1
Master Thesis
MSc Finance and MSc International Financial Management
The private company discount controlled for macro-economic and
industry variables in Germany and the United Kingdom
By Jorg Ruben Meier
s1908863
University of Groningen
Faculty of Economics and Business
First supervisor: dr. W. (Wim) Westerman
Second supervisor: prof. dr. W. (Wolfgang) Bessler
Groningen, 8
thof January 2016
2
The private company discount controlled for macro-economic and
industry variables in Germany and the United Kingdom
Jorg Ruben Meier
Abstract
This paper investigates whether the size of the private company discount decreases when controlled for macro-economic and industry variables. Disaggregating the private company discount is desired, because a more stable discount is essential in improving firm valuations. The private company discounts are derived from firm acquisitions in Germany and in the United Kingdom. The results show that a lower standard private company discount should be applied in an equity and enterprise valuation of German and English firms. Subsequently, supplements for the macro-economic and industry cycle conditions can be included. The results are more pronounced for firms in the United Kingdom.
JEL Classification: G30, G34
Keywords: Private company discount; Discount for lack of marketability; Firm valuation; Private firms; Valuation multiples
1. Introduction
3
Assets should not be categorized into only liquid and illiquid but should allow for a continuum, where the degree of illiquidity varies between assets (Damodaran 2005). BVR (2001) defines liquidity as "the ability to quickly convert property to cash or pay a liability". In general, an asset is considered to be liquid if it is marketable and has a stable market price. Firms usually lack the requirement of a stable market price due to an infrequent trade of ownership and therefore have to be described in terms of marketability. BVR (2001) defines marketability as "the ability to quickly convert property to cash at minimal cost". Paglia and Harjoto (2010) add "with a high degree of certainty of realizing the anticipated amount of proceeds" to make the definition more accurate. To elucidate, marketability can be seen as a requirement for liquidity. In other words, an asset is illiquid when it is marketable, however when an asset is illiquid, it does not inevitably mean that it is non-marketable. According to Bajaj et al. (2001), investors pay more for an asset that is easily marketable than for an identical asset that is non-marketable, and therefore a discount should reflect the relative absence of marketability. However, usual valuation methodologies do not explicitly account for the marketability of an asset. Therefore the asset is valued as marketable and reduced with a discount for the lack of marketability (DLOM) (Bajaj et al. 2001).4
controlled for these variables in the past. This paper fills that gap and derives a standard marketability discount that includes supplements for macro-economic and industry cycle conditions. Furthermore, previous research mainly focuses on the Unites States of America (USA) and only a few papers focus on Europe (Elnathan et al. 2010; Klein & Scheibel 2012). Despite the economic differences of these regions, professionals still apply the same discount level to private firms. Considering the lack of research conducted on the European market, this paper focuses on Germany and the United Kingdom (UK) and answers the following research question:"What is the size of the private company discount when controlled for macro-economic and industry variables in Germany and the United Kingdom and does the discount size change over time?"
The remainder of this paper is organized as follows. Section 2 reviews the prior literature on the marketability discount and describes the principal hypotheses. Section 3 provides a data description and outlines the model and variable construction in greater detail. Section 4 presents the empirical results of all statistical tests and the conclusion regarding the hypotheses. Section 5 concludes and offers implications for future research.
2. Literature review
2.1. The marketability discount
Many papers are devoted to investigate the marketability discount and several methodologies are used to estimate the discount. According to Albuquerque and Schroth (2015), the main reason for the many different methodologies used is that the estimation of marketability is difficult because it is very hard to measure what the price should be without the frictions. However, many papers are dedicated to conduct research with regard to the marketability discount and many different methodologies are used of which a majority will be discussed below.
2.1.1. Prior empirical estimates of the marketability discount
5
firms (Damodaran 2005; Block 2007; Paglia & Harjoto 2010). Thirdly, private equity investors can be compensated via shares because they provide advice and oversight and sometimes commit to invest additional capital once targets are achieved (Koeplin et al. 2000; Bajaj et al. 2001; Damodaran 2005; Block 2007; Paglia & Harjoto 2010). Lastly, the methodology assumes that the restricted shares cannot be transformed into a liquid security and that assumption does not hold with, for example, equity swaps and zero-cost collars (Bajaj et al. 2001). Consequently, the marketability discount is in most cases considerably smaller than estimated via this methodology.Another frequently used methodology is the initial public offerings (IPO) approach. This approach compares the price of an asset during a period when it is relatively nonmarketable, before the IPO, to a period in which it is more marketable, after the IPO. However, there are several points of criticism on the use of this methodology. Firstly, the resulting discounts appear to be incredibly large with percentages running up to 75% (Bajaj et al. 2001; Damodaran 2005). Furthermore, the methodology appears to have some real survivorship bias, because only successful firms will issue shares through an IPO, and the transactions possibly contain compensation for the services of insiders, because equity transactions before an IPO are very different types of transactions than after an IPO (Koeplin et al. 2000; Bajaj et al. 2001; Block 2007; Hall 2008; Paglia & Harjoto 2010). Because this methodology is seriously flawed and produces estimates that overstate the marketability discount it is less used nowadays.
6
2.1.2. The acquisition approachThe acquisition approach is gaining attention since its introduction in 2000 and represents an independent assessment of the private company discount (Bajaj et al. 2001). It is the most frequently used methodology when conducting research with regard to the marketability discount for a controlling interest in the firm. Moreover, previous methodologies involve only listed firms and not private firms while a more direct methodology to calculate the marketability discount is preferred (Zanni 2013). The acquisition approach introduced by Koeplin et al. (2000) compares the acquisition prices of private firms to the acquisition prices of comparable listed firms to estimate the marketability discount. First, each private firm acquisition is matched with a listed firm acquisition and then the discount is calculated by the following valuation multiples: enterprise value to sales, to earnings before interest, taxes, depreciation and amortization (EBITDA), to earnings before interest and taxes (EBIT), and to assets. The multiple approach generally serves as a reality check for derived DCF values and is a standard technique frequently used in practice. Although it does not give precise valuation inputs, it provides evidence on acquisition prices of comparable transactions (Koeplin et al. 2000). They advocate their use of the multiple method on the results of Kaplan and Ruback (1996), because they find that the DCF method yields the most reliable estimates, but that the multiple method results in the lowest valuation errors. Koeplin et al. (2000) find a discount of 20%-30% for private firms in the United States of America and a discount of 40%-50% for non-U.S. private firms.
