PRIVATE EQUITY IN PORTUGAL - AN ANALYSIS

Transcription

1 33 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS PRIVATE EQUITY IN PORTUGAL - AN ANALYSIS OF THE PORTFOLIO COMPANIES OPERATING PERFORMANCE JOSÉ PEDRO MENDES * E MIGUEL SOUSA ** 1. INTRODUCTION The private equity (PE) industry has experienced impressive growth and increased significance over the last decade. First as a mainly US phenomenon, it then spread to Europe and today is a worldwide industry. The number and value of transactions have increased tremendously and its industry scope has broadened significantly. Despite these facts it has received little attention in the academic literature. Since the first wave of studies in the late 80s and early 90s little research has been done until very recent years. The reason is closely linked to its nature, involving private transactions, which makes information gathering quite often a challenging and difficult task. In Portugal research has been centralized in the venture capital industry or private equity without any clear distinction between venture capital and buyout activity. This paper examines buyout activity in Portugal backed by private equity firms not directly or indirectly run by the Portuguese government or other public entities. A self-collected sample of 101 buyouts completed between 1996 and 2010 by a private equity firm was constructed and historical performance analyzed to the most recent available data (2011). By presenting evidence on pre- and post-buyout portfolio companies operating performance, this paper answers a major question in the PE industry - are PE firms successful in improving the operating performance of their portfolio companies? To our best knowledge this is the first study ever made specifically relating to buyout activity in Portugal. While previous studies of Portuguese targets have focused on venture capital or private equity in its most general definition, this paper focuses exclusively on buyout activity in Portugal. In addition, this study presents evidence regarding a peripheral southern European economy with characteristics very distinct from those of the US, UK and northern Europe, where the vast majority of academic literature has been focused. Finally, as all transactions are private-to-private, this paper contributes to a better understanding of the effects of buyouts with a PE sponsorship over private companies, an area still quite unexplored internationally. The remainder of this paper is structured as follows. In section 2, a review of the academic literature is presented. Section 3 explains the methodology and the sample selection process and section 4 presents the empirical results regarding the portfolio companies operating performance, before and after the buyout. Section 5 concludes. * Faculdade de Economia, Universidade do Porto; ** Faculdade de Economia, Universidade do Porto e CEF.UP, Said Business School, University of Oxford

2 34 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS 2. LITERATURE REVIEW 2.1. Private equity and sources of value creation Since early studies in the late 80 s and 90 s up to today s studies the value creation process and the influence of the PE firms has been a major topic in academic private equity related literature. There is a general consensus that, on average, private equity activity does create value for their portfolio companies (Kaplan, 1989b; Guo et al., 2011) and that portfolio companies have a better financial performance when they go public compared to other IPOs companies (Degeorge and Zeckhauser, 1993; Hotlhausen and Larcker, 1996; Cao and Lerner, 2009; Levis, 2010). The increase in wealth comes from different sources such as tax benefits, superior corporate governance models and better management practices, as well as operational restructuring and increased efficiency. Very briefly, tax benefits arise from the very highly leveraged financial structures (Kaplan, 1989a; Guo et al., 2011), which also play an important role in reducing agency costs (Jensen, 1989). Better corporate governance models and better management practices, such as the participation of managers and directors on equity capital and incentives closely tied to performance (Jensen, 1989; Nikoskelainen and Wright, 2007; Leslie and Oyer, 2009), and improved control and reporting mechanisms (Bloom et al., 2009) are imposed by private equity investors when they invest in the companies. Operational restructuring also plays an important role in value creation, as is demonstrated in several academic articles that show operating performance and efficiency improvements by portfolio companies during the holding period, when compared with industry and/or peer firms (Kaplan, 1989b; Smith, 1990; Murray et al., 2006; Cressy et al., 2007; Davis et al., 2009) Operating performance Kaplan (1989b) using a sample of MBOs that occurred between 1980 and 1986, finds that, after the MBO, the companies increase their operating income (before depreciation) and their net cash flow while at the same time they reduce their capital expenditures (CAPEX). Comparing the three years after the buyout with the pre-buyout year he presents evidence of a significant increase in the operating income to assets and to sales as well as the net cash flow net of industry changes. Value creation evidence was also presented as investors earned a combined median total market-adjusted return of 77%. Hypotheses of wealth transfers from employees to PE investors and inside information advantages are rejected while better incentives and reduced agency costs are pointed out as the main reasons for the improved performance. Using a different methodology, Smith (1990) reaches the same conclusions. Using a sample of 58 MBOs which occurred between 1977 and 1986, he finds significant and sustained increases in the operating returns. Once again the improved management incentives are pointed out as the explanation and other hypotheses such as layoffs and cutbacks in discretionary expenses are rejected. Consistently, Lichtenberg and Siegel (1990) uncover a positive effect from LBOs on the total factor productivity at plant level for LBOs which occurred between 1983 and With reverse LBOs 1 occurring after 1983, Muscarella and Vetsuypens (1990) detect, during the period under private equity ownership, an improvement in profitability measures, both in absolute terms and when compared to a random sample. 1- A reverse LBO occurs when a company that has been taken private in the past through a LBO goes public again.

