Private equity and inorganic growth strategies : the effects of add-on acquisition on deal performance

(1)

Private Equity and Inorganic Growth Strategies:

The Effects of Add-On Acquisition on Deal Performance

By Belle van Oostenbruggen Master Thesis

Msc Finance: Quantitative Finance Thesis supervisor: dr. J.K. (Jens) Martin

(2)

Statement of Originality

This document is written by Student Belle van Oostenbruggen who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

In the last decade, the private equity (PE) market has experienced increasing

inflows of capital. Although private investors are progressively seeking for PE funds to diversify their investment returns, the actual value creation techniques that PE investors apply are still to be uncovered. As one of the primary growth channels, inorganic growth strategies have recently been acknowledged as an essential driver of deal performance. Here, add-on acquisitions are used to increase the market potential of portfolio firms. Despite recent studies trying to grasp the synergy power of these strategies, only a few have been able to explore the effects in more detail. This study uses a unique hand-collected sample of 1,126 private equity deals, of which entry years ranges between 1979 and 2008. Findings suggest that increasing numbers of add-on acquisitions positively affect deal performance. Although

difficult to isolate, results also indicate that cross-industry targets enhance PE deal performance significantly, mainly in the case of primary buyouts. Secondary

buyouts, on the contrary, appear to benefit most from add-on similarities regarding industry and country characteristics.

Keywords: private equity, inorganic growth strategies, add-on acquisitions,

(4)

1. Introduction

Over the past several years, the private equity (PE) industry has been proliferating, both in size and importance. Due to low interest rates and economic stability, the PE industry experiences growing inflows of fresh capital, often resulting in a situation in which PE investors are raising more money than they can spend (Espinoza, 2018). However, from the PE investor’s perspective, the competition within the industry has become fierce, with high asset prices, limited investment opportunities and increasing dry powder1_{– translating into unallocated cash}

reserves. With a total market value of funds at an all-time high of 2.83 trillion dollars as of June 20172_{, PE has been acknowledged as an essential investment}

vehicle nowadays.

PE has been characterised as a driver of economic growth and competitiveness. Bernstein, Lerner, Sorensen and Strömberg (2017) find that industries with high PE activity have grown more quickly regarding total production and employment. Besides, Frontier Economics (2013) concludes that PE contributes to the creation of 5,600 new businesses in Europe every year. Furthermore, they state that

PE-backed companies are 50% less likely to fail compared to a control group of similar non-PE backed companies.

On the other hand, PE is also often linked to net job losses and the considerable use of leverage. David, Haltiwanger, Handley, Jarmin, Lerner and Miranda (2014) analyse buyouts of 3,200 target firms between 1980 and 2005. Here, they compare financial data before and after the acquisition by PE firms and find that

employment shrunk by six per cent over five years, relative to a comparable control group. Also, debt is often used to finance the buyout because it allows the investor to control the majority of stock without having to invest dramatically. Besides, return on investment increases significantly when using a high percentage of debt.

1_{KPMG Alternative investments 3.0 (2018)}

(7)

In this way, portfolio companies can be left with unhealthy high ratios of debt after the exit of the PE investor.

Although its persistently increasing volume, it is still relatively unclear how PE investors create value and obtain their returns. Here, a general distinction can be made between organic (internal) and inorganic (external) growth strategies. In the case of the former, the PE firm pursues a strategy in which the operations of the portfolio firm itself are expanded. Using an inorganic growth strategy, on the contrary, the PE firm relies on the acquisitions of related companies (so-called add-on acquisitiadd-ons) to fuse them with the portfolio firm. The underlying motivatiadd-on for so-called mergers and acquisitions (M&A) activity is often the case of synergy, the potential financial benefits of combining multiple businesses.

This study aims to contribute to the existing literature by further examining the effects of inorganic growth strategies on PE deal performance. To operationalise this question, several aspects regarding inorganic growth strategies will be tested, such as the effect of entry timing, holding period and (dis)similarities of add-on

acquisitions on deal performance. The dataset of Degeorge, Martin, and Phalippou (2016) is used as a starting point, after which the method of Nikoskelainen and Wright (2007) is applied to retrieve add-on data. This data collection process is crucial to do the research and thereby offer unique new insights regarding the topic of this study.

The set-up of this master thesis is as follows: first, findings from previous and relevant research on the topic are discussed (section 2). At the end of this section, also the hypotheses are stated that are based on findings so far. Section 3 continues continue with the methods that are used to retrieve the data set, which can be viewed as a considerable large part of this study. Here, also the descriptive

statistics of the final sample can be found. In the fourth section, the methods that are used to test the stated hypotheses are described. After that, the empirical results are presented in section 5. At the end of this study (section 6 and 7), the discussion and conclusion describe implications for future research and make some final remarks regarding the research findings.

(8)

2. Literature Review

Despite the PE asset class growing at fast pace, research regarding fund

performance and growth tactics still provide inconclusive results. The main obstacle in examining the effects of PE ownership on deal performance is the lack of publicly available data on the subject. This lack of information is primarily due to the fact that PE managers are generally not obliged to any reporting requirements. In the next section, the limited and relevant research findings on PE will be discussed. Afterwards, the hypotheses will be stated.

2.1 Private equity

The main idea behind PE is to buy a certain amount of equity (stake) of a private company and exit, after a certain period, while realising a profit. During the so-called holding period, different strategies can be applied to grow and optimise the portfolio company's business and thus increase the value of the assets. The holding period generally takes up three to five years, after which there are various ways to exit the deal. The reason for this relatively short time is the way in which PE funds are arranged: they often invest for institutional investors and return their

investments within a predetermined period.

Private equity consists of a wide range of alternative investment methods. It includes funds specialised on (leveraged) buyout, growth equity, venture capital (VC), but is also used to describe angel (private) investors. These different types of PE investment vehicles generally have different approaches and target companies in which they invest. For example, VC funds generally invest in start-up companies in high growth sectors, whereas growth equity funds often engage in more mature businesses that are seeking ways to scale their operations and enter into new

markets. In this research paper, the term PE fund can generally be interpreted as a (leveraged) buyout fund that invests in relatively mature firms.

The term buyout (sometimes also referred to as BO) generally refers to gaining controlling interest in a company. Here, a distinction can be made between

(9)

management buyouts (MBO) and leveraged buyouts (LBO). In the former, the majority of shares of a particular company are acquired by their management team. In the case of an LBO, the majority of shares are bought by investors that use a significant amount of debt to finance their investment. One of the reasons to use debt as a financial resource stems from the free cash flow theory: leverage reduces the agency costs that arise when managers have to deal with excess cash flows (Jensen, 1986). Because leverage obligates the firm to pay periodic interest costs, this will reduce the need for monitoring from a stakeholders' perspective. Besides, the threat of bankruptcy (by not being able to pay interest) encourages the company to improve its operating efficiency and thereby increase the value of the firm. Other reasons to use debt are to create a tax shield by allowing the portfolio firm to deduct interest payments as expenses (Kaplan, 1989; Graham, 2000), and increasing the rate of return from a PE investor’s perspective.