7
management role (Koeplin et al. 2000; Bajaj et al. 2001; Block 2007; Paglia & Harjoto 2010). Lastly, private firm acquisitions are associated with higher due diligence costs due to the lack of available and reliable information and the acquisition price can be lower as a compensation for these costs (Block 2007; Capron & Shen 2007; Finnerty 2012).2.1.3. The private company discount
To solve the largest problem with the differences in characteristics between private and listed firms, Kooli et al. (2003) use reference portfolios of comparable listed firms instead of matching private firms to individual listed firms. They assume that the universe of listed deals from which a benchmark multiple is constructed is less subject to arbitrary factors and more rationally priced. This assumption is based on the advocacy of Brav et al. (2000) to use a reference portfolio over the matching firm procedure to measure abnormal returns following equity issuances. They find a substantial discount of 34% with the earnings multiple, 20% with the cash flows multiple and 17% with the sales multiple attached to private firm acquisition prices. Furthermore, their cross-sectional analysis shows that large private firms with a high growth rate have a smaller discount which indicates that the discount varies with the firm's characteristics.
Despite the firm characteristics problem is partly evaded by using portfolios, the management compensation and higher due diligence costs show that the acquisition approach is not a clean measure of the marketability discount. The downward pressure on the acquisition price might overestimate the discount estimates and therefore actually measures the upper bound of the discount which is also called the private company discount (Koeplin et al. 2000; Bajaj et al. 2001; Kooli et al. 2003; Block 2007; Paglia & Harjoto 2010). Considering only few minor adjustments are necessary with the application of the acquisition approach, it seems to be the most appropriate methodology to calculate the marketability discount.
The acquisition approach is used by Klein and Scheibel (2012) in a research on the European market instead of the market in the United States of America. They find a discount of approximately 5% for private firms, which is lower than the potential costs for an IPO, creating a valuable option of selling minority stakes in the market. Furthermore, they find that specific items, such as profitability and size, have no influence on the discount when using the acquisition approach. The smaller estimates in the European market raise the question whether it is justifiable to apply a private company discount in the valuation of private firms in Germany and the United Kingdom. Therefore, the first alternative hypothesis is as follows:
8
2.2. Industry differencesBesides different discounts applied in different regions, some papers also investigate the differences between industries within the same country. Block (2007) uses the acquisition approach as well but he uses a more current database and decomposes the discount by industry to investigate the differences between industries. He finds differences ranging from 8%-10% for financial serves and real estate firms to 30%-40% for manufacturing and construction firms. Furthermore, Paglia and Harjoto (2010) also use the acquisition approach to break down the discount by industry, but they use current market values of invested capital multiples for listed firms within the same six digit industry classification instead of trading multiples. They argue that the resulting larger sample provides a cleaner measure of the marketability discount and find that all multiples are lower for private firms, even up to 75% with the mean sales multiple. Moreover, their decomposed discount shows that the marketability discount is lower (30%-41%) for firms in the healthcare industry and higher (66%-68%) for professional services and transportation firms. Furthermore, they find that firms that with higher net sales, higher profit margins and a positive net income have a lower discount and that a decrease in the stock market for firms with smaller market capitalization leads to an increase in the discount for private firms. Block (2007) and Paglia and Harjoto (2010) find large differences in estimates which indicates that the marketability discount will also differ between firms in different industries in Germany and United Kingdom and therefore the following second alternative hypothesis is formulated:
H2: The private company discount applied in the acquisition price of private firms in Germany and the United Kingdom differs between industries.
2.3. Country differences
9
Kingdom is most often classified as an Anglo-Saxon/liberal country. Continental/conservative countries are considered to have high social expenditure because they see the welfare state as a compensator of first resort, while the Anglo-Saxon/liberal countries focus on the short-term and maximizing shareholder value and therefore give the adoption of equalizing instruments low priority (Hall & Soskice 2001). Unfortunately, not all industries in these countries are well-suited for such research. According to Boyle (2011) and Busse and Blümel (2014), the healthcare industry in respectively the United Kingdom and Germany is highly regulated, which also applies to the energy and utilities industry in Germany (OECD 2014) and in the United Kingdom (Corry et al. 2011). Furthermore, Burgstahler and Eames (2003) argue that financial services and real estate firms are highly regulated industries and therefore subject to more complex earnings management incentives. Koeplin et al. (2000) exclude financial firms and regulated utilities from their research and De Franco et al. (2011) argue that it is better to exclude regulated industries, because these might have a confounding effect on the result. A consequence might be that the industry is less marketable in general and therefore the marketability discount estimates might be exaggerated. Moreover, the natural resources industry of both countries is not taken into consideration either, because it contributes less than 1% to the total GDP which can imply that this sector is less affected by changes in the macro-economic variables (WorldBank 2015).10
sterling. People used to look at the United Kingdom as an industrial economy, but this view is inaccurate nowadays (CIA 2014). The industrial sector only contributes 20.5% to the English GDP while the professional services and transportation sector contributes 78.9%. The contribution of the agricultural sector is negligible (CIA 2014). Despite the economies of Germany and the United Kingdom seem quite comparable in general, more acquisitions occurred in the United Kingdom than in Germany, implying that the English economy is more liquid in general. Moreover, the Anglo-Saxon culture in the United Kingdom makes the management of the firm aim to maximize shareholder value, implying that they aim to minimize the marketability discount applied by professionals. Therefore the third alternative hypothesis is formulated as follows:H3: The private company discount applied in the acquisition price of private firms in Germany is larger than in the United Kingdom.