3 35 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS Companies put into practice restructuring activities, such as divestments and reorganization of production facilities, in order to optimize the firm s asset mix. Also, Degeorge and Zeckhauser (1993) find that the reverse LBOs which occurred between 1983 and 1987 substantially outperform comparison firms before the IPO. Finally, Holthausen and Larcker (1996), using a sample of 90 reverse LBOs which occurred between 1983 and 1988, attain conclusions similar to those achieved by their predecessors. They find that the operating performance of reverse LBOs is significantly better than the industry before the IPO and that a strong relationship exists between performance and the percentage of equity owned by the operating management and other insiders. Declines in operating performance after an IPO are associated with the decline of managerial equity ownership and are expressed through increases in working capital levels and capital expenditures. More recently, researchers have started to include in their samples the latest transactions including many which have occurred during the 21st century buyout bust. Using a sample of US buyouts which occurred between 1990 and 2006, Guo et al. (2011) show that LBOs still create value. Median market and risk adjusted returns to pre- or (post-) buyout capital are estimated at 72.5% (40.9%), distress cases included. However, operating gains seem to be substantially smaller than those documented for deals that occurred in the 80s, and not significantly better than industry or benchmark firms. Operating gains alone are not sufficient to explain the positive average returns achieved by private equity investors. In fact, they only account for 22.9% of the returns with the rest being attributed to tax benefits (33.8%) and industry valuation multiples (17.7%). Davis et al. (2009) also document a significant increase in PRIVATE EQUITY IN PORTUGAL... : 35 productivity. The productivity growth differential between buyouts and comparable firms is around 2% within two years after the transaction. Lerner et al. (2009) show an increase in patent quality after buyout and no deterioration of the level of R&D. They reject the hypothesis that PE-backed firms sacrifice long-run investments but rather verify a beneficial refocusing of R&D activities (patent portfolios). After the first wave of studies, academic research has focused more on European buyout activity, mainly in UK, largely due to data availability. Once again most of these papers find LBOs to be associated with significant operating and productivity improvements. Murray et al. (2006) find that UK buyouts occurring between 1980 and 1998 experienced a significant operating performance improvement before the IPOs and that the improvement is the result more of a better use of the assets to generate sales than through higher margins. Although no significant differences are found between PE-backed and non-pe-backed deals, they find different performances among deals backed by prestigious and non-prestigious PEfirms. Also for the UK, Wier et al. (2008) document that public-to-private (PTP) transactions 2, between 1998 and 2004, experience a deterioration of performance relative to the pre-buyout situation but this change in performance is not worse than the change in firms that remained public. There is even some evidence of performance improvements when related with firms that remained public. Moreover, deals backed by PE firms perform better than the industry and not worse than deals without a PE firm intervention. At plant-level, Harris et al. (2005) find evidence of substantial increases in labor and total productivity for MBOs and, once again, the changes in incentive and governance mechanisms are indicated to be the reason for 2- A public-to-private transaction, also known as PTP, consists of the acquisition of a listed company by a private equity investor.

4 36 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS the enhanced economic efficiency. A similar firm-level productivity improvement is reported by Amess (2002). Cressy et al. (2007) analyze PE-backed buyouts which occurred in the UK between 1995 and 2002, and find that the operating profitability is greater than that of comparable non-buyout companies by 4.5%. Moreover, industry specialization by the PE firm raises this differential by an additional 8.5%. They also found evidence that investment selection plays an important role, which together with financial engineering techniques may overpass managerial incentives in raising performance. In Western Europe, Acharya et al. (2010) find that ownership by large and mature PE firms has a positive impact on the operating performance of their portfolio companies relative to the sector operating performance and that there is a strong connection between ownership by large and mature PE firms and the value creation process. In Sweden, Bergstrom et al. (2007) also find that PE sponsored buyouts experience significant increases in operating profitability and find no evidence that the value created on a company level is transferred from its employees. Contrastingly, Desbrières and Schatt (2002) find deterioration in the short term performance of French firms involved in LBOs throughout Achleitner et al. (2010) notice that, in 206 buyouts that occurred in Europe between 1991 and 2005, two-thirds of the value created is the result of operational and market effects (changes in EBITDA multiples) and one-third is due to leverage. Furthermore the leverage effect appears to be stronger for larger buyouts whereas for smaller ones the increase of revenue plays a more important role. Finally, in an interesting working paper Hsu et al. (2010) analyze the stock prices of industry competitors around any announcement of PE-backed investments. PE investments in an industry decrease competitors stock prices and their operating performance. Also the withdrawal of a previously announced PE investment leads to the opposite outcome as competitors stock prices increase. These results support the hypothesis that performance differences are driven, at least partially, by the advantages brought to the targets by PE investors. 2.3 Private-to-private transactions Despite the fact that private-to-private transactions account for a significant portion of all private equity market little research has been devoted to them, as the great majority of the previous empirical studies have been biased towards reverse LBOs and public-to-private transactions. This fact comes largely as a result of data unavailability. One important exception is Chung (2011), who analysed private-to-private buyouts in the UK. The motivations and economic forces driving private-to-private LBOs is different than publicto-private buyouts and the sources of value, namely reduced agency costs, will not be the same. In private companies the ownership structure most of the time is highly concentrated and quite often the owners themselves serve as managers of their companies. Therefore the agency problems resulting from incentive misalignment between owners and managers are very unlikely to exist to the extent they do in public companies and so do not represent an important potential source of value. Instead, the motivations behind the acquisitions of private companies by PE firms may be to mitigate inefficiencies which are the consequence of investment constraints that small companies typically face. Among these constraints are limited or costly financing conditions, risk aversion by property/management, a lack of knowledge,

5 37 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS expertise and financial resources to make valuable investments, and different goals, such as preserving wealth, keeping the business stable, maintaining stable income flows and providing employment opportunities for their descendants. PE firms mitigate these investment constraints through: reducing the ownership concentration and thus the risk exposure by owners; permitting better access to credit; advanced management skills and networking; and more risk tolerance and greater capacity to undertake risky investments. An outstanding finding is that, unlike public companies, private targets acquired by PE firms grow substantially after the buyout. The typical targets are companies with large profitability and growth opportunities. In fact, as found by Chung (2011), rather than improving operating efficiency these firms grow in size and make greater investments. Companies experienced an increase in EBITDA - industryadjusted EBITDA grows by 12% during the three post-buyout years - but margins do not improve and these companies face a decline in the average profitability after the buyout. Similar results are reached by Boucly et al. (2009) using a sample of French LBOs, which present evidence that portfolio companies after the buyouts foster growth by alleviating financial constraints, particularly those faced by small and medium privately held enterprises. In conclusion, it can be said that PE firms reorganize their portfolio companies and reduce their inefficiencies - agency problems in case of public companies and investment constraints in the case of private companies (Chung, 2011). 3. METHODOLOGY AND SAMPLE SELECTION 3.1. Methodology PRIVATE EQUITY IN PORTUGAL... : 37 In this paper, four variables are used to assess the evolution of company before and after the buyout. Book value of total assets and total sales are used as proxy for size. CAPEX measures the investment activity by the portfolio company. Finally, EBITDA 3 is used as proxy of cash-flow generated by the company 4. All variables are presented before interests and taxes, therefore controlling for effects resulting from leverage or other financial decisions. In order to assess the operating performance before and after the buyout three ratios are used: return on sales as a measure of profitability, return on assets as a measure of productivity and asset turnover as a measure of efficiency. Profitability and productivity measures are calculated by deflating the EBITDA by the end-ofperiod total sales and the end-of-period total assets respectively. These first two ratios aim to measure the ability of the company to improve or maintain its profitability margin and the productivity in asset utilization (the return on assets). These two indicators have been extensively used in past studies (for example Kaplan, 1989b; Jain and Kini 1994; Murray et al., 2006; Guo et al. 2011) and partially control for acquisitions and divestures after the buyout. The third ratio - asset turnover (Total Sales/Total Assets) - aims to assess the company s efficiency in using its assets to generate sales and to separate the variation of the productivity into margin effect or greater effectiveness in producing sales (Murray et al., 2006). 3- Earnings before interest, taxes, depreciation and amortization 4- EBITDA is often used in academic research and by practitioners as a proxy of the cash flow. A correct measure of cash-flow subtracts from the EBITDA the changes in working capital and the CAPEX. Changes in working capital are not subtracted because working capital data is not available for the majority of portfolio companies and, also, previous related studies (Kaplan, 1989b; Jain and Kini, 1994) did not include changes in working capital in their cash flows measures. CAPEX was already analyzed separately.

6 38 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS Changes in portfolio companies size and operating performance are measured over a three year period after the buyout (year +1 to year +3) relative to the last fiscal year before the acquisition (year -1). A similar analysis is also made for the pre-buyout year relative to the previous year (year -2). The overall change between the pre-buyout year and the first three full post-buyout years (3y avg.) is also presented. The fiscal year of the acquisition (year 0) is not considered as it includes both pre- and postbuyout operations, making it difficult to differentiate between pre- and post-buyout performance. Post-buyout operating performance change is calculated according to the formula, where i symbolizes the company, t the fiscal year of the buyout and j the fiscal year after the buyout. To analyze the operating performance before the buyout the formula is equal to 5.. The results are reported as both unadjusted and adjusted by industry changes during the same period. Industry data was extracted from the Central de Balanços by the Bank of Portugal. Central de Balanços provides the average Income Statement and Functional Balance Sheet according to industry classification. Unfortunately, industry data provided by Central de Balanços suffers from severe limitations that narrow substantially the possibilities of analysis. Before 2005 industry data was obtained through non-mandatory annual surveys. However since 2006 these surveys have become compulsory. This change makes Income Statements and Balance Sheets not directly comparable between these two periods. As a consequence some variables, such as CAPEX and net cash flow cannot be calculated for any industry for the year 2006 since the calculation is partly based on the comparisons of the same variables between 2006 and Moreover, industry adjustments cannot be performed for any variable during this period. However, industry adjustments were possible in the case of operating performance measures since industry values are relative to size and so don t exhibit significant differences and discrepancies between the two periods. Industry matching was based on the company s four digit CAE Rev. 3 code and industry adjustment values are obtained by subtracting the changes in the portfolio companies operating performance ratios from the industry average change in the same ratio during the same period. Through this adjustment it is possible to control for both time period and industry effects. Finally, a Wilcoxon sign rank test is performed to test whether the median percentage change in operating performance and industry-adjusted operating performance variables are significantly different from zero Sample selection In order to select the sample of all private equity buyout transactions that took place in Portugal and involved Portuguese target companies, the Capital IQ database was used. Capital IQ is a database commercialized and managed by Standards & Poor s and provides information on the portfolio company name, acquisition date and the buyers/investors The difference between the ratios (e.g.: EBITDA/Total Sales) in the year t+j and the year t-1 was also calculated as an alternative measure in order to avoid the exclusion of observations for which the variable in the year t-1 is equal to zero or negative. This methodology was previously used in Jain and Kini (1994). However, no substantial differences were found between the results produced by these two different methodologies. 6- Capital IQ description and potential biases on sample selection are discussed in detail in the Appendix.