When it comes to the seller side of a PE buyout deal, a distinction can be made between two types of sellers: in a primary buyout (PBO), the seller in a PE deal is generally an individual investor, the government or the company itself. On the contrary, in a secondary buyout (SBO), a PE firm sells its stake (interest) in a company to another PE firm. Here, the question has arisen to which extent another PE firm can add value to a company that has already been in the hands of another PE firm for a considerable time.

After a predetermined time, the PE firm will exit the equity to gain (hopefully positive) returns and distribute a share of it back to its investors. Here, multiple exit strategies are possible, such as a trade sale, initial public offering (IPO) or the earlier mentioned secondary buyout to another PE firm. In a trade sale, the PE firm sells its stake to a so-called trade buyer, which is a party that operates in the same industry as the exited portfolio company. In an initial public offering (IPO), the shares of the (private) portfolio firm are offered to the public for sale.

The PE firm generally has two sources of income. Firstly, it raises profits by taking management fees before any investment has been made (usually 1.5% to

(10)

2.0%). Secondly, performance fees of approximately 20 per cent are subtracted from any realised profits (value difference between entry and time of exit) after exit.

2.2 Performance

The results regarding PE fund performance seem somewhat ambiguous. Harris, Jenkinson, and Kaplan (2014) have studied the performance of nearly 1,400 U.S. PE funds formed between 1984 and 2008. They found that these funds outperform the S&P 500 with at least 20 per cent over the life of the fund. Annually, this corresponds to 3 per cent per year. Also, Robinson and Sensoy (2016) investigated PE cash flows that were obtained from 837 buyout- and VC funds. They found that, on average, buyouts outrun the market by 18 per cent (here, they also used the S&P 500 as a benchmark). However, Phalippou (2013) argued that the S&P 500 does not meet the requirements of an appropriate benchmark, as the enterprise value (EV) of PE companies is substantially smaller compared to those of the listed companies in the S&P 500 index. Therefore, he used the small-cap value index as a benchmark instead. Using a sample of 392 buyout funds between 1993 and 2010, Phalippou estimated that these funds underperform the small-cap index with 3.1% per year. Next, Kaplan and Schoar (2005) found that the performance of the 746 PE funds in their sample increased with both fund size and experience of the general partner.

Regarding consistency of returns, Braun, Jenkinson and Stoff (2017) also found the persistence of PE returns to be decreasing over the years, which would be primarily caused by the maturing sector and accompanying competitiveness.

Regarding the case of SBOs, Achleitner and Figge (2014) also tested if, compared to PBO deals, any significant differences in equity returns could be detected. To do so, they used a sample of 2,456 buyout deals, of which 448 were labelled as SBO. They concluded that secondary buyouts generate only slightly lower deal returns compared to primary buyouts. Bonini (2015) found similar

results, even in the case of already positive SBO deal returns. However, Degeorge et al. (2016) concluded that this is predominantly the case when PE investors are under pressure to spend.

(11)

2.3 Determinants of return

2.3.1 Strategies

Besides the effects of PE activity on the economy and deal returns, researchers (and investors) are mostly interested in the components that push PE performance. Gompers, Kaplan and Mukharlyamov (2016) tried to answer this question by interviewing 79 PE investors, with joined assets under management of over 750 billion dollars. When asked about their strategies regarding value creation, the following methods were mentioned (in descending order of importance): increasing revenues, improving governance and incentives, providing assistance during the chosen exit route, carrying out add-on acquisitions, reinstating the management team and cutting costs.

As already mentioned, a clear distinction can be made between organic and inorganic growth strategies. Regarding organic growth, one can think of methods to improve operations (and increase earnings) by adding additional financial resources and increase management skills. Acharya, Gottschalg, Hahn and Kehoe (2012) used a combined dataset of 395 PE deals to determine if PE houses indeed use value creation techniques to improve the performance of portfolio companies. They found a positive correlation between EBITDA margins and deal performance. Also,

increased revenues were found to be positively related to deal performance. On the contrary, Guo, Hotchkiss and Song (2011) found quite the opposite results. As they examined the effect of PE ownership on firm value in 192 LBO deals between 1990 and 2006, they found only little improvements in operating performance compared to a benchmark of similar pre-buyout firms.

Another method to increase business operations is to acquire other add-on companies during the holding period of the PE firm, known as inorganic growth strategy. Hammer, Knauer, Pflücke and Schwetzler (2017) found that the case of add-on acquisitions is more likely in certain circumstances, such as when the PE sponsor is experienced with reputational capital, the portfolio firm being relatively large and the firm’s industry being moderately fragmented. Also, in case of SBOs in

(12)

which add-ons have already been acquired in the previous buyout, it is likely the case that the new buyer continues the inorganic growth strategy.

Regarding firm performance, Valkama, Maula, Nikoskelainen and Wright (2013) examined the subsample of 321 buyouts in the UK that took place between 1995 and 2004. They found that both add-on acquisition by the portfolio company and the portfolio company size have a significantly positive effect on firm

performance, measured both by enterprise value (EV) and internal rates of return (IRR). Also, Nikoskelainen and Wright (2007) examined the effect of add-on

acquisition on leveraged buyouts. As they hand-collected detailed data on 321 exited buyouts in the UK, they were able to investigate the effects of a comprehensive set of variables on deal performance. Among other things, they found add-on

acquisitions and target size to have a positive effect on deal returns.

The buy-and-build strategy is often used when applying an inorganic growth strategy. Here, the PE firm acquirers an often well-established firm that will

function as a so-called platform. Generally, this platform firm has some competitive assets and capabilities in place that can be taken advantage of in follow-up

acquisitions. The platform firm is generally bought early for this strategy to become successful.

2.3.2 Mergers and acquisitions

According to Bloomberg (2018), global merger and acquisition (M&A) activity reached 828.6 billion dollars in the first quarter of 2018. Compared to the same period one year ago, this means an increase of 10.3 per cent. From a shareholder theory perspective, managers should always undertake those actions that are in the best interest of the firm’s shareholders (Jensen, 2002). Here, increasing the market value of the firm (that is, all financial claims on the firm) in the long run is, in general, a desirable outcome. Therefore, based on this theory, any activities regarding M&A should be the result of value maximising decision-making.

Advantages of mergers and acquisitions can arise in a variety of ways. For instance, from an efficiency perspective, M&A activities can lead to synergies and economies

(13)

of scale. Besides, M&A could provide diversifying benefits in the case of

cross-industry takeovers, as this could lead to a reduction in risk exposure. Also, merging activities can increase a firms’ market power and thereby reduce competition.

Devos, Kadapakkam and Krishnamurthy (2009) have analysed the relative effects of three types of benefits regarding merging firms: i) productive efficiencies, ii) tax benefits and iii) increased market power. Their sample consisted of 264 mergers of unregulated industrial firms in between the years 1980 and 2004. In their sample, they found a highly significant average total synergy of 10.03 per cent. Of this high percentage in total gains, they also uncovered the relative effects of the three benefits mentioned above: only 1.64 per cent could be allocated to tax gains, whereas operating synergies caused the most substantial portion (8.38 per cent).