2.4. The marketability discount determinants
Although much research is conducted with regard to the appropriate level of the marketability discount, only a small number of papers are dedicated to conducting research with regard to the determinants of the marketability discount. Officer (2007) investigates whether the size of the private company discount is influenced by the illiquidity of the target firm. He finds a private company discount of 15%-30%, but also that liquidity constraints of the parent, aggregate debt market liquidity and information asymmetry are related to acquisition discounts. Since private firms do not have access to the stock market, De Franco et al. (2011) argue that this liquidity explanation overlaps with the general notion of the lack of marketability. Their research includes asset acquisitions next to stock acquisitions in the investigation of the private company discount. They find a private company discount of 20%-40% and argue that it is partially explained by the information quality facing the buyer, considering that they find a large deal value reduction when not hiring one of the Big 4 auditors. This earnings quality explanation is based on the lower quality of external monitoring and lower innate earnings quality with private firms. Furthermore, Elnathan et al. (2010) investigate whether the professionals' position (buy or sell side) matters for final firm value. They use the acquisition approach to calculate the discount, but they apply a valuation multiple provided by valuation experts when valuing the firm instead of the more regularly used trading multiple. They find evidence for a private company discount when the buyer has commissioned the valuation and that higher valuations are more common for private firms than for listed firms, when the seller is the commissioner. Therefore, they predict that the private company discount can at least be partially explained by the compliance of the expert with the interests of the valuation commissioner.
11
marketability discount. They measure the marketability discount with the restricted stock approach and use the estimates and economy-wide, industry- and firm-specific variables of Albuquerque and Schroth (2015) to calculate the appropriate discount for each case. They calculate the partial correlation between the marketability discount and the determinant and square it to see which proportion of each determinant explains the residual variation in the marketability discount. They find that the marketability discount varies significantly with the external market conditions surrounding the acquisition and that it is most highly correlated with economy-wide or business cycle determinants of liquidity. All economy-wide variables are significant, but the GDP growth and market return variables are most highly correlated with the marketability discount, because periods of high GDP growth and market return are periods of increased liquidity and lower discounts. The market volatility variable explains a smaller part of the residual variation in the marketability discount, but is more important than the proxies for funding costs and funding liquidity. Of the industry and firm-specific variables, the industry M&A activity variable is most significant, followed by the target minus industry leverage variable. Periods with high industry M&A activity seem to be related to an increased demand for industry assets and low discounts. The target volatility and industry asset specificity variables have partial correlations less than 5% but are still marginally significant, which is not the case for the industry market-to-book ratio and block(holding)-to-industry size variables. Considering the importance of economy-wide and business cycle determinants for the marketability discount, the question can be raised what the discount will be when controlled for these macro-economic and industry variables and if the marketability discount is not only a consequence of the economic cycle of the industry and country. The influence of the macro-economic and industry control variables are tested separately to prevent that the macro-economic variables affect the industry variables. Therefore the fourth and fifth alternative hypotheses are formulated as follows:H4: A lower private company discount should be applied in the acquisition price of a private firm after the discount is controlled for macro-economic variables.
H5: A lower private company discount should be applied in the acquisition price of a private firm after the discount is controlled for industry variables.
2.5. Time differences
12
H6: The size of the private company discount is constant over time after it is controlled for macro-economic and industry determinants.3. Methodology
3.1. Data description
13
Table 1Industrial distribution Germany and the United Kingdom.
Germany United Kingdom
Industry SIC Code Private Listed Private Listed
Natural resources 01-09 1 - 4 2
Manufacturing and construction* 10-33 56 32 318 62
Energy and utilities 35-39 11 10 71 22
Wholesale and retail trade* 45-47 12 17 194 36
Information technology and communication* 58-63 16 19 168 36
Financial services and real estate 64-68 4 10 91 24
Professional services and transportation* 69-84 11 17 340 66
Healthcare 86-88 8 5 88 9
Total acquisitions relevant industries 95 85 1,020 200
Total acquisitions 119 110 1,274 257
*Relevant industry
3.2. Model and variable construction 3.2.1. Model construction
Following Kooli et al. (2003), control portfolios of listed firm acquisitions are constructed, because these benchmark portfolios are more rationally priced and less subject to arbitrary factors. Moreover, Brav et al. (2000) advocate the use of a reference portfolio over the matching firm procedure to measure abnormal returns following equity issuances. The portfolios are constructed based on the revenues and contain listed firm acquisitions that are done in the same industry and have the closing date of the acquisition in the same year. If no acquisitions of comparable listed firms are done in the same year, the portfolio with the nearest closing date is used as control portfolio. After the acquisitions are matched to the portfolio with the best fit to the firm's profile, the discount is calculated as
(1) where i is the industry, k is the valuation multiple, s is the size portfolio and t is the time period.
14
enterprise valuation and based on the PE multiple for an equity valuation. The discount is positive if the trading multiple of private firms is less than the median trading multiple of listed firms and negative if the trading multiple of private firms is larger than the median trading multiple of listed firms. A negative private company discount means that a premium is applied in the acquisition price of a private firm instead of a discount. After the private company discount is derived for each private firm, the discount is controlled for macro-economic and industry variables to investigate the influence on the discount. To prevent the influence of the macro-economic variables affecting the industry variables, the impact of the macro-economic variables and industry variables are separately tested as respectively (2)where i is the industry, k is the valuation multiple, s is the size portfolio and t is the time period. (3)
where i is the industry, k is the valuation multiple, s is the size portfolio and t is the time period.
3.2.2. Variable construction
15
providers when market returns are low (Gromb & Vayanos 2002; Brunnermeier & Pedersen 2009) and investors have stronger balance sheets and are less likely to face liquidity shocks during a period of high GDP growth. Based on the results of KPMG (2014) and Albuquerque and Schroth (2015), the last macro-economic control variable is expected to have a positive relation with the discount. The yield curve slope variable acts as a proxy for the marginal cost of funding and higher cost of funding can lead to higher acquiring costs and therefore a reduced interest in private firms, which increases the discount. Similar to the yield curve slope variable, all industry control variables are expected to have a positive relation with the private company discount. Between the industry M&A activity variable and the discount a negative association can be expected as well because an increased supply of industry-specific assets can increase the discount. However, a positive association is expected based on the results of Albuquerque and Schroth (2015) and the advocacy of a high liquidity for industry-specific assets that could explain the increased M&A activity (Schlingemann et al. 2002; Ortiz-Molina & Phillips 2014). The variables industry asset specificity and industry market-to-book ratio serve as a proxy for respectively the amount of industry-specific knowledge and the time-series variation in investment opportunities in the same industry. Based on the results of KPMG (2014) and Albuquerque and Schroth (2015) more potential buyers for firms with generic productive assets and more growth options available in the industry are expected to lower the private company discount. Lastly, a higher discount is expected with a larger industry leverage ratio, because industries with higher leverage ratios are more constrained in borrowing.3.3. Descriptive statistics 3.3.1. Data characteristics
16
Table 2Descriptive statistics Germany and United Kingdom.
