004 156 223 (4.268*** 0.000 59 116 (0.173 0.862 21 16 (2.159** 0.031 6 17 (2.051** 0.040 23 16 (0.698 0.485 47 58 ROE (0.998 0.318 197 188 (2.371*** 0.018 68 98 (0.776 0.437 22 15 (1.194 0.232 9 11 (1.329 0.184 16 13 (0.379 0.704 52 51 LEV1 (2.410** 0.016 197 167 (5.193*** 0.000 106 56 (1.500 0.133 21 17 (2.743*** 0.006 15 7 (2.037** 0.042 12 26 (1.906* 0.057 43 61 LEV2 (2.549** 0.011 176 129 (1.928** 0.054 83 61 (0.294 0.769 22 15 (1.755* 0.079 15 6 (1.455 0.145 20 17 (0.099 0.921 36 30 LEV3 (1.567 0.117 194 169 (3.788*** 0.000 104 58 (2.458** 0.014 27 11 (2.386** 0.017 14 8 (2.646*** 0.008 11 27 (2.684*** 0.007 38 65 LAB1 (0.691 0.499 138 138 (1.600 0.109 76 79 (0.744 0.456 12 15 (1.045 0.296 10 10 * * * * (0.816 0.414 40 34 LAB2 (4.471*** 0.000 190 132 (5.288*** 0.000 103 55 (2.214** 0.027 18 13 (1.704* 0.088 13 9 (0.229 0.819 15 20 (0.077 0.938 41 35

M. J. Arcas & P. Bachiller

Zones classificationa French

German

Scandinavian

British

Eastern

Note: Varc, variable of the control firm; Varp, variable of the privatized firm. This table presents the Wilcoxon statistic for rank differences. a La Porta et al. (1997) classification. *Statistically significant at the 10% level. **Statistically significant at the 5% level. ***Statistically significant at the 1% level. *, Indicates insufficient observations to estimate.

Performance and Capital Structure of Privatized Firms in Europe

117

private firms tend to use more leverage than privatized firms do. Nevertheless, in the German zone, this result is only valid for the leverage calculated by subtracting the trade credit and equivalent (LEV3), so we cannot say anything about the other ratios. In the British and Eastern zones, LEV1 and LEV2 indicate that privatized companies are characterized by a higher leverage compared to private firms. With respect to labour intensity, the variable LAB1 (employees to sales) is not significant in any zone. Employees to assets (LAB2) indicates higher labour intensity in private companies in the French, German and Scandinavian zones. In the British and Eastern zones there are no differences between the two groups. From these results, we can see that there are important differences between zones. The French zone companies have similar characteristics to the Scandinavian zone firms: higher profitability for privatized firms and higher leverage and employment intensity for private firms. In the German zone, differences in profitability are not significant, but leverage and labour intensity have the same behaviour as in the French and Scandinavian zones. In the British and Eastern zones the characteristics of the firms are very different to the other zones. With regard to profitability, in the Eastern zone the evidence is weaker but similar to the other zones (the privatized firms are more profitable). In the British zone the opposite occurs, privatized companies generate less ROA than control firms. Leverage in the two zones is significantly less in private firms.4 Finally, there are no differences in labour intensity in the firms of these countries. These zonal differences in the results are due to the different processes and aims of privatization in the countries of the sample. Gonzalo et al. (2003) show that the objectives pursued by the governments in the privatization process vary between countries. Their results reveal that privatization in developed countries (French, German, British and Scandinavian zones) can be viewed as a way to reduce the public deficit or indebtedness, while in the Eastern zone the aim is to reduce the State’s weight in the economy. Also, the privatization process has been different in each country. In the Eastern countries and the British zone the privatized firms during the years of our study are smaller than in the rest of Europe. In the UK, the main companies were privatized before 1995. In the Eastern countries privatization has started recently. Moreover, in the Eastern countries, unlike other OECD countries, the government did not give subsidies to SOEs before privatization. In some countries, SOEs were restructured to be more profitable and subsequently to be sold to private companies. This is the case of France (Berne & Pogorel, 2004) and Spain (Verges, 2000). Finally, in the UK there ´ is evidence that efficiency improved in some companies 3 years before the privatization (Boussofiane et al., 1997). Our results differ from those of Dewenter and Malatesta (2001) where, at international level, it is observed that SOEs are less profitable than private firms. In the same way, Megginson et al. (1994) conclude that privatized firms improve their operating efficiency after being privatized. This difference may be due to the fact that in our sample the control group includes smaller firms than the privatized group (see Table 3). Most of the latter have operated in a natural monopoly of the network where, consequently, new firms must get into debt to overcome barriers to entry and so are less profitable than privatized firms.

