Companies
Evidence from the EU-15
OLIVER CORNELIS
1Master Thesis MScBA Finance March 2011
This study analyzes the transaction values of public and private companies in the EU-15 in the period from 2001 to 2009. EBITDA and EBIT multiples are used to compare and analyze transaction values. The major finding of this study is that EBITDA multiples of private companies are significantly higher than of public companies, but that these higher EBITDA multiples are not explained by the public status of the target company when other fundamental factors are taken into account. There is no significant difference between the EBIT multiples of public and private companies. Apart from that, the results indicate that there are significant differences between transaction values of public and private firms in a number of years, countries and industries.
Key words: transaction multiples, fundamental factors, public status effect JEL Classifications: F30; G12; L33
1
I. INTRODUCTION
In this paper I analyze if differences exist between transaction values of public and private companies. To compare and analyze the transaction values of public and private firms I use EBITDA and EBIT multiples. I test whether differences in transaction values of public and private companies are explained by growth, size, profitability, solvency and other fundamental factors, but the main research objective of this paper is to find out whether the public status of the target company is a significant fundamental factor in explaining its transaction value. In this respect, a public company is a company that is listed on a stock exchange. I use a dataset of acquisitions from the 15 countries that were members of the European Union before May 20042. Acquisitions that took place between 1 January 2001 and 31 December 2009 are selected and I use both univariate and multivariate tests to analyze differences in transaction values of public and private targets.
One of the reasons why the value of a private firm might be different than the value of a similar public firm is the relative scarce amount of information available about the private firm. This information asymmetry leads to uncertainty about the value of the private firm and this might result in lower transaction prices for private companies. Besides that, investors value the degree to which an asset can be converted to cash quickly. This means that, in general, investors will pay more for an asset that is readily marketable than for an otherwise identical asset that is not readily marketable (Bajaj et al., 2001). Shares of a public company are in general better marketable than shares of a private company and this could lead to higher prices for public companies. The owners of private companies often have a large, concentrated and illiquid position in their firm. When they invested a large part of their wealth in the firm, this illiquid position limits portfolio diversification for the owners, and this in turn increases the unsystematic risk of their portfolios. Selling the company offers them a solution to this problem and therefore the owners might be willing to sell the company at a lower price than similar public companies. Next to that, private companies might be faced by a limitation of capital availability and this has negative implications for the private firm‟s growth prospects. To overcome these capital constraints and to avoid the costs that are entailed in an IPO, the owners of the private firm might sell the company. When private companies are faced by severe capital constraints, this could lead to lower prices when they are sold. Agency
problems occur in a public firm when there is a conflict of interest between the owners and the management of the firm. If investors apply a discount for expected agency problems this leads to lower prices for public companies (Hartmann-Wendels et al. 2009). The ability of an acquisition to create economic value is another important factor that can lead to different transaction values of public and private companies. An acquisition can create value by combining a specialized resource of the target firm with a higher valued operating strategy of the bidding firm (Chatterjee, 1986). Target companies with more synergy opportunities that can be exploited are likely to have a higher transaction value. Apart from the characteristics of the target firm, the characteristics of the acquiring firm play a role in the determination of the transaction value as well. Well-diversified companies have a lower cost of capital than ill-diversified companies (Kerins et al. 2003). When well-ill-diversified public companies value a target company, a lower cost of capital leads to a higher value when future free cash flows are discounted at the cost of capital. So if a company is acquired by a well-diversified company, this leads to a higher price than when it is acquired by an ill-diversified company, because the cost of capital for the well-diversified is lower.
The results of the univariate tests show that private targets have a significantly higher EBITDA multiple than public companies and that there is no significant difference between the EBIT multiple of public and private firms. Apart from that, the univariate tests results show that there is a significant difference between EBITDA and EBIT multiples of public and private companies in a number of years, countries and industries. The multivariate analysis shows that the public status of the target company is not a significant variable in explaining transaction value when other fundamental factors are taken into account. I find that growth and solvency have a significantly positive impact on EBITDA and EBIT multiples, while profitability and size have a significantly negative impact on both multiples. Cash payments have a significantly negative impact on EBITDA multiples, but no significant impact on EBIT multiples. I also analyze year, country and industry effects and find that a number of year, country and industry dummies significantly affect transaction multiples.
literature studies the public status effect in the United States, whereas I analyze transaction values of public and private companies in 15 European countries. The main contribution of this study is that I find that the public status of the target firm does not have a significant influence on the transaction value of the target firm when other fundamental factors are taken into account. Both, academics and practitioners suggest that one should apply a discount when comparing the transaction value of a private firm to a similar public firm. This study, however, shows that the public status of the target firm does not have a significant influence on the transaction value when other fundamental factors are taken into account.
II. LITERATURE REVIEW
II.I Firm Value
The main research objective of this paper is to find out whether the public status of the target firm is a significant fundamental factor in explaining transaction values of public and private firms. The transaction values of similar public and private can be different for a number of reasons. First, I will discuss the reasons why a private company might have a lower value than a similar public firm and after that I will discuss the reasons for a higher transaction value of private companies. Moreover, I will discuss existing literature about the value of public and private firms.
lead to lower prices for private companies. To overcome these capital constraints and to avoid the costs that are entailed in an IPO, the owners of a private firm might accept a lower price. Lastly, the fact that managers of private firms are often also the owners of the company can lead to lower prices for private firms when the owners are compensated in the acquisition by other means (Koeplin et al. 2000). An example of an alternative compensation is an employment contract, when this employment contract represents above-market compensation.
Agency costs form a reason why the value of a public firm might be lower than the value of a similar private firm. Agency costs arise when there is a conflict of interest between the management and the owners of a firm. Agency problems often occur in public firms where there is a separation between ownership and control. In private firms, agency problems are often absent, because the owners are usually also in control of the company. When possible agency problems are taken into account, public firms are sold at a discount relative to similar private firms. Hartmann-Wendels et al. (2009) find that investors use firm value discounts to cope with expected agency costs. For each additional agency risk factor that investors mention within the investment proposal a firm‟s value drops by 20%.
The ability of an acquisition to create economic value is another important factor that can lead to different transaction prices for public and private companies. An acquisition can create value by combining a specialized resource of the target firm with a higher valued operating strategy of the bidding firm. The ability to create value depends on the amount of specialized resource held by the target firm relative to the total amount present in the economy, the problems of implementing the resource, and the availability of opportunities to utilize this resource (Chatterjee, 1986). Chatterjee (1986) divides possible synergies into three broad classes (i) synergies related to a lower cost of capital (ii) synergies related to lower production costs, and (iii) synergies related to the ability to charge higher prices. In general, empirical literature demonstrates that acquisitions create economic value (Jensen and Ruback, 1983). Public and private companies can have a different ability to create economic value from synergies, and this can lead to different transaction prices for public and private firms. Target companies with more synergy opportunities that can be exploited are likely to have a higher transaction price.
lower cost of capital than ill-diversified companies (Kerins et al. 2003). When a well-diversified company values a target company, the value is likely to be higher, because the future cash flows of the target company are discounted at a relatively low cost of capital. In contrast, when a company is acquired by an ill-diversified company, the transaction value is lower, because the future cash flows are discounted at a relatively high cost of capital. This is one way in which the characteristics of the acquiring firm play a role in the determination of the transaction value.
Study description Sample requirements Sample years Sample Description Results
Bajaj et al. (2001) Analysis of the marketability discount of private placements The company must be listed on the NYSE, AMEX 1990-1995 88 observations Private placements were issued at a of equity. The marketability discount is determined by or NASDAQ and private placements are taken from discount of 22% relative to the market comparing the price of privately placed shares to the the Securities Data Corporation (SDC) database and price of the firm's publicly traded shares market price of the firm's publicly traded shares verified by the Dow Jones News Retrieval Service
Block (2007) Comparison of transaction value of private and public Bid must involve at least 50% of the target's equity. 1999-2006 91 transactions The transaction value of a private company target firms. Similar public and private companies have Transaction originate from the SDC database or is in general 20-25% lower than the value the same SIC code, are from the same country and were Wall Street Journal of a similar public company. The private acquired in the same year. The transactions are also company discount is highest for manufacturing
broken down by eight industries firms and lowest for financial firms
Gupta and Misra (2010) Analysis of the equity multiples of acquisitions of banks. Acquisitions of banks originate from the SDC database. 1981-2005 1,049 acquisitions Median multiples of private targets are 17% The equity multiples of public and private target banks Target must acquire 100% of the target's equity. Deals lower than of private targets and mean are compared. were the deal value is not publicly disclosed are removed. multiples are 16% lower.
Harjoto and Paglia (2010) Analysis of the marketability discount of private companies Transactions originate from Pratt's Stats database. Private 1993-2008 674 matched-pairs Marketability discounts for private firms relative to similar public firms. Similar public and private firms have annual sales high than $ 10 million. Public firms up to 70%. Discounts for private firms companies have the same SIC code, are acquired in the must be listed on the NYSE, AMEX or NASDAQ. are highest for professional services and same year and have comparable annual sales Financial firms are excluded lowest in the healthcare sector Koeplin et al. (2000) Comparison of transaction value of private and public Bid must involve at least 50% of the target's equity 1984-1998 84 U.S. transactions Within the U.S., discounts of 20-30% for
targets firms. Similar public and private companies have equity, and all acquisitions of financial and regulated and 108 foreign private companies compared to similar the same SIC code, are from the same country and were firms are excluded. Transaction originate from transactions public companies. Outside the U.S., private
acquired in the same year SDC database firms are sold at a discount of 40-50%
compared to similar public companies Officer (2007) Comparison of transaction value of private and public Transactions originate from the SDC database. Bid must 1979-2003 12,716 succesfull and Discounts of 15% for stand-alone private
target firms. Similar public and private companies have involve at least 50% of the target's equity, deal value is unsuccesfull bids companies and of 30% for subsidiaries of the same SIC code, a deal value within the 20% range, at least $ 50 million and must be all-cash, all-common stock public firms, relative to similar public and are acquired within three years of each other or a mix of cash and common stock companies
TABLE I Overview of prior research
II.II Fundamental factors
of an acquiring firm offers stock when they believe that the firm is overvalued. This causes the market reaction to the acquisition to be negative. In contrast, when acquiring firms offer stock to acquire private companies, the asymmetric information problem is mitigated through the disclosure of private information to the target firm. Moreover, the owners of the private target firm have an incentive to assess the acquiring firm‟s offer carefully, because they will end up holding a large amount of the acquiring firm‟s stock after the acquisition. Their willingness to accept a stock offers reveals favorable information about the bidding firm to the market, resulting in a positive stock price reaction. Another reason why a stock offer might cause a positive market reaction was provided by Chang (1998). He examines bidder returns at the announcement of a takeover proposal when the target firm is a private company. Chang (1998) argues that owners of private targets may join the board of the acquiring firm, and the addition of a large block holder yields valuation gains from improved monitoring. Moreover, he finds that gains for the acquiring firm are higher in stock offers than in cash offers when a block holder is created. Another reason why it is important to include the method of payment in the model is that cash offers and stock offers have different tax implications. Cash offers generate tax obligations for the target firm‟s stockholders, whereas stock offers are, in general tax-free acquisitions, so that any capital gains realized by the target firm‟s stockholders are deferred until the stock is sold. Due to this difference in tax treatment, the bidding firm may have to pay a higher price in the case of cash offer to offset the tax burden of the selling stockholders (Wansley et al., 1983). Lastly, stockholders of a target company might prefer cash when they sell the company from a liquidity point of view and therefore they might be willing to accept a lower price when the transaction is paid with cash.
chosen from three different pools of firms: firms from the same country, firms from the same region, and firms from the 30 countries in the OECD. Their dataset contains 67,433 firm-year observations between 1993 and 2002. Dittmann and Weiner (2005) find that for most European countries value estimations are more accurate when comparable firms are chosen from the EU-15 countries or from the 30 OECD countries. For transactions in the United States, United Kingdom, Denmark or Greece, comparable firms should even be from the same country. Liu et al. (2002) find that differences in firm value across different countries might originate from differences in the level of alignment of financial and tax accounting. Industry is another fundamental factor that drives differences in transaction values of public and private companies. Different industries are likely to have different risk characteristics and earnings growth perspectives. Although I control for risk and revenue growth in my analysis, industry might still influence transaction value in a way that is not captured by other fundamentals and therefore I also control for industry. Moreover, to the extent that firm value is affected by accounting methods, selecting companies from the same industry increases the comparability of firm values, because firms in the same industry often apply similar accounting methods (Alford. 1992). The year in which an acquisition took place is the next fundamental factor that may drive transaction value. Valuations differ over time for all companies, influenced by the state of the economy.
To compare and analyze the transaction values of public and private companies, I use EBITDA and EBIT multiples. I choose to use EBITDA and EBIT multiples, because the values of EBITDA and EBIT are independent of the capital structure of the target company. This is in contrast to after-tax earnings, which are influenced by the capital structure. Model (1) is constructed in which the EBITDA and EBIT multiple are the dependent variables and the fundamental factors are the independent variables. The public status of a target company is included in the model as one of the fundamental factors to test its significance. With a significant coefficient of the public status dummy, the transaction values of public and private companies are significantly different. The model is specified below and the definitions and expected signs are presented in table II.
Variable Expected Sign Description
TVi The transaction value
GROWTHi + The expected growth rate of the target firm
PROFITi + The expected profitability of the target firm
SIZEi + Size of the target firm
SOLVi + Solvency of the target firm
METHODi + Method of payment dummy
PUBLICi - Public status dummy
YEARi +/- Year dummy
CTRYi +/- Country dummy
INDi +/- Industry dummy
TABLE II
Variables used in the empirical analysis
III. DATA
The dataset of acquisitions originates from the Zephyr database. From this database 8,069 acquisitions are obtained. From this set of transactions, 7,231 involve a private target and 838 a public target. For acquisitions involving a private target company, I use the following selection criteria: (i) the acquisition is completed between 1 January 2001 and 31 December 2009; (ii) both the acquiring firm and target firm have their head office in one of the EU-15 countries; (iii) the transaction price is at least EUR 1 million; (iv) the acquiring firm does not have a stake in the target firm prior to the acquisition and must acquire a 100% stake in the target firm. The selection criteria for acquisitions of public companies are less stringent, because applying the same selection criteria to transactions of public companies would result in a dataset with too few public companies. To overcome this problem and to make a better comparison between transaction multiples of private and public firms possible, the following selection criteria are used for acquisitions of public firms: (i) the acquisition is completed between 1 January 2001 and 31 December 2009; (ii) the target firm has its head office in one of the EU-15 countries; (iii) the transaction price is at least EUR 1 million; (iv) the acquiring firm must acquire a stake of at least 50% in the target firm (following the approach of Koeplin et al., 2000).
Table III presents the descriptive statistics of the transaction multiples. I find that the median EBITDA multiple for private firms is higher than for public firms, while the median EBIT multiple for private firms is lower. The significance of these differences will be tested in section five. The skewness statistic is positive for all multiples indicating that the distribution of the multiples is skewed to the left. The kurtosis measure shows that the distribution is leptokurtic relative to a normal distribution. The high Jarque-Bera values indicate that the transaction multiples are not normally distributed.
Total Public Private Total Public Private
Mean 13.83 11.73 14.14 22.76 22.77 22.75 Median 9.34 8.54 9.48 13.31 13.79 13.30 Minimum 1.19 1.27 1.19 1.34 2.20 1.34 Maximum 91.27 73.53 91.27 204.34 204.34 190.23 Standard Deviation 13.96 12.02 14.19 27.11 26.35 27.22 Skewness 2.69 2.71 2.67 3.05 3.12 3.04 Kurtosis 11.35 11.33 11.21 14.13 16.27 13.85 Jarque-Bera 6,607.15* 832.31* 5,619.69* 10,779.09* 1,809.52* 9,055.06 number of observations 1,608 202 1,406 1,608 202 1,406 EBITDA EBIT TABLE III
Descriptive statistics transaction multiples
This table contains descriptive statistics for the transaction multiples. The multiples are over the period 2001-2009
* significant at the 0.01 level; ** significant at the 0.05 level; *** significant at the 0.1 level
Year # public # private # total EBITDA EBIT 2001 6 16 22 8.12 16.41 2002 7 59 66 9.43 14.17 2003 10 115 125 6.07 8.69 2004 11 160 171 9.67 13.02 2005 26 223 249 10.70 13.82 2006 25 268 293 9.23 13.04 2007 48 338 386 10.35 15.59 2008 36 210 246 8.58 12.65 2009 33 17 50 6.48 10.77 Total 202 1,406 1,608 9.34 13.31 Test statistic 55.91* 40.33* p-value 0.00 0.00 TABLE IV
Distribution of transactions across years
This table shows the number of transactions and the median transaction multiple per year
Kruskall Wallis
* significant at the 0.01 level; ** significant at the 0.05 level; *** significant at the 0.1 level
Country # public # private # total EBITDA EBIT
Austria 1 4 5 9.84 18.52 Belgium 4 33 37 11.74 15.34 Denmark 7 17 24 10.33 21.02 Finland 6 25 31 7.53 11.14 France 10 212 222 8.65 12.45 Germany 45 24 69 8.23 11.94 Greece - 3 3 5.80 7.78 Ireland 1 - 1 8.14 15.63 Italy 19 105 124 8.88 14.35 Luxembourg - 1 1 8.44 19.64 The Netherlands 5 17 22 6.16 10.18 Portugal 1 14 15 12.28 22.36 Spain 11 131 142 10.44 14.53 Sweden 28 75 103 7.68 11.20 United Kingdom 64 745 809 9.73 13.00 Total 202 1,406 1,608 9.34 13.31 Test statistic 32.74* 32.42* p-value 0.00 0.00 TABLE V
Distribution of transactions across countries
Kruskal Wallis
This table shows the number of target companies and the median transaction multiples per country
Table V shows the number of transactions and the median transaction multiples per country. The table shows that around 50% of the target companies are from the United Kingdom and that there are very few firms from Greece, Ireland and Luxembourg. Germany is the only country with more public than private target companies in the dataset. The Kruskal Wallis test shows that the differences in multiples between countries are significant.
Industry # public # private # total EBITDA EBIT
Agriculture - 3 3 6.22 8.12
Chemicals and materials 3 28 31 9.19 13.22
Business and industrial products 20 126 146 7.44 10.34
Business and industrial services 35 276 311 9.03 12.22
Construction 20 136 156 7.50 10.37
Transportation 8 68 76 7.81 12.50
Consumer goods and retail 23 223 246 8.49 12.81
Consumer services: other 12 82 94 11.96 18.37
Energy and environment 7 38 45 11.25 21.28
Financial services 21 71 92 10.03 13.05
Real Estate 10 33 43 14.63 19.69
Communications 18 152 170 9.81 14.36
Computer and electronics 22 110 132 10.82 15.69
Life sciences 3 60 63 12.66 19.23
Total 202 1,406 1,608 9.34 13.31
Test statistic 77.57* 73.09*
p-value 0.00 0.00
TABLE VI
Distribution of transactions across industries
This table shows the number of target companies and median transaction multiples per industry
Kruskal Wallis
* significant at the 0.01 level; ** significant at the 0.05 level; *** significant at the 0.1 level
multiples of transaction paid alternatively. There is no significant difference between the EBIT multiple of transactions paid with cash and transactions paid alternatively.
Method of paymemt # public # private # total EBITDA EBIT
Cash 149 531 680 9.10 12.87 Alternative 53 875 928 9.46 13.84 Total 202 1,406 1,608 9.34 13.31 Test statistic 2.55* 1.49 p-value 0.01 0.13 Mann-Whitney
This table shows the number of transactions and median transaction multiples per method of payment TABLE VII
Distribution of transactions across method of payment
* significant at the 0.01 level; ** significant at the 0.05 level; *** significant at the 0.1 level
The compounded annual growth rate (CAGR) of operating revenue over the last three years prior to the acquisition is the proxy for the expected future growth rate of a target firm. The CAGR is used because it decreases the effect of volatility of periodic returns. Following Dittmann and Weiner (2005) and Gupta and Misra (2010), the proxy of the expected future profitability is the Return On Assets (ROA) in which the ROA is defined as profit before tax divided by total assets. In contrast to growth, I choose to use the ROA in the last year before the acquisition, because profitability is less volatile than revenue. I proxy size by the logarithm of total assets one year before the target company was acquired (King and Segal, 2003). Lastly, the solvency of a firm is determined by the solvency ratio of the target company one year before it was acquired. The solvency ratio is defined as total shareholders‟ funds divided by total assets. The operating revenue, return on assets, total assets and solvency ratio figures are obtained from Amadeus. Table IX contains the descriptive statistics for the firm specific fundamental factors. The high Jarque-Bera statistics for all firm specific fundamentals indicate that they are not normally distributed.
Growth Profitability Size Solvency Mean 0.18 0.14 4.29 0.37 Median 0.09 0.10 4.20 0.36 Minimum -0.60 -0.65 2.29 -0.82 Maximum 4.29 0.97 7.03 1.00 Standard Deviation 0.36 0.15 0.00 0.24 Skewness 3.86 1.53 0.77 -0.27 Kurtosis 26.25 8.05 3.16 4.21 Jarque-Bera 40,221.78* 2,341.52* 64.17* 117.39* number of observations 1,608 1,608 1,608 1,608
over the past three years, size is measured by the logarithm of total assets, profitability by Return On Assets and solvency by the solvency ratio
TABLE IX
Descriptive statistics fundamental factors
This table contains desciptive statistics for growth, profitability, size and solvency in the period 2001- 2009. Growth is defined as the CAGR of operating revenue
* significant at the 0.01 level; ** significant at the 0.05 level; *** significant at the 0.1 level
The Pearson correlation coefficients for the explanatory variables are presented in Appendix I. The coefficients of most of the variables are statistically insignificant. However, the correlation of profitability with size and solvency might cause problems when they are included in the same analysis. The correlation of profitability with size and solvency can be explained by the fact that these three variables are all related to the total assets of the target firm. The next section discusses the problem of multicollinearity in more detail.
IV. METHODOLOGY
In this section, I explain the methodology that I use to analyze the transaction values. I use both univariate and multivariate analyses, to analyze the differences in transaction values of public and private companies in the EU-15 in the period from 1 January 2001 to 31 December 2009.
IV.I Univariate Analysis
To test if differences exist between transaction values of public and private targets the Mann-Whitney U test is used. I use the Mann-Mann-Whitney U test, because this test does not rely on the assumption of normality and the descriptive statistics in the previous section show that the transaction multiples are not normally distributed. The median values of the EBITDA and EBIT multiples of public and private target firms are computed and then analyzed by the Mann-Whitney U test. The Mann-Whitney U test, tests the null hypothesis that the medians of the two subsamples in the test are the same. I also use the Mann-Whitney U test to test for differences between fundamental factors of public and private firms. Lastly, the Mann-Whitney U test is also used to analyze the transaction values of public and private companies per year, country and industry.
IV.II Multivariate analysis
A multivariate analysis are performed to test whether differences in transaction multiples exist when fundamental factors are taken into account. Compared to the univariate test, the multivariate test adds explanatory power, because it controls for the interaction between fundamental factors. The Mann-Whitney U tests only tests if there are differences between transaction multiples of public and private companies and whether these differences are significant. A multivariate analysis, on the other hand, relates differences in transaction multiples directly to differences in fundamental factors.
payment and public status together with one year, country or industry dummy. This means that a separate regression analysis is run for every year, country and industry. The transaction multiples are estimated by the Ordinary Least Squares (OLS) method. The OLS has a number of assumptions that are listed in table X.
Technical notation Interpretation
(1) E(ut) = 0 The errors have zero mean
(2) var (ut) = σ 2
< ∞ The variance of the errors is constant and finite over all values of xt
(3) cov (ui, uj)= 0 The errors are linearly independent of one another
(4) cov (ut, xt) = 0 There is no relationship between the error and corresponding x variate
(5) ut= N(0, σ2) Ut is normally distributed
to the Ordinary Least Squares (OLS) method TABLE X
Assumptions of the OLS method
This table contains the assumtions that are made relating
term, the OLS estimator will not be consistent, because the results assign explanatory power to the explanatory variables, where in reality the explanatory power is arising from the correlation between the error term and the dependent variable (Brooks, 2002). The last assumption is that the error terms are normally distributed. When the error terms are not normally distributed, a possible solution could be to delete outliers from the dataset.
V. RESULTS
V.I Univariate Test Results
I run univariate tests to analyze whether the median EBITDA and EBIT transaction multiples of public and private target firms significantly differ. Apart from that, I also run a univariate analysis to determine whether the median values of growth, profitability, size and solvency differ significantly between public and private targets. The results of these tests are presented in table XI. The Mann-Whitney statistic in table XI indicates that EBITDA transaction multiples of private targets are significantly higher than EBITDA multiples of public targets. On the other hand, EBIT multiples of public targets are higher than of private targets but this difference is not significant. Table XI also presents the results of the Mann-Whitney U tests that are run to determine the significance of the difference between median values of growth, profitability, size and solvency of public and private targets. The results indicate that there is no significant difference between the median growth rates of public and private targets. There is, however, a significant difference between the median values of profitability, size and solvency of public and private targets. The median public firm is significantly larger and has a significantly higher solvency ratio. Private firms, on the other hand, have a significantly higher median value of profitability. From the results in table XI we can derive that the significantly higher EBITDA multiples of private firms could be explained by the fact that the median private firm is significantly more profitable. In contrast, the fact that the median private company is significantly smaller and less solvent should have a negative effect on the transaction multiples of private firms, according to the literature.
EBITDA EBIT Growth Profitability Size Solvency
Public targets 8.54 13.79 0.07 0.05 5.13 0.41
Private targets 9.48 13.30 0.09 0.11 4.10 0.35
Test statistic 3.14* 0.32 0.50 8.09* 15.87* 2.73*
p-value 0.00 0.75 0.62 0.00 0.00 0.01
Mann-Whitney
determine whether the values of public and private targets are significantly different TABLE XI
Univariate test results
This table contains the median values of the EBITDA and EBIT multiple, growth, profitability size and solvency of public and private targets. The Mann-Whitney test is used to
I also analyze the transaction multiples of public and private companies per year, country and industry, because I might miss valuable information when I only analyze the multiples in the complete dataset. The Mann-Whitney U test is run to determine whether EBITDA and EBIT multiples of public and private targets differ significantly in a certain year, country or industry. The results in table XII indicate that private targets have a significantly higher EBITDA multiple in 2005 and 2009 and a significantly higher EBIT multiple in 2009. In 2007, however, the EBIT multiples of private firms are significantly lower than of public firms. Table XII also shows that EBITDA and EBIT multiples of private firms are significantly higher in Denmark, while they are significantly lower in Finland and Spain. In The Netherlands only the EBITDA multiple of private targets is significantly lower than of public targets. Lastly, table XII also shows the results of comparison of transaction multiples of private and public firms per industry. EBITDA en EBIT multiples of private firms are significantly lower in Construction, while they are significantly higher in Real estate. EBITDA multiples of private firms are also higher in Chemicals and materials, Business and industrial services, Transportation, Consumer services: other, Energy and environment and Life sciences. EBIT multiples of private firms, on the other hand, are significantly smaller in Financial Services. The actual EBITDA and EBIT multiples of public and private targets per year, country and industry are presented in Appendix II.
Year EBITDA EBIT Country EBITDA EBIT Industry EBITDA EBIT
2001 0 0 Austria 0 0 Agriculture 0 0
2002 0 0 Belgium 0 0 Chemicals and materials + 0
2003 0 0 Denmark + + Business and industrial products 0 0 2004 0 0 Finland - - Business and industrial services + 0
2005 + 0 France 0 0 Construction -
-2006 0 0 Germany 0 0 Transportation + 0
2007 0 - Greece 0 0 Consumer goods and retail 0 0
2008 0 0 Ireland 0 0 Consumer services: other + 0
2009 + + Italy 0 0 Energy and environment + 0
Luxembourg 0 0 Financial services 0
-Netherlands - 0 Real estate + +
Portugal 0 0 Communications 0 0
Spain - - Computer and electronics 0 0
Sweden 0 0 Life sciences + 0
United Kingdom 0 0 TABLE XII
This table shows the results of the univariate tests that are run to analyze the multiples of public and private companies per year per country and industry. Where - stands for a significantly lower multiple for private targets, 0 stands for not significant
V.II Multivariate Test Results
I run multivariate tests to analyze transaction multiples of public and private companies when fundamental factors are taken into account. First, I present the results of the regression analysis that includes growth, profitability, size, solvency, method of payment and public status. After that, the results of the year, country and industry dummy variables are presented and discussed.
Table XII presents the estimation results of model (1) that is estimated using OLS. The coefficient for growth is significantly positive in both the EBITDA and EBIT multiple model and this positive relation is consistent with the theory that as long as the Return On Invested Capital (ROIC) of a firm is higher than its Weighted Average Cost of Capital (WACC), higher growth results in a higher transaction value (Koller et al., 2005). Table XII also shows that solvency has significantly positive impact on EBITDA and EBIT transaction multiples. This significantly positive coefficients means that target firms with higher solvency ratios, have higher transaction values. The results of the solvency coefficients are as expected and are consistent with the theory that a higher solvency ratio leads to lower risk levels and that lower risk levels result in higher transaction values. The method of payment dummy variable has a negative coefficient in both the EBITDA and EBIT multiple model, but only the coefficient in the EBITDA multiple model is significant. The negative coefficient of the method of payment dummy is as expected and means that cash offers lead to significantly lower EBITDA transaction multiples. The R2 statistic for the EBITDA model is 0.05 and for the EBIT model it is 0.08.
coefficient of size in the EBITDA and EBIT model, however, implies that larger firms have lower transaction multiples. Damodaran, however, does not find a clear relationship between firm size and multiples in his analysis of all publicly traded company in the United States as of January 2011.
Year EBITDA EBIT
Constant 25.92* 47.35* Growth 4.51* 3.28*** Profitability -17.41* -58.67* Size -2.68* -4.32* Solvency 3.29*** 6.83*** Public status 0.55 -0.35 Method of payment -1.47** 1.29 R2 0.05 0.08 TABLE XIII Multivariate test results
variable is the transaction multiple. The year, country and industry dummies are not included in this regression. Model (1) is estimated using White's heteroskedasticity-consistent standard errors
This table shows the results of model (1), which is estimated using OLS. The dependent
* significant at the 0.01 level; ** significant at the 0.05 level; *** significant at the 0.1 level
analyze the transaction values of private and public companies controlling for a number of fundamental factors.
Year EBITDA EBIT Country EBITDA EBIT Industry EBITDA EBIT
2001 0 0 Austria 0 0 Agriculture 0
-2002 0 0 Belgium 0 0 Chemicals and materials 0
-2003 - - Denmark 0 + Business and industrial products -
-2004 0 0 Finland 0 0 Business and industrial services -
-2005 0 0 France 0 0 Construction 0
-2006 0 0 Germany 0 0 Transportation 0 0
2007 + 0 Greece - 0 Consumer goods and retail 0
-2008 0 0 Ireland 0 0 Consumer services: other + +
2009 0 - Italy 0 0 Energy and environment 0 0
Luxembourg - - Financial services + 0
Netherlands - 0 Real estate + +
Portugal 0 0 Communications 0 +
Spain 0 + Computer and electronics + 0
Sweden - - Life sciences + +
United Kingdom 0 0
regression is run in which growth, profitability, size, solvency, the public status and the method of payment are also included
TABLE XIV
This table shows the results of the multivariate tests that are run to analyze the multiples of public and private companies per year, per country and industry. Where - stands for a significantly negative dummy coefficient, 0 stands for not significant
and + stands for a significantly positive dummy coefficient. For every year, country and industry a separate regression
Multivariate test results per year, country and industry
VI. CONCLUSION
In this study I analyze transaction values of public and private companies. The value of a private and public might differ for a number of reasons. For example, the lack of marketability of the private firm can lead to a private company discount, while the presence of agency problems in public firms can result in lower transaction values for public companies. In this study I analyze whether the public status of a target company is a significant factor in explaining its transaction value.
The univariate tests show that private companies have significantly higher EBITDA multiples than public firms, but that there is no significant difference between the EBIT multiples of public and private firms. I also find that EBITDA and EBIT transaction multiples significantly differ in a number of years, countries and industries. Next to that, the results of the univariate tests show that the median private firm is significantly more profitable, smaller and less solvent. Multivariate tests are run to analyze whether the differences in transaction values of public and private firms can be explained by fundamental factors and whether the public status of the target firm is a significant fundamental factor. The multivariate tests results indicate that the public status of the target company is not a significant fundamental factor in explaining EBITDA and EBIT transaction multiples. The multivariate tests also show that growth and solvency have a significantly positive impact on EBITDA and EBIT multiples and that profitability and size have a significantly negative impact on both multiples. A separate regression is run for every year, country and industry dummy variable and I find that a number of year, country and industry dummy variables are significant in explaining transaction multiples.
REFERENCES
Alford, A.W., 1992, “The Effect of the Set of Comparable Firms on the Accuracy of the Price-Earnings Valuation Method”, Journal of Accounting Research, 30 (1), 94-108
Bajaj, M., Denis D.J., Ferris, S.F., Sarin, A., 2001, “Firm Value and Marketability
Discounts”, working paper, SSRN
Block, S., 2007, “The Liquidity Discount in Valuing Privately Owned Companies”, Journal
of Applied Finance, Fall/Winter 2007, 33-40
Bhojraj, S., Lee, C.M.C., 2003, “International Valuation Using Smart Multiples”. Working
paper, Cornell University, Ithaca
Brooks, C., 2002, Introductory Econometrics for Finance, 1st edition, Cambridge University
Press, Cambridge
Chang, S., 1998, “Takeover of Privately Held Targets, Methods of Payment, and Bidder Returns”. Journal of Finance, 53 (2), 773-784
Chatterjee, S., 1986, “Types of Synergy and Economic Value: The Impact of Acquisitions on Merging and Rival Firms”, Strategic Management Journal, 7, 119-139
Damodaran, A., 2002, Investment Valuation, John Wiley & Sons, New York
Damodaran, A. (Website) visited at: http://pages.stern.nyu.edu/~adamodar/.
Dittmann, I., Weiner, C., 2005, “Selecting Comparables for the Valuation of European Firms”, working paper, Humboldt University, Berlin
Faccio, M., McConnell, J.J., Stolin, D., 2006, “Returns to Acquirers of Public and Private Targets”, Journal of Financial and Quantative Analysis, 41 (1), 197-220
Gupta, A., Misra, L., 2010, “Listing Effects and the Private Company Discount in Bank Acquisitions”, Handbook of Quantitative Finance and Risk Management, Springer Verlag Hansen, R.G., Lott, J.R., 1996, “Externalities and Corporate Objectives in a World with
Diversified Shareholder/Consumers”, Journal of Financial and Quantitative Analysis, 31 (1), 43-68
Harjoto, M.A., Paglia, J.K., 2010, “The Discount for Lack of Marketability in Privately Owned Companies: A Multiples Approach”, Journal of Business Valuation and Economic
Loss Analysis, 5 (1), Article 5
Hartmann-Wendels, T., Keienburg, G., Sievers, S., 2009, “Agency Risk and Firm Valuation:
An Empirical Analysis of Venture Capitalists‟ Private Expectations”, working paper, SSRN Jensen, M.C., 1989, “Eclipse of the Public Corporation”, Harvard Business Review, 67, 61-75 Jensen, M.C., Ruback, R.S., 1983, “The Market For Corporate Control: The Scientific
Evidence”, Journal of Financial Economics, 11, 5-50
Kerins, F., Smith, J., Smith, R, 2003, “Opportunity Cost of Capital for Venture Capital Investors and Entrepreneurs”, Journal of Financial and Quantitative Analysis, 39 (2), 385-405
Kim, M., Ritter, J.R., 2003, “Valuing IPOs”, Journal of Financial Economics, 53, (3), 409-437
King, M.R., Segal, D., 2003, “Valuation of Canadian- vs. U.S.-Public Equity: Is There a Discount? “, Bank of Canada working paper, 2003-6
Koeplin, J., Sarin, A., Shapiro, A.C., 2000, “The Private Company Discount”, Journal of
Applied Corporate Finance, 12 (4), 94-101
La Porta, R., Lopez-de-Silanes, F., Shleifer, A. Vishny, R., 1998, “Law and Finance”, Journal
of Political Economy, 106 (6), 1113-1155
Lie, E., Lie, H.J., 2002, “Multiples Used to Estimate Corporate Value”, Financial Analysts
Journal, 58 (2), 44-54
Liu, J., Nissim, D., Thomas, J., 2002, “Equity Valuation Using Multiples”, Journal of
Accounting Research, 40, 135-172
Myers, S., Majluf, N., 1984, “Corporate Financing and Investment Decisions when Firms have Information that Investors do not have”, Journal of Financial Economic, 13, 187-221 Officer, M., 2007, “The price of corporate liquidity: Acquisition discounts for private targets”,
Journal of Financial Economics, 83, 571-598
Officer, M., Poulsen, A.B., Stegemoller M., “Target-firm information asymmetry and
acquirer returns”, Review of Finance Forthcoming
Schreiner, A., Spremann, K., 2007, “Multiples and Their Valuation Accuracy in European Equity Markets”, working paper, University of St. Gallen, St. Gallen.
Travlos, N.G., 1987, “Corporate Takeover Bids, Methods of Payment, and Bidding Firms‟ Stock Returns”, Journal of Finance, 42 (4), 943-963
APPENDIX I: Correlation matrix
Growth Profitability Size Solvency
Growth 1.00
Profitability 0.10 1.00
Size -0.04 -0.41* 1.00
Solvency -0.11 0.23 -0.08 1.00
TABLE 1.1
This table shows the Pearson correlation coefficients of the explanatory variables
APPENDIX II: Transaction multiples per year, country and industry
Year
public private public private
2001 7.54 8.12 11.19 17.43 2002 11.96 8.66 16.39 13.35 2003 5.70 6.13 14.36 8.25 2004 5.37 9.78 12.71 13.26 2005 9.69 10.70 13.57 13.99 2006 9.34 9.15 13.31 13.02 2007 12.01 10.19 18.28 14.86 2008 8.07 8.93 11.63 12.79 2009 3.98 12.41 7.41 19.91 Total 8.54 9.48 13.79 13.30 TABLE 2.1
Public and private multiples per year
This table shows the median transaction multiples of public and private companies per year
EBITDA EBIT
Country
public private public private
Austria 4.46 19.99 7.91 24.33 Belgium 13.07 11.74 35.86 15.34 Denmark 4.03 13.47 11.86 22.90 Finland 30.99 6.57 56.01 8.70 France 8.42 8.67 10.68 12.66 Germany 5.78 9.52 11.94 12.56 Greece - 5.80 - 7.78 Ireland 8.14 - 15.63 -Italy 7.90 9.06 16.71 14.15 Luxembourg - 8.44 - 19.64 The Netherlands 11.96 5.97 16.74 9.13 Portugal 3.50 12.54 6.85 23.95 Spain 15.96 10.16 21.29 14.20 Sweden 6.60 7.69 11.54 10.87 United Kingdom 9.66 9.73 13.15 13.00 Total 8.54 9.48 13.79 13.30 TABLE 2.2
Public and private multiples per country
This table shows the median transaction multiples of public and private companies per country
Country
public private public private
Agriculture - 6.22 - 8.12
Chemicals and materials 4.46 10.36 11.44 14.62 Business and industrial products 6.60 7.52 11.31 10.26 Business and industrial services 7.95 9.21 11.77 12.47
Construction 11.66 7.23 18.21 9.60
Transportation 5.54 8.44 13.25 21.50 Consumer goods and retail 8.49 8.49 11.95 12.92 Consumer services: other 7.83 12.39 14.81 18.52 Energy and environment 6.96 12.84 19.50 21.36 Financial services 11.77 9.66 18.22 12.05
Real Estate 7.80 17.44 9.17 21.40
Communications 7.44 9.90 16.72 14.00 Computer and electronics 10.15 11.67 16.89 14.67 Life sciences 7.63 13.05 13.81 19.29
Total 8.54 9.48 13.79 13.30
TABLE 2.3
Public and private multiples per industry
This table shows the median transaction multiples of public and private companies per industry
APPENDIX III: Coefficients year, country and industry dummy variables
Year EBITDA EBIT Country EBITDA EBIT Industry EBITDA EBIT
2001 -2.57 3.21 Austria 5.40 26.00 Agriculture -5.71 -15.68*
2002 0.71 -1.65 Belgium 0.96 6.52 Chemicals and materials -0.41 -6.59* 2003 -4.96* -5.97* Denmark 5.61 13.18*** Business and industrial products -4.51* -6.19* 2004 -0.62 -0.51 Finland -0.97 -2.68 Business and industrial services -1.79** -3.84*
2005 0.77 -0.65 France 0.74 -0.25 Construction -1.01 -4.40**
2006 -0.49 2.52 Germany -1.47 -2.92 Transportation -1.26 6.65
2007 1.59*** 1.11 Greece -6.57*** -10.00 Consumer goods and retail -2.75 -3.89* 2008 0.19 1.32 Ireland -0,58 -2.23 Consumer services: other 2.94*** 5.94*** 2009 -2.09 -5.95*** Italy -1.80 -0.55 Energy and environment 2.36 8.59
Luxembourg -5.36* -6.86* Financial services 3.79** 2.89 Netherlands -3.43*** -2.89 Real Estate 9.68* 8.21***
Portugal 4.36 15.23 Communications 1.74 4.20*
Spain 2.52 6.12** Computer and electronics 2.35*** 2.43 Sweden -4.25* -7.44* Life sciences 4.03** 9.10** United Kingdom 0.27 -1.38
White's heteroskedasticity-consistent standard errors
TABLE 3.1
This table shows the coefficients of the dummy variables for year, country and industry. The coefficients are estimated by running a separate regression for every year, country and industry in which growth, profitability, size,
solvency, the public status and the method of payment are also included. The coefficients are estimated using
Coefficients year, country and industry dummies