Germany United Kingdom
EBIT PE Revenue EBIT PE Revenue
Mean 0.22 0.25 0.09 0.25 0.21 0.31 Median 0.31 0.38 0.12 0.32 0.27 0.42 Maximum 0.98 0.99 0.97 1.00 1.00 0.98 Minimum -0.95 -0.97 -0.95 -0.98 -1.00 -1.00 Std 0.43 0.48 0.52 0.38 0.44 0.50 Skewness -0.80 -0.75 -0.19 -0.72 -0.62 -0.76 Kurtosis 3.12 2.65 1.97 3.24 2.78 2.67 N 95 95 95 1,020 1,020 1,020
17
Table 3Descriptive statistics German industries. Manufacturing and construction Wholesale and retail Information technology and communication Professional services and transportation
EBIT Revenue EBIT Revenue EBIT Revenue EBIT Revenue
Mean 0.15 0.01 0.42 0.19 0.22 0.13 0.40 0.24 Median 0.24 0.00 0.39 0.36 0.40 0.08 0.49 0.46 Maximum 0.78 0.79 0.93 0.83 0.98 0.97 0.82 0.94 Minimum -0.95 -0.95 -0.15 -0.77 -0.74 -0.65 -0.36 -0.38 Std 0.42 0.48 0.30 0.54 0.53 0.62 0.38 0.48 Skewness -0.91 -0.31 -0.10 -0.50 -0.49 0.13 -0.87 -0.06 Kurtosis 3.31 2.12 2.36 1.77 2.04 1.50 2.57 1.39 N 56 56 12 12 14 14 11 11
The numbers of discounts available in every industry in the United Kingdom is much larger than in Germany. Therefore the measurement differences are more limited than in Germany. The information technology and communication industry shows the largest discount, with a median EBIT discount of 41% and a median revenue discount of 54% and the wholesale and retail trade industry shows the lowest discount, with a median EBIT discount of 26% and a median revenue discount of 32%. The discount derived with the EBIT measurement is 26% and derived with the revenue measurement 42% in the manufacturing and construction industry. In the professional services and transportation industry the discounts are 34% with the EBIT measurement and 47% with the revenue measurement. In line with previous research, the standard deviations in almost every industry are higher than the coefficient, except for the EBIT measurement in the manufacturing and construction industry. Moreover, the skewness and kurtosis of each industry is in line with the total United Kingdom data sample, implying that the disturbances are non-normally distributed in every industry as well.
Table 4
Descriptive statistics English industries. Manufacturing and construction Wholesale and retail Information technology and communication Professional services and transportation
EBIT Revenue EBIT Revenue EBIT Revenue EBIT Revenue
18
3.3.2. Ordinary least squares (OLS) assumptionsAs tables 3 and 4 show, the skewness and kurtosis coefficients reveal that the parameters for both countries are non-normally distributed. Considering most parameters in the model are skewed and have excess kurtosis, a logarithmic specification is preferred (Brooks 2008; KPMG 2014; Albuquerque & Schroth 2015). However, the logarithm specification of the parameters does not provide very different results and therefore are omitted from all statistical tests. Furthermore, the removal of outliers does not solve the non-normality problem either, because White's heteroskedasticity test shows that non-normality is caused by heteroskedasticity. Therefore regressions with heteroskedasticity-robust standard errors are performed to gain the minimum variance among the class of unbiased estimators again (Brooks 2008). Although possible, further statistical corrections are not applied, because there is no financial theory suggesting that the relation between the variables requires a non-linear model. Tables 5 and 6 show the correlation matrix of the control variables employed in the models. Both tables show no remarkable correlations or indicators of potential multicollinearity between the control variables, because the macro-economic control variable market volatility is excluded. Statistical tests show that no other violations of the OLS assumptions, such as autocorrelation, endogeneity or a non-linear functional form are present in the model.
Table 5
Correlation matrix Germany.
19
Table 6Correlation matrix United Kingdom.
GDP growth Market return Yield curve slope Industry leverage Industry M&A activity Industry asset specificity Industry market-to-book ratio GDP growth 1 Market return 0.235 1 Yield curve slope 0.327 -0.161 1 Industry leverage -0.066 0.027 -0.089 1 Industry M&A activity -0.022 0.128 -0.063 0.125 1 Industry asset specificity -0.180 0.115 -0.297 0.087 0.024 1 Industry market -to-book ratio 0.028 0.103 -0.092 0.125 0.062 0.438 1 4. Empirical results 4.1. Main results
The standard OLS regressions are performed through the application of the respective multiples as dependent variables and the private company discount as independent variable. The resulting regression constant is considered to be the appropriate private company discount, because it is deemed as the fixed part of the discount. The private company discount can be tailored to the private firm by entering the private firm multiple and the median of the comparable portfolio multiple in the equation.
20
private company discounts in the United Kingdom are 31% with the EBIT, 25% with the PE and 49% with the revenue measurement. The corresponding standard deviations of the discount are 2% with the EBIT, 3% with the PE and 3% with the revenue measurement and the adjusted r-squared of the English model is respectively 89%, 83% and 59%. The lower predictive power of the revenue measurement indicates that it gives weaker estimations of the private company discount and therefore a lower weight should be attached to the resulting estimates.Table 7
Regressions results Germany and United Kingdom.
