Private Equity investments in the Netherlands
An analysis of the factors that explain the impact on shareholder wealth
Author: Duco Simon Luitse
-University of Groningen- Faculty of Economics and Business
Master Business Administration, specialization Finance
2
Private Equity investments in the Netherlands
An analysis of the factors that explain the impact on shareholder wealth
Duco Simon Luitse1
-Abstract-
This study examines the relation between private equity (PE) investments in Dutch listed companies and the shareholder wealth effects. PE firms that take a minority share
(between 5 percent till 50 percent) experience the distinct corporate governance structure in the Netherlands with regard to the protection of minority shareholders. Simultaneously they can benefit from these rights. This is the first study to investigate the effects of PE investments on shareholders’ wealth in the Netherlands. An unique dataset was
constructed comprising 95 PE investments found in the period from January 1999 till November 2010 and was tested via an event study methodology. This study shows a significant positive effect of PE investments in Dutch listed companies of 1.216% around the event date (t = -1 till t = +1). Through ordinary least squares regression analyses the effect of different variables on the wealth effects is measured and distinguish between: the size of a stake, the age of PE firm, family ownership, PE country of residence, PE reputation and undervaluation of target firm. The results show that the main sources of shareholder wealth gains are reputation, as measured by the size of the total capital invested by a PE firm, and the PE country of residence.
Key words: wealth effects, event study, private equity, Netherlands
3
Preface
This paper will be my final work at the University of Groningen. I look back on a
memorable time in Groningen. I made new friends for life, learned many new things, but moreover this was the place where I have grown as a person. I could not have succeeded the Msc BA - Finance without the support of a few people. First my loving parents, who never lost faith in me, and were my biggest supporters along the way. Second, my friends and family. And last but not least, my girlfriend, who’s loving support helped me through the hard times.
The support of dr. W. Westerman was vital in the completion of this paper. I would like to thank him for his ever valuable and constructive directions, but also look back on pleasant conversations which rose above the content of this paper.
4
Table of content
Page
Table of content with respect to Tables and figures ...5
Part 1. Introduction... 6
Part 2. Literature review... 9
2.1 PE, a short introduction... 9
2.2 Corporate governance structure in the Netherlands... 12
2.3 Previous literature on PE shareholder activism... 14
2.4 Dutch target firm, PE firm ...17
Part 3. Data... 22
Part 4. Methodology... 26
4.1 Event study methodology... 26
4.1.1 Market Model... 27
4.2 Tests... 29
4.2.1 Normality test... 30
4.2.2 Parametric test... 31
4.2.3 Nonparametric Tests... 32
4.3 Ordinary Least Squares Regressions... 33
Part 5. Results...35
5.1 Event study results... 35
5.2 Result of OLS regression analysis... 37
Part 6. Conclusions and recommendations... 42
6.1 Conclusions... 42
6.2 Recommendations... 45
Part 7. References... 47
5
Table of content with respect to Tables and figures
Page
Tables
1. Top 5 biggest global PE deals over the past 25 years... 10
2. Top 5 biggest Dutch PE deals over the past 25 years ... 10
3. Overview of venture capital firms, private equity firms and hedge funds... 11
4. Shareholders’ rights in the Netherlands... 14
5. Summary of the results of previous event studies... 16
6. The distribution of the stakes... 23
7. Country of origin of the PE firms investing in Dutch listed firms... 25
8. Characteristics for 95 initial PE investments in Dutch listed companies... 25
9. Results of the normality test... 31
10. The cumulative return for different event periods... 36
11. Correlations between PE firm, target firm and deal characteristics... 38
12. Results of Ordinal Least Square Regression Analysis... 41
13. Overview of hypotheses and results... 43
14. Daily average abnormal returns and cumulative average abnormal returns... 52
15. Results of Ordinal Least Square Regression Analysis (Step-wise)... 53
Figures 1. Distributions of the announcements by year... 24
6
1. Introduction
In a recent published study by the International Center for Finance Research of the IESE Business School in association with Ernst&Young2, The Netherlands was marked as the ninth most favorable country in the world for private equity (PE) and venture capitalist investments. The main reason the literature gives for this result is the fact that investors are well protected and a good corporate governance structure is in place. An example of this point is the implementation of the Dutch ‘Code-Tabaksblat’ in 2004. This corporate governance code had an important element that strived to strengthen the power of control and, at the same time, the protection of the shareholders. The strengthened position of shareholders was willingly used by active shareholders.
Shareholder activism had already become more visible for the Dutch audience in
particular during the past decade. The case of the Dutch industrial company Stork, which was privatized by the English PE company Candover at the end of 2007, is a striking example. Before Stork was privatized, a fierce takeover battle took place between Candover and the hedge funds Centaurus and Paulson. These three shareholders, which each possessed a sufficient amount of shares, actively interfered with the strategy of Stork. By using their shareholders’ rights all three parties were extremely powerful to pursue their objectives. Finally, Candover won the battle.3 The extensive media coverage of the battle together with the extensive government interference led to a public
discussion whether shareholders had yet not become too powerful. Due the distinctive characteristics of the corporate governance system in the Netherlands, shareholders already have the right to place an item on the agenda of a general meeting of shareholders through the holdings of only 1% of the company’s outstanding shares.
Gillan and Starks (2007) observe that during the last decade hedge funds and PE firms have assumed prominence in the activist arena, and claim that these parties have become
2 http://www.ey.com/NL/nl/Newsroom/News-releases/PR_Nederland-het-negende-aantrekkelijkste-land_100210
7 increasingly important players in the financial markets. Particularly in their capacity as active monitors of corporate performance and agents of change. Wright and Robbie (1998) add to this that PE firms deliver a degree expertise. PE firms can have a significant impact on the business of a company, because they deliver much closer supervision of management and thereby allow the company to be better governed, creating large wealth gains for investors (Masulis et al., 2009).
This paper will focus on the short term wealth effects of initial PE minority investments in Dutch listed companies and results in the following main research question:
‘How do Dutch target firm, PE firm and deal characteristics affect the wealth effects of Dutch listed stocks when PE investments are considered?’
In order to answer this question this paper builds on a comprehensive study of Achleitner et al. (2010) which investigated the wealth effects of PE investor purchases of shares in German quoted companies. Similar to Achleitner et al. (2010) and common in PE literature, this paper will employ an event study methodology to measure the effects. This study differs from other existing literature, including that of Achleitner et al. (2010), because it incorporates PE firm, target firm and deal characteristics to explain the short term announcement effect. In this way a wide range of questions will be answered. Does the reputation of a PE firm have an influence on the announcement effect? Does the size of the stake and the age of a PE firm matter here? Does a foreign PE firm bring about more positive wealth effects than a native one? And, does an initial stake of a family owned PE firm produce other results than a non-family owned PE firm? In order to answer these questions an unique dataset was constructed and tested by the ordinary least squares regressions method (OLS), a method that is used by the most leading scientists in the field.