118

M. J. Arcas & P. Bachiller

Multivariate Analysis To study the factors that determine the leverage of the firms in greater depth, we carry out a multivariate regression which includes the following determinants: privatized or non-privatized firm, regulated or non-regulated firm, total assets, tangibility and ROA,5 as can be seen in equation (1). To avoid the effect of outliers on the results, observations whose standardized absolute value is greater than three have been eliminated. As can be seen in the first line of Table 7, the results in the total sample show a negative and significant coefficient for PRIV, which means that this variable is inversely correlated with leverage, that is that privatized firms are less leveraged than private firms. However, the coefficient for REG is positive and significant, so regulated companies use more leverage than non-regulated firms. The coefficient of return on assets (ROA) confirms the inverse relationship between profitability and leverage; a more profitable company will have less leverage. This is supported in the pecking order theory, which suggests that firms prefer to use internal equity to external funds. Most papers conclude this, for instance, the studies of Ozkan (2001) and Hall et al. (2004), for British enterprises, and the study of De Miguel and Pindado (2001) for Spanish companies. Finally, in the total sample, size and tangibility are not correlated with leverage. The three dummy variables indicative of geographical zones are significant, which justifies carrying out the study by zones. When we carry out the study by zones, we observe that the French and Eastern zones have a similar behaviour. Privatization, regulation, size and profitability variables are significant and have the same sign in both zones. Similar to the regression for the total sample, there is a direct relation between leverage and regulation, and an inverse one with the privatization and profitability variables. Moreover, size is positive and significant; a larger firm will use more leverage than a smaller firm. These results are the same as those found by Wald (1999). However, Hall et al. (2004) corroborate a direct relationship between long-term leverage and size in Spanish companies, but not in the other countries from the French zone, where this variable is not significant in their study. In the German zone, previous results are repeated, except the fact that the tangibility variable is significant and positive, implying that a firm with more tangible assets will use more debt. These results agree with Rajan and Zingales (1995) in the G7 study, in which Germany shows a direct relationship between tangibility and leverage. However, Hall et al. (2004) assert that the sign of this variable for German companies depends on the long- or short-term debt. In our analysis, size is positively related to debt. This result coincides with Wald (1999) in the case of German companies. In the Scandinavian zone, all the variables apart from size are significant in the estimation and have the same meaning as in previous cases, except the TAN variable that confirms that a firm with more tangible assets will use less leverage. In British countries, only the variables PRIV and SIZE are statistically significant and size has the opposite sign to the other zones, that is, larger British firms have less leverage. This is because British capital markets are very developed and these firms will prefer to issue stocks in order to obtain financing. Rajan and Zingales (1995) conclude the same but Ozkan (2001) finds weak evidence that British companies’ size exerts a positive effect on their debt

Performance and Capital Structure of Privatized Firms in Europe

Table 7. Multivariate regression results
Regression PRIV REG SIZE TAN (0.019 ((0.693) 0.022 (0.674) ROA ZFRAN ZGERM ZESC ZBRIT F-value 90.343 147.018 31.530 16.212 4.817 R2 .517 .679 .681 .623 .246 R2 corrected .511 .674 .659 .585 .195