Germany United Kingdom
EBITDA EBIT PE Revenue EBITDA EBIT PE Revenue
Private multiple -0.10*** (0.01) -0.06*** (0.00) -0.04*** (0.00) -0.45*** (0.07) -0.09*** (0.00) -0.06*** (0.00) -0.04*** (0.00) -0.42*** (0.02) Listed multiple 0.08*** (0.01) 0.05*** (0.01) 0.02*** (0.00) 0.00*** (0.00) 0.06*** (0.00) 0.04*** (0.00) 0.03*** (0.00) 0.13*** (0.01) C 0.08 (0.06) 0.15* (0.09) 0.43*** (0.10) 0.47*** (0.07) 0.24*** (0.03) 0.31*** (0.02) 0.25*** (0.03) 0.49*** (0.03) N 95 95 95 95 1,020 1,020 1,020 1,020
Note: * denotes significance at a 90% confidence interval, ** at a 95% confidence interval, and *** at a 99% confidence interval.
4.2. Industry results
To test if the private company discount differs between industries, mean equality tests are applied on all three measurements in the German and English data sample. The results on the German data sample show that the EBIT measurement has a marginal significance of 0.09 on the F-test and the results on the other two measurements are insignificant. This weak support implies that the second null hypothesis can only be rejected based on the EBIT measurement in the German model. The support for industry differences is stronger in the English data sample, with a strong significance (p-value<0.01) on the F-test with the EBIT measurement and a medium level of significance (p-value 0.03) on the F-test with the revenue measurement, but an insignificant result on the F-test with the PE measurement. The results imply that the second null hypothesis can be rejected based on the EBIT and revenue measurement in the English model. Considering the insignificance with the PE measurement in both models, the industry regressions are only based on the EBIT and revenue measurements and measured via regressions with heteroskedasticity-robust standard errors. Furthermore, the sequence of the industry discounts is statistically tested by additional mean equality tests of each industry combination.
21
significance (p-level <0.01) and a marginal significance (p-level 0.08) and the EBIT discount in the wholesale and retail trade industry has a marginal level of significance (p-level 0.02). The EBIT and revenue discounts are 55% and 31% in the professional services and transportation industry and the EBIT discount is 25% in the wholesale and retail trade industry. The corresponding standard deviations are similar to the standard deviations in the general German model with respectively 11%, 15% and 8% much smaller than the coefficients. Although the other discounts are insignificant, the differences between the measurements in the same industry are remarkable. In the information technology and communication industry the discount derived by the EBIT measurement gives a discount of 14%, while the discount derived by the revenue measurement shows a premium of 23%. Both discounts derived in the manufacturing and construction industry show premiums as well, but considering the insignificance of these discounts no conclusions can be drawn based on these results. No additional tests are performed on the sequence of the industry discounts in Germany due to the odd coefficients and low significance levels in most industries, which are assumed to be attributable to the small data sample in each industry. The adjusted r-squares of the discounts with significance are 86% and 91% with the EBIT and revenue measurement in the professional services and transportation industry and 93% with the EBIT measurement in the wholesale and retail trade industry.Table 8
Regressions results German industries. Manufacturing and construction Wholesale and retail Information technology and communication Professional services and transportation
EBIT Revenue EBIT Revenue EBIT Revenue EBIT Revenue
Private multiple -0.07*** (0.00) -0.92*** (0.06) -0.06*** (0.01) -0.83*** (0.09) -0.05** (0.00) -0.05 (0.00) -0.09*** (0.01) -1.74* (0.17) Listed multiple 0.09*** (0.01) 0.91*** (0.08) 0.04*** (0.00) 0.81*** (0.14) 0.04*** (0.01) 0.45*** (0.06) 0.04*** (0.01) 1.21*** (0.30) C -0.20 (0.14) -0.02 (0.07) 0.25** (0.08) 0.03 (0.13) 0.14 (0.14) -0.23 (0.17) 0.55*** (0.11) 0.31* (0.15) N 56 56 12 12 14 14 11 11
Note: * denotes significance at a 90% confidence interval, ** at a 95% confidence interval, and *** at a 99% confidence interval.
22
the revenue measurement. The professional services and transportation industry shows the second largest discount of 26% with the EBIT measurement and 17% with the revenue measurement. The information technology and communication industry shows the largest discount of 28% with the EBIT measurement and 35% with the revenue measurement. Compared to the general English model, the industry constants are lower but the multiple coefficients are higher implying that more fluctuations occur within an industry than within a country. This is probably attributable to the diversification effect that occurs on a macro-economic level that mitigates any industry shock and creates a more stable discount. The results of the mean equality test with the EBIT measurements show that the largest and second largest discounts are indeed applied in the information technology and communication and in the professional services and transportation industry, but shows no support for a difference between the manufacturing and construction and the wholesale and retail trade industry. The results with the revenue measurement only show support for larger discounts applied in the information technology and communication and in the professional services and transportation industry and smaller discounts applied in the manufacturing and construction industry than in the wholesale and retail trade industry. The corresponding standard deviations of the discounts are in line with the general English model concentrating around 5% for every measurement in every industry. The smaller standard deviations present in the English model indicate that the data sample is more precise than the German model (Brooks 2008). The adjusted r-squares with the EBIT measurement range from 92% in the information technology and communication industry to 98% in the wholesale and retail trade industry, while the adjusted r-squares of the revenue measurement range from 57% in the information technology and communication industry to 91% in the manufacturing and construction industry. The predictive power of each measurement indicates that the EBIT measurement provides better estimations of the private company discount.Table 9
Regressions results English industries. Manufacturing and construction Wholesale and retail Information technology and communication Professional services and transportation
EBIT Revenue EBIT Revenue EBIT Revenue EBIT Revenue
Private multiple -0.07*** (0.00) -0.60** (0.01) -0.07*** (0.00) -0.72*** (0.03) -0.04*** (0.00) -0.27** (0.04) -0.06*** (0.00) -0.48** (0.02) Listed multiple 0.06*** (0.00) 0.52*** (0.03) 0.06*** (0.00) 0.57*** (0.04) 0.03*** (0.00) 0.09*** (0.01) 0.05*** (0.00) 0.37*** (0.02) C 0.20*** (0.03) 0.11** (0.05) 0.24*** (0.04) 0.14*** (0.05) 0.28*** (0.04) 0.35*** (0.06) 0.26*** (0.03) 0.17*** (0.04) N 318 318 194 194 168 168 340 340
23
4.3. Country resultsTo test if the same private company discount is applied in the acquisition price of private firms in Germany and in the United Kingdom, a mean equality test is applied on all three discount measurements in a combined data sample of all German and English private firm acquisitions. The results of the t-test with the EBIT and PE measurements are insignificant, but the result with the revenue measurement has a strong significance (p-value <0.01). The results imply that the third null hypothesis can be rejected based on the revenue measurement, but cannot be rejected based on the EBIT and PE measurements. The revenue results show that the coefficients of the private multiples are similar in both countries, whereas the coefficients of the listed multiples are more different between both countries. The discounts are opposing to the expectations, because the discount of 49% in the United Kingdom is larger than the discount of 47% in Germany. The opposing result is probably attributable to the weak estimations provided by the revenue measurement and the small German data sample. The standard deviation in the German model is higher than the standard deviation in the English model, but the adjusted r-squared of 35% in the German model is lower than the predictive power of 59% of the English model.
4.4. Determinants results
4.4.1. Control variable characteristics
After the presence of the private company discount is found, singular regressions including control variables are performed to test the influence on the discount. Subsequently, all significant industry control variables are added to the multiple regressions to control for the industry effects and all significant macro-economic control variables are added to a separate multiple regression to control for the macro-economic effects. As mentioned before, the macro-economic and industry variables are tested separately to prevent that the variables affect each other.
24
measurement, but are insignificant for the other measurements. The market return variable has significance with the EBIT measurement (p-value 0.01) and a marginal significance with the revenue measurement (p-value of 0.07), but is insignificant with the PE measurement. The singular regressions including an industry variable show that all industry variables are at least marginally significant in one measurement. The industry M&A activity and industry market-to-book ratio variables have a strong significance (p-values<0.01) in respectively the EBIT and revenue measurement for the industry M&A activity variable and the PE and revenue measurement for the industry market-to-book ratio variable, but the last measurement is insignificant for both variables. The industry asset specificity variable has a strong significance (p-values<0.01) in the revenue measurement and the industry leverage variable has a marginal significance (p-value 0.09) with the PE measurement, but for both variables the other two measurements are insignificant. Considering all control variables have at least a marginal significance in one measurement in Germany or in the United Kingdom, all control variables are included in the multiple regressions for both countries.4.4.2. Macro-economic variables
25
25% to 15%. The control variables GDP growth and market return have a relation with the discount, however the yield curve slope variable lack a relation with the discount again. Contrary to the expectations, the relation between the GDP growth variable and the discount is positive, implying that a higher GDP growth leads to a higher discount. Apparently, the preference for cash during a period of high GDP growth is dominant over the liquidity shock theory. In line with expectations, the relation between the discount and the market return variable shows a negative relation which indicates that a lower discount is applied when the stock market goes up. The adjusted r-squared of the German PE and English EBIT measurements do not change, but the adjusted r-squared of the German revenue measurement increases from 35% to 73% and of the English PE measurement increases with 1% to 84%. The increased adjusted r-squares imply that the macro-economic variables do add to the determination of the discount derived with the revenue measurement.Table 10
German regressions including macro-economic control variables.
Germany United Kingdom
EBIT PE Revenue EBIT PE Revenue
Private multiple -0.06*** (0.00) -0.04*** (0.00) -0.68*** (0.05) -0.06*** (0.00) -0.04*** (0.00) -0.43*** (0.02) Listed multiple 0.05*** (0.01) 0.02*** (0.00) 0.44*** (0.06) 0.04*** (0.00) 0.03*** (0.00) 0.13*** (0.01) GDP growth 2.09 (2.63) -2.00 (2.57) -7.13* (3.92) 1.32 (0.90) 7.84*** (1.83) 2.15 (2.88) Market return 0.00 (0.01) 0.00 (0.02) -0.02 (0.02) -21.57*** (7.77) -25.8* (13.29) 30.98 (21.23) Yield curve slope 18.46 (18.92) 10.37 (23.69) -10.10 (3.69) 0.00 (0.00) 0.00 (0.01) 0.00 (0.01) C 0.14 (0.10) 0.45*** (0.1) 0.33*** (0.12) 0.31*** (0.03) 0.15*** (0.04) 0.48 (0.06) N 95 95 95 1,020 1,020 1,020
Note: * denotes significance at a 90% confidence interval, ** at a 95% confidence interval, and *** at a 99% confidence interval.
4.4.3. Industry variables
27
Table 11English regressions including macro-economic control variables.
Germany United Kingdom
EBIT PE Revenue EBIT PE Revenue
Private multiple -0.06*** (0.00) -0.04*** (0.00) -0.71*** (0.05) -0.06*** (0.00) -0.04*** (0.00) -0.43*** (0.02) Listed multiple 0.05*** (0.01) 0.02*** (0.00) 0.46*** (0.05) 0.04*** (0.00) 0.03*** (0.00) 0.13*** (0.01) Industry leverage -0.33 (0.30) -0.36 (0.40) -0.15 (0.39) -0.03 (0.06) 0.13 (0.09) -0.01 (0.18) Industry M&A activity 0.14 (0.42) 0.24 (0.55) 0.04 (0.73) 0.07*** (0.02) 0.03 (0.03) 0.18*** (0.06) Industry asset specificity 0.14 (0.09) 0.25 (0.15) -0.01 (0.14) 0.03 (0.06) -0.09 (0.09) 0.34** (0.14) Industry market -to-book ratio 0.06** (0.03) 0.00 (0.04) -0.13** (0.05) -0.01 (0.01) 0.02*** (0.01) 0.03* (0.01) C 0.04 (0.10) 0.37*** (0.13) 0.43*** (0.10) 0.30*** (0.02) 0.22*** (0.03) 0.40*** (0.03) N 95 95 95 1,020 1,020 1,020
Note: * denotes significance at a 90% confidence interval, ** at a 95% confidence interval, and *** at a 99% confidence interval.
4.5. Time variables
28
5. Discussion and conclusionThis paper contributes to the growing literature regarding the private company discount by examining the size of the private company discount after it is controlled for macro-economic and industry variables in Germany and the United Kingdom and to see whether this discount changes over time. Disaggregating the private company discount is desired, because a more stable discount is essential in improving firm valuations.
First, the application of a private company discount in Germany and the United Kingdom is examined. Compared to previous research, the estimates show lower discounts with lower standard deviations, implying that the coefficients are more precise on average (Brooks 2008). The lowest weight should be attached to the revenue measurement due to its lower predictive power. In Germany, the private company discount derived with the EBIT measurement is 15% and derived with the PE measurement is 43%. In the United kingdom the private company discount derived with the EBIT measurement is 31% and derived with the PE measurement is 25%. The discount differs per industry in both countries. However the English industry discounts are considered more valuable due to the weak statistical support in the German model. Moreover, the EBIT measurement is statistically more valuable than the revenue measurement. Therefore the discount is 20% in the manufacturing and construction industry, 24% in the wholesale and retail trade industry, 26% in the professional services and transportation industry and 28% and the information technology and communication industry. Compared to previous research, the discounts and standard deviations are lower. Furthermore, the industry sequence of higher discounts applied in the information technology and communication and in the professional services and transportation industry than in the manufacturing and construction and in the wholesale and retail trade industry is similar to the industry sequence of Paglia and Harjoto (2010). A comparison of the German and English discounts shows that the same private company discount is applied in both countries, considering less weight should be attached to the only supporting measurement. The result is contradictory to the theoretical expectations, but indicates that the discount is apparently more dependent on the industry than on the home country. This might be due to the globalisation effect on 'open' economies, but future research with a broader comparison of similar and opposing countries is needed before any conclusions can be drawn.
29
between the discount and the control variables. The English discount derived with the EBIT and PE measurements are 31% and 15% when controlled for the macro-economic variables and 30% and 22% when controlled for industry variables. The English private company discounts after controlling for macro-economic and industry variables are lower, because they are partly determined by the control variables GDP growth, market return, industry M&A activity and industry market-to-book ratio.It is recommended that management applies a standard private company discount between 30% and 31% when an enterprise valuation is done and a standard private company discount between 15% and 22% when an equity valuation is done. As the private company discount can be seen as the upper bound of the marketability discount, the marketability discount cannot be higher than 31% in an enterprise valuation and 22% in an equity valuation. Subsequently, supplements for the macro-economic and industry cycle conditions can be included in which an macro-economic cycle with upward movements will result in a higher discount. However, further research with regard to the precise increase related to each variable is needed. The bankruptcy of Lehman Brothers in September 2008 and the Internet bubble in 2001 show that the size of the private company discounts in Germany and in the United Kingdom remain unstable over time. Therefore, it is recommended that the time frame of future research is extended and that this research is conducted on a more regular basis.
The results in this paper should be interpreted as a rule of thumb, because the marketability discount is likely to differ per firm. Moreover, caution is required when interpreting the results, because the results can stem from limitations of this paper. Firstly, the acquisition approach has limitations as mentioned before. Furthermore, the selection criteria set to the acquisitions reduce the data samples, while these are already small compared to the total number of acquisitions done. In addition, the English data sample is more than ten times as large as the German data sample. The small and unequal size of the data samples decreases the robustness of the test results and might have led to biases in the results. Subsequently, the classification into eight different industries based on the UK SIC code is quite debatable. The multiples in an industry classification can differ between industries and therefore decrease the robustness of the test results as well. Therefore, further research into the industry classification for the marketability discount is needed.
6. References
Albuquerque, R., Schroth, E., 2015. The value of control and the costs of illiquidity. The Journal of Finance 70, 2899-2900.
30
Bajaj, M., Denis, D.J., Ferris, S.P., Sarin, A., 2001. Firm value and marketability discounts. Journal ofCorporation Law 27, 89.
Block, S., 2007. The liquidity discount in valuing privately owned companies. Journal of Applied Finance 17, 33-40.
Boyle, S., 2011. Health systems in transition: health system review United Kingdom (England). European Observatory on Health Systems and Policies 13.
Brav, A., Geczy, C., Gompers, P.A., 2000. Is the abnormal return following equity issuances anomalous? Journal of Financial Economics 56, 209-249.
Brooks, C., 2008. Introductory econometrics for finance. Cambridge University Press, New York.
Brunnermeier, M.K., Pedersen, L.H., 2009. Market liquidity and funding liquidity. Review of Financial studies 22, 2201-2238.
Burgstahler, D.C., Eames, M.J., 2003. Earnings management to avoid losses and earnings decreases: are analysts fooled? Contemporary Accounting Research 20, 253-294.
Busse, R., Blümel, M., 2014. Health systems in transition: health system review Germany. European Observatory on Health Systems and Policies 16.
BVR, 2001. International glossary of business valuation terms. http://www.bvresources.com/freedownloads/intglossaryBVTerms2001.pdf.
Capron, L., Shen, J.-C., 2007. Acquisitions of private vs. public firms: private information, target selection, and acquirer returns. Strategic Management Journal 28, 891-911.
CIA, 2014. The world factbook. Central Intelligence Agency, Washington, DC. https://www.cia.gov/library/publications/the-world-factbook/geos/gm.html.
Corry, D., Valero, A., Van Reenen, J., 2011. UK economic performance since 1997: growth, productivity and jobs. Centre for Economic Performance, London.
31
Damodaran, A., 2005. Marketability and value: measuring the illiquidity discount. Unpublishedworking paper. Stern School of Business, New York.
De Franco, G., Gavious, I., Jin, J.Y., Richardson, G.D., 2011. Do private company targets that hire big 4 auditors receive higher proceeds? Contemporary Accounting Research 28, 215-262.
DeAngelo, H., DeAngelo, L., Wruck, K.H., 2002. Asset liquidity, debt covenants, and managerial discretion in financial distress: the collapse of LA Gear. Journal of Financial Economics 64, 3-34.
Dyck, A., Zingales, L., 2004. Private benefits of control: an international comparison. The Journal of Finance 59, 537-600.
Dustmann, C., Fitzenberger, B., Schönberg, U., Spitz-Oener, A., 2014. From sick man of Europe to economic superstar: Germany's resurgent economy. The Journal of Economic Perspectives 28, 167-188.
Elnathan, D., Gavious, I., Hauser, S., 2010. An analysis of private versus public firm valuations and the contribution of financial experts. The International Journal of Accounting 45, 387-412.
Finnerty, J.D., 2012. An average-strike put option model of the marketability discount. The Journal of Derivatives 19, 53-69.
Fuller, K., Netter, J., Stegemoller, M., 2002. What do returns to acquiring firms tell us? Evidence from firms that make many acquisitions. The Journal of Finance 57, 1763-1793.
Gromb, D., Vayanos, D., 2002. Equilibrium and welfare in markets with financially constrained arbitrageurs. Journal of financial Economics 66, 361-407.
Hall, L.S., 2008. Is there a “best” lack of marketability discount model? https://www.bvmarketdata.com/pdf/IsThereaBestDiscountModel.pdf.
32
Jindra, J., Moeller, T., 2015. Target financial independence and takeover pricing. The Journal ofFinancial Research 38, 379–413.
Kaplan, S.N., Ruback, R.S., 1996. The market pricing of cash flow forecasts: discounted cash flow vs. the method of “comparables”. Journal of Applied Corporate Finance 8, 45-60.
Klein, C., Scheibel, M., 2012. The private company discount from a European perspective: an analysis based on the acquisition approach for comparable transactions of European target companies. The Journal of Private Equity 16, 74-82.
Koeplin, J., Sarin, A., Shapiro, A.C., 2000. The private company discount. Journal of Applied Corporate Finance 12, 94-101.
Kooli, M., Kortas, M., L'her, J.-F., 2003. A new examination of the private company discount: the acquisition approach. The Journal of Private Equity 6, 48-55.
KPMG, 2014. The marketability discount of controlling blocks of shares. KPMG’s Global Valuation Institute. https://assets.kpmg.com/content/dam/kpmg/pdf/2015/09/marketability-discount-controlling-blocks-shares.pdf.
OECD, 2014. Economic survey of Germany 2014. http://www.oecd.org/eco/Germany-Overview-2014.pdf.
Officer, M.S., 2007. The price of corporate liquidity: acquisition discounts for unlisted targets. Journal of Financial Economics 83, 571-598.
Ortiz-Molina, H., Phillips, G.M., 2014. Real asset illiquidity and the cost of capital. Journal of Financial and Quantitative Analysis 49, 1-32.
Paglia, J.K., Harjoto, M., 2010. The discount for lack of marketability in privately owned companies: a multiples approach. Journal of Business Valuation and Economic Loss Analysis 5.
Reuer, J.J., Ragozzino, R., 2008. Adverse selection and M&A design: The roles of alliances and IPOs. Journal of Economic Behavior & Organization 66, 195-212.
33
Strömberg, P., 2000. Conflicts of interest and market illiquidity in bankruptcy auctions: theory andtests. The Journal of Finance 55, 2641-2692.
WorldBank, 2015. World development indicators. The World Bank Group. http://databank.worldbank.org/data/reports.aspx?source=2&type=metadata&series=NY.GDP. MKTP.CD#, visited on Saturday 7 November 2015, at 11:23.
34
7. Appendices
7.1. Appendix A
Fig. 1. Number of acquisitions per year in Germany.
35
7.2. Appendix B
Table 12
Industry classifications.
Germany United Kingdom Germany United Kingdom
Natural resources Information technology and communication*
Agriculture 1 3 Computer software 8 80
Mining - 1 Computer: semiconductors 2 2
Total 1 4 Internet / ecommerce 10 28
Media 5 101
Manufacturing and construction* Telecommunications: carriers 1 17
Automotive 13 24 Total 26 228
Biotechnology 2 7
Chemicals and materials 11 31 Financial services and real estate
Computer: hardware 1 12 Financial services 1 69
Industrial automation 4 11 Real estate 3 22
Industrial products and services 39 161 Total 4 91
Industrial: electronics 8 15
Manufacturing (other) 7 60 Professional services and transportation*
Telecommunications: hardware - 12 Transportation 4 70
Construction 6 117 Computer services 10 54
Defence - 13 Government - -
Total 91 463 Leisure 1 124
Other - -
Energy and utilities Services (other) 8 319
Energy 8 53 Total 23 567
Utilities (other) 3 18
Total 11 71 Healthcare
Medical 4 64
Wholesale and retail trade* Medical: pharmaceuticals 4 24
Consumer: foods 2 79 Total 8 88
Consumer: other 16 104
Consumer: retail 5 134
36
7.3. Appendix C
Table 13
Variable description.
Type Variable name Variable description
Multiples Enterprise value The sum of the implied equity plus the net debt of the target firm at the closing date of the acquisition. EBITDA multiple The enterprise value divided by the EBITDA of the target firm at the closing date of the acquisition. EBIT multiple The enterprise value divided by the EBIT of the target firm at the closing date of the acquisition.
PE multiple The offer price per share divided by the earnings per share at the closing date of the acquisition. If one or more values are missing the implied equity value divided by the earnings at the closing date of the acquisition.
Revenue multiple The enterprise value divided by the revenues of the target firm at the closing date of the acquisition. Macro-economic
control variables
Market return The annualized average daily return (in %) for all stocks on the CDAX and FTSE All-Share Index for the last month before the closing date of the acquisition.
GDP growth The average annual growth rate (in %) of the German and English GDP per capita in the last quarter before the closing date of the acquisition.
Yield curve slope The difference between the interest rate on the ten-year and three-month German and English government bonds for the last month before the closing date of the acquisition.
Industry control variables
Industry leverage The median proportion of long-term debt to total assets of all listed firms in the same Mergermarket sector classification and in the year prior to the year of the closing date of the acquisition.
Industry asset specificity
The median machinery and equipment to total assets ratio of all listed firms in the same Mergermarket sector classification and in the year prior to the year of the closing date of the acquisition, as defined by Strömberg (2000).
Industry M&A activity
The total M&A activity as compared to the total M&A activity (in %) in the country in the same Mergermarket sector classification in the last quarter before the closing date of the acquisition.
Industry market-to-book ratio