8 triangulation was used on a large scale to provide this paper with the most accurate and sufficient data. Important databases that were consulted were Zephyr, CapitalIQ and the database of the Authority Financial Markets (AFM). Beside the use of databases, specialized websites on the PE industry provided this paper with valuable information. Finally, experts in the field helped to fill in the gaps and contributed, through their level of expertise, to the success of this paper. The data was collected from January 1999 through November 2010. An eleven year window was chosen in order to give the results a more robust dimension. Lastly, this paper attempted to give an academic face to the increasingly common professional literature.
To the author’s best knowledge, this is the first study to investigate the effect of PE investments on listed companies in the Netherlands. The results of this paper can be of great importance to three major stakeholders in PE investments. The first group of stakeholders are individual and institutional investors who invest in Dutch listed
companies. The potential of positive and significant wealth effects may signal an investor that it is financially interesting to buy Dutch stocks that are on the PE radar. The results will also give a more in-depth view on which PE firm, target firm and deal characteristics an investor must look for. For the second stakeholder, a director of a Dutch listed
company, it is also interesting to know what the effects are on their stock price as a result of a PE investment. Finally, for the third group of stakeholders, the PE industry itself, this study can give an answer to whether investments in Dutch listed companies hide
attractive investment opportunities.
The remainder of this paper is organized as follows. The paper will start by describing in chapter two the structure of PE firms in general and the corporate governance structure in the Netherlands. In that same chapter former empirical research performed on PE
activism will be discussed and the hypotheses will be given. The data and descriptive statistics will be explained in chapter three and the methodology is discussed in chapter four. In chapter five the results are discussed and finally in chapter six the paper will be concluded.
9
2. Literature review
Part two of this paper is scheduled as follows. First, a general introduction concerning the PE industry will be presented. The corporate governance structure in the Netherlands will be described next. Finally, a thorough description of previous literature and PE activism in particular, will be given.
2.1 PE, a short introduction
Though the levels of investments fluctuate, from 2005 to 2009, an average of €4 billion was invested yearly by national and international PE firms in Dutch companies.4 In 2009, PE accounts for 19% of the Dutch gross domestic product and for 6% of the employment in the private sector.5 With large and active PE firms such as The Blackstone Group, The Carlyle Group and Kohlberg Kravis Roberts & Co (KKR), in recent years an increase is witnessed around the globe in the number and size of deals done by PE firms (Froud and Williams 2007). Steve Kaplan, a well-known and respected researcher of PE firms, said the following:
“Because of the consistency of performance by its best firms, PE has established itself as a -permanent- asset class; there’s little doubt in my mind that PE is here to stay.”
(Kaplan (2007), pp.8)
But what is exactly a PE firm and how do they work, and what distinguishes them from other parties like venture capital and hedge funds?
Private equity investments tend to evolve in waves and the three most fierceful buyout waves were experienced in the late 1980s, the mid-1990s and from 2005 until 2007. 6 As can be seen from Table 1 and Table 2, especially during the last wave the biggest deals were completed. With respect to the value of the biggest PE deals done worldwide, the Netherlands seems be playing only a modest role.
4 http://www.nvp.nl/docs/ondernemend_vermogen_2009.pdf 5 http://www.nvp.nl/feiten/factsheet.php
10
Table I
Top 5 biggest global PE deals over the past 25 years (values includes assumed debt). 7
Nr. Company Price (billion) Acquirer Year
1 Equity Office Properties Trust $38,9 Blackstone 2007
2 Hospital Corp. Of America $32,7 Bain, KKR, Merril Lynch 2006
3 RJR Nabisco* $31,1 KKR 1989
4 Harrah's Entertainment $27,8 Apollo, Texas Pacific 2006
5 Clear Channel Communications $26,7 Bain, Thomas H. Lee 2006 * Corrected for inflation, the RJR Nabisco deal remains the biggest of the past 25 years.
Table II
Top 5 biggest Dutch PE deals over the past 25 years (values includes assumed debt). 8
Nr. Company Price (billion) Acquirer Year
1 VNU € 7,5 Alpinvest, Blackstone, Carlyle,
KKR, Hellman&Friedman, Thomas H. Lee
2006
2 Philips Semiconductors € 6,6 KKR, Silver Lake Partners, Alpinvest
2006
3 Essent Kabelcom € 2,6 Cinven, Warburg Pincus 2006
4 Casema € 2,1 Cinven, Warburg Pincus 2006
5 Gouden Gids (world directories) € 2,1 Apax, Cinven 2004
PE firms raise equity capital through a PE fund. The limited partners that invest in a PE fund are big public pension funds, insurance companies or wealthy individuals. The money that is raised is then invested in a wide variety of 10 to 20 companies or operating businesses with the PE firm acting as the general partner with usually 1% of the total capital provided. The investments of these funds are characterized by their high leverage and sometimes 70% of the cost of purchase is financed with debt. A PE firm mainly aims to exit its investment within three to five years. The use of this high leverage and the focus on the short-term are one of main reasons why PE firms are in the centre of public discussion whether or not their practice is legitimate and ethical responsible. Because once the companies’ income starts to decline it is possible that companies are not able to
11 pay the high interest payments on the debt anymore and fall into financial distress. Another reason why PE firms are criticized is because they too much have a short-term focus and are not sufficiently committed with the interests of all stakeholders, including staff, on the long term. Jensen (1989) makes a stand for PE firms and explains why PE firms use such high leverage. He argues that by paying future cash flows for debt
obligations creates a strong commitment by management not to spend these cash flows on negative present value projects. Furthermore, the increased risk of financial distress, caused by the debt payments, enhances the managers’ incentives to make the company more efficient and eventually increase shareholder value. The latter opinion is not always shared. In the view of Froud and Williams (2007), PE represents a reshuffle of ownership claims for value capture which then allows value extraction, primarily for the benefit of the few who are situated as PE principals or senior managers in the operating business. The term PE is frequently mixed up with terms as venture capital and hedge fund. Klein and Zur (2009) state that hedge funds, PE funds and venture capital funds roughly fall under a similar investment umbrella. They differ mainly through their investment
strategies. In Table 3 an overview is given with respect to the investment strategies of PE firms, venture capital firms (VC) and hedge funds (HF).
Table III
Overview of the main characteristics of venture capital firms, private equity firms and hedge funds.