2.3.3 Intra-industry acquisition

The rationale behind intra-industrial acquisitions is the idea that because of the merging firms' size, they can benefit from economies of scale, market power and valuation improvements in their respective industry. In their paper, Hammer et al. (2017) researched the likelihood of PE deals to engage in intra-industrial

acquisitions. They found that PE investor- and portfolio firm’ experience in previous intra-industrial acquisitions increases the case of industry-specific inorganic growth strategies.

On the contrary, as mentioned by Humphery-Jenner (2013), cross-industrial diversification could also lead to increased knowledge sharing and networking gains. Also, the diversifying character of inter-industrial mergers can reduce risk.

2.3.4 Cross-border expansion

Also, so-called cross-border mergers and acquisitions are becoming more and more important, accounting for over 46 per cent of all M&A activity in 20173_{. Espenlaub,}

Khurshed and Mohamed (2014) have studied a subsample of 4502 VC investments,

(14)

both intra-country and cross-border. Here, they found that cross-border investments are exited significantly sooner, whereby similar effects have been found in both deals that exit via trade sale and IPO. However, they also pointed to the fact that this effect is primarily driven by the results of cross-border activities in North

America. Also, Hammer et al. (2017) found that both PE investor- and portfolio firm’ experience in recent cross-border activity increases the case of international add-on acquisitions in a certain PE deal.

2.3.5 Holding period

The holding period (also referred to as duration) of a deal is the time in which the PE firm is invested in the portfolio company. As a fund has a predetermined

lifetime, the time to implement suitable (organic and inorganic) growth strategies is only limited. Therefore, one could imagine that the holding period is a crucial aspect determining the successfulness of the PE firm intervention. The average holding period of PE funds generally varies broadly, although often indicated to be between three to five years (Acharya et al., 2013, Degeorge et al., 2016, Kaplan & Schoar, 2005).

Lopez-De-Silanes, Phalippou and Gottschalg (2012) examined the effect of duration on deal returns. By comparing the average IRR of investments with different holding periods, they found a significantly negative relationship between holding period and deal performance. This is primarily due to the idea that it is more difficult to obtain a high IRR over a more extended period.

Regarding the difference between the holding periods of PBOs versus SBOs, Degeorge et al. (2013) also found that, in general, the holding period of SBOs is longer than that of other types of buyouts. A reason for this could be the fact that SBO deals generally apply inorganic growth strategies, as the organic growth approaches have most likely already been applied by previous PE investors.

In conclusion, existing literature implies a positive relationship between add-on acquisition and deal performance. Because relatively few studies have looked at the

(15)

effect if intra-industrial and cross-border add-on activity on deal performance, this study will offer new insights into the matter.

2.4 Hypotheses

After reviewing relevant past research findings on the topic, the following seven hypotheses have been derived. These predictions will be tested to extend the understandings of inorganic growth strategies and PE deal performance.

Firstly, this study will try to confirm some of the assumptions regarding the specific deal types that choose inorganic growth strategies. The first hypothesis predicts that primarily SBOs and deals with relatively long holding periods engage in add-on activity. Regarding SBOs, this is highly likely because of the idea that previous PE investors have already achieved efficiency gains. Therefore, the ‘buying’ PE managers in SBOs deals will consider inorganic growth strategies to further increase the value of their investment (Hammer et al., 2016). Secondly, as inorganic growth strategies generally require time to be executed, deals with short holding periods (i.e., less than two years) are considered to mostly focus on internal growth strategies. Hence, the first hypothesis is stated as follows:

Hypothesis I: The propensity of deals to apply inorganic growth strategies is higher among those that are labelled as SBO and have a relatively extended holding

period.

Secondly, this study will try to verify the positive relationship between

inorganic growth strategies and deal performance that was found in past literature (Hammer et al., 2016; Nikoskelainen & Wright, 2007; Valkama et al., 2013).

Consequently, the second hypothesis is stated as follows:

Hypothesis II: Deals applying inorganic growth strategies generate higher deal returns relative to deals applying organic growth strategies.

(16)

Next, the presumably adverse effects of holding period will be tested (Lopez-De-Silanes et al., 2012). This, negative correlation is expected because high growth rates (IRR) are more difficult to obtain for more extended periods of time. Therefore, the third hypothesis is stated as follows:

Hypothesis III: In general, the holding period of a PE deal correlates negatively with deal performance.

The fourth hypothesis will test the theory regarding the implementation time that is generally required to make inorganic growth strategies successful. Here, different spects will be considered. Firstly, the effects of portfolio firms that are bought early, i.e. within 2,5 years of the fund’s vintage year, and apply inorganic growth strategies will be tested on deal performance. The early versus late

distinction will be based on the used method of Degeorge et al. (2016). Secondly, the findings of Achleitner and Figge (2014) will be assessed regarding the (slightly) lower deal returns of SBOs compared to PBOs. Thirdly, as past findings suggest that SBOs generally pursue inorganic growth strategies (Hammer et al., 2015), the successfulness of SBOs engaging in add-on activity will also be added to the

analysis. Lastly, the different combinations of deal type and entry timing will be added to gain more insights. Thus, the fourth hypothesis is stated as follows:

Hypothesis IV: Portfolio firms that pursue inorganic growth strategies and are bought early generate higher deal returns compared to those that are bought late,

especially in the case of SBOs.

Regarding the fifth hypothesis, the synergy effects of intra-industry merger and acquisition activities are to be confirmed (Devos et al., 2009). Add-ons in which the firm to be merged with is active in the same industry will be referred to as intra-industry add-ons in this study. In all other cases, add-ons will be labelled as

(17)

cross-industry acquisitions. Because of the relatively similar business models of the portfolio firm and intra-industry add-on(s), the process of merging the two firms should be considered more straightforward and, therefore, economic benefits should increase (Hammer et al., 2017). Also, the case of relatively early acquisition and extended holding period should enlarge these positive effects, as the time to execute the inorganic growth strategy successfully increases. Therefore, the fifth hypothesis is stated as follows:

Hypothesis V: Intra-industrial add-on acquisitions correlate positively with deal performance, especially in the case of deals that are bought early and

experience relatively long holding periods.

The sixth hypothesis tests the effect of international acquisitions on deal return. As Humphery-Jenner et al. (2016) concluded, PE investors applying this strategy generally have more experience in international M&A activities and often enjoy an extensive network in the add-on country. Therefore, the case of cross-border

acquisitions in PE deals is expected to be generally successful. Hence, the sixth hypothesis is stated as follows:

Hypothesis VI: Deals engaging in cross-border inorganic growth strategies perform better in comparison with deals engaging in intra-country inorganic growth

strategies, especially in the case of deals that are entered early and have a relatively extended deal duration.

Lastly, the composition of the different types of add-on acquisitions in PE deals will be tested. Because of the lack of research regarding this topic, the hypothesis is based on the following assumption: engaging in similar add-on acquisitions is

generally assumed to be easier to implement, as the similarities between different ons are more pronounced. Therefore, deals that only focus on (dis)similar

(18)

add-ons are expected to obtain higher deal returns. Therefore, the seventh hypothesis is as follows:

Hypothesis VII: Deals that focus on only one type of add-on acquisition generate higher deal returns compared to those that engage in mixed add-on

(19)

3. Data and descriptive statistics

Due to the lack of data available on the subject, a large part of this study involved collecting the appropriate data. At the heart of this study lies the unique data set of Degeorge et al. (2016), containing deal specific information that is not publicly available but crucial to do this research. Information on add-on activity has been collected manually from data source Capital IQ4_{. Below, the construction of the core}

data sample is explained. Also, the descriptive statistics regarding the dataset are discussed.

3.1 Data sources

The collected data set of Degeorge et al. (2016) serves as the point of departure in this study. Here, the researchers primarily obtained data from Private Placement Memorandums (PPMs), legal documents that are given to potential investors ahead of the sale of stocks or securities in a particular private company. In this dataset, deal level information can be found such as i) the name of PE house, ii) the name of the portfolio company, iii) amount of money invested, iv) type of deal (SBO, BO, Buy-and-build), v) the acquiring month and year, vi) the exit month and year, and vii) the fund size. Also, the dependent variable (deal performance) is subtracted from this dataset, measured as the gross internal rate of return (IRR). This variable can be viewed as a measurement of profitability of potential investments that is often used in research papers.

However, this dataset does not include any M&A information regarding the portfolio firms. That is, there is no information about any listed add-on activities that a portfolio firm undertook during the holding period of the PE house.

Therefore, the Capital IQ database has been used to collect this type of data. Using Capital IQ, information can be obtained on all transactions of both listed and

unlisted firms that were available in their database. As portfolio firms are generally

(20)

on the buy-side in case of any add-on acquisition, tracing back any inorganic growth behaviour via this database is an appropriate method. Among others, important deal information that can be found in this database are i) the name of PE house, ii) the name of the portfolio company, iii) amount of money invested, iv) the acquiring month and year, vi) the exit month and year, v) information regarding add-on firms (i.e., industry and country characteristics) and vi) exit method.

3.2 Data collection process and preparation

The data collection process included different phases. As already mentioned, the data set of Degeorge et al. (2016) is used as a backbone. First, the full list of 10,183 deals had to be complemented with the corresponding Capital IQ identification number. In this first step, different problems arose. Firstly, as the deals in this data set took place in between 1969 and 2010, one can imagine that quite a substantial number of companies in the sample changed their entity name at least once.

Therefore, finding the corresponding entity name was sometimes a complicated and lengthy process. The second factor that made this job somewhat complicated was that some Private Placement Memorandums (PPMs) did not provide the correct legal entity name or the dataset itself included errors, which made the task of

finding the correct company profile more difficult. Third and lastly, there were quite some companies in the data set that carried seemingly generic entity names. This fact also caused quite some extra lookup work, as there were sometimes over ten different companies with almost identical business names. In the end, of 7,571 deals both portfolio company and PE investor were identified in the Capital IQ database, and therefore the corresponding identification number could be retrieved.

The second challenge facing the data set of Degeorge et al. (2016) was to match the portfolio company and PE investor to find the right transaction. In every deal, there is generally an entry and exit transaction; that is, a date on which the PE investor acquirers the portfolio company and a date on which the investor exits its initial investment and sells its stake again. If one of these two transactions could not be found in the Capital IQ database, the given holding period from the data set

(21)

could not be verified, and therefore the specific deal was not eligible for this

analysis. From the 7,571 deals, only 2,257 deals met this requirement. At this stage, all duplicates were dropped as well, which led to the final data set consisting of 2,007 deals.

Afterwards, data on add-on acquisitions was collected. A transaction can be defined as ‘add-on’ (hence, an inorganic growth strategy) if a portfolio company acquires another company during the PE holding period. Conveniently, the portfolio firm is always mentioned on the ‘buyer’-side in case of an add-on transaction and, therefore, these deals can easily be traced. If no add-on acquisitions were found in the used data source, it was assumed that the PE firm only applied organic growth strategies during the deal. Also, add-on acquisitions would only be added to the database in the case that the data source would report a closing date. Firstly, all transactions were collected of which the target company was tagged as the buyer and of which the transaction type was labelled ‘merger/acquisition'. Secondly, those transactions were filtered that took place within the holding period of the PE firm. That is, any transaction date that did not fall within the holding period of the PE deal was filtered out of the dataset. In the end, the transaction details from the final set of add-ons were retrieved, such as: i) the name of portfolio firm, ii) industry of portfolio firm, iii) amount of money invested and iv) country of incorporation.

Next, dummy variables were created that indicated if a portfolio firm applied an inorganic growth strategy (acquired at least one add-on company). If so, the number of add-ons were added in a separate column too. Besides, details on the add-on acquisitions were collected, such as industry and country of origin.

Lastly, two filters were applied to come up with the final data set. The first filter was applied to ensure that the entry- and exit date on the PPM did not differ substantially from those that were subtracted from the Capital IQ database. As a rule of thumb, the maximum difference allowed between the two dates was set at 11 months, as several deals only contained the entry- and exit year in the original dataset. In this way, the deals would not be filtered out if they only matched the entry- and exit year but were missing the corresponding month data. The second

(22)

filter tested for the availability of the internal rate of return (IRR) measurement, as this would be used as the dependent variable in future analyses. After applying these two final filters, 1,126 deals were available for inquiry.

3.3 Descriptive statistics

After the data collection has been completed, the remaining sample consisted of 1,126 observations, of which 236 deals showed add-on activity.

In the first table, the descriptive statistics can be found in which the original dataset of Degeorge et al. (2016) is compared to the collected sample. Here, the main steps of the data collection process are shown with the corresponding numbers of deals that met the different requirements.

Table 1

In the second table, deal characteristics can be found regarding entry year and add-on activity. In the used sample, deals can be found with entry year ranging from 1979 to 2008. However, the biggest portion of deals has been completed in the years 1992 to 2005 (with at least 30 deals per year). Regarding inorganic growth strategy, the most substantial portion of add-on activity took place between 1996 and 2004, whereas PE portfolio firms acquired almost zero add-ons before 1990. The average holding period of the sample (46.34 or just over three years) is in line with previous findings.

Table 2

In the third table, the descriptive statistics regarding the holding period, entry channel and investment timing can be found. Here, it can be viewed that add-on activities seem to occur more often when the holding period of deals increases. However, the average amount of add-ons remains to be stable.

Table 3

Table 4 displays the different entry channels of the complete set of deals. Here, the most substantial part of deals can be viewed as primary buy-outs. Also, the main exit route of the used sample is categorised as trade sale, whereas selling the stake to another PE firm comes second.