Total sample (0.663 (n0871) ((25.724)*** French (n0387) German (n084) Scandinavian (n064) British (n086) Eastern (n0250) (0.804 ((24.743)*** (0.779 ((10.937)*** (0.767 ((7.497)***

0.124 (0.013 (4.576)*** ((0.482) 0.093 (2.814)*** 0.210 (3.126)*** 0.293 (3.081)*** 0.116 (3.673)*** 0.346 (4.388)*** 0.063 (0.613) (0.243 ((1.920)**

(0.131 (0.230 (0.119 (0.142 (0.043 ((5.105)*** ((5.556)*** ((3.520)*** ((3.270)*** ((1.361) (0.136 ((4.181)***

0.204 (0.141 (2.842)*** ((2.048)** (0.238 (0.309 ((2.610)** ((3.330)*** 0.129 (0.920) 0.133 ((1.107) (0.136 ((2.455)**

(0.426 0.010 ((3.662)*** ((0.381) (0.640 ((11.330)*** 0.111 (1.855)*

0.141 (0.096 (2.459)** ((1.629)

28.314

.421

.406

Note: The table shows estimated coefficients of the regression. The t-statistic in brackets. *Statistically significant at the 10% level. **Statistically significant at the 5% level. ***Statistically significant at the 1% level.

119

120

M. J. Arcas & P. Bachiller

ratios. In this zone, the results are different to the other zones due to the specific characteristics of the analyzed companies, such as smaller size (Table 4). Results of the regression show that the PRIV variable is inversely related to leverage, that is, a privatized firm will be less leveraged than a private firm for every zone. The REG and ROA variables are also significant in all the countries’ regressions except the British and Eastern zones. Results show that regulated and less profitable companies are more leveraged than non-regulated and more profitable companies. Size has a direct relation to leverage in the French, German and Eastern zones; this coincides with the bankruptcy costs theory that asserts larger firms support more leverage. Finally, tangibility has a different behaviour according to the zone and leads to contradictory results. We have repeated the analysis with the total debt and long-term debt variables, with similar results. With total debt, PRIV and REG maintain the same results in the regressions as in our multivariate analysis. With long-term debt, only PRIV is significant and denotes an inverse relationship between leverage and privatized companies. Conclusions The objective of this paper is to analyze whether there are differences in performance between private firms and recently privatized firms in the EU, as well as to determine whether ownership (state-owned versus private) and regulation affect capital structure. To fulfil these objectives, we have investigated the differences in the profitability, leverage and labour intensity of privatized firms in the EU and we have also analyzed the factors that determine capital structure. For the total sample, we have found evidence that privatized firms are more profitable, less leveraged and less labour intensive than private firms. Contrary to previous studies, our results show that privatized firms are not less efficient than firms with private ownership and are consistent with Verges (2000). ´ Nevertheless, results differ for each geographical zone. In the French and Scandinavian zones, privatized firms are more profitable, less leveraged and have less labour intensity than private companies. In the German zone, there are only differences in leverage and labour intensity. In the British zone the results are quite different, as profitability is lower and leverage is higher for privatized firms than for private firms. Finally, in the Eastern zone, privatized companies have more profitability and more leverage. These different results between zones may be because of different reasons. First, due to different economic and legal environments. Although our study refers only to the EU, there are legal and economic differences between countries. For instance, Hyclak and Kina (1994) show that the privatization experience in the Eastern zone is different because of the lack of developed capital market. Another reason is that the privatization process has been different in each country. The privatization process has taken place at different times in each country. The wave of privatizations started earlier in the UK than in other EU countries, and in the Eastern zone it is more recent. Also, the goals of privatization vary between countries (Gonzalo et al., 2003). With respect to the capital structure, privatization, regulation and profitability determine leverage. However, the results also differ between zones. In all zones, the

Performance and Capital Structure of Privatized Firms in Europe

121

privatized firms are significantly less leveraged and the regulated firms are significantly more leveraged. Profitability is higher for less leveraged firms in all zones except for the British. Size is not significantly related to capital structure for the total sample; however, our results are different depending on the zone. In the French, German and Eastern zones, larger firms are more leveraged than smaller firms; however, in the British zone, the opposite happens. Finally, tangibility shows weak results. As we use accounting figures, one limitation of this empirical study may be the differences in accounting between the countries. To control for this, we use matchedpairs in our sample within companies of the same zone. Another limitation is that the privatization process is differently developed in the different countries. These limitations require the results to be interpreted with caution. With respect to previous research, this paper adds new evidence about the privatization process and firms’ capital structure in the EU. Furthermore, this paper provides an analysis of the effects of the privatization process on firm performance for all the EU by analysing the differences between privatized and non-privatized firms using a matched-pairs sample which allows for comparisons between companies of similar characteristics.

Acknowledgements
This study has been carried out with the financial support of the Spanish National R'D'I plan through the research project SEJ2007-62215-ECON-FEDER.

Notes
1

We express in millions so that the coefficients of regression are not practically zero because there are differences of size between leverage and total assets. 2 We have divided the countries according to the legal character. In the French zone we include Belgium, Holland, Italy, France, Portugal and Spain; in the German zone Germany and Austria; in the Scandinavian zone Denmark, Sweden; in the British zone Ireland and UK and in the Eastern zone Slovakia, Hungary, Latvia, Poland and Malta, which does not belong to this zone but entered the EU at the same time. 3 As the OECD report points out, the greatest global amount raised from privatization in the UK during the 1990s was in 1991. 4 For a detailed study of the debt in developing countries, see Correia Da Silva et al. (2004) 5 We have replicated this analysis with other leverage measures and the results obtained have been similar but with less statistical significance.

References
Arrondo, R. and Gomez-Anson, S. (2003) A study of Spanish firms’ security issue decision under ´ ´ asymmetric information and agency costs, Applied Financial Economics, 13, pp. 771Á782. Berne, M. and Pogorel, G. (2004) Privatisation Experiences in France, CESifo Working Paper no. 1195, Category 9: Industrial Organisation, May 2004 (Munich: CESifo). Bevan, A. A. and Danbolt, J. (2002) Capital structure and its determinants in the UK Á a descompositional analysis, Applied Financial Economics, 12, pp. 159Á170. Boussofiane, A., Martin, S. and Parker, D. (1997) The impact on technical efficiency of the UK privatisation programme, Applied Economics, 29, pp. 297Á310. Boycko, M., Shleifer, A. and Vishny, R. W. (1996) A theory of privatisation, The Economic Journal, 106, March, pp. 309Á319.

122

M. J. Arcas & P. Bachiller

Cohen, S. (2001) A strategic framework for devolving responsibility and functions from government to the private sector, Public Administration Review, 61(4), pp. 432Á440. Correia Da Silva, L., Estache, A. and Jarvela, S. (2004) Is Debt Replacing Equity in Regulated Privatised ¨ ¨ Infrastructure in Developing Countries? World Bank Policy Research, Working Paper 3374, August 2004 (Washington, DC: The World Bank). De Miguel, A. and Pindado, J. (2001) Determinants of capital structure: New evidence from Spanish panel data, Journal of Corporate Finance, 7, pp. 77Á99. Dewenter, K. L. and Malatesta, P. H. (2001) State-owned and privately owned firms: An empirical analysis of profitability, leverage, and labour intensity, The American Economic Review, 91(1), pp. 320Á334. Goldstein, A. (2003) Privatisation in Italy 1993Á2002: Goals, Institutions, Outcomes, and Outstanding Issues, CESifo Working Paper no. 912, Category 1: Public Finance, April 2003 (Munich: CESifo). Gonzalo, J. A., Pina, V. and Torres, L. (2003) Objectives and valuation of companies in privatisation processes, Public Management Review, 5(2), pp. 177Á196. Hall, G. C., Hutchinson, P. J. and Nicos, M. (2004) Determinants of the capital structures of European SMEs, Journal of Business Finance & Accounting, 31(5Á6), pp. 711Á728. Hodges, R. (1997) Competition and efficiency after privatisation: The role of the NAO, Public Money & Management, 17(1), pp. 35Á42. Hyclak, T. J. and King, A. E. (1994) The privatisation experience in Eastern Europe, World Economy, 17(4), pp. 529Á550. Jensen, J. and Meckling, M. (1976) Theory of the firm: Managerial behavior, agency costs and capital structure, Journal of Financial Economics, 3, pp. 305Á360. Jensen, G. R., Solberg, D. P. and Zorn, T. S. (1992) Simultaneous determination of insider ownership, debt, and dividend policies, Journal of Financial and Quantitative Analysis, 27, June, pp. 247Á263. Knieps, G. (2004) Privatisation of Network Industries in Germany: A Disaggregated Approach, CESifo Working Paper no. 1188, Category 9: Industrial Organisation, May 2004 (Munich: CESifo). Kole, S. R. and Mulherin, J. H. (1997) The government as a shareholder: A case from the United States, Journal of Law & Economics, 40(1), pp. 1Á22. La Porta, R., Lopez De Silanes, F., Shleifer, A. and Vishny, R. W. (1997) Legal determinants of external finance, The Journal of Finance, LII(3), pp. 1131Á1150. Megginson, W. L., Nash, R. C. and Van Randenbourgh, M. (1994) The financial and operating performance of newly privatised firms: An international empirical analysis, The Journal of Finance, XLIX(2), pp. 403Á452. Newbery, D. M. (1997) Privatisation and liberalisation of networks utilities, European Economic Review, 41, pp. 357Á383. Organization for Economic Cooperation and Development (OECD) (2001) Recent Privatisation Trends, Financial Markets Trends, No. 79 (Paris: OECD). Organization for Economic Cooperation and Development (OECD) (2003) Privatising State-owned Enterprises. An Overview of Policies and Practices in OECD Countries (Paris: OECD). Omran, M. (2004) Performance consequences of privatising Egyptian state-owned enterprises: The effect of post-privatisation ownership structure on firm performance, World Development, 32(6), pp. 1019. Ozkan, A. (2001) Determinants of capital structure and adjustment to long run target: Evidence from UK company panel data, Journal of Business Finance & Accounting, 28(1&2), January/March, pp. 175Á198. Parker, D. (1998) Reducing regulatory risk: The case for a new regulatory contract with the privatised utilities, Public Money & Management, 18(4), pp. 51Á57. Rajan, R. G. and Zingales, L. (1995) What we know about capital structure? Some evidence from international data, Journal of Finance, L(5), pp. 1421Á1460. Ryen, G. T., Vasconcelos, G. M. and Kish, R. J. (1997) Capital structure decisions: What have we learned?, Business Horizons, 40(5), pp. 41Á50. Springdal, K. and Mador, M. (2004) Organisational changes resulting from the privatisation of state enterprises, Public Money & Management, 24(1), pp. 69Á80. Tian, L. (2001) Government Shareholding and the Value of China’s Modern Firms, Working Paper London Business School, August 2001 (London: London Business School). Titman, S. and Wessels, R. (1988) The determinants of capital structure choice, Journal of Finance, XLIII(1), pp. 1Á19.

Performance and Capital Structure of Privatized Firms in Europe

123

Verges, J. (2000) Privatisations in Spain: Process, policies and goals, European Journal of Law and ´ Economics, 9(3), pp. 255Á280. Vickers, J. and Yarrow, C. (1988) Privatisation and Economics Analysis (Cambridge, MA: The MIT Press). Villalonga, B. (2000) Privatisation and efficiency: differentiating ownership effects from political, organisational, and dynamic effects, Journal of Economic Behavior & Organisation, 42, pp. 43Á74. Wald, J. K. (1999) How firms characteristics affect capital structure: An international comparison, The Journal of Financial Research, XXII(2), pp. 161Á187.