Venture Capital Private Equity Hedge Fund
Time to hold investments 3 to 7 years 3 to 5 years 6 to 18 months Interference with management High level of interference with
management
High level of interference with management
Low level of interference with management
Profile of target Start-up (technology) with growth potential
Mature Business with stable cash flows
Multiple strategies, i.e.; short selling, activist, arbitrage. Finance strategy Investment paid by equity Investment highly leveraged Through the short time to hold the
investments are more liquid. But in some cases also highly leveraged. Ownership Significant level of ownership
after investment
Majority stake or significant minority stake after investment
12
2.2 Corporate governance structure in the Netherlands
La Porta et al. (2002) state that better shareholder protection is empirically associated with a higher valuation of corporate assets. Shleifer and Vishny (1986) argue that if shareholders rights are not well protected by the law, investors holding a significant amount of shares can act as a substitute to monitor the management, because they have both the power and the incentives. PE firms can fill in this gap perfectly. This in accordance with Masulis et al. (2009) who described PE as a model that allows corporations to be better governed. The Netherlands used to be a typical case of a Continental European country with poor legal protection of small shareholders, and where the position of shareholders have historically been weak (Timmerman et al. (2002) and La Porta et al. (1998)).
Only recently a shift is witnessed where the position of shareholders is strengthened at the expense of the board of directors and the supervisory board. 9 As mentioned in the introduction, the ‘Code-Tabaksblat’(Code), introduced in 2004, is one of the examples of a change in the corporate governance structure in the Netherlands. Other European countries introduced a similar corporate governance code, like the United Kingdom (‘Higgs’), Germany (‘Cromme’), France (‘Vienot’) and Belgium (‘Lippens’).
The Code is based on the underlying philosophy of ‘comply or explain’ and contains principles to improve transparency in the financial statements, better accountability of the board and a strengthening of control and protection of shareholders.10 Further, the Code contained an important element that a company should state in its annual report on how it covers the principles of the Code applied in the past year.11
With the sell of Stork and in 2007, fueled by the actions of activist PE shareholders, major questions were raised whether the balance of power had not been beaten by the benefit of shareholders. In the end of 2009, the Dutch government submissioned a bill based on the recommendations of the committee-Frijns which would serve as an addition to the Code and ensured a renewed balance between the rights and powers of the boards
9 From now on ‘the board’ will refer to the board of directors and the supervisory board.
10http://www.minfin.nl/Onderwerpen/Financi%C3%ABle_markten/Corporate_governance/Corporate_gove rnance_code_code_Tabaksblat
13 and the shareholders.12 The bill consisted of three main recommendations. It introduced a new, lower threshold of three percent (instead of five) for the reporting of substantial shareholdings. The bill also increased the threshold for exercising the right to put items on the agenda of a shareholders meeting from one to three percent. Third, shareholders who pass the three percent threshold should indicate their intentions.
Another important feature that has an effect on the corporate governance in the Netherlands is the Dutch two-tier model. In the Dutch two-tier management model, companies have a management board that performs the executive function, and a non-executive supervisory board which advises and supervises the management board. The law requires the supervisory board to serve the firm’s interest. The two-tier model has the advantage that it forms a reliable mechanism to ward off unfriendly takeover attempts for publicly listed companies. However, the two-tier board is also criticized by legal scholars for the fact that it is not in line with international practices and it is peculiar and
complex.13
According to the Dutch law on financial supervision (Wet Financieel Toezicht), investors in the Netherlands have to declare if their stockholdings in a company reach 5 percent or more of the subscribed capital. As an investor accumulates stocks in a specific company he passes certain threshold which automatically gives him more right to say. Table 4 presents a schematic overview of shareholders’ rights in the Netherlands.
14
Table IV
Shareholders’ rights in the Netherlands.
Index List
A * Access to the general meeting. B * Voting right.
C * Sometimes: Veto.
D * Agenda Setting.
E * Block a decision of the board concerning a merger or split of the company by convening a general meeting.
F * Sell-out right of a public offer.
G * Convening a general meeting with judicial authorization. H * Survey right.
I * Decision to exclude pre-emptive, merger, demerger, capital reduction adopted if at least 50%
of the issued capital is represented.
J * Decision to exclude pre-emptive, merger, demerger, capital reduction adopted if less than 50%
of the issued capital is represented.
Percentage Rights
1 share A / B / C
1% of the subscribed capital A / B / C / D
5% of the subscribed capital A / B / C / D / E / F
10% of the subscribed capital A / B / C / D / E / F / G / H 25% of the subscribed capital A / B / C / D / E / F / G / H / I 33% of the subscribed capital A / B / C / D / E / F / G / H / J
Source: Peters (2009)
2.3 Literature on PE shareholder activism
15 announcement alone that a PE firm takes a stake in a listed company can have an effect on the stock price of that same listed company. Research with respect to PE shareholder activism is still very young, however there are a number of scientists who have
investigated the consequences of shareholder activism. In this section, leading research will be discussed.
Almost all research published in the past 25 years shows that an announcement of a PE investment in a exchange-listed stock leads to a certain level of wealth effects for the targets’ shareholder. The reasons given for this significant effect are versatile. According to Bethel et al. (1998), who investigated the effects of activist block share purchases in the U.S., PE proves to have a significant influence in elevating a company’s stock price in the short-term if it takes a stake of 5% or more in a listed company. The main reason they give for this effect is that as a result of PE shareholder activism, asset divestures will increase, mergers and acquisitions will decrease and industry-adjusted operating profitability will rise. Moreover, an increase in operating performance and a reduction of capital expenditures are also suggested as possible reasons for the significant results found (Brav et al., 2008 and Klein and Zur, 2009).
In a recently published article, Achleitner et al. (2010) measured the wealth effects of PE investor purchases of shares in German quoted companies. They found that PE investors, with a stake of 25% or more, generate positive wealth effects for target shareholders around the event date. This result is mainly due to the increase in monitoring of the board by a large shareholder, namely the PE firm. These results are supported by Mientzer et al. (2008) who also performed a study on the German stock market. They found that positive announcement effects are triggered if a PE-company acquires at least 5% of the voting rights. Proxy variables for agency costs explain the market reaction for investments of PE firms.
16 that a company has been targeted by a shareholder activist. Gillan and Starks (2007) admit that some other studies found positive short-term market reactions. As a result, this paper follows the suggestions made by the literature who found significant effects and this leads to the following hypothesis:
H1- Hypothesis; ‘initial PE investments in Dutch listed companies lead to significant positive wealth effects.’
In Table 5 a summary is presented of previous described event studies who found significant effects of PE participations on company stock performance.
Table V
Summary of the results of previous event studies on the announcement effect of PE investments in exchange-listed companies.
Author (publishing year) Number of firms used Period of study Country Days before and
after the first announcement [0] Average cumulative abnormal return Bethel et al. (1998) 244 1980-1989 U.S. [-30 ; +30] 14.2% Brav et al. (2008) 236 2001-2006 U.S. [-20 ; +20] 8.4% Mietzner (2008) 159 1993-2007 DE [-20 ; +20] 3.55%
17
2.4 Dutch target firm, PE firm and deal characteristics
The unique dataset constructed for this paper made it possible to investigate the effect of different target firm, PE firm and deal characteristics which all can have an influence on the wealth effects and should therefore be examined.