(23)

Table 4

Lastly, the fifth table displays the mean and standard deviations of the

independent variable of this research: the performance measure Gross IRR. Here, the results of a t-test can be viewed, where the mean Gross IRR has been calculated in different subsamples that hold the same characteristics (entry channel,

investment timing and holding period). The dependent variable has been winsorized at 1% level before starting the regression analyses. As can be viewed, no significant difference can be found between the deal performance of PBO versus SBO deals. Besides, the investment timing and relative holding period appear to have no significant effect on IRR.

Table 5

In conclusion, no distinct pattern is yet to be observed regarding the effect of add-on acquisition on deal performance. Therefore, regression analyses will be used to examine the sample more thoroughly.

(24)

4. Methodology

The objective of this study is to analyse different aspects regarding inorganic growth strategies and the performance of PE deals. In this part, the statistical methods that are used in this study will be described. Also, the key variables will be featured.

In this study, a probit regression has been run to test for add-on behaviour. Using this type of regression, the probability that a deal with specific features falls within the inorganic growth category is tested. Next, the ordinary least square (OLS) method satisfies the statistical needs to analyse the obtained data set

further. Here, the parameters of the chosen variables are estimated by minimising the sum of squared residuals between a predicted linear function and the given data points. In addition to the key dependent and independent variables used, additional control variables and fixed effects have been added to increase the reliability of the results.

4.1 Testing the add-on propensity hypothesis

Firstly, a probit regression has been run to test the propensity of certain types of deals to apply inorganic growth strategies.

!""# = %&+ %()*+#+ %,-./01# + %23456# + %7)*+-./01# + %8)*+3.9:#+ %;<*+-./01#

+ =459/40 ?./@.A0:B + C@D:" :CC:=9B

The probit regression stated above consists of only dummy variables. Here, )*+#

indicates whether the seller in the PE transaction is another PE investor (such as in the case of a secondary or tertiary buyout). In the case of a primary buyout, the dummy variable is therefore equal to zero. Furthermore, the variable EarlyJ

indicates the timing of the PE entry date: if this takes places within the first 2.5 years of the fund lifetime (that is, after the vintage year), the dummy variable equals one. In all other cases, the dummy variable takes the value of zero. The

(25)

variable 3456# will test the propensity of deals with a holding period of more than

two years to apply inorganic growth strategies. Also, combination dummies are added that indicate the combination of deal type and the timing ()*+-./01#,

)*+3.9:_# and <*+-./01_#). A dummy regarding the case of a PBO deal that is acquired late has been omitted because of collinearity issues. This aspect has been taken into account in the other regressions too.

Regarding control variables, the following variables have been added in every regression: holding period (in months), investment size (in U.S. dollars or USD), fund size (in USD) and a dummy for club deal. The holding period can be easily obtained by calculating the difference (in months or years) between the entry and exit date of the deal. Regarding the second control variable, some adjustments had to be made. As the investment size is generally expressed in the local currency, they have been converted to USD using data on exchange rates from Compustat5_{. The}

fund size is added to control for the level of experience of the PE investor. Besides, the club deal dummy corresponds to the case in which at least two different PE firms are involved in the deal. Therefore, this variable is also included to control for PE management experience. In the following regression descriptions, the control variables will not be mentioned explicitly anymore.

The fixed effects that have been included in this regression can be categorised as company-, country- and industry fixed effects. Firstly, the PE house has been added to control for PE investor specific characteristics. Also, the countries of the PE house and portfolio company have been added to take country-specific issues into account. Lastly, in the probit regression, there has been controlled for the industry of the portfolio firm.

4.2 Testing the inorganic growth hypothesis

Afterwards, the effects of applying an inorganic growth strategy will be tested. Here, a dummy variable has been constructed. The variable !""# takes the value

(26)

one if one or more add-on acquisitions take place in which the portfolio firm is on the buyer side of the deal during the holding period. If this is not the case, the dummy variable is equal to zero. As this is the primary variable of interest, this variable has not been tested isolated.

4.3 Testing the duration hypothesis

To test the hypothesis regarding holding period effects, also this variable has also been added to the other regression analyses. Here, the holding period consists of the number of months between the entry and exit date of the portfolio firm.

4.4 Testing the secondary buyout and timing hypothesis

Effects of investment timing and buyout firms has been tested in the following regression. Here, different sets of variables has been added to the regression in separate stages.

LMM_# = %_&+ %₍!""_# + %_,!""_NOPQ,#+ %₂-./01_# + %₇)*+_#+ %₈)*+-./01_#+ %_;)*+3.9:_# + %_R<*+-./01_# + %_S!""_NOPQ,# ∗ -./01_#+ %_U!""_NOPQ,# ∗ )*+_#+ %_(&!""_NOPQ,# ∗ )*+-./01_#+ %₍₍!""_NOPQ,#∗ )*+3.9:_#+ %_(,!""_NOPQ,#∗ <*+-./01_#

+ =459/40B ?./@.A0:B + C@D:" :CC:=9B

Here, the dummy variable !""_# and the total amount of add-on acquisitions (!""NOPQ,#) test for inorganic growth strategies as mentioned above. Besides, the

dummy Early_J has been included to test for investment timing. With this, an interaction term has also been added to the regression to test for the combined effect of add-on activity and early entry on deal performance. Furthermore, the dummy )*+# has been included, as well as dummy’s that indicates the combination

of deal type and investment timing. Also, interaction terms have been used to test for the combined effect of add-on activity and these combination dummies.

The control variables are similar to those mentioned before. The fixed effects that have been used in the next couple of regressions can be split up into four

(27)

different categories: company fixed effects, country fixed effects, industry fixed effects and time fixed effects. Regarding the first, the PE house variable is added to control for any PE firm-specific characteristics. Secondly, the portfolio company- and PE house country are appended to control for country fixed effects. Thirdly, the target firm's industry has been added as a fixed effect too. Lastly, time fixed effects control for any market-wide changes in time, such as an economic crisis. Therefore, the entry year of the PE deal has been included.

4.5 Testing the intra-industry hypothesis

After that, the fifth hypothesis regarding synergy effects has been tested, in which industry characteristics of add-ons are of primary interest.

LMM# = %&+ %(L59/.L5"VB9/1# + %,L59/.L5"VB9/1NOPQ,# + %2W/4BBL5"VB9/1#

+ %7W/4BBL5"VB9/1NOPQ,# + %8-./01#+ %;W/4BBL5"VB9/1NOPQ,#∗ -./01#

+ %_R-./01&)ℎ4/9_# + %_S-./01&3456_#+ %_U3.9:&3456_#

+ %(&W/4BBL5"VB9/1NOPQ,#∗ -./01&)ℎ4/9# + %((W/4BBL5"VB9/1NOPQ,#

∗ -./01&3456# + %(,W/4BBL5"VB9/1NOPQ,# ∗ 3.9:&3456#+ =459/40 ?./@.A0:B

+ C@D:" :CC:=9B

Here, L59/.L5"VB9/1_# is a dummy variable that is equal to one when the portfolio firm acquired at least one intra-industry add-on during the deal’s lifetime. In any other case, the dummy is set equal to zero. Here, also the total amount of intra-industry add-ons have been included (L59/.L5"VB9/1NOPQ,#). Next, cross-industry

add-on acquisitiadd-on variables have been added, both as a dummy (W/4BBL5"VB9/1#) and a

continuous variable (W/4BBL5"VB9/1_NOPQ,#). Also, dummy variables have been added that match the investment timing (early versus late) and holding period (short versus long). Here, as a rule of thumb, deals can be viewed as having a considerably short holding period if there is a maximum of two years (or 24 months) between the entry and exit date of the PE deal. Interaction terms have also been added to

(28)

increase the understandings regarding cross-industry acquisitions and investment timing, combined with duration.

Regarding control variables and fixed effects, the same variables have been used as mentioned above (section 4.4).

4.6 Testing the cross-border hypothesis

After that, analysis has been done regarding cross-border add-on activity. Here, the following regression has been run.

LMM_# = %_&+ %₍L59/.W4V59/1!""_# + %_,L59/.W4V59/1!""_NOPQ,#+ %₍W/4BBW4V59/1!""_# + %,W/4BBW4V59/1!""NOPQ,# + %2-./01#+ %7W/4BBW4V59/1!""# ∗ -./01#

+ %₈-./01&)ℎ4/9_# + %_;-./01&3456_# + %_R3.9:&3456_#

+ %_SW/4BBW4V59/1!""_# ∗ -./01&)ℎ4/9_#+ %_UW/4BBW4V59/1!""_#

∗ -./01&3456#+ %(&W/4BBW4V59/1!""#∗ 3.9:&3456# + =459/40 ?./@.A0:B

+ C@D:" :CC:=9B

Here, L59/.W4V59/1!""# is a dummy variable indicating if at least one add-on

acquisition is based in the same country as the portfolio firm. Also, the continuous variable that consists of the number of intra-country add-ons is presented in the regression (L59/.W4V59/1!""NOPQ,#). Besides, the dummy (W/4BBW4V59/1!""#) and a

continuous variable (W/4BBW4V59/1!""_NOPQ,#) indicate whether cross-country add-on acquisitions have been used in a certain PE deal. The rest of the variables that have been used in this regression are similar to those of section 4.4 and 4.5.

4.7 Testing the composition hypothesis

Lastly, the composition of add-on strategies has been examined.

LMM# = %&+ %(+501L59/.L5"VB9/1# + %,+501W/4BBL5"VB9/1# + %2Z@D:"L5"VB9/1#

+ %₇+501L59/.W4V59/1_# + %₈+501W/4BBW4V59/1_# + %_;Z@D:"W4V59/1_# + =459/40 ?./@.A0:B + C@D:" :CC:=9B

(29)

Here, +501L59/.L5"VB9/1_# is a dummy indicating if a deal only applies intra-industry inorganic growth strategies. The +501W/4BBL5"VB9/1# indicates the opposite: deals

that only engage in cross-industry add-on activity. The rest of the deals are

gathered in the dummy Z@D:"L5"VB9/1#. These are the deals that acquire both

intra-industry and cross-intra-industry add-ons. Similar dummies have been set up for effects of intra-country versus cross-country characteristics of add-on acquisitions. These regressions have been run on three subsamples: i) all add-on deals, ii) PBO deals and iii) SBO deals.

(30)

5. Empirical Results

The hypotheses have been studied using the regression models stated in the previous section. The regression results are presented in five different tables (see Appendix, table 7-11). In table 7, the estimated coefficients are shown regarding add-on activity, investment timing and SBO performance. In table 8, the results are exhibited regarding add-on industry effects. Here, both the effects of intra-industry and cross-industry add-ons have been tested. Also, the effects of investment timing and holding period have been included in the analysis. Afterwards, the results regarding the international scope of add-on activity (table 9) are discussed. Lastly, in tables 10 and 11, different subsamples have been tested regarding industry and country effects of add-on activity. In all regressions, the standard errors have been clustered by both deal entry year and involved PE house. Also, every regression included both control variables and fixed effects.

5.1 Add-on activity

The first hypothesis states that SBO deals and deals that experience a long holding period have a higher probability of engaging in inorganic growth strategies. These predictions are based on the fact that SBO deals are assumed to be internally optimised already, and therefore value creation techniques focus more on inorganic growth strategies. Besides, more extended holding periods are generally needed to successfully acquire add-on firms.

The propensity of specific deals to use inorganic growth strategies has been tested using a probit regression in table 6. Here, the results imply that deals labelled as SBO (whereby also tertiary buyouts are included) show a higher

probability of engaging in inorganic growth strategies. The effect has been found to be significant at a 10% level. This finding is in line with expectations. Also, deals that are held for a relatively extended time, i.e. more than 24 months, show a higher probability of add-on activity (with a positive coefficient at 5% significance).

(31)

As these deals have a relatively long portfolio lifetime, the process of acquiring another firm is easier to fulfil. Therefore, the first hypothesis is accepted. All other variables show no significant effects on the propensity to engage in add-on

activities.

The second hypothesis regarding the effect of add-on acquisition on deal performance has been examined next. Here, it is expected that inorganic growth strategies lead to higher deal returns.

In table 7, the first two specifications display the regression results testing the effect of inorganic growth strategies on deal performance. In both regressions, the mentioned control variables have been added to the regression analysis. The only difference between the two is the use of the fixed effects (included in specification 2). As can be seen, the effect of add-on activity on deal performance appears to be

positive but insignificant in the first two regressions. In words, this means that deals with inorganic growth strategies show, on average, no significantly higher internal rate of return compared to deals that apply organic growth strategies.

Looking at the third and fourth specification of table 7, also the effect of the number of add-on acquisitions on deal performance has been examined. Here, significant positive estimates have been found, both with and without adding fixed effects to the regression. These findings are significant at 5% and 10% levels, respectively. Therefore, the number of add-ons is positively related to deal performance.

At first glance, it appears to be the case that inorganic growth strategies do not have a significant positive effect on deal performance. However, the estimator turns significantly positive when the number of add-on transactions are considered.

The effect of investment timing regarding deals applying inorganic growth strategies has been tested in the fifth specification of table 7. As the third hypothesis predicted, deals that are acquired early on in the fund’s lifetime are

(32)

expected to obtain higher deal returns compared to deals that are acquired late. Looking at the results, however, no significant effect can be found regarding the deal performance of portfolio companies that are acquired in the first 2,5 years of the fund’s lifetime. Because buy-and-build strategies can be viewed as those deals that are acquired early and acquire a relatively large amount of add-on acquisitions, especially the interaction term in specification five is of interest. As the interaction term in this specification shows no significant estimator, the effect of this strategy on deal performance appears to be neutral.

5.2 Secondary buyout

In specifications 6 to 8 of, the results regarding the effect of SBOs on deal performance can be seen. Firstly, the results of Achleitner and Figge (2014) regarding the effects of SBOs on deal performance show more cogent findings. As the coefficient of the SBO variable shows to be negative and significant at a 1% level, it can be concluded that SBO deals appear to have lower deal performance overall.

Next, SBOs that apply inorganic growth strategies have been examined more thoroughly in specification 7. Here, the estimated coefficient of the SBO dummy remains to be significantly negative (at 10% level). Besides, applying inorganic growth strategies in SBO deals appears to be one of the drivers of this negative correlation. These findings are interesting indeed, as inorganic growth would generally be thought of as the most straightforward approach for SBO deals to obtain positive returns.

In specification 8, results regarding deal type and investment timing can be viewed. Here, results are rather conspicuous. Firstly, SBO deals remain to show lower deal performance, regardless of the investment timing. On the contrary, the case of early investment in PBO deals doesn't appear to have any significant effect on deal performance. Also, none of the interaction variables is found to be

significant. That is, although inorganic growth strategies appear to be one of the drivers of the negative correlation of SBO deals on performance, the investment

(33)

timing does not appear to be of influence. After reviewing both specification 5 and 8, one can state that no support has been found regarding the fourth hypothesis.

In conclusion, the effect of add-on activity on deal performance remains

somewhat ambiguous. Although all estimated coefficients regarding the two add-on activity variables show positive estimates, only three out of seven estimates tested significant. Here, it can be stated that applying an inorganic growth strategy does not necessarily increase deal performance. However, increasing the number of add-ons during the deal’s lifetime does appear to boost return. Also, SBO deals show significantly lower deal returns compared to PBOs, whereby add-on activity can also be viewed as a significant driver. Lastly, the effect of investment timing does not influence the performance of deals applying inorganic growth strategies.

5.3 Intra-industry activity

The fifth hypothesis predicts the positive effect of intra-industry add-on activity. From an intuitive perspective, synergy effects should become more pronounced in mergers between firms with similar industry backgrounds. Here, both the case of vertical and horizontal integration benefits would become more obvious.

In table 8, the regression results regarding these intra-industry synergy effects can be found. In the first four specifications, the separate effects of both

intra-industry add-ons and cross-intra-industry add-ons have been tested. Here, it is important to note that deals can engage in both types of industry add-ons. Surprisingly, it appears to be the case that cross-industry acquisitions, instead of intra-industry add-ons, effect PE deal performance positively. Also here, the effect is dependent on the number of cross-industry add-ons (specification 4). In words, an additional cross-industry add-on increases deal returns with 13 percentage points (significant at 10% level).

In specification 5, this effect has been tested more thoroughly. Here, also the investment timing has been included. Similar to add-on deals in general, also here

(34)

the estimated coefficients turn out to be insignificant. That is, investment timing does not appear to have a significant effect on the successfulness of cross-industry add-on acquisitions.

In the last column of table 8, investment timing and holding period are also tested, together with the number of cross-industry add-ons. Here, it can be viewed that deals that are acquired early and experience short duration have significantly higher deal performance compared to the rest of the sample. In words, the IRR increases with 65 percentage points in case of early investment and a maximum holding period of two years. This effect has been found to be significant at 5% level. This finding can be attributed to deals that are labelled as ‘quick flips’, in which PE firms exit platform firms again in less than two years. However, interacting this variable with the number of cross-industry add-ons results in an insignificant effect. Thereby, the buy-and-build strategy for this type of PE deals cannot be confirmed. That is, it is not possible to identify cross-industry inorganic growth strategy as the specific action that is undertaken to increase the performance of these short-lived deals. Additionally, cross-industry add-on acquisition correlates significantly negative within deals that are bought early and hold for a considerable time. This finding would suggest that buy-and-build deals that engage in cross-industry add-on activities experience lower deal returns. Besides, the dummy indicating deals that are bought late and experience long holding periods also shows to have a significant positive effect on deal returns (at 1-per cent level). In words, these deals enjoy 75 percentage point higher IRR.

To conclude, cross-industry add-on activity appears to have a positive effect on deal performance, unless the deal is bought early and is kept within the PE

portfolio for a considerable time. Additionally, the number of cross-industry add-ons enhances the deal performance considerably too. These results are not in line with the previously stated predictions, and therefore the fifth hypothesis is rejected.

(35)

5.4 Cross-border activity

In hypothesis 6, deals that engage in cross-border add-on activity are predicted to perform better compared to both deals with local add-on activity and those that don't apply any inorganic growth strategy. One of the critical drivers according to theory, is that the international scope of add-ons increases the market value of the portfolio firm. Also, as PE managers are expected to have significant experience in border acquisitions and enjoy an extensive international network, cross-national add-ons are likely to be more successful.

In table 9, the effects of international add-on activity are displayed. In the first two regression, the case of intra-country add-on activity has been examined,

together with controls and fixed effects. Here, although engaging in intra-country add-on activities appears to have a significantly negative effect on deal returns, increasing the number of intra-country add-ons does imply a positive effect on deal performance whatsoever. Regarding cross-country add-ons (specification 3-4), the results are quite the opposite: as the dummy variable is significantly positive at a 10-per cent level, the enhancing effects seem to disappear when the number of cross-country add-ons is considered. These results could be based on the fact that market value increases significantly when portfolio firms become international, but the synergy effects of additional cross-border acquisitions are more difficult to obtain. Therefore, the effects of the number of cross-country add-ons remains neutral. Also, when other variables are added to the regression (specification 5-6), the significant estimator of cross-country add-on activity remains positive but insignificant. Regarding the interaction terms, one significant finding can be found: those deals that engage in cross-country add-on acquisition, are bought early and are short-lived, experience significantly greater deal returns compared to the rest of the sample. That is, IRR increased with 2.35% (which has been found to be

significant at 10-per cent level). Therefore, the greater deal performance of these short-lived deals can be partly explained by cross-border add-on activity.

(36)

The overall findings partly confirm the sixth hypothesis. Although international add-on activity does increase the deal performance on average, it is not dependent on the number of cross-country add-on acquisitions. Also, the combination of early investment and short holding period significantly increases the effect of

international add-on expansion. These findings could point out to a particular PE strategy in these types of deals.

5.5 Composition effects

Looking at tables 10 and 11, the effects of industry and country add-ons have been examined in a more isolated setting. As portfolio companies can engage in both types of add-ons during the PE holding period, it is difficult to isolate the effects of one of the two types on deal performance. Therefore, those deals that apply only one of the two methods are taken into account in these regressions.

In table 10, the overall results are presented regarding so-called ‘add-on deals’. Here, it can be viewed that intra-industry add-on deals experience significantly lowers deal returns, compared to the other types of add-on deals (that is, both mixed and cross-industry deals). In words, deal performance drops with 30 percentage points when only intra-industry add-on activity takes place. On the contrary, portfolio firms that only engage in cross-industry add-on acquisitions correlate positively with IRR at a 5-per cent level. For mixed industry add-ons, no significant effect within add-on deals on deal performance can be found. Also, regarding

country effects of add-on acquisitions, no significant effect can be found in table 10 (specification 4-6).

In table 11, results regarding differences in PBO and SBO deals that apply inorganic growth strategies are shown. As the results suggest, engaging in only cross-industry add-on activity results in significantly higher deal performance of PBO deals. The 33-percentage point increase in deal performance has been found to be significant at 10% level. Other than that, no significant effects have been found regarding different add-on compositions in PBO deals.

(37)

Regarding SBO deals, the results are more pronounced. Applying a mix of both similar- and cross-industry add-ons results in significantly lower deal returns compared to the other two tested groups (only similar or only cross-industry add-on activity). Similar findings have been found regarding mixed country add-on activity. Besides, significant positive estimates have been found regarding the effects of acquiring only intra-industry or intra-country add-ons in SBO deals.

In conclusion, the findings support the seventh hypothesis. The significant effects found point to the fact that deals that focus primarily on only one type of add-on acquisitions obtain higher deal returns.

5.6 Holding period

Lastly, the hypothesis regarding the holding period of a deal has been examined in all previously discussed regressions. Here, a significantly negative effect has been found in all estimated outputs. In words, an increase in duration of one month reduces deal performance with two percentage points (significant at 1% level).

Therefore, the third hypothesis regarding the negative correlation between holding period and deal performance can be confirmed.

(38)

6. Discussion

Although this study appears to offer some new and interesting insights into the subject of this thesis, it is essential to consider the issues regarding the

experimental design too.

The first issue regarding the research that has been done is the used sample itself. Firstly, the dataset of Degeorge et al. (2016), that has been used as a starting point, consisted of hand-collected data itself. Potential threats to this way of

gathering data are possible inaccuracy (typos) and variation regarding the degree to which collection guidelines are met. As the data on add-on acquisitions has been collected in similar ways, this increased the chance of an inaccurate representation of the deals that are examined. Also, because of the use of merely one database regarding inorganic growth strategies (namely, Capital IQ), the possibility of

missing or inaccurate information is reasonable. Additionally, because of the lack of reporting requirements for PE managers, databases are often biased towards data input of only those that do report. Therefore, it is highly likely that PE managers tend to report their results only in case of successful deal activity, which can result in a biased representation.

A second drawback regarding the research applied is the use of the gross IRR as a measure of deal performance. One of the underlying assumptions of the IRR is that fund inflows and outflows can be characterised by the same risk. However, in real life this does not appear to be the case: inflows are generally viewed as riskier compared to outflows, resulting in the fact that the performance rate becomes overestimated easily. Secondly, the IRR is sensitive to cash flow timings. As PE funds generally deal with variable cash flows, both in amount and timing, the IRR measure gets unstable fast. Thirdly, one of the fundamental principles of the IRR is the idea that excess cash can be reinvested at the IRR (generated up to that point in time) at any given time. This principle would only hold when PE managers can both invest and borrow at the IRR at any time during the deal's lifetime, which is an impossible assumption.

(39)

Thirdly, the case of omitted variable bias is always a reasonable threat. Because of the limited use of variables in the regression analyses, there is always a risk of not including all appropriate variables at hand. Also, there are always variables missing from the dataset in the first place, as it is nearly impossible to collect all the relevant variables that can be of use. For instance, it was challenging to obtain financial data (Revenues, EBIT, EBITDA) on both portfolio firms and add-ons in the Capital IQ database. Since it is entirely reasonable to suggest that add-on

proportion is an important variable to control for, this decreases the power of the estimated regression outputs.

In addition, the case of reversed causality is a possible issue in almost every study. This phenomenon occurs when the dependent variable causes changes in the independent variable. Regarding this study, it is possible that deal performance increases the use of add-on acquisitions, as PE managers become overconfident during the holding period of the deal. Here, it could be the case that only managers that are overseeing already well-performing deals choose inorganic growth

strategies to increase their returns.

Lastly, the bounded amount of data points decreases the likelihood of the given estimators to be true. In the limited timeframe of this study, data has only been collected on 213 deals applying inorganic growth strategies. Therefore, it is hard to make any firm conclusions about the results that have been obtained.

(40)

7. Conclusion

The strategies that successful PE investors apply to obtain their results are still tricky to uncover. Because of the lack of appropriate data at hand, only a few

researchers so far have been able to dig deeper into the subject to uncover accurate insights. This thesis has tried to contribute to this field of study by examining the effects of different types of add-on acquisitions on PE deal performance. To do so, a new sample has been constructed, consisting of 1,126 PE deals (of which 236

inorganic growth buyouts were identified). Here, the effect of add-on acquisition on deal performance has been examined more thoroughly.

The findings of this study are mixed. As expected, SBO deals and other deals that experience relatively long holding periods have a higher probability of applying inorganic growth strategies. Although using add-on activity on its own does not safeguard greater returns, increasing the number of add-ons does contribute to a higher deal performance. Besides, SBO deal returns are generally lower compared to PBOs. Regarding the type of add-on, mostly cross-industry acquisitions appear to correlate positively with deal returns. Next to this, applying cross-country inorganic growth strategies combined with early investment and a short holding period

increases the deal returns significantly too. Also, cross-country inorganic growth strategies correlate positively with deal performance, especially in the case of short-lived deals. Lastly, different strategies regarding add-ons appear to be successful in PBOs compared to SBOs.

As the results are mixed, it is still too early to determine the exact effect of inorganic growth strategies on deal returns. In future research, there are a couple of aspects that should be examined more thoroughly. Firstly, increasing the number of data points that include financial indicators, such as revenues and EBITDA, should be obtained to control for the size of add-on acquisitions. Secondly, extended

datasets should be obtained to increase the preciseness of certain tested effects. Thirdly, focusing on the amount and timing of add-on acquisitions, a better view

(41)

regarding the true effect of inorganic growth strategies on deal returns could be provided.

(42)

8. Appendix: Glossary of terms

Add-on acquisition: a company that is added to a company within the PE firms’ portfolio.

Buy-and-build strategy: a strategy in which a firm is bought to function as a

platform for later add-on acquisitions. This strategy belongs to the inorganic growth category.

Conglomerate merger: a merger between firms that are operating in unrelated industries.

Cross-country add-on: an add-on acquisition in which the acquired firm does not share the same country background as the portfolio firm.

Cross-industry add-on: an add-on acquisition in which the acquired firm does not share the same industry background as the portfolio firm.

Holding period: calculated as the difference in time between entry and exit date. The holding period is expressed in months. In this study, the holding period is categorised as short in the case that a portfolio firm is held for a maximum of 24 months.

Internal rate of return (IRR): a measurement of profitability of potential

investments. The IRR is the discount rate that makes the net present value of all future cash flows (that stem from the particular investment project) equal to zero. In this study, the return is based on private placement memorandums (PPMs).

Intra-country add-on: an add-on acquisition in which the acquired firm shares the same country background as the portfolio firm.