7 39 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS PRIVATE EQUITY IN PORTUGAL... : 39 In order for a transaction be included in the sample three requirements had to be fulfilled: i) the portfolio company must have its headquarters in Portugal; ii) there must be a clearly identifiable private equity sponsor(s); iii) the sponsor(s) mustn t be directly or indirectly run by the government or other public entities. While the first requirement is quite obvious since the aim of this paper is to examine the operating performance of Portuguese targets, the other two may require a more detailed explanation. For a PE sponsor to be considered valid and the deal included in the sample, the PE firm should have as its core business investments in mature companies. These include investments such as LBOs, MBOs, MBIs, expansion and replacement capital and turnaround investments. In this way, venture capital investments were excluded. The nationality of the PE firm is irrelevant for the purposes of this study, meaning that Portuguese and foreign sponsors are included in the sample. Transactions where at least one of the PE firms involved was considered to be a governmental affiliate were not included in the sample. Government and other public entities may pursue goals other than strictly wealthmaximizing objectives, such as territorial development and employment growth (Cumming and Macintosh, 2006; Lerner, 1999) or, as in the Portuguese case, may aim to support investment activities mainly in small and medium size companies (Duarte, 2006). In order to keep a strictly economic rationale and avoid potential biases these deals were not included in the sample 7. In a market where government is a strong player this certainly diminishes the sample size and exclude many other transactions, but the inclusion of these transactions would not be consistent with the objective of investigating the pure buyout activity in Portugal. Finally, in the rare case of secondary buyouts only the first transaction, when the portfolio company turned into a PE sponsorship structure for the first time, was included in the sample. The reason relies on the distinctive characteristics and motivations for PE sponsors to undertake this deal type (Jenkinson and Sousa, 2010), which can be expected to influence portfolio companies operating performance in a way that is different to the buyouts of independent companies. It was possible to identify 101 transactions which occurred between 1996 and 2010 that fulfilled the required conditions. This unique sample covers the vast majority of buyout deals in Portugal without governmental sponsorship and therefore is very representative of Portuguese pure buyout activity. The sample includes only private to private transactions. At this point, accounting data from the target companies was collected from SABI, a commercial database managed by Bureau Van Dijk, which contains financial information for over 20,000 Portuguese companies from onward. From the initial sample of 101 companies it was possible to obtain information for only 82 companies. Relevant information included the financial statements, year of foundation and industry classification. As all transactions are private-to-private the data collecting process is particularly sensitive to data availability by SABI. SABI presents some limitations, the first being the range of its period of coverage. In order to analyze the impact of a PE buyout on the portfolio company s 7- We are not making any assumptions that governmental intervention is either negative or positive. The exclusion of this type of transaction is merely justified by different deal rationales. 8- SABI description and potential database related biases are discussed in detail in the Appendix.

8 40 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS operating performance, accounting data from the year before the acquisition is fundamental, which means that transactions before 1999 are automatically excluded. This and others handicaps narrow significantly the number of portfolio companies upon which the performance measures can be calculated and may have created a bias toward more recent transactions (after 2000). In all the analysis presented the number of observations will be indicated with the evidence Descriptive statistics Table 1 reports the yearly distribution of the buyouts included in the final sample. As shown the Portuguese buyout industry is still quite recent. Transactions in the 90 s were rare and the first boom only occurred in However, it is only after 2004 that a sustained buyout activity arises. Along with the international trend, the years between 2004 and 2007 were good years for the buyout industry in Portugal in terms of deals closed. However, a remarkable year occurs in 2008 when, during the peak of the financial crisis and in sharp contrast to Portugal s global peers, a record number of investments (19) are accomplished. The last two years 2009 and 2010 mark a slowdown in the number of deals closed. Table 1: Buyout Year Distribution Ye ar Numbe r of trans actions Total 101

9 41 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS Table 2 provides transaction statistics on all the 101 buyouts, focusing on characteristics such as deal type (non-syndicated vs. syndicated), origin country of PE firms and the number of deals on which each PE firm was involved. In Table 2, Panel A, it can be seen that the vast majority of investments (79.21%) are undertaken by individual Portuguese PE firms. The Portuguese buyout market is still a domestic one where 88.12% of all transactions are carried out exclusively by Portuguese PE firms. Foreign individual buyouts are very rare (4.95%) and more than half of the total foreign investments are carried in syndication with a Portuguese firm(s), suggesting that foreign PE houses look for a national partner in order to undertake these investments. There is no record of a PRIVATE EQUITY IN PORTUGAL... : 41 syndicated transaction carried out entirely by foreign PE firms. Syndicated deals are not the common practice and only account for 15.84% of the total of transactions. They are expected in the presence of a foreign PE investor (as previously noticed) and in the case of exclusively Portuguese deals for particularly large transactions. Table 2, Panel B, ranks the PE firms according to the number of deals completed. Among the 101 transactions 18 different PE firms were identified with a total of 118 participations. All the top 5 firms reported are Portuguese and clearly dominate the market. In particular, the top 4 PE houses are the biggest and most important players, who alone took almost 75% of all deals. Table 2: Transaction characteristics Number of deals % Panel A: Deal type and country of Private Equity Firms Non-syndicated Portugal 80 79,21 Others 5 4,95 Total 85 84,16 Syndicated Portugal only 9 8,91 Portugal and others 7 6,93 Total 16 15,84 Total deals Panel B: Private Equity Firms by number of deals Espírito Santo Capital, S.C.R., S.A ,19 Inter-Risco, S.C.R., S.A ,8 Change Partners - SCR, S.A ,8 Explorer Investments, S.C.R, S.A ,1 ECS - Sociedade de Capital de Risco, SA 6 5,08 Others (13) 26 22,03 Total

10 42 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS Another significant fact is that the two leading PE firms (Espírito Santo Capital, S.C.R., S.A. and Inter-Risco, S.C.R., S.A. 9 ) are bank affiliate institutions, illustrating the importance of banks and their bank affiliate firms in the Portuguese private equity industry. Overall, these results put in evidence a market still predominately domestic, with little competition and dominance by four major PE houses, and where banks and their affiliate PE firms are more important than independent PE firms. Table 3 describes the distribution of 86 portfolio companies across industries. 10 The buyouts occurred in a variety of industries and encompass almost every sector of activity. The predominant industry is business and industrial products (27.91%), immediately followed by business and industrial services (19.77%) and consumer goods and retail (11.63%). Table 3: Portfolio companies industry distribution Industry Portfolio companies Business and industrial products 24 Business and industrial services 17 Consumer goods and retail 10 Consumer services: Other 8 Energy and environment 7 Communications 6 Computer and consumer electronics 5 Construction 3 Life sciences 3 Chemicals and materials 1 Table 4 presents statistics on some portfolio companies characteristics prior to the acquisition. The last full year before the completion of the buyout (year -1) is used as the reference year since the acquisition may occur at any point during year 0 and so financial statements may already incorporate buyout effects, making interpretation of that year difficult and potentially ambiguous and inaccurate. The final column (N) represents the number of observations contained in the sample. 9- Inter-Risco was separated from Banco Português de Investimento and turned into an independent PE firm controlled by the management team in November Industry classification is based on the Classificação Portuguesa de Actividades Económicas (CAE) Rev. 3.

11 43 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS PRIVATE EQUITY IN PORTUGAL... : 43 Table 4: Portfolio companies characteristics Variable Median Mean Min Max St. Dev. N Firm age (y) 12 17, ,12 82 Total assets Total Sales EBITDA CAPEX All values in thousands of euros The median 11 age of the buyout targets was 12 years at the time of buyout. In the year before the buyout the portfolio companies had a median (mean) of total assets of 8.6 million ( 26 million) and median (mean) total sales of 7.8 million ( 16.7 million). The largest target had a total assets book value of 539 million in comparison to the smallest whose total assets had a value of just 225 thousand. EBITDA and CAPEX 12 exhibit positive median and mean values before the acquisition. As expected, all variables exhibit significantly smaller values to those seen in previous operating performance studies, particularly those that focus exclusively on public-to-private transactions (Kaplan, 1989b; Murray et al., 2006; Boucly et al., 2009). However, the figures are similar to the values found by Chung (2011) who analyses private-to-private buyouts in the UK. Naturally, the small scale of the Portuguese economy and the very nature of the targets (private companies) result in quite a modest deal size when compared to international studies. Moreover, while transactions of billions are not rare in the US or the UK, in Portugal the largest target had a book value of total assets of approximately half billion euros. 4. OPERATING PERFORMANCE OF BUYOUT COMPANIES 4.1. Pre-buyout operating performance Table 5 reports the median growth in total assets, total sales, EBITDA and CAPEX in the last full year before the buyout. It can be seen that the book value of both total assets and total sales grow significantly and substantially (6.98% and 7.90% respectively) before the acquisition. Even more noticeable, target companies experience a significant increase in EBITDA (21.72%). This fact is particularly important and cogent for private equity investors looking for high cash flow generating companies in order to fulfill high debt service payments resulting from the leverage structures. CAPEX decreases before the acquisition (28.20%) and although this decrease is not statistically significant it may indicate that the portfolio companies are facing financial constraints to investment before the buyout. 11-Median rather than the mean will be used throughout this paper due to the potential dominance of outliers that dominate the means and make the distribution very skewed. This outlier dominance may be particularly severe in the presence of a small sample, as is the case here, and can be seen in the differences between the minimum and maximum values. 12- CAPEX data was not available in the financial statements so the values presented are calculated according to the formula CAPEX = Fixed Assets net of Amortizations (t) - Fixed Assets net of Amortizations (t-1) + Amortizations (t). Although this is not an exact measure of CAPEX, these values are expected to be an accurate proxy of the true amount of the investments implemented.

12 44 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS Table 5: Portfolio companies growth before buyout (year -2 to -1) Variable Median change (%) N Book value of total assets 6,98 *** 47 Total Sales 7,90 *** 46 EBITDA 21,72 * 42 CAPEX -28,20 38 *, **, *** significantly different from zero at 10%, 5% and 1% respectively Table 6 presents the portfolio companies operating performance change in the last year before the buyout. Although the results indicate that portfolio companies experienced an increase in all operating performance measures in the year before their acquisition, none of these results is statistically significant and not significantly different to that seen in their industry counterparts. Table 6: Changes in portfolio companies operating performance before buyout (year -2 to -1) Profitability Productivity Efficiency EBITDA/Total Sales EBITDA/Total Assets Total Sales/Total Assets Median change (%) 3,18 1,51 3,28 N Median ind-adj change (%) 7,84-1,79 3,07 N *, **, *** significantly different from zero at 10%, 5% and 1% respectively The results presented suggest that private equity sponsors seem to choose companies with a great potential to generate EBITDA but with financial constraints to investment. These findings bear strong resemblances to those seen in Chung (2011), where private targets display a significant growth in EBITDA but a quite modest operating performance before the buyout Post-buyout operating performance Change in levels Table 7 reports the median changes in total assets, total sales, EBITDA, CAPEX and net cash flow for the three post-buyout years. The results show a steady and significant growth in total assets, reaching its peak at 67.34% in the third post-buyout year. The median increase in total assets over the three year period equals 41.01% and is significant at the 1% level. Total Sales also exhibit a growth in all three post-buyout years and a median three year period average growth of 10.4%. These results are significant at least at the 10% level 13. Although not statically significant, portfolio companies exhibit a decrease in EBITDA over the three year period of -5.35%. CAPEX also decreases in the year +1 and year +2, followed by a strong growth in year +3. However, once again these results are not statically significant. 13- Although industry adjustment was not possible due to the industry data constraints previously explained, it is possible to infer that, considering the poor performance that the Portuguese economy exhibited during the 21st century, these changes would most probably be superior to those experienced in the industry.

13 45 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS PRIVATE EQUITY IN PORTUGAL... : 45 Table 7: Changes in portfolio companies growth after buyout *, **, *** significantly different from zero at 10%, 5% and 1% respectively Median change (%) Year relative to buyout: -1 to 3y avg. -1 to +1-1 to +2-1 to +3 Total assets 41,01 *** 20,55 *** 45,37 *** 67,34 *** N Total Sales 10,4 *** 12,35 ** 19,79 ** 12,74 * N EBITDA -5,35-1,1 13,74-29,09 N CAPEX 13,72-5,1-36,46 32,02 N Overall, these findings partially confirm the results of Chung (2011) and Boucly et al. (2009) that private targets become larger after the buyout, and are contrary to Kaplan (1989b) s findings that targets size decreases after the buyout. Like Chung (2011), assets and sales both experience a significant growth during the three post-buyout years Profitability According to Table 8, the portfolio companies EBITDA margin decreases after the completion of the buyout. The median decrease in the EBITDA margin is equal to 17.64% across the three post-buyout years (-0.87%, -5.22%, % in years +1, +2 and +3 after the buyout, respectively). This average decrease is statistically significant at a 10% level. The industry-adjusted change for the three year average is still negative (-23.57%) and is statistically significant at the 5% level, revealing that the negative performance of portfolio companies after the buyout is only partially justified by a decrease in the industry profitability. Table 8: Changes in portfolio companies profitability Profitability -1 to 3y avg. -1 to +1-1 to +2-1 to +3 EBITDA/Total Sales Median change (%) -17,64 * -0,87-5,22-30,57 N Median ind-adj change (%) -23,57 ** -1,08-11,51-18,91 N *, **, *** significantly different from zero at 10%, 5% and 1% respectively These findings present similarities with more recent studies that focused on later transactions which show poor and significantly smaller operating and margin gains than those documented in earlier research (Guo et al., 2011; Wier et al., 2008). Moreover, these results are strongly connected to those reached in Chung (2011) for private-to-private transactions that evidenced a strong decline on average profitability of the portfolio companies over time after the completion of the buyout and a performance below the industry.

14 46 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS Productivity Even if profitability deteriorates companies can still create value by increasing the productivity of their assets, for example by selling nonproductive assets (Guo et al., 2011). Table 9 presents the portfolio companies productivity after the buyout, measured by the return on assets change. Table 9: Changes in portfolio companies productivity Productivity -1 to 3y avg. -1 to +1-1 to +2-1 to +3 EBITDA/Total Assets Median change (%) -39,50 ** -24,93-31,65 * -55,70 *** N Median ind-adj change (%) -23,19 ** -34,21-30,20 ** -49,73 ** N *, **, *** significantly different from zero at 10%, 5% and 1% respectively From Table 9, a persistent and substantial decline of productivity is clear across all three post-buyout years. In the three year period the return on assets of portfolio companies falls %, statistically significant at the 5% level. Indeed, after the acquisition the productivity falls dramatically in all three post-buyout years: %, %, % in years +1, +2 and +3, respectively. Adjusted by industry change for the same period, the decrease of productivity after the buyout is even more severe in the first year (-34.21%), while in the second and third year the negative performance is only slightly explained by industry effects. For the overall period, the decrease in portfolio companies productivity, after the buyout, is much worse than in comparable companies as the industry-adjusted three year average median equals %. Once more these findings are contrary to those in Kaplan (1989b) but are consistent with the poor operating performance of portfolio companies found in Guo et al. (2011) and Wier et al. (2008). Once again, as in the case of profitability, the productivity results are very similar to those found in Chung (2011) where portfolio companies exhibit drastic declines in productivity in all three post-buyout years Efficiency Analyzing the asset turnover allows us to discern if the changes in productivity are only the result of lower margins or also the result of a decrease in portfolio companies efficiency in using its assets to generate sales. Although table 10 shows essentially a fall in asset turnover after the buyout, none of these results is statistically significant.

15 47 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS PRIVATE EQUITY IN PORTUGAL... : 47 Table 10: Changes in portfolio companies efficiency Asset Turnover -1 to 3y avg. -1 to +1-1 to +2-1 to +3 Total Sales/Total Assets Median change (%) -6,71-7,10-13,67-23,64 N Median ind-adj change (%) 4,14-9,07-0,80-22,66 N *, **, *** significantly different from zero at 10%, 5% and 1% respectively While this result is not consistent with Kaplan (1989b) and Murray et al. (2006), who demonstrate that the improvement of asset efficiencies is an important source of value creation, the improvement is mostly generated by asset disposals. Although the companies in our sample don t improve (or deteriorate) their asset efficiency after the buyout, they significantly increase their assets. To sum up, it is possible to say that the decrease in portfolio companies performance after the buyout is essentially the result of a strong decline in the companies margins. 5. CONCLUSIONS This paper studies the Portuguese private equity industry and examines the operating performance of portfolio companies from a selfcollected sample of 101 buyouts completed by private equity firms between 1996 and with little competition, dominated and controlled by four major PE houses and where banks and bank affiliate institutions are the main players. Evidence on pre- and post-buyout operating performance is presented. Before the buyout, the private targets experienced a strong and significant increase in EBITDA and, after the buyout, portfolio companies increase their assets and sales significantly. These findings are consistent with Chung (2011), suggesting that private equity investors look for private companies with larger profitability and growth opportunities. After the buyout, although portfolio companies increase their assets and sales significantly, their operating performance is very poor. EBITDA decreases after the buyout whereas CAPEX only increases in the third year after the buyout. Analyzing transactions which occurred in Portugal it s possible to conclude that the Portuguese private equity market is a market still far from its maturity. Sustainable buyout activity only emerged after 2004 and most of the investments have been carried out by individual and domestic PE firms. Furthermore, it is a market Moreover, the analysis of the three operating performance measures return on sales as measure of profitability, return on assets as a measure of productivity, and asset turnover as measure of efficiency leads to the conclusion that the operating performance does not improve and even deteriorates after the buyout.

16 48 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS All three measures exhibit an inferior performance when compared to the pre-buyout year. In particular, the decline of profitability and productivity proved to be statistically significant. Industry-adjustment reveals that the operating performance is below the industry, suggesting that the low performance is only slightly explained by time and industry effects. The findings for these Portuguese portfolio companies seem to confirm recent research focused on latter transactions (Guo et al., 2011 and Wier et al., 2008) that don`t encounter significant improvements in operating performance, particularly when compared to the operating performance improvements found for the first boom of buyouts (Kaplan, 1989b). In addition, these results clearly confirm the poor operating performance of private targets already found in Chung (2011).

17 49 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS PRIVATE EQUITY IN PORTUGAL... : 49 REFERENCES Acharya, V., Hahn, M. and Kehoe, C. (2010) Corporate Governance and Value Creation: Evidence from Private Equity, Working Paper, NYU Stern School of Business. Achleitner, A., Braun, R., Engel, N., Figge, C. and Tappeiner, F. (2010), Value Creation Drivers in Private Equity Buyouts: Empirical Evidence from Europe, The Journal of Private Equity, Vol. 13, No. 2, p Amess, K. (2002), Management buyouts and firm-level productivity: evidence from a panel of UK manufacturing firms, Scottish Journal of Political Economy, Vol. 49, No. 3, p Bergström, C., M. Grubb, and S. Jonsson (2007), The operating impact of buyouts in Sweden: A study of value creation, The Journal of Private Equity, Vol. 11, No. 1, p Bloom, N., Raffaella, S. and Reenen, J. V. (2009), Do private equity owned firms have better management practices?, Working Paper, London School of Economics and Political Science. Boucly, Q., Sraer, D. and Thesmar, D. (2009), Job creating LBOs, Working Paper, HEC Paris. Cao, J. and Lerner, J. (2009), The performance of reverse leveraged buyouts, Journal of Financial Economics, Vol. 91, No. 2, p Chung, J. (2011), Leveraged Buyouts of Private Companies, Working Paper, The Chinese University of Hong Kong. Cressy, R., Munari, F. and Malipiero, A. (2007), Playing to their strengths? Evidence that specialization in the private equity industry confers competitive advantage, Journal of Corporate Finance, Vol. 13, No. 4, p Cumming, D., Macintosh, J.G. (2006), Crowding out private equity: Canadian evidence, Journal of Business Venturing, Vol. 21, p Davis, S., Haltiwanger, J., Jarmin, R., Lerner, J. and Miranda, J. (2009), Private equity, jobs and productivity, The Global Economic Impact of Private Equity Report 2009, World Economic Forum. Degeorge, F. and Zeckhauser, R. (1993), The reverse LBO decision and firm performance: Theory and evidence, Journal of Finance, Vol. 48, No. 4, p Desbrieres, P. and A. Schatt (2002): The Impacts of LBOs on the Performance of Acquired Firms: The French Case", Journal of Business, Finance and Accounting, Vol. 29, No. 5, p Duarte, P. (2006), Capital de Risco Análise da Indústria em Portugal, Dissertação submetida como requisito parcial para obtenção do grau demestre em Finanças, ISCTE. Guo, S., Hotchkiss, E. and Song, W. (2011), Do buyouts (still) create value?, Journal of Finance, Vol. 66, No. 2, p Harris, R., Siegel, D. S. and Wright, M. (2005), Assessing the impact of management buyouts on economic efficiency: Plant-level evidence from the United Kingdom, The Review of Economics and Statistics, Vol. 87, No. 1, p Holthausen, R. W. and Larcker, D. F. (1996), The financial performance of reverse leveraged buyouts, Journal of Financial Economics, Vol. 42, No. 3, p Hsu, H., Reed, A. and Rocholl, J. (2010), Competitive effects of Private Equity investments, Working Paper, EFM 2010 Symposium of Entrepreneurial Finance & Venture Capital Markets.

18 50 : CADERNOS DO MERCADO DE VALORES MOBILIÁRIOS REFERENCES Jain, B. A. and Kini, O. (1994), The post-issue operating performance of IPO firms, Journal of Finance, Vol. 49, No. 5, p Jenkinson, T. and Sousa, M. (2010), Keep taking the private equity medicine? How operating performance differs between secondary deals and companies that return to public markets, Working Paper, University of Oxford. Jensen, M. C. (1989), Eclipse of the public corporation, Harvard Business Review, September-October 1989, p Kaplan, S. N. (1989a), Evidence on taxes as a source of value, Journal of Finance, Vol. 44, No. 3, p Kaplan, S. N. (1989b), The effects of management buyouts on operating performance and value, Journal of Financial Economics, Vol. 24, No. 2, p Lerner, J. (1999), The Government as venture capitalist: the long-run impact of the SBIR program, Journal of Business, Vol. 72, p Lerner, J., Sorensen, M. and Stromberg, P. (2009), Private equity and long run investment: the case of innovation, Working Paper, EFA 2009 Bergen Meetings Paper. Leslie, P. and Oyer, P. (2009), Managerial Incentives and Value Creation: Evidence from Private Equity, Working Paper, EFA 2009 Bergen Meetings Paper. Levis, M. (2010), The performance of private equity backed IPOs, Working Paper, City University, London. Lichtenberg, F. R. and Siegel, D. (1990), The effects of leveraged buyouts on productivity and related aspects of firm behavior, Journal of Financial Economics, Vol. 27, No. 1, p Murray, G., Niu, D. and Harris, R. (2006), The Operating performance of buyout IPOs in the UK and the influence of private equity financing, Working Paper, University of Exeter. Muscarella, C.J. and M.R. Vetsuypens (1990): Efficiency and Organizational Structure: A Study of Reverse LBOs, Journal of Finance, Vol. 45, p Nikoskelainen, E. and Wright, M. (2007), The impact of corporate governance mechanisms on value increase in leveraged buyouts, Journal of Corporate Finance, Vol.13, No. 4, p Smith, A. J. (1990), Corporate ownership structure and performance: The case of management buyouts, Journal of Financial Economics, Vol. 27, No. 1, p Strömberg, P. (2008), The New Demography of Private Equity, Working Paper, Swedish Institute for Financial Research. Weir, Jones, P. and Wright, M. (2008), Public to Private Transactions, Private Equity and Performance in the U.K.: An Empirical Analysis of the Impact of Going Private, Working Paper, Robert Gordon University - Aberdeen Business School.