Size of stake
Whereas Mientzer et al. (2008) and Bethel et al. (1998) considered a significant effect only if the PE firm has bought a stake of 5 percent or more of the target company, Wruck (1989) makes a clear distinction with respect to the size of the stake. When the level of ownership concentration after the sale is low (0 percent to 5 percent) or high (at least 25% or more), the relation between changes in the firm value at the announcement and changes in ownership concentration is accompanied by a 4.5 percent average positive abnormal return. In a middle range of ownership concentration (5 percent to 25 percent), however, this relation is negative.
As an explanation for this result, Wruck (1989) suggest that the ability of incumbent shareholders to become entrenched outweighs any benefits of having a blockholder in place. The results of Wruck (1989) are partly supported by a study of Thomsen et al. (2006) who investigated the influence of PE shareholder activism for firms from
Continental Europe. They concluded that there was a significant negative effect between blockholder ownership (of 10 percent or more) and firm value.
According to their research, blockholder ownership apparently has tended to exceed the level which maximizes firm value from the viewpoint of minority investors. Blockholders eventually use their control rights to extract private benefits for their own. Nonetheless, Shleifer and Vishny (1997) argue that large shareholders provide improved monitoring and would therefore outweigh the costs of the extraction of these private benefits.
18 From the information above can be concluded that a relation between the size of the stake and the wealth effects is possible. Only the sign of the relationship is unknown, and as a result this leads to the following hypothesis:
H2- Hypothesis: ‘the size of the stake of initial PE investments in Dutch listed companies have a significant influence on the wealth effects.’
Age
In a working paper by Loderer and Waelchli (2010) the correlation is investigated between the age of a firm and its performance. They find that as firms grow older, their profitability seems to decline. Older firms suffer from ossification of their routines, non-learning processes, blindness, and conservatism, which cause poor performance and decline (Durand and Coeurderoy, 2001). For the PE industry this would imply that older PE firms show a lower performance compared to younger PE firms. From this, the basic assumption can be made that as PE firms grow older they are less successful. If older PE firms are less successful, this would implement that a significant negative effect is expected with respect to the wealth effects. Consequently, the hypothesis states:
H3- Hypothesis: ‘the age of a PE firm, who invests in Dutch listed companies, will have a significant negative influence on the wealth effects.’
Reputation
Dunbar and Schwalbach (2000), who investigated corporate reputation and firm
performance in Germany, found that corporate reputation can have a significant positive influence on performance. Furthermore, they conclude that the size of a company can have a positive influence on the reputation of a company. This paper follows the
19 The amount of employees under contract is the second measure of size. This measure is based on the basic assumption that a larger firm has more employees. Based on the suggestions of Fombrun (1996), the number of employees is strongly associated with a positive reputation.
Finally, size is based on two lists published by Private Equity International (PEI) in 2007 and 2008. Based on their capital raised, these lists14 represent 50 of the largest buyout funds worldwide over the past five years. The rankings are based on the amount of PE direct-investment capital raised by each firm over a period of five years. The results of Dunbar and Schwalbach (2000) are complemented by the findings of Roberts and
Dowling (2002) which shows that firms with relatively good reputations are better able to sustain superior profit outcomes over time. This study will investigate whether the
reputation of a PE firm, specified by the three measures above, will have a positive influence on the wealth effects. Consequently this leads to three different hypotheses:
H4a- Hypothesis: ‘the number of employees a PE firm has under contract, which invests in Dutch listed companies, will experience a significant positive influence on the wealth effects.’
H4b- Hypothesis: ‘the total amount of capital invested by a PE firm, which invests in Dutch listed companies, will experience a significant positive influence on the wealth effects.’
H4c- Hypothesis: ‘PE firms who invests in Dutch listed companies, and listed on the PEI list (2007 or 2008), will experience significant positive wealth effects.’
Country
A well-documented characteristic of international investment portfolios is the strong bias in favor of domestic securities (Coval and Moskowitz, 1999 and Kang and Stulz, 1997). As a reason Coval and Moskowitz (1999) suggest that asymmetric information between local and nonlocal investors may drive the preference for geographically proximate
20 investments. It can be assumed that if an investor holds more information, they are in the position to make better judgments with respect to their investments and as a result be more successful. In line with this argument, this paper will test whether Dutch PE firms are more successful than foreign PE firms in their investments in Dutch listed companies and as a consequence have a higher significant positive influence on the wealth effects.
H5- Hypothesis: ‘native PE firm investments in Dutch listed companies will create a more significant positive influence on the wealth effects than investments made by foreign PE firms.’
Family owned PE firms
Anderson and Reeb (2003) investigated founding-family ownership and firm performance in the S&P 500 and conclusded that family firms perform better than
nonfamily firms. In the dataset used by this paper a great share of the PE firms are owned and controlled by the founding-family. Demsetz and Lehn (1985) note that concentrated investors - like family owners - have substantial economic incentives to monitor
managers closely, thus reducing agency conflicts and maximizing firm value. This paper therefore expects a positive relationship between family owned PE firms and the
announcement effect.
H6- Hypothesis: ‘investments in Dutch listed companies made by family owned PE firms have a higher significant positive influence on the wealth effects than nonfamily-owned PE firm investments.’
Undervaluation
21 part is to identify an undervalued stock. PE firms can identify an undervalued stock if they might have private information about the intrinsic value of the company.
Consequently, undervalued firms are interesting for PE firms because they can use their expertise to improve the performance of the company by, for example, improving efficiency, divesting poorly performing parts of the business, and setting higher
performance targets (Achleitner et al., 2010). PE firms may use their right to say, which comes with the significant stockholdings, to actively interfere with a companies’ business in order to achieve maximization of shareholder’s value. The undervaluation of stocks is therefore an interesting topic to investigate and this leads to the following hypothesis:
22
3. Data
Researchers who investigate PE firms experience that it is difficult to get access to specific information concerning PE firms. Especially, PE firms with a closed nature, such as privately owned PE firms, are not keen on disclosing key information. However, this paper succeeded to construct an unique dataset that includes 95 initial investments in Dutch exchange-listed15 companies made by PE firms. The data covers the period from January 1999 through November 2010 and has been obtained from multiple databases and through professionals in the private equity industry.
“Good research practice obligates the researcher to triangulate, that is, to use multiple methods, data sources, and researchers to enhance the validity of research findings.”
