Organic versus Inorganic Growth Strategies
The Role of Add-On Acquisitions in Private Equity Deals.
Lars Krahnstöver
11086645
7th June 2016
Statement of Originality
This document is written by Lars Krahnstöver who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is 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.
Abstract:
Since decades, researchers have tried to uncover the detailed mechanisms that drive private equity (PE) deals. Still, little is known about the concrete optimizing processes of portfolio firms that often lead to astonishing investment returns. Opening this black box further means to detect hidden growth channels and to understand their fundamental functionalities. In recent years, inorganic growth strategies have been identified as important growth levers in PE deals. Using a company as a platform for later follow-up acquisitions has been proven to be a promising growth strategy. However, the interrelationship of inorganic growth with further key deal performance drivers has only been insufficiently investigated. This study analyzes 264 buyout deals with investment inceptions between 2001 and 2008. The results suggest that enhancements in financials are more important in deals with inorganic growth. Furthermore, secondary buyouts (SBO) seem to perform similar when they pursue organic or inorganic growth strategies. The beneficial concept of organic growth appears to only hold for deals with sufficient length of holding period. Finally, the higher proportions of add-on volume seem to enhance deal performance.
Table of contents
1. Introduction ... 5
!
2. Literature review & hypotheses ... 10
!
2.1. Private equity ... 10
!
2.2. Secondary buyouts ... 12
!
2.3. Holding period ... 14
!
2.4. Add-on acquisitions ... 14
!
2.5. Hypotheses ... 15
!
3. Data and descriptive statistics ... 18
!
3.1. Data collection process and preparation ... 18
!
3.2. Descriptive statistics and prior analysis ... 21
!
4. Methodology ... 24
!
4.1. Add-on acquisitions and changes in financials ... 24
!
4.2. Add-on acquisitions and secondary buyouts ... 27
!
4.3. Add-on acquisitions and holding period ... 28
!
4.4. Proportion of add-on acquisitions ... 29
!
5. Empirical results ... 30
!
5.1. Add-on activity ... 31!
5.2. Changes in financials ... 32!
5.3. Secondary buyouts ... 34!
5.4. Holding period ... 35!
5.5. Proportion of add-ons ... 36!
6. Discussion ... 38!
7. Conclusion ... 41!
8. Appendix: Variable description ... 42
!
1. Introduction
Private equity (PE) firms are confronted with increasing competition and changing market conditions. As a response to the late effects of the recent financial crisis, central banks are flooding the financial markets with fresh capital. Cheap money has triggered a greater demand for alternative investment opportunities other than the bank account and has consequently affected a wide range of investment areas. As PE has yielded high returns over decades, many additional investors have been attracted to the PE market. With an increasing number of PE firms looking for profitable investment opportunities, competition for these investments is rising. As an overall result, the PE sector has grown so big, that its investment decisions influence the growth speed of whole industries regarding productivity and employment (Bernstein, Lerner, Sørensen, and Strömberg, 2014). In brief, the business model of PE firms is to acquire stakes in companies or perform a full takeover, optimize the target’s business and sell the company at a profit after usually three to five years. Since optimizing can include job cuts, the PE industry has always faced controversial debates about its business practices that are criticized to not follow the right moral standards. In fact, buyouts are on average followed by net job losses (Davis, Haltiwanger, Handley, Jarmin, Lerner, and Miranda, 2014). However, buyouts also tend to increase total factor productivity (Davis et al., 2014), which is considered as one of the advantages of PE firms. But does the PE management really do a better job than other management constellations?
The simple answer is yes – and no. For instance, Wilson, Wright, Siegel, and Scholes (2012) show that PE-backed companies benefit from the influence of the PE firm. The authors find that financially and economically, these companies have an edge over non-PE-backed companies before and even during the recent financial crisis. Moreover, Harris, Jenkinson, and Kaplan (2014) show in their study, including almost 1400 U.S. buyouts, that buyout returns seem to consistently outperform the public market. This evidence is supported by Robinson and Sensoy (2011), who examined buyouts of PE funds over the period from 1984 to 2010. They find that over fund lifetime, buyouts outperform the S&P 500 index by an average of 18%, netted of fees. Likewise, Kaserer and Diller (2004) demonstrate that returns of European PE funds
outperform relative to the MSCI Europe Equity and J.P.Morgan Government Bond Index from 1989 to 2003. A different perspective is presented by Phalippou (2014). He argues that PE buyouts focus mostly on small and value companies. In his view, a more appropriate benchmark than major indices are hence small cap indices. While his study confirms that buyouts outperform the S&P 500 on average, he also shows that, compared to small and value indices, buyouts perform worse on average. In addition to that, there are studies disproving the outperformance hypothesis. For instance, Kaplan and Schoar (2005) find that fund returns do not beat the public market on average. Moreover, they show that fund returns vary a lot regarding their performances, highlighting the importance of fund selection skills for investors. Likewise, Phalippou and Gottschalg (2009) cast further doubt on the success story of PE buyouts by suggesting annual underperformance of 6% after adjusting for risk and fees. One risk of purchasing a portfolio company is the risk of overpricing. In fact, Axelson, Jenkinson, Strömberg, and Weisbach (2013) find that buyout leverage and pricing are mainly driven by economic wide credit conditions. They provide evidence that deal leverage is negatively related to fund return, arguing that easier access to financial resources increases the likelihood of overpaying.
As shown above, results concerning the performance of PE firms are mixed. However, as mentioned before, PE firms have grown in numbers during the last years and have a great impact on the economy. Understanding in greater detail what drives PE performance is hence crucial.
The above discussed increases in competition in the PE market and the growing maturity of the PE market have resulted in decreased persistency in returns for PE fund managers (Braun, Jenkinson, and Stoff, 2015). One result of this change is the trend towards secondary buyouts (SBOs). In the most general way, a SBO is a buyout in which one PE firm sells its portfolio company to another PE firm. If the seller in a PE deal is not a PE firm, e.g. government or other stakeholders, the deal is called primary buyout (PBO). SBOs have received a great deal of attention in recent years, since it is the most rapidly growing segment in the industry of PE (Arcot, Fluck, Gaspar, and Hege, 2015) and already accounts for over 60% of all buyout activity (Bonini, 2015).
The question that arises with increases in frequency and volume of SBOs is how the value creation process differs from that of PBOs. Assuming that PE firms in general work similar, what is left to optimize for the next PE firm after the portfolio company was already subject to restructuring and roll-up processes by the previous PE firm? Of course, market timing is an argument that might advocate for high returns even for SBOs. But most capital structure adjustments and operating performances are likely to be optimized by the prior owner. The latter assumption is supported by findings of e.g. Wang (2012), and Jenkinson and Sousa (2012), who emphasize that PE firms manage to increase operation performances in PBOs to a greater extent relative to SBOs. So what growth method is left for PE firms to substantially increase enterprise value of the underlying portfolio firm?
The answer here is inorganic growth, also called external growth. Inorganic growth describes the strategy to grow a portfolio company by acquiring related businesses, so called ons, and integrate them into the original company. A term related to add-on acquisitiadd-ons is the buy-and-build strategy. The aim of this strategy is to buy a certain company that functions as a platform for later follow-on acquisitions and to create growth during the integration process. This strategy usually demands for longer investment horizons to yield further synergies throughout the combined firm. The market share and volume of add-on transactions suggests a strong trend towards this strategy, driving growth measures from organic to inorganic strategies: In 2015, the global transaction volume of add-on acquisitions by PE-backed portfolio firms reached USD 267 billion.1 In the U.S. PE market, 68% of all closed transactions between USD
25 million and USD 100 million in the first quarter of 2016 are attributed to add-on acquisitions.2
Even though add-on transactions seem to exhibit a rapid rise in popularity, there are only few academic studies that dive deeper into this topic. The reason for the lack of studies investigating aspects of SBOs and buy-and-build strategies is due to a limitation of sufficient data. To analyze the impact of add-on acquisition on e.g. deal
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1 Source: Bain & Company, Global Private Equity Report 2016
performance, one needs to have detailed deal-level information that is in general not publicly available.
There are, however, a few studies that have started to examine add-on transactions and SBOs. The study by Nikoskelainen and Wright (2007) is one of the first studies that analyzed the role of additional acquisitions to platform companies during the PE holding period. Findings of the study support the positive effect of add-on transactions in PE deals, with the authors calling inorganic growth strategies a key driver for deal performance. Hammer, Knauer, Pflücke, and Schwetzler (2015) conducted a large-scale study with 9,548 PE deals to investigate the effects of add-on acquisitions. They support the assumption that additional acquisitions during the holding period of the PE firm are highly beneficial for the overall deal performance. Furthermore, they find a higher likelihood for add-on acquisitions for SBOs compared to other buyouts, which might explain the simultaneous rise of SBO deals and buyouts that pursue a buy-and-build strategy.
While studies on add-on transactions and SBOs have started to examine the positive effects of these transactions and have linked them to SBOs, many other aspects have not been investigated so far. This study hence aims at extending the relatively new finding regarding add-on acquisitions by examining further questions related to PE deals. By adding publicly available add-on transaction information to a subsample of an existing data set used in the study by Degeorge, Martin, and Phalippou, (2016), this study exploits a unique data set of 264 exited buyout deals worldwide that were acquired by a PE firm between 2001 and 2008. After adopting the method of Nikoskelainen and Wright (2007) to separate deals with and without on activity during the PE holding period, this study analyzes the impact of add-ons on deal performance in this new sample.
The main aim of this study is to investigate the effect of inorganic growth in greater detail. To do so, this study combines the effect of add-on activity with other major effects that have only been considered isolated up to now. First, this study analyzes the relationship of changes in operational performance and add-on acquisitions. Second, add-on activity is related to the entry channel, that is either SBO or PBO. Based on the finding by and Hammer et al. (2016) and Nikoskelainen and Wright
(2007) that add-on transactions are beneficial and more likely to occur in SBOs, it can be assumed that SBOs perform better when they are subject to inorganic growth strategies compared to SBOs that solely pursue organic growth measurements. Third, the combined effect of short holding period and inorganic growth is analyzed in more detail. Analyzing this effect also contributes to the question if buy-and-build strategies are favorable, even if the portfolio firm is only hold shortly. As mentioned above, it is assumed that buy-and-build strategies demand for longer investment horizons to realize synergies throughout the combined firm.
Lastly, the proportion of add-ons itself will be subject to a closer analysis. Since no previous study managed to effectively control for the proportion of added business size during the holding period, this study introduces an approach that uses the revenue and earning streams as a proxy for add-on proportion.
The subsequent structure of this study is as follows: Section 2 summarizes important literature regarding all research fields that are addressed in this analysis and derives the research hypotheses. Hereinafter, section 3 continuous to outline the data collection process and presents key insights into the descriptive statistics. The fourth section explains the methodology that this study is based of. Empirical results are presented in section 5. Section 6 discusses implications of the findings and gives an outlook for future research. The last section concludes the paper.
2. Literature review & hypotheses
2.1. Private equity
There are three major groups that are involved in PE investments, namely PE firms, venture capital firms and angel investors. Even if the types of investment strategies might differ, the overall business idea is similar, which is in brief: Buying equity of a company, helping to develop its business and exit the stake at a profit. Among others, a major difference of PE firms compared to venture capital firms and angel investors is the point in time when they invest into a company. While venture capitalists and angel investors usually focus on companies in the earlier stages of the business lifecycle, PE firms invest in more mature firms. However, PE firms are not passive financial sponsors, but are actively engaged in company management decisions. Investments can range form minority shares to a complete buyout. The term buyout is further differentiated in management buyouts (MBO) and leveraged buyouts (LBO). MBO means, that the management of a certain company takes over the majority of shares from the previous owner. On the other hand, LBO generally means that a company or asset is acquired by one or more investors who fund their investment mostly by debt. The basic idea of a LBO is to take over control of the company, transfer the debt to the target company and pay the debt back with the target’s future cash flows. This enables investors to acquire companies with relatively small amounts of equity. This method has several advantages as it decreases bankruptcy costs and decreases the cost of capital, since debt financing is usually cheaper than equity financing. LBOs exhibited their first major boom in the late 80s, but decreased in popularity again until the period of 2004 to 2007, when the frequency and volume of LBOs had skyrocketed (Shivdasani and Wang, 2011). Due to favorable market conditions, it seems that buyouts are on the rise again and seem to aim at the pre financial crisis level.
PE firms are not limited to one fund. Often, several funds are raised by one PE firm that might differ in investment criteria like company sizes, regions, industries, etc. On average, funds are planned to stay active for approximately 10 years. Due to
unfavorable exit market conditions etc., lifetime of funds might be extended for some years. It is a common practice to raise additional money for a subsequent fund during the lifetime of the previous one. Hence, performance of the subsequent fund is highly correlated with the ability to raise additional resources (Kaplan and Schoar, 2005). PE firms make profits by taking fees (approx. 1.5% to 2%) before investing the money raised in funds and by collecting a stake of approximately 20% of the investment profits after exit (Arcot et al., 2015).
As already mentioned in the introduction, there is great ambivalence with regard to the effectiveness of PE business models. As one advantage of PE firms it has been suggested, that they provide portfolio companies with additional financial resources, and hence reduce situations of underinvestment (Cohn and Towery, 2013). Regarding improvements in operational performance, further support for PE firms comes from Acharya, Gottschalg, Hahn, and Kehoe (2013), who analyzed 395 deals of target firms in Western Europe between 1991 and 2007. Besides the positive correlation of improvements in revenues with deal performance they also find that EBITDA margin is highly correlated with deal performance. The results of Kaserer (2011), who evaluated a sample of 332 mid-market buyout deals performed in Europe between 1990 and 2011, supports the finding of Acharya et al. (2013).3 Aside from
demonstrating a highly significant and positive influence of revenues on deal performance, Kaserer (2011) further suggests that about two third of deal returns can be attributed to improvements in earnings. However, he finds that EBITDA multiple enhancements are negligible with respect to deal performance, which is contrary to the study by Guo, Hotchkiss, and Song (2011). They analyzed 192 U.S. buyouts between 1990 and 2006 and find that that the relationship of the EBITDA multiple and deal performance is highly positively correlated. On the other hand, they only find small improvements in other operating performance measures. Results of the study Cohn and Towery (2013), based on an analysis of 317 LBOs, which took place between 1995 and 2007, provide only few evidence of improvements in operation performance. In an
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
3 In addition, one can also find several earlier studies claiming that PE firms manage to increase
operational performance relative to their peers (Guo, Hotchkiss, and Song 2011; Kaplan, 1989; Smith, 1990).
overall review, Kaplan and Strömberg (2008) conclude that, on average, literature indicates a positive relationship of PE buyouts and enhancements in operation performance.
When speaking about PE deals, a crucial element of the business cycle is the exit, since the portfolio firm value at that point in time mainly drives deal performance. Hence, the importance of the exit channel in PE deals has been often highlighted. For instance, Guo et al. (2011) find that exiting through an initial public offering (IPO), which means selling the company shares through the public market, is the most lucrative exit route for PE firms. Conversely, Katz (2009) suggests that this relation also holds the other way around. He analyzed 147 IPOs (123 PE-backed and 24 non-PE-backed) between 1980 and 2005 and finds that besides increased earning quality of PE-backed IPOs, companies that are mainly owned by PE firms at IPO, have also higher long-term stock price performances.
2.2. Secondary buyouts
As mentioned in the introduction, a buyout of a PE firm is considered as secondary buyout (SBO), if the seller(s) of the portfolio company is another PE firm (or group) that assigns the majority of portfolio company shares to the buying PE firm. In turn, primary buyouts are transactions that do not have a PE firm on the sell side. The popularity of SBOs has increased rapidly in recent years and already accounts for more than half of all buyouts (Bonini, 2015).
However, in the context of SBOs, the question of whether SBOs generate lower deal performance relative to PBOs has been raised frequently. One major issue with SBOs is that the additional value through the buyout for the improvement process of the portfolio company is questionable, since most capital structure adjustments and operating performances are likely to be optimized by the prior owner. Additionally, it has been argued that a SBO is only beneficial for the buying PE firm, when the PE firms skill set is complementary to the selling PE firm (Degeorge et al., 2016).
Support for these concerns of lower deal performances in SBO transactions comes from Achleitner and Figge (2014), who examined data of 2,456 buyouts closed
between 1990 and 2010, including 448 SBOs. Their findings indicate slightly lower equity returns for SBOs compared to other deals. Moreover, Jenkinson and Sousa (2012) compared performance of 308 deals that were either exited through an SBO or IPO between 2000 and 2007. Their findings indicate significant lower performance over the subsequent three years after exit, when the exit was performed through an SBO. Bonini (2015) compares the returns of PE investors regarding PBOs and SBOs and concludes that even if returns are positive for SBO, deal returns for investors are higher in PBO deals. Moreover, Wang (2012) argues that under favorable debt market conditions, SBOs are more likely to occur and connects the access to cheaper money with lower deal performances for SBO transactions due to overpricing.
Another strong influence factor on deal performance is promoted by Arcot et al. (2015), who explore the effect of buying and selling pressure on deal performance. They find that the pressure to spend capital yields in worse performance. The definition of pressure refers to the age of the fund. Funds are usually planned to exist for about 10 to 12 years. Managers in more mature funds that still have uncommitted capital left, might face more pressure to invest this capital. On the other hand, managers is funds that last e.g. for already 9 years, managers might feel pressured to exit their investments to pay off their investors. In the study by Degeorge et al. (2016), an investment into a portfolio company is defined as bought late, if the fund invests in the company later than 2.5 years after fund foundation. If the difference of closing date and fund vintage date is 2.5 years or less, the investment is considered as bought early. Consequently, deals that invest early are assumed to face less buying or selling pressure, this the fund lifetime is still considered to be in the early stage.
While PE funds under selling pressure tend to divest for lower multiples, funds under buying pressure tend to overprice their investments and use less debt (Arcot et al., 2015). The study by Degeorge et al. (2016) confirms that SBOs perform worse with pressured PE fund on the buy side, supporting the overall assumption of value
The essence of this section is that SBOs seem to underperform on average, compared to other buyouts. One major reason for this might be the effect of pressure on deal performance. Hence this issue will also be subject in this analysis.
2.3. Holding period
The term holding period in this study is almost self-explanatory. It describes the time that the PE firm is invested in the portfolio company, in other words, the time from entry to (final) exit transaction. As indicated earlier, the average holing period for buyouts can vary in over different time horizons. Kaplan and Strömberg (2008) analyzed LBOs from 1970 to 2007. According to their data, the average holding period for those buyouts is approximately six years. The study by Degeorge et al. (2016) on the other hand finds average holding periods of about four years. Additionally the authors show that there is no significant difference in holding period for PBO or SBO deals. Contrary to this, in an earlier study Degeorge et al. (2013) demonstrate that the holding period of SBOs is in fact longer than the one for other buyouts. This finding extents results by Hammer et al. (2015) that SBOs are more often subject to inorganic growth strategies, which usually demand for longer holding periods.
Research suggests that the length of holding period matters for deal performance (Phalippou, 2009). For instance, the study by Lopez-De-Silanes, Phalippou, and Gottschalg (2012) suggests a negative relationship of deal performance and holding period. However, as mentioned in the introduction, one can assume that inorganic growth strategies usually demand for longer holding periods, since the implementation of add-ons into the platform company requires time. Holding period should hence have a differential effect on deal performance depending on whether the growth strategy is organic or inorganic.
2.4. Add-on acquisitions
As outlined earlier, the major aim of this study is to determine the effect of add-on acquisitions on deal performance itself, but also to examine the interaction of add-ons in combination with other deal characteristics. An add-on acquisition is defined as
acquisition made by the portfolio company, while being under PE management. The aim is to acquire further companies and to integrate them into the platform company.
The specific focus on inorganic growth is relatively new in academic literature, which is probably due to the restricted data access. The study by Hammer et al. (2016) is the only study to examine difference of organic and inorganic growth strategies on a large scale sample. They analyzed 9.548 PE transactions between 1997 and 2012. Their findings show the positive effect of inorganic growth on deal performance. As already mentioned, the authors further show that add-on acquisitions are more likely to occur in SBO deals, suggesting that add-ons might be more beneficial for SBOs than for PBOs.
Another study that examined the effect of inorganic growth in more detail is the study by Nikoskelainen and Wright (2007). They distinguished between deals with and without external growth under PE management. Their analysis of 321 exited buyouts in the United Kingdom between 1995 and 2004 points out that add-ons are a key driver for deal performance. Add-on acquisitions especially help to increase the targets overall enterprise value significantly. The authors conclude that the advancing consolidation is some industries is associated with the increasing use of buy-and-build strategies.
In summary, the few existing studies related to inorganic growth strategies and deal performance suggest a positive relation of add-on acquisitions and deal performance, which might be even more pronounced for SBOs.
2.5. Hypotheses
As mentioned in the introduction, although inorganic growth in the PE market has risen in importance, studies investigating effects on deal performance are still scarce. Based on literature reviewed in the previous section, this study will derive five hypotheses that help to understand the effect of add-on acquisitions to a greater extent.
The first aim of the study will be to confirm the positive impact of add-on acquisitions on deal performance in general (see Hammer et al., 2016; Nikoskelainen and Wright, 2007). The first hypothesis is hence as follows:
Hypothesis I: Inorganic growth strategies lead to higher deal returns.
That improvements in revenues, EBITDA and EBIT, i.e. financials of the portfolio firm, are associated with higher deal performance has been explained in the previous sections. Given the short holding periods of portfolio firms, potential sellers might have difficulties to evaluate the process and quality of the add-on integration, especially when several add-ons are acquired. Hence, this study assumes that improvements in financials are even more important for deals with an inorganic growth strategy, to convince the potential buyer of the success of the acquisitions. For instance, acquisitions are often reasoned by e.g. cross selling potentials and other synergy effects. If true, then one would especially expect that deals involved in add-on acquisitions experience on average higher relative enhancements in financials, when the add-on acquisition is a good choice. This situation refers to the problem of asymmetric information.4 Since portfolio firms are held private, potential buyers
might only have limited access to reliable data. Thus, the study expects that for those deals the increase in financials is even more important, since it functions as a positive signal to the market. Therefore, the second hypothesis is as follows:
Hypothesis II: For deals pursuing inorganic growth strategies, increases in financials
are more important for deal performance than for deals pursuing organic growth strategies.
As stated by several studies, SBOs tend to underperform relative to PBOs and hence are often claimed to be value destroying (Bonini, 2015; Jenkinson and Sousa,
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
4 For more information about asymmetric information regarding transactions see e.g. Krishnaswami and
2012; Wang, 2012).5 In addition, SBOs are subject to inorganic growth more often
than other buyouts (Hammer et al., 2016). The question that is still unanswered is if add-on acquisitions also improve deal performance within the group of SBO deals. In addition, the relationship between add-on in SBO deals with deal performance might also depend on selling pressure. As mentioned before, SBOs perform worse under both, selling and buying pressure. Hence, the third hypothesis is as follows:
Hypothesis III: SBOs perform better with inorganic growth strategies compared to
SBOs with organic growth strategies, especially when they are bought early, i.e. when they experience lower pressure.
The next research question concerns the combined effect of inorganic growth and a short holding period. As mentioned in the literature review, when viewed isolated, both effects are assumed to be beneficial for deal performance. However, as outlined before, a qualitative and successful company integration of add-on transactions needs time. It can hence be assumed that longer holding periods are beneficial for inorganic growth strategies. Therefore, the fourth hypothesis is as follows:
Hypothesis IV: The interaction of an inorganic growth strategy with a short holding
period is negatively correlated with deal performance.
While research on add-on acquisitions has proven its beneficial effect on deal performance, no previous study has related the exact proportion of add-ons to deal performance. Assuming that PE firms pick their add-on acquisitions carefully and have the knowledge to integrate their add-on targets well, the last hypothesis is hence as follows:
Hypothesis V: The higher the proportion of add-ons relative to the platform
company, the better the deal performance.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
5 Under some circumstances, e.g. when buying and selling PE firm have complementary skill sets, SBOs
might outperform other buyouts (Degeorge et al., 2016). However, this study relies on the general finding of underperformance of SBO deals.
3. Data and descriptive statistics
The data for this study is based on the data set used in the study by Degeorge et al. (2016). This original data set is mainly gathered by manually evaluating a great amount of Private Placement Memorandums (PPMs). To this initial data set, information about add-on deals was added to create an extended data set. Since the data collection had to be done manually, the data collection process and the final data set account for a major part of this master thesis. In the following, the data collection process is explained in more detail.
3.1. Data collection process and preparation
The data set of Degeorge et al. (2016) provides the starting point for the data collection process for this study. This original data set includes a variety of deal level information that is not publicly available but essential for conducting this analysis. The measurement for deal performance is the internal rate of return from the PE perspective, gross of fees. This information is crucial for the analyses of this study since it is the depended variable of all regressions. Furthermore, the data set includes other sensible information about e.g. the investment size of each deal etc. Besides this information, the deal selection also functions as a framework for this study. Out of the whole set of deals, this study focuses on all deals that were made in 2001 or later and that were categorized as exited. This limits the initial sample of several thousand deals to 1,563 deals that were closed between 2001 and 2010. The latest exit year for one of those deals in this sample is the year 2013.
However, the data does not provide any information about merger and acquisition activity of those companies. Hence, with regard to data collection, the main challenge was to find all information of these portfolio companies. This includes the entry and the exit deal, as well as all acquired add-ons during the time of PE-ownership. Among others, these information include i) deal announcement date ii) deal closed date, iii) target name, iv) buyer name, v) seller name, vi) transaction value, vii) target revenue, viii) target EBITDA, ix) target EBIT, x) target industry and xi) target country. The data sources that were used to gather these information are Capital IQ and Thomson One. In a first attempt, all available data of each entry and exit deal of the 1,563
deals were scanned and collected manually from the data source Capital IQ. Furthermore, the same information was collected for each add-on transaction that was made by the portfolio company. A transaction is defined as add-on transaction if the portfolio company acquired another company during the holding period6. The whole
procedure is repeated with the data source of Thomson One. In this study, Thomson One is always only of secondary priority. Thus, if both platforms provide differing data for the same variable e.g. revenue, the figure from Capital IQ has priority.
Since PE firms are in general conservative when it comes to information transparency, both, Capital IQ and Thomson One provide only little information with respect to most PE deals. For some deals, there was no information at all, for others there was only information for the entry or the exit deal. For almost no deal, full information was found, especially with respect to financial data. For example there were several deals that had e.g. revenue data and no EBITDA or EBIT data at entry and the other way around for the exit deal. Scarcity for add-on acquisitions at closed date is even higher. On the other hand, deals for which neither the entry nor the exit deal was listed in either of these two data banks, but for which there was information on add-on acquisitions, were included in the data set. The information scarcity is a great downside for the analysis process because it limits the number of observations and the analyzing possibilities for this study.
Regarding collection of add-on data, several issues arose due to ambiguity between the data sources Capital IQ and Thomson One. If there was no add-on listed in none of those two data sources, the deal was assumed to only have used an organic growth strategy. On the other hand, if e.g. one data source provided three closed add-ons and the other one lists four, then it was assumed that the total amount of closed deals was four. In addition to differences in number of add-ons between Capital IQ and Thomson One, both data sources sometimes provided conflicting information on specific add-ons, making it difficult to verify results from one data source with the other. As a result, the following selection mechanism was chosen: If in the just described example only two add-ons from each data bank could be matched with each
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
6 The closed dates in Capital IQ and Thomson One tend to differ slightly with respect to the original
data set. The reference point for defining a transaction as add-on transaction is always made based on the dates in the original data set.
other (based on name, closed date, industry, country etc.) then each potential unique on was taken into the edited data set. In that case, the deal would count five add-ons. Another ambiguity is the definition of closed deal itself. An add-on transaction was only considered as closed, if at least one data source provided a closed date for this deal. In some cases, only one of those two sources provided a closed date.
Since the data for add-on acquisitions was hand collected, some decisions on specific cases that came up during the collection process had to be made. For instance, the data sources often listed deals that seemed to fit to the initial data set at the first glance. However, some deal details tended to differ regarding e.g. the concrete entry or exit month. Based on the other information like buyer and seller name etc., these deals were either included or excluded in the data set.
In order to ensure the highest possible data quality, the final sample only considers deals where at least some information about the entry deal and the exit deal was available in at least one of the two data sources. The assumption behind this is as follows: Finding the entry and the exit, is a reasonable argument for a certain level of coverage of the underlying portfolio firm. If only information about the entry deal, the exit deal, or non of those two deals was available, the assumption that add-on information were provided correctly is more difficult to argue for. Hence, all deals that lack in entry and/or exit deal information are excluded from the analysis, even if the data source have add-on information available. In addition to this, there is a further requirement for deals to be considered. This is, that a deal has to have at least one of the three main financials (revenue, EBITDA, EBIT) available in both, the entry and the exit deal. Again, the assumption is that financials are usually sensible data, that PE firms tend to not publish for their portfolio companies. Having one of those three figures at entry and exit assures further quality for add-on data. In summary, these requirements increase the validity of the analysis, but also decrease the number of observations to 308 deals. A final limitation is the availability of deal performance information. Since 44 of the 308 deals have no information about the internal rate of return (IRR), these deals were also excluded from the analysis. Finally, 264 deals were left to start the analysis process.
3.2. Descriptive statistics and prior analysis
Of the 264 deals included in the final sample, 153 deals used only organic growth strategies and 111 deals acquired at least one add-on under the holding period of the PE firm.
Table 1 offers a first look at the distribution of deals with respect to their entry year, the length of holding period and add-ons. The sample covers deals with entry years from 2001 to 2008, whereby deals at the end of the observation period are less pronounced. The relation of deals with organic versus inorganic growth strategies is in general constant over the entry years and the holding period. The medians of closed add-on acquisitions are mostly zero, except for deals with the entry year 2006 and deals with holding periods longer than four years. The mean of the amount of add-on acquisitions ranges form 0.33 to 1.76. Mean values for these transactions do not reveal a clear trend regarding the entry year. However, regarding the holding period, deals that were hold longer by the PE firm tend to perform more add-on acquisition. This finding is quite logical since e.g. PE firms that intent to grow by external measures might plan to exit later in order to extract more synergies from add-ons and increase firm value further before selling.
Another interesting finding is that there seem to be no clear preferences when to make add-on acquisitions relative to the holding period. The overall distribution shows that approximately 30%, 34% and 36% of closed add-on acquisitions are in the first, second and last third, respectively. However, almost half of the add-ons of deals starting in 2003 and 2006 were closed in the last third of the holding period. A similar picture can be drawn for deals that were hold between four and five years. In this case, even slightly more than half of the add-ons were closed after the first two thirds of holding period had already passed.
Table 1
Figure 1 shows deal performance of buyouts relative to the number of add-ons acquisitions. The highest achieved performance of about 250% occurs for deals that were not involved in additional add-on activity, but also for deals that performed up to three add-on acquisitions. What is striking is that almost all deals exhibit positive deal performances. It is further striking that a handful of deals, which closed an exceptional high amount of add-on deals, performed relatively poor compared to the
average peer group. This might account for the negative relationship of closed add-ons and deal performance, that is indicated in figure 1. For deals up to six add-ons, the relationship between add-ons and deal performance tends to be slightly positive. This indication raises the question, if there is actually a positive correlation of deal performance and closed add-on acquisitions, which turns into a reverse trend if the number of add-ons increases over a certain threshold of deals. However, to analyze this specification in more detail a greater number of observations would be needed to get robust results. Hence, this ambivalent relationship is not subject in this study, but might be an interesting issue to look at in further studies. In total, the figure offers a first indication that deals with inorganic growth strategies might not be superior to organic growth strategies.
Figure 1
Figure 2 shows the relation between deal performance and the deal holding period. The overall trend for those two variables is clearly a negative one. Deals that exhibit on average shorter holding periods seem to perform better than deals with relatively long holding periods. This finding supports the results of (Lopez-De-Silanes et al., 2012), who also argue for a negative relationship of those two factors.
Figure 2
Table 2 compares deal performance between different subsamples with respect to differences in organic and inorganic growth strategies. Once a portfolio company acquired at least one add-on, the deal is categorized as inorganic grower. This definition is similar to the definition in the study by Nkoskelainen and Wright (2007). The subsamples are generated with respect to main characteristics of the categories entry channel, investment timing and exit route. Each subsample is additionally divided into deals with and without add-on activity. By comparing the mean performance of those groups, one gets a first glimpse into the issue of performance differences for deals with different growth strategies. Findings suggest that the chosen growth strategy does not have a significant impact on deal performance when considering the full sample. Along with that, the results also do not suggest differences within the different groups of entry channel, i.e. add-on acquisitions seem to not improve deal performance within SBO deals. When it comes to investment timing, however, add-ons seem to play a significant role. Funds that are older than 2.5 years
seem to have, on average, 16% higher deal returns (Gross IRR) compared to their peer group without any add-on activity. The difference is significant on a 10% level. Portfolio companies that were bought early reveal no significant difference in performance with regard to add-on activity.
While add-on activity seems to be not significant within groups of deals that are exited via IPO or trade sale, inorganic growth increases deal performance for portfolio companies, which are exited via SBO by 17%. This difference is again significant at the 10% level.
Table 2
Overall, descriptive statistics do not draw a clear picture of the role of add-on acquisitions in PE deals, but also do not disprove its relevance.
4. Methodology
This study aims to analyze different aspects regarding deal performance and inorganic growth. This section gives an overview about the statistical methods used during the analysis and highlights the key indicators that are the basis for the assessment of the different hypotheses. The data is structured cross sectional and the statistical method used is the ordinary least square (OLS) method. The first hypothesis that wants to prove the positive effect of add-on acquisitions on deal performance is not listed isolated. Since this study uses a dummy variable to indicate organic or inorganic growth, similar to the study by Nikoskelainen and Wright (2007), the methodology to test this hypothesis is not explained in an isolated section, but included to the other sections.
4.1. Add-on acquisitions and changes in financials
The second hypothesis aims at the relationship of the growth strategy and changes in key financials of the related portfolio company. It is argued that improvements are even more important for deal performance, when the deal is involved in add-on acquisitions, