ABSTRACT. Keywords: Equity premium, discount premium, private transactions, Eurozone transactions. JEL Classification: G34, G39 RESUMEN

Transcription

1 EQUITY DISCOUNT PREMIUM BETWEEN PUBLIC AND PRIVATE COMPANIES IN THE EUROZONE 33 Jaime Álvarez* y Paula Herrera** Equity discount premium between public and private companies in the Eurozone Prima de descuento entre las empresas cotizadas y no cotizadas en la Eurozona ABSTRACT Non-listed firms analyzed by investors show poor firm marketability and financial information. Due to these factors, there is a higher perceived risk and, as a consequence, the private-held firms are sold at a discount over similar listed firms and over firms that generate acceptable levels of financial information within the Euro area. This paper concludes that the discount on privately-own companies across industries is economically reasonable and statistically significant. Keywords: Equity premium, discount premium, private transactions, Eurozone transactions. JEL Classification: G34, G39 RESUMEN En el análisis de valoración de las empresas no cotizadas los inversores potenciales consideran la falta de información financiera y liquidez. Esto hace que se incremente el riesgo percibido y, por tanto, se generen primasde descuento con relación a las empresas cotizadas y sobre las que se maneja todo tipo de información dentro de la Eurozona. Este trabajo estudia las transacciones realizadas en diferentes sectores con el resultado de que dicho descuento es razonable y estadísticamente significativo. Palabras Clave: Prima de descuento, transacciones privadas, empresas no cotizadas, transacciones en la Zona Euro. Código JEL: G34, G39 Recibido: 20 de junio de 2012 Aceptado: 3 de septiembre de 2012 * Facultad de Ciencias Económicas y Empresariales. Universidad Complutense de Madrid (España). Contacto: jaime@jaimealvarez.com ** Universidad de los Andes (Colombia).

2 34 ANÁLISIS FINANCIERO 1. INTRODUCTION When valuing assets and investments, investors analyze several key factors, like risk, size, liquidity, and interest rates. Nevertheless, the process is more complex when the firm is not listed, due to a lack of liquidity, company-specific information, and, most important, an observable market price. As a consequence, specific and appropriate valuation methods should be considered. This lack of marketability and information make those firms less interesting to potential investors. Thus, the lower the level of appeal to investors in a particular stock, the higher the expected return. As a consequence, these type of firms should be sold at a discount over similar listed firms. As the discounted cash flow (DCF) model requires some parameters not available to the unlisted firm, the analyst requires a complementary model using estimations through comparable firms. A well known model to value private companies is the Relative Valuation Method (RVM), which, besides its simplicity, incorporates some market values. In this way, the value of an unlisted firm is approximated by the prices paid for similar firms, after being standardized for performance measures. The accuracy and reliability of the method is a function of how similar the comparables firms are. However, there are many factors that affect the decision to buy or sell stocks of a particular listed firm, differing considerably in terms of value creation (size, earnings, cash flows, and risk). Accordingly, the valuation process should take into account these different features. Many studies, using different measurement methods, provide evidence of a liquidity discount. Nevertheless, few studies have tried to estimate the value of a firm based on others nonstandard features besides liquidity. Koeplin et al (2000) tried to estimate what they called Private Company Discount, just for the American market. They acknowledge that any private placement discount should be valued differently than listed firms. Accordingly, the discount should be associated with other factors besides liquidity, and thus, taken into account. One way to find out the private company discount is to calculate the difference between the lower price paid by a closely held firm, and the price paid by a portfolio of listed firms, both groups with similar value features. To examine the Private Company Discount embedded in the Euro Zone, we compare historical transaction multiples of privately held companies with transactions multiples of similar publicly held firms. The values used in the multiples are Deal Equity Value and Deal Enterprise Value against Profit before and after tax, EBITDA, EBIT, Operating Income, Total Assets and Shareholders funds. This article proceeds as follows. Section I presents a literature review, Section II describes the sample data set, Section III explains the methodology, Section IV describes and analyzes the empirical results. Finally, Section V concludes. I. LITERATURE REVIEW Potential investors of unlisted firms don t have a direct access to an observable equity price posted on any platform. If the investor decides to value a specific private company by comparing it to industry-like listed firms, the accuracy of that valuation will depend on how similar are both firms in terms of value characteristics like risk, growth rate, capital structure, the size and timing of cash flows, and liquidity. However, the fundamentals of listed companies within a specific industry may differ considerably, thus a discount may be applied to account for these differences in characteristics. The most outstanding characteristic is marketability, meaning the ease at which an asset could be traded since the market is not frictionless or free of trading costs. Marketability could be defined as the degree to which an asset can be converted into cash quickly with almost no transaction costs (Bajaj et al, 2001). The estimation of the marketability discount has thoroughly been studied and measured using different models whose results differ substantially (Koeplin et al., 2000; Chiming Wu, 2010; Officer, 2006; Bajaj et al, 2001). Other studies (Hibbert et al, 2009) attempted to estimate a liquidity premium or marketability discount exclusively. In relative terms it refers to the price discount or excess return that bears an asset when it is compared against a security with higher level of liquidity, and with equivalent key characteristics like credit and market risk, sector of activity, etc. The methods to estimate the marketability discount could be grouped in i) Pre-IPO approach ii) restricted stock approach, and iii) comparable acquisitions approach.

3 EQUITY DISCOUNT PREMIUM BETWEEN PUBLIC AND PRIVATE COMPANIES IN THE EUROZONE 35 In the first approach, the discount is estimated by comparing the Initial Public Offering (IPO) price of a firm with the price of previous transactions when the firm was private. Emory (1997) found an average marketability discount of 44% (median of 43%) for transaction prices relative to the IPO prices, being the average discount steady through the period Likewise, Williamette Management Associates estimated an average marketability discount ranging from 32% to 74%. The Pre-IPO approach has several deficiencies in the estimation of the marketability discount. First, the transactions before the IPO might differ substantially from the ones embedded in the IPO itself. Pre-IPO dealings probably are issued for insiders, like managers or other investors that provide monitoring or administration services (Damodaran, 2002a). Therefore, part of the price discount for private pre- IPO transactions reflects some type of management compensation for providing those services. Second, only successful firms or the ones with high expectations of improvement carried the IPO ahead, resulting in sample bias, since unsuccessful firms are not included. Also, changes in the characteristics of values, and in the macroeconomic environment between private transactions and IPO dates, affect the prices, and thus the discount estimation. The restricted stock approach compares the difference in prices between market and restricted stocks (common shares not marketable during a period of time). The lack of liquidity of the restricted type explains the difference in prices. At Bajaj et al (2001) is shown a review of 8 authors that estimated the average marketability discount within a 20% - 30% range, inlcuiding his own estimation of 22,21%. Silber (1991) reported an average discount of 33,75% within a -12,7% - 84% range. He suggested that the discount should not be the same for all types of companies. Wruck (1989), and Hertzel & Smith (1983) found mean marketability discounts of 17,6% and 13,5% respectively 1. Nevertheless, the majority of the restricted stock issues are placed to insiders and accredited investors who typically are committed to management and monitoring activities of the firms, thus the discount reflects some compensation for services provided and monitoring costs. Additionally, Bajaj et al. (2001) found significant factors affecting the discount from a cross-sectional analysis. Those factors were percentage of shares issued over total shares after issue, business risk measured by volatility, financial distress, and total income. The comparable acquisitions approach compares the acquisition price multiples of private and public held firms. As it was the case for the other two approaches, the discount reduction accounts for more factors than just marketability. Koeplin et al (2000) used the following simple approach: They reported that private targets were, in domestic and foreign transactions, significantly smaller in net sales and assets than their public peers. Furthermore, the cumulative growth rate for the 3 years before the acquisition was in domestic transactions higher for private targets than public targets. They showed that private companies sell at a significant discount based on EBIT and EBITDA 2. In case of domestic transactions, the mean was 28.26% and 20.30% respectively. For foreign targets, average was 43.87% and 53.85%. As for sales multiples, the discount was not significant for neither domestic nor foreign transactions. Keoplin et al. (2000) suggested that to value revenue, different industries follow different procedures. Additionally, attempting to control for the difference in companies characteristics, Koeplin et al. (2000) estimated regressions separately for domestic and foreign transactions. The independent variables were size, growth, industry, and a dummy variable indicating privately / publicly held. They found that private company discount is still significant for earnings multiples. Officer (2006) compared the multiples paid for unlisted targets 3 with portfolios of comparable publicly held firms. The multiples used were price to equity book value, price to earnings, deal value to EBITDA and deal value to sales. Officer found that multiples of unlisted firms had an average discount of 15% -30% relative to the multiples paid to control related publicly held firms. Although the approach is similar to Koeplin et al. (2000), Officer attributes the discount to the lack of liquidity. A flaw in the comparable acquisitions approach is that it relies in the systematic differences in features between private and public firms, ie, closely held companies have smaller assets and revenues, and higher earnings growth rates. Additionally, Bajaj et al. (2001) show that in contrast to a public transaction of a comparable firm, private transactions for Pre-IPO, and restricted stock approaches are acquired by insiders who receive compensation for services or employment contracts,