(Mathison (1988), pp. 13)
In this research data triangulation was used regularly in order to construct a complete and unique dataset. One data source that was consulted was Zephyr. This database contains M&A, IPO and venture capital deals with links to detailed financial company
information. Another important source of data is obtained from Capital IQ. This professional database is often used by the investment banking industry and contains multiple financial and deal information. Entrance to this database was provided through Morgan Stanley. The Authority Financial Markets (Autoriteit Financiële Markten (AFM)), a Dutch governmental organization that records all substantial participations in Dutch listed companies, was the third database with key information on PE
participations. In addition to the use of databases, websites did also act as a valuable source of information. However, a risk is hidden in randomly using websites. Because there are numerous websites on PE firms and deals characteristics, this paper carefully choose only to use highly reliable websites that are approved and supported by either governmental organizations or the PE industry itself. These websites include the official website of the European private equity and Venture Capital Association (www.evca.eu),
23 the official website of the Dutch organization of private equity firms (www.nvp.nl) and a leading financial information group that keeps record of all private equity developments (www.peimedia.com). Specific PE firm information was also obtained from the PE company website. The website of the Financial Newspaper (Het Financieele Dagblad) provided access to Dow Jones Newswires and was therefore an important source of information. All stock data was gathered from the database of Thomson Datastream. This paper made one important adjustment to the selection criteria of Achleitner et al. (2010). They added PE firm investments to their dataset only if the level of the stake exceeded 25 percent of the subscribed capital. This study adds PE firm investments if the stake is at least 3 percent but not more than 50 percent of the subscribed capital. The reason for this adjustment is twofold. First, due to the distinct characteristics of the corporate governance structure in the Netherlands, as earlier described in section 2.2, shareholders with a relatively low percentage of total outstanding shares already have a certain amount of right to say. Second, if a shareholder holds a stake of 50% or more it is in fact the owner of the company. And this paper is only focused on PE minority
investments.
Table VI
The distribution of the stakes among the various groups based on the rights of shareholders in the Netherlands as illustrated in Table 4.
Percentage of stake Number Percentage
>5% 2 2,1% 5%-10% 84 88,4% 10%-25% 6 6,3% 25%-33% 2 2,1% 33%- 50% 1 1,1% All 95 100%
24 Through a manual search of the above described websites an additional number of 66 events were found. These events were always cross-checked with the database of the AFM to enhance their validity. Finally, 5 events were deleted because they showed overlap in the event-window. No contamination, such as other share price sensitive information announced around the announcement date, was found. It should be noted that, similar to Achleitner et al. (2010) and Brav et al. (2008), this study did not include any investments made by banks, pension funds, insurance companies, pension funds and individuals in the dataset. Table 6 shows the distribution of the stakes of all 95
announcements and Figure I illustrate a clear peak in the number of announcements in the years 2005 and 2006. This observation is in line with the third fierceful buyout wave that was experienced worldwide from 2005 until 2007 as earlier discussed in section 2.1.
Figure I
Distributions of the announcements by year.
25 not in all cases sufficient. A specialist of Ecart Invest, a The Haque based private equity firm, and an associate of Morgan Stanley helped to fill in absent information. Their extensive knowledge of the private equity industry was of great importance in this iterative process. Table 7 shows the country of origin of the PE firms.
Table VII
Country of origin of the PE firms investing in Dutch listed firms.
Country Number Percentage
The Netherlands 69 72,6%
U.S. 21 22,1%
U.K. 5 5,3%
All 95 100%
In Table 8 a summary is presented of the characteristics of the sample. In case of the total capital invested by a PE firm (Capital) it can be concluded that the sample is skewed. This is indicated by the discrepancy between the mean and median values. With respect to the level of undervaluation, another surprising element is seen. In case of both the mean and the median, the values indicate that the Dutch target firms’ stocks were actually ‘overvalued’ at the time of the PE investments. This is shown by the fact that the figures are greater than one.
Table VIII
Characteristics for 95 initial PE investments in Dutch listed companies during the period from January 1999 till November 2010.
Variable Mean Median Standard deviation Min Max
26
4. Methodology
4.1 Event study methodology
An event study will identify whether the purchase of a stake in a Dutch listed company by a PE investor has a significant influence on the value of the listed company. The event is marked as the announcement of the purchase of the stake.
When an event study is used it is an important condition that there is market rationality, only then the market will reflect all information in the market prices.
There are certain important steps to be undertaken which will act as the foundations of a solid event study. The first step is to define the event itself. In this study the event of interest will be the announcement day of the purchase of the stake by the PE investor. Once selected the event, the period over which the security prices of the firm involved in this event will be examined, this is better known as the event window (MacKinlay 1997). Following Achleitner et al. (2010), Brav et al. (2008) and Mietzner (2008) the length of the event window will be 41 days, from t = -20 till t = +20, (relative to the
announcement day t = 0). This paper will also use a different set of event windows to test the results. Finally, an estimation window will be set to test for the normal return prior to the event window. According to MacKinlay (1997) and Achleitner et al. (2010) an estimation window of over 200 days prior to the event will be sufficient. In this study the estimation window will be set at t = -220 till t = -21. The estimation window and event window are graphically shown in figure II.
Figure II
27 It is presumed that the event has a certain impact on the share price of the Dutch listed company. However, this event is not the only factor which has an impact on the share price. General market movements also influence the share price. To filter out this effect, the actual individual bidder’s return is measured compared to the normal market return. In this way the abnormal return (AR) is derived and is described as follows:
it itit
R
E
R
AR
(1)it
AR , R and it E
Rit are respectively the abnormal return, the actual return and thenormal return for firm i on day t .
4.1.1 Market Model
Brown and Warner (1980, 1985) describe three main approaches in the event study methodology. First comes the mean adjusted returns model, which assumes that the ex
ante expected return for a given firm i is equal to a constant. Second, the market adjusted
returns model, which assumes that ex ante expected returns are equal across securities, but not necessarily constant for a given firm i . The third model is the market and risk adjusted returns model. This model, better known as the market model, presumes that some version of the Capital Asset Pricing Model generates expected returns. MacKinlay (1997) explains that the market model represents a possible improvement over the constant mean return models (first two models). By removing the portion of the return that is related to variation in the market's return, the variance of the abnormal return is reduced. As a result, there is a high chance that the use of the market model can lead to an increased ability to detect event effects.This paper will make use of the market model based on the recommendations of MacKinlay (1997) and based on the fact that Achleitner et al. (2010) uses the market model too. The market model is derived as follows:
28 Where R and it R are, respectively, the period-mt t returns on firm i and the market
portfolio. The zero mean disturbance term is presented asit. Finally,i and a are the i
parameters of the market model. The market portfolio is a representative broad based stock index which reflects the overall market fluctuations. In this paper the MSCI world index is used as a proxy for the market portfolio. The abnormal return for the market model is: mt i i it it
R
R
AR
ˆ
ˆ
(3)The abnormal return for firm i on day t is presented asAR and is a product of the it
difference between the return on day t (R ) and the expected return without the event it
(ˆiˆiRmt). The coefficients ˆ and i ˆ are OLS estimates obtained from regressions of i
firmi’s daily returns on the market return over the estimation period from t = -220 till t = -21. This is exactly the same as is done by Achleitner et al. (2010).
For each day of the event window of 41 days, daily abnormal AR ’s are then calculated:
N i it tAR
N
AR
11
(4)where N is the total numbers of events. In this paper a large estimation window is used and consequently the variance can be derived:
N i t tN
AR
1 2 21
)
var(
(5)29
2 1 2 1,
)
(
AR
tCAR
(6)Eventually the variance of cumulative average abnormal returns can be calculated by:
2 1 2 1,
))
(
var(
AR
tCAR
(7)and, in order to set the covariance to zero, it is important for the event windows of the events to not show any overlap.
4.2 Tests
In order to test the statistical significance of abnormal ARs and CARs a parametric and nonparametric test will be performed. The most common parametric test used in other event studies (Achleiter et al., 2010 and Bethel et al., 1998) is a t-test. The use and implication of the two-sided t-test will be further explained in this chapter. The
nonparametric rank test used in this paper, is explained by Corrado (1989) and suggested by MacKinlay (1997). The nonparametric test Achleitner et al. (2010) used was the standardised cross-sectional test, suggested by Boehmer, Masumeci, and Poulsen (1991). However, Seiler (2000) suggests that both the nonparametric rank test and the
standardised cross-sectional test are robust towards event-induced variance increases that bias tests for mean abnormal returns in event studies. Thus, based on the
recommendations of MacKinlay (1997) and Seiler (2000) this study chooses to use the tests of Corrado (1989).
30 thumb non-normally distributed variables have no observable impact on event study methodology and results. Therefore, in this study the results of both the parametric and nonparametric tests will be considered.
4.2.1 Normality test
The Jarque-Bera (JB) test is used here to examine whether the average abnormal returns are normally distributed. Jarque and Bera (1980, pp. 255) state: “Violation of the
normality assumption may lead the investigator to inaccurate inferential statements”. As
a rule, a parametric test, i.e. t-test, can be validly used in an event study if the average abnormal returns (AR ) in the estimation window are normally distributed. The Jarque-t
Bera test is used to examine whether the data is normally distributed and is described as follows: JB = 4 ) 3 ( 6 2 2 K S N (8)
31
Table IX
Results of the normality test.
Dutch targets
Skewness 0,7189
Kurtosis 0,7551
Jarque Bera 59,2229
Observations 95
Table 9 gives the outcome of the normality test and shows that a parametric test will give the most valid explanation for the results.
4.2.2 Parametric test
In leading scientific research and the previous mentioned event studies the most common used parametric test is the t-test. Like Achleitner et al. (2010) this paper will use a two-sided t-test, for the reason that the literature does not present a uniform picture of the sign of the abnormal returns. As a consequence the null hypothesis is rejected for values of the test statistic falling into either tail of the sample distribution. Mackinlay (1997) gives the following equations to test the null hypothesis for the AR and t CAR(1,2):
~ (0,1) var AR 1/2 N AR t t (9)And for the cumulative abnormal returns, CAR(1,2):
) 1 , 0 ( ~ )) , ( var( ) , ( 2 / 1 2 1 2 1 N CAR CAR (10)
32
4.2.3 Nonparametric Tests
Corrado (1989) studied a nonparametric rank test for abnormal security-price
performance that is preferable to the parametric t-test for a broad spectrum of fat-tailed security-returns distributions. Because daily security return distributions are characterized by a highly nonnormal distribution, when abnormal performance is present, the use of the rank test is more powerful than other parametric tests (Corrado, 1989).
The rank test of Corrado (1989) can be specified as follows:
N i i K S L K N T 1 0 ) ( 2 1 1 (11) Standard deviation:
1 2 2 1 2 1 1 1 ) ( t t t N i it L K N L K S (12)The ‘rank’,K , of every abnormal return is determined using both the estimation window it
and the event window. The L specifies the length of this period. The number of events is specified by N. Finally, t is the number of days previous to the event day, -220 in this 2
paper, and t1 is the number of days following the event day, + 20 in this paper. By using the nonparametric rank test, the distribution of the sample is transformed in a uniform distribution of rank value, regardless of any asymmetry in the original distribution (Corrado, 1989).
33
4.3 Ordinary Least Squares Regressions
The influence of different PE firm, target firm and deal characteristics, discussed earlier in section 2.3 and 2.4, on the cross-sectional differences in the wealth effects will be tested through ordinary least squares (OLS) regressions. The control variables will be size and age. Similarly as in Achleitner et al. (2010), the control variable C-SIZE is defined as the natural logarithm of the target firm’s total assets in the year preceding the PE
investment. The control variable C-AGE is defined as the natural logarithm of the
difference between the year of the initial participation and the target firm’s founding year. Due to the great differences between the minimum and maximum values16 in case of the variables AGE, STAKE, CAPITAL17 and EMPLOYEES18 this paper will use the natural logarithm. The variable UNDERVAL19 will not be scaled. Dummies are used to test the effects of the variables: D-COUNTRY, D-FAMILY and D-PEI.
The OLS regression is as follows:
0 ) 1 , 1 ( i
CAR + 1 AGE + 2 STAKE + 3 CAPITAL +
4
EMPLOYEES + 5 UNDERVAL + 6 C-SIZE (13)
+ 7 C-AGE + 8 D-COUNTRY + 9 D-FAMILY + 10 D-PEI + 11 D-AEX + 12 D-AMX + 13 D-ASX + e i
16 See Table 8
17 In case of data constraints the following method is used. All data on the total capital invested is gathered from the year 2010. These figures are corrected and counted back, using the PE firms countries’ specific growth rate, until the year of the announcement. The database of the International Monetary Fund website provided crucial information on the growth rate. In the end, all data is scaled using the Logarithm (LOG). 18 All data concerning the number of employees is gathered from the year 2010 and scaled using the LOG. 19 Undervaluation is measured as the ratio of the stock price 60 trading days preceding the
34
i
CAR = the 3-day cumulated abnormal return for company i on days -1, 0 and
1.
AGE = the age of the PE firm at the time of the initial participation. STAKE = the level of the stake of the initial participation.
CAPITAL = the amount of total capital invested by the PE firm. EMPLOYEES= the amount of employees the PE firm has under contract. UNDERVAL = the level of undervaluation of the stock of the Dutch target firm. C-SIZE = the natural logarithm of the total assets the Dutch target firm had under
management at the time of the announcement.
C-AGE = the natural logarithm of the age of the Dutch target firm at the time of the announcement.
D-COUNTRY = ‘1’ if the headquarter of the PE firm is situated in the Netherlands, ‘0’ if otherwise.
D-FAMILY = ‘1’ if the PE firm is a family-owned firm, ‘0’ if otherwise.
D-PEI = ‘1’ if a PE firm is listed on the PEI 2007 or 2008 index, ‘0’ if otherwise.
i
35
5. Results
In this part the results will be presented and evaluated. First, the results with respect to the event study will be presented. Later, the results of the OLS regression analysis will be discussed.
5.1 Event study results
Table 14 (see Appendix) presents the results of the event study for all 95 initial participations. Column 2 shows the daily abnormal returns (AR) in %, column 3 and 4 show the significance tests, respectively the t-test and the Corrado test statistic.
Eventually, in Column 5 and 6 the cumulative daily average abnormal return (CAR) is shown together with its significance test, the t-test.
In addition, from this Table can be concluded that on the day of the announcement (t=0) no significant result is observed, but one day after the announcement (t= 1) an average abnormal return of 0,755% is earned. This outcome is significant for both the parametric t-test (1%) and the nonparametric Corrado-test (10%). It is sometimes difficult to state at what time the information with regard to the announcement reaches the market.
Therefore, the 3-day CAR (-1; 1) is used here to give a more robust explanation of the impact of the announcement. The 3-day CAR (-1; 1) amounts to 1,216% and is
36 Further, one significant abnormal return is visible prior and one after the event period (-1; 1). On day -17 and day +6 the Corrado test shows significant result on a 5% level. That a significant result is detected before the event period can be an indicator of leakage of information about PE investments. This is in violation with the assumption of an information-efficient market.
In addition, this explanation is further supported by Table 10. Here the results are shown for different event periods. The 3-day CAR (-1; 1) shows the highest cumulative change (1,216%) and is significant on a 5% level (t-test). But the event period (-20; 0) shows the highest significance values. However, this study will use the 3-day CAR (-1; 1) since leading scientific PE literature, that apply an event study methodology, show that a short event window surrounding the announcement date confines most of the announcement effect of the PE investment.
Table X
The cumulative return for different event periods.
Event Window CAR (%) t-test
(-1 ; 1) 1,216 2,5045** (-2 ; 2) 0,766 1,050 (0 ; 1) 0,977 2,014** (0 ; 10) 0,977 0,366 (0 ; 20) -0,167 -0,033 (-10 ; 0) 1,049 4,265*** (-20 ; 0) 1,053 4,276*** (-20 ; 20) 0,663 0,128
* Significant at the 0,10 level ** Significant at the 0,05 level *** Significant at the 0,01 level
37
5.2 Result of OLS regression analysis
An OLS regression analysis will be used to investigate the impact of different Dutch target firm, PE firm and deal characteristics on the wealth effects. The dependent variable used here is the 3-day CAR (-1; 1). Table 11 shows the Pearson correlation between Dutch target firm, PE firm and deal characteristics. Despite the fact that the correlation coefficient between the dummy D-COUNTRY and the dummy D-FAMILY shows signs of multicollinearity, it remains an interesting topic to investigate.20 Further noTable is the relative high positive correlation between the amount of employees a PE firm has under contract and the total capital under management by a PE firm.
Table 12 gives the answer to the H2-hypothesis whether the size of a stake has a significant influence on the wealth effects. Whereas the sign is positive here, the t-test shows no significant result. Therefore, the H2-hypothesis should be rejected. This contradicts the results found by Wruck (1989) and Thomsen et al. (2006), but confirms those of Achleitner et al. (2010). The H3-hypothesis suggested that the age of a PE firm is negatively associated with the wealth effects. The positive sign found here is not in line with the suggestions of Durand and Coeurderoy (2001) and Loderer and Waelchli (2010), and therefore the H3-hypothesis should be rejected. A reason for this contradicting result can be due to the fact that not specifically the age of a PE firm is of a decisive factor but more the age and experience of its employees. For example, Monolith, a 5-year old PE firm which was included in this study’s dataset, has employees under contract with an average of 10-years experience in the field of PE investments.
In order to measure whether the reputation of a PE firm has an influence on the wealth effects, this paper constructed three indicators for reputation, i.e. the number of
employees under contract, the total amount of capital invested and a notation on the PEI 2007 and/or 2008 list. Hypothesis-H4a stated that the number of employees a PE firm has under contract, which invests in Dutch listed companies, will have a significant positive influence on the wealth effects.
Age Stake Capital Employees Underval C-Size C-Age D-Country D-Family D-PEI Age(Log) 1 Stake(Log) 0,042 1 Capital(Log) 0,041 0,099 1 Employees(Log) -0,028 -0,013 0,619*** 1 Underval -0,033 0,015 0,178** -0,105 1 C-Size –0,226** 0,018 0,556*** 0,461*** 0,089 1 C-Age 0,017 0,009 0,230** 0,042 0,007 0,349*** 1 D-Country 0,271*** 0,026 –0,574*** –0,509*** -0,094 –0,594*** 0,001 1 D-Family 0,400*** -0,087 –0,443*** –0,367*** -0,161 –0,450*** 0,024 0,861*** 1 D-PEI -0,057 -0,020 0,639*** 0,424*** 0,214** 0,218** 0,061 –0,342*** –0,294*** 1 Independent Variables: Control Variables: Age(Log) = age of a PE firm(Log) C-Size =
Control Variable for market capitalizaion (Log)
Stake(Log)
= the level of the initial stake(Log)
C-Age =
Control Variable for age of the firm (log)
Capital(Log)
= the total amount of capital invested by PE firm(Log)
Employees(Log)
= the amount of employees a PE firm employs (Log)
Dummies:
Underval =
the level of undervaluation of the stock of the Dutch firm(Log)
D-Country
= dummy Country of residence PE firm
D-Family
= dummy Family PE firm
Sample size
(N) = 95
D-PEI
= dummy PEI notation PE firm
38
*** Significant at 1 %, ** Significant at 5 %, * Significant at 10%
Table XI
39 The result from Table 12 shows that the sign is indeed positive, however, the result is insignificant. One conclusion from this result that might drawn is that the number of employees a PE firm has under contract is not a good indicator for the level of reputation. Another suggestion is that reputation has no effect on the wealth effects at all. Yet, the latter point is not in line with the results from the H4b-hypothesis. Though the outcome of Table 12 shows no significant result, the step-wise regression (see Appendix: Table 15) confirms that the total amount of capital invested has a significant positive influence on the wealth effects. Therefore, there is prove that the H4b-hypothesis should be
accepted. These unique findings contradict the results of Achleitner et al. (2010), who did not find any significant result with respect to the variable reputation. Consequently this conclusion supports the suggestion made by Dunbar and Schwalbach (2000) that size is a good measure for reputation. The third measure used for reputation was the fact whether a PE firm was mentioned on the PEI 2007 and/or 2008 list. The result shown in Table 12 point to a significant negative effect of a listing on the PEI 2007 and/or 2008 list on the wealth effects. This study expected a positive sign and as a consequence the H4c-hypothesis is not accepted. It is assumed that this is due to the fact that there were not many PE firms in the sample listed on the PEI list, which may have biased the results. The fifth hypothesis described that native PE firm investments in Dutch listed companies would experience a more significant positive influence on the wealth effects than
investments made by foreign PE firms. As a result of information asymmetry, investments made by Dutch PE firms in Dutch listed companies will have a have a stronger significant influence on the wealth effects than investments made by foreign PE firms. The outcome of the test, shown in Table 12, confirms this statement and
40 therefore fit to investigate whether the results of Anderson and Reeb (2003) hold for PE firms. The OLS regression analysis shows a significant result on a 1% level. But in stead of a positive coefficient, which was expected, a negative coefficient is observed.
Therefore, the H6-hypothesis is not accepted. This study found additional work which supported the contradicting result. Work of Faccio et al. (2001) noted that family firms are relatively poor performers due to conflicts that arise as a family tries to manage an enterprise. Furthermore, Dyer (2006) examined the ‘family effect’ on firm performance and found that in three of the nine investigated studies family owned firms showed relatively poor performance compared to non-family owned firms. Whereas in four cases family firms outperformed non-family firms. As a main argument, Dyer (2006) suggested that the different methodological approaches employed across the studies might account for the contradictory findings.
Lastly, similar to Achleitner et al. (2010), this paper investigated whether the level of undervaluation of the stocks of Dutch target companies will have a significant positive influence on the wealth effects created by PE firm investments. Undervalued stocks are interesting for PE firms because they can use their expertise to boost a company’s value. In this paper however, no significant effect was found between the undervaluation of Dutch target firms’ stocks and the wealth effects. As a consequence the H7-hypothese is rejected.
The explanatory power of the regression analysis is expressed by the R² and adjusted-R². The R² values found in this paper differ from 10 percent to 16 percent which are
Coeff. t. Age(Log) -,098 ,826 Stake(Log) ? ,053 ,501 Capital(Log) + ,147 ,784 Employees(Log) + ,179 1,270 Underval + ,106 ,966 C-Size -,056 -,367 C-Age -,027 -,239 D-Country + ,539 2,141** D-Family + -,639 –2,836*** D-PEI + -,242 –1,743* Sample Size (N) 95 R - Squared ,157 Adj R- Squared ,057 F-Statistic 1,565 Prob (F=Statistic) 0,131 Independent Variables: Control Variables: Age(Log) = age of a PE firm(Log) C-Size =
Control Variable for market capitalizaion (Log)
Stake(Log)
= the level of the initial stake(Log)
C-Age =
Control Variable for age of the firm (log)
Capital(Log)
= the total amount of capital invested by PE firm(Log)
Employees(Log)
= the amount of employees a PE firm employs (Log)
Dummies:
Underval =
the level of undervaluation of the stock of the Dutch firm
D-Country
= dummy Country of residence PE firm
D-Family
= dummy Family PE firm
D-PEI
= dummy PEI notation PE fir
m
41
Independent Variables
Table XII
Results of Ordinal Least Square Regression Analysis: Modelling the relationship between
PE firm, target firm and deal characteristics and firm performance (CAR)
.
*** Significant at 1 %, ** Significant at 5 %, * Significant at 10
%
Predicted sign
42
6. Conclusions and recommendations
In the first part of this section the main research question will be answered and the paper will further be concluded. Finally, recommendations for future research is discussed.
6.1 Conclusions
This paper examined how Dutch target firm, PE firm and deal characteristics affect the wealth effects of Dutch listed stocks, when PE investments are considered.
An event study methodology was used to investigate whether an initial stake led to positive wealth effects. For the research of the target firm, PE firm and deal
characteristics an OLS regression analysis was employed. The unique sample consisted of 95 initial investments taken by PE firms in Dutch listed companies in the period from January 1999 through November 2010. The results of the event study show that there is a significant relationship between initial PE firm investments in Dutch listed companies and the wealth effects. In addition, in the event window (-1; +1) a significant positive return of 1,216 % is found. These results are consistent with the findings of Achleitner et al. (2010) however; this study found less strong positive wealth effects for target
shareholders. The fact that this study added initial stakes of 3 percent till 50 percent may have affected the reaction on the target firms´ share price and eventually resulted in lower positive wealth effects compared to Achleitner et al. (2010).
Determinant Hypothesis Expected Sign Actual Sign Accepted Significant at Remarks Size
‘the size of the stake of initial PE investments in Dutch listed companies have a significant influence on the wealth effects.’
? + No ---Age
‘the age of a PE firm, who invests in Dutch listed companies, will have a significant negative influence on the wealth effects.’
-+ No ---Reputation
‘the number of employees a PE firm has under contract, which invests in Dutch listed companies, will experience a significant positive influence on the wealth effects.’
+ + No ---"
‘the total amount of capital invested by a PE firm, which invests in Dutch listed companies, will experience a significant positive influence on the wealth effects.’
+ + Yes 10% ---"
‘PE firms who invests in Dutch listed companies, and listed on the PEI list (2007 or 2008), will experience significant positive wealth effects.’
+ -No ---Country
‘native PE firm investments in Dutch listed companies will create a more significant positive influence on the wealth effects than investments made by foreign PE firms.’
+ + Yes 5% ---Family
‘investments in Dutch listed companies made by famil
y
owned PE firms have a higher significant positive influence on the wealth effects than nonfamily-owned PE firm investments.’
+
-No
---The result of the OLS regression analysis shows a negative result, which is significant on a 1%-level.
Undervaluation
‘wealth effects created by PE firm investments are positively related to the level of undervaluation of the stocks of Dutch listed target companies.’
+ + No ---43 Table XIII
44 This contradicts the results of Loderer and Waelchli (2010), and it may implement that the impact on the wealth effects of both young as old PE firm’ investments are considered the same by the market. The reputation of a PE firm, which is represented by the total amount of capital invested by a PE firm, showed to have a significant influence on the wealth effects. Achleitner et al. (2010) found no significant results, and therefore this finding is new in the PE literature. Investment decisions made by repuTable PE firms in the Netherlands seem therefore to be followed and rewarded by the market. It must be said that the two other determinations of reputation, i.e. the number of employees a PE firm has under contract and a notation on the PEI 2007 and/or 2008 list, showed no significant results. Coval and Moskowitz (1999) argued that local investors have more information than nonlocal investors concerning their investment opportunities. As a result an information asymmetry exists in favor of the local investor, and can lead to better judgments with respect to investments made. Consistent with the above, this paper found that investments by Dutch PE firms lead to stronger and higher wealth effects when compared to investments by foreign PE firms. Again, this finding is new in the PE literature. With respect to whether it would have an influence on the wealth effects if a PE firm is family-owned or nonfamily-owned, this paper found significant negative results. This would implement that investments made by family-owned PE firms lead to significant lower wealth effects compared to investments made by nonfamily-owned PE firms. This finding is not supported by the suggestions of Anderson and Reeb (2003), who in turn did find a significant positive relationship. Dyer (2006) argues that different methodological approaches employed across the studies might account for the
contradictory findings. Finally, this paper tried to explore if the undervaluation of Dutch target firms´ stocks have an influence on the wealth effects. In their research, Achleitner et al. (2010) found a significant positive relationship. This paper, however, found no significant results. This is remarkable, because PE firms have a reputation to buy undervalued stocks and boost their value within 3 to 5 years in order to sell their stake with a profit.