4 36 ANÁLISIS FINANCIERO resulting in a reduction of price. Nevertheless, those differences are not relevant to our case, because the goal is to capture the total reduction in price of a private held firm compared to a public held one, and where marketability is just one of the factors accounted for. In fact, applying just a liquidity discount is only appropriate when valuing firms based on multiples of comparable transactions. To value a firm, it is common to use the discount cash flow (DCF) method along with a comparable multiples approach (CMA, also called relative valuation), because the market values similar companies in similar ways. Nevertheless, the DCF valuation method fits better to publicly held than privately held companies. This is due mainly to the difficulties to measure the market risk of unlisted firms and, therefore, their cost of capital. The less reliable these factors are, the less reliable is also the value of the estimated firm. There are two main reasons to apply the CMA to our work. First, the use of this method is quite spread among analysts of unlisted firms. Three reasons: easy to understand, fewer explicit assumptions, and the incorporation of the market factor. According to Damodaran (2002b), relative valuations yield values that are closer to the market price than DCF s. Kaplan and Ruback (1996) found that the relative method in the same industry have lower valuation errors, although the DCF method provides better reliable value estimations. However, the value of the firm should be adjusted somehow, since the characteristics of private and public firms are different. IPOs, private equity, and venture capital transactions, covering all the European Stock Exchange spectrum, and the targets represent almost all European countries. The data included and transactions of public and private held targets respectively, from September 1996 to April The data was depurated to eliminate all transactions with missing information as industry, date and deal type. The original idea was to estimate private company discounts based on matching private transaction with public acquisition within the same country, industry division, and similar time period. Unfortunately, due to the large private unmatched transactions, the sample became significantly reduced. For example, France is the country with more private and listed transactions, but only 130 pairs fit the conditions. In consequence, the results for each country could fall into a sort of selection biased, failing to be statistically significant. To overcome these obstacles, it was decided to analyze the private company discount for the total Euro Zone instead. The second reason to use the CMA instead of the DCF method is embedded in the analysis. The DCF method is not very suitable to obtain some conclusions based on hundred of firms across the Euro Zone. Apply the DCF method to so many firms would generate many drawbacks. This is due to the fact that there are many variables playing in the field: CF calculations, expectations, risks, etc., that surely would be a burden in the process of DCF valuation, generating more noise than information. The relative valuation method, on the other hand is more useful and practical, generating less errors, and, therefore, easier to identify facts and reach conclusions. II. THE SAMPLE The source of the data is Zephyr, a Bureau Van Dijk data base with information ranging from mergers and acquisitions to

5 EQUITY DISCOUNT PREMIUM BETWEEN PUBLIC AND PRIVATE COMPANIES IN THE EUROZONE 37 The sample displayed various categories. As shown in 1, the deal types can be classified in two groups regarding the controlling interest. Transactions with a controlling interest is just 1.41% of total listed companies while in closely held companies is 63.94%. We define acquisition following the Zephyr 4 as Any deal where the acquirer ends up with 50% or more of the equity of the Target is coded as an Acquisition as the Acquirer now has control of the Target. Even if the acquired stake is very small; if the final stake is 50% or above the deal is classed as an Acquisition. The Acquirer has purchased a number of shares in the Target and the resulting stake held is less than 50%. Be aware that a stake of only 2% could be classified as an acquisition if the Acquirer s overall stake reaches 50% or above. It could be expected that when new investors buy participations in unlisted firms, they have the intention to be involved directly in managing it, or indirectly, in monitoring it. On the contrary, investors in public, listed firms, usually hold a portfolio of diversified firms, letting managers and board directors to run the firms. Thus, when investors are involved in the administration of the firm, two things could occur: i) the acquired price has taken a discount to compensate management and/or monitor services; ii) a premium is paid for acquiring a controlling interest 5 and, therefore, reducing discount on the private company. Following these lines, the sample was broken down in two groups based on transactions ending or not in a controlling interest. (1 and 2).

6 38 ANÁLISIS FINANCIERO All transactions have been sort out following three dimensions: country (Europe), year ( ), and industry (classified by 2 digit NACE code list). For each dimension, the number of deals has been further broken down in controlling or not controlling interest, and private and public transactions (see Table 2). The appendix shows the results by industry. Some sectors have been eliminated to accommodate data, being reduced from 84 to 52 sectors. As shown in Table 2, these selection criteria reduced the sample to 83 and 5,310 public transactions, and 950 and 2,342 private transactions in controlling and no controlling interest respectively. Although the sample is significantly reduced mostly in the controlling interest segment, almost all countries in the Euro Zone are well covered. Table 3 reports some statistics on Operating revenue, EBIT- DA, EBIT, Profit before & after taxes, Total Assets, and Shareholders Funds for the four variables studied. The mean and median are higher for public than private companies. In fact, the mean of the listed firms is statistically different than the one of the unlisted at a significance level of 1%. The exception is Total Assets in the controlling interest sample, where the significance level is 6%. This could be attributed to a high level of variance for Total Assets, so that the mean plus one standard deviation in private transactions exceeds the mean for public targets. Although median and mean are unbiased estimators of the central tendency, the mean is highly sensitive to extreme outliers, whereas the median is hardly sensitive and, therefore, a

7 EQUITY DISCOUNT PREMIUM BETWEEN PUBLIC AND PRIVATE COMPANIES IN THE EUROZONE 39 more robust metric. Observing both metrics for all groups, there is an outstanding difference between them, and this is factored in the estimation of the private company discount. Finally, Table 3 shows a large standard deviation for most of the variables, indicating a great dispersion from the average. III. METHODOLOGY To estimate a liquidity discount for unlisted firms in the Euro Zone, we developed a method following Koeplin et al. (2000), and Officer (2006). Valuation Multiples Used The main idea is to approximate the value of a firm through the use of multiples of publicly held comparable firm with similar growth rates, industry, risk, cash flow generation, and size. This methodology standardizes the prices paid by transforming them into fundamental measures or «value characteristics» of the firm acquired. The ratios or multiples obtained are then applied to any firm being valued. For example, if the price paid for a comparable firm B (similar to A) was 100 and the net assets of B were of 50, then the multiple price to net assets for B is 2,00 ( 100 / 50). Thus, for A who has 25 of net assets, the value is 50 EUR ( 25 x 2,00), A s net assets x Multiple price to net assets of comparable firm B. Usually prices are standardized relative to revenues, book value (such as net assets or stock/equity s book value) and earnings. It is also common in some industries to use additional specific parameters such as employees, clients, etc., each one with some advantages and pitfalls (see Mascareñas, 2005). The multiples selected must be precise and accurate in reflecting the essentials of the firm, the cash flow generated by the firm, with its associated risk, and the growth rate. In other words, they should be potential measures of value for both public and private firms. However, it is hard to decide the most appropriate multiples without knowing which specific firm is going to be valued, and for the purpose of estimating the private company discount, many measures of performance should be used, and this is the reason this paper makes use of all multiples available from Zephyr. Zephyr defines Deal Equity Value as: If stated, equity value equals deal value. If not specified, the value equals the value of shares given and the estimated price per share. Deal Enterprise Value: If stated, enterprise value equals deal value. If not specified, use the following definition: equity value + short term & long term interest bearing debts cash & equivalents. When the multiple turns negative, the results have not significance. The meaning of each multiple is different due to the influence that the accounting variables have upon it. P/E multiples are influenced by the mix debt and equity, or capital structure because earnings are computed after interest expenses and taxes. Price to EBITDA multiples are influenced by capital intensity, whereas Price to EBIT by depreciation. Both, nevertheless, are independent of the capital structure. However, these multiples do not reflect how the firm is controlling costs, and it varies extensively among industries. Private Company Discount The mean and median of multiples based on all unlisted firm transactions were compared to those listed firms under the same industry and year. In comparison to Koeplin et al. (2000), who match just one private transaction with a public comparable company, or Officer (2007), with a portfolio of comparables, we take a different path. Thus, with the intention to avoid an excessive sample size reduction and selection bias, a group of similar public and private firms are compared. With the intention to only count multiples that were economically reasonable, negative parameter values and their associated multiples were deleted. The private company discount was estimated by matching listed and unlisted firms for each of the two controlling segments, and for every sector, multiple, and year. The discount was estimated by the following formula (see Koeplin et al, 2000): Hypothesis Tests We carried out the Student T Test to each multiple of unlisted firms to check if there is a statistical significance in the

8 40 ANÁLISIS FINANCIERO mean estimation. This type of test is used to judge goodness of the estimator compared to the data in the sample when the variance is unknown. In other words, to test if the private company discount is likely to occur by chance or not. The null hypothesis used for multiple i is: Ho:mean private equity discount for multiple i=0 In the other hand, the alternative hypothesis is: Ha:mean private equity discount for multiple i /= 0 Therefore, it is used a two tailed t-test expecting a rejection for the null hypothesis for a given significance level. IV. RESULTS & ANALISYS Table 4 exhibits the mean, median and standard deviation of deal equity value relative to each fundamental. It can be observed the difference between means and medians. Medians are always lower than means in both sample types (controlling and not controlling), but it is even more outstanding in private firms transactions. Thus, it can be inferred the presence of high outliers in both samples. In addition, the standard deviation is usually higher than the mean (some reach 300%); ie, see deal equity value to EBITDA in the not controlling interest, and to EBIT in the controlling interest sample respectively, reinforcing the thought of considerable outliers in the sample.

9 EQUITY DISCOUNT PREMIUM BETWEEN PUBLIC AND PRIVATE COMPANIES IN THE EUROZONE 41 The most outstanding and surprising - fact about table 4 is that these deal equity value multiples suggest that the private company discount is instead a premium. In both sample types, almost all medians and means are higher in privately held transaction multiples than in listed firms. Additionally, in private firms the standard deviation of multiples is higher. Also, when a controlling interest is acquired, the mean earnings multiples are higher than when it is not acquired, but this is not the case for the median. Also, this result is not as clear for mean transactions of private firms, although it is for the median, perhaps, reflecting the presence of outliers in transactions of private firms since means are more sensitive to them than medians. For example, only the mean of deal equity to EBIT is higher when a controlling interest is acquired. But for the median, the operating income multiple, and all earnings multiples (EBIT- DA, EBIT, and profit before and after tax) are higher when a premium is paid for the controlling interest. The statistics of deal enterprise value multiples are shown in Table 5. In this case, the sample is smaller because the corresponding data are not available for all transactions. Particularly, the smaller group in deal enterprise value multiples is in the financial sector.

10 42 ANÁLISIS FINANCIERO In this case, the statistics show quite similar results. First, medians are lower than means in both samples, but high standard deviations, particularly in privately held targets. Second, in general terms, mean and median multiples are higher in the acquisition of private firms. Third, roughly, all mean multiples are higher when a controlling interest is acquired. Table 6 shows the estimates of private company discounts for the Euro Zone. The mean and median are all negative but one. This conclusion is similar to the one found in our previous analysis of the multiples for transactions in both sample types: there is no private company discount. Instead, the results suggest a premium. Mean discounts are very large although, these cases have not statistical significance, others have and a premium of this size is not rational. This issue is addressed afterward. For the first multiple, the mean private company discounts are -766% and -626% with median of -22% and -6%, respectively. Mean (median) discounts from deal equity value to EBIT- DA reach -711% (-40%), and from deal enterprise value multiples 515% (-29%). At last, for total asset multiple, the resulting private company discounts have means (medians) of - 726% (-30%) as of deal equity value and -432% (-7%) as of deal enterprise value. Mean discounts are still very large, but medians are below 100% though in a wide range and sill negative.

11 EQUITY DISCOUNT PREMIUM BETWEEN PUBLIC AND PRIVATE COMPANIES IN THE EUROZONE 43 Considering the group Without controlling interest, only the mean discounts of operating revenue, EBIT, and total assets are statistically significant at 99% level for both, deal equity and enterprise value. Something similar can be seen in the Controlling interest acquired group. But, only deal equity value multiples to operating revenue and total assets, and deal enterprise value multiples to profit before and after tax are statistically significant with confidence level of 99%. However, for confidence levels of 95% the following mean discounts are statistically significant: deal equity value multiple to EBITDA, and deal enterprise value multiples to operating revenue, EBITDA, EBIT and total assets. Regarding shareholders funds multiples; they are significant at a level of 10% in both deal value types. Comparing both groups across fundamentals (Figure 1), every mean and median deal equity multiple is higher than its corresponding enterprise equity multiple: thin lines are below their corresponding thick lines. This is reasonable, given that usually deal equity value is lower than deal enterprise value. On the other hand, nothing can be concluded about a lower discount when a controlling interest is acquired or not. Therefore, when the stake acquired does not include a controlling interest, means & medians deal equity and enterprise value multiples all have similar patterns. But when a controlling interest is acquired, then, there is barely any correlation. At any rate, this is less obvious than in the previous case, and changes are more variable. Also, the patterns between mean deal types (for means and medians) are less similar. As it can be seen in the analysis of targets features (section II), and in the descriptive statistics of the multiples used, the fact that mean discounts are extremely larger than median discounts suggests again the presence of large outliers. But, even though medians are in the negative two digit bracket, there is no economical reason to think in a premium instead of a discount for private equity. An explanation is that this estimation of the private company discount for the Euro zone does not have any comparable value within the same country. Instead, many transactions are compared for many countries with different macro & micro economic features and risk levels. Fur-

12 44 ANÁLISIS FINANCIERO ther, it cannot be ignored the possibility of faulty data and, therefore, an erroneous estimation of multiples. With the intention of controlling some of these possible drawbacks, the private company discount was computed again, this time taking into account only the previously estimated positive private company discounts for each sector and year. exhibits the new results. Although the sample is significantly reduced, all mean private equity discounts are statistically significant at level of 1% or 5%. As expected, the discounts are positive after acquiring a controlling interest. Additionally, all means (medians) are in a range of 37% - 77% (33%-79%), much lower than those previously found. Since mean and median ranges are quite similar, then, we could assume that the private company discount for the euro zone is in that range. Figure 2 plots the results in Table 7. Mean private equity discounts (Panel A) follow similar patterns between deal types, but for median discounts, the likeness is beyond doubt. It is remarkable that, when a controlling interest is acquired, median discounts (Panel B) of the EBIT multiples are higher than 70%, while median discounts of other multiples are below 64%. Mean discounts with multiples EBIT tend to be higher for both Panels A & B. This may suggest that when buying a private firm, the market penalizes more the EBIT multiple.

13 EQUITY DISCOUNT PREMIUM BETWEEN PUBLIC AND PRIVATE COMPANIES IN THE EUROZONE 45 V. CONCLUSIONS The private company discount computed for the Euro Zone is not economically reasonable. First, a negative private company discount suggests a premium over public firms. This goes against the standard theory of a discount price paid for a firm with less liquidity, lower size, and higher risk. Second, mean values are extremely large (reaching a 25,000% premium). Thus, when applying private equity discounts to transactions of listed firms, the resulting multiples are very large. For example, the multiple deal equity value to operating revenue of a public firm is equal to 3. Applying a discount of %, the resulting multiple is 303 (3 x (1-( %)). This is not a credible multiple for any firm. On the contrary, median discounts (above -100%) are more reasonable in value, although they are still negative. The large values, and the gap between means and medians can be explained by the presence of large outliers, since the mean is more sensitive to them. These results can also be attributed to the fact that this study compares transactions across all countries in the Euro Zone. In other words, in order to compute the private company discount of a private firm, a one-toone comparison between firms in different countries should be arranged. When outliers and negative values for mean and median private company discounts are left out, then, the discount is economically reasonable and statistically significant for all deal value types and multiples. It was found that mean private company discounts (medians) range between 37% (33%) and 77% (79%). Although all countries in the Euro zone use the same currency, they differ in economic features (business cycle, GDP ) and risk levels. Besides, as we cover many countries across Europe with different accounting and fiscal systems, there is the possibility of faulty information and, therefore, miscalcu-

14 46 ANÁLISIS FINANCIERO lation of multiples in the Zephyr s database. Therefore, these two factors could account for both, the large figures, and the large percentage of the discount found. An additional limitation is that those who buy privately held firms tend to be high ranking officers and private equity funds. As they provide services like managing and monitoring, part of the discount reflects the compensation for those services. And, although this study controls for differences based in the controlling stake, no significant differences between each sample type (controlling & not) were found. REFERENCES Anson, M., 2010: Measuring a Premium for Liquidity Risk. The Journal of Private Equity. Vol 13 n 2. Spring. Pp: Bajaj, M,, Denis, D. J., Ferris, S.P., Sarin, A., 2001: Firm Value and Marketability Discounts. Journal of Corporation Law. Vol 27. Pp: Brännström, S., Färdigh, S., 2009: The myth of the private company discount - New discoveries on the pricing of private companies in Sweden. In GUPEA - Gothenburg University Publications Electronic Archive. Chiming W., 2010: Liquidity Discount in Valuing unlisted targets Japanese M&A market evidence. In BAII 2011 International Conference On Business and Information. Emory, J., 1997: The Value of Marketability as Illustrated in Initial Public Offerings of Common Stock. Business Valuation Review. Vol 16 n 3. September. Pp: Hibbert, J., Kirchner, A., Kretzschmar, G., Li, R., McNeil, A., 2009: Liquidity premium Literature Review of theoretical and empirical evidence. In Research & Insight Liquidity Premium - Barrie & Hibbert. Kaplan, S., Ruback, R., 1996: The Market Pricing of Cash Flow Forecasts: Discounted Cash Flow vs. the Method of Comparables. Journal of Applied Corporate Finance. Vol 8 n 4. Winter. Pp: Koeplin, J., Sarin, A., Shapiro, A., 2000: The Private Company Discount. Journal of Applied Corporate Finance. Vol 12 n 4. Winter. Pp: Mascareñas, J., 2005: Fusiones y Adquisiciones de Empresas. McGrow Hill. Madrid. Officer, M. S., 2006: The Price of Corporate Liquidity: Acquisition Discounts for Unlisted Targets. Journal of financial Economics. Vol. 83 n. 3. Pp: Silber, W. L., 1991: Discounts of Restricted Stock: The Impact of Illiquidity on Stock Prices. Financial Analyst Journal. Vol 47 n 4. July-August. Pp: Notes 1.- Quoted in Bajaj, et al. (2001) at EBITDA: earnings before interest, taxes, depreciation & amortization 3.- Officer (2006) defines unlisted targets as stand alone private corporations, and not listed subsidiaries of publicly held firms. 4.- Bureau Van Dijk. Glossary. In Zephyr User Guide. N&newsess=1 at Deal Classifications. Date of access: May 04, See Damodaran (2002a) at 1; Bajaj et al. (2001) at 97, and Koeplin et al (2000) at 95.

15 EQUITY DISCOUNT PREMIUM BETWEEN PUBLIC AND PRIVATE COMPANIES IN THE EUROZONE 47 APPENDIX - DEALS AVAILABLE FOR EACH ECONOMIC ACTIVITY. SOME SECTORS HAVE BEEN ELIMINATED TO ACCOMMODATE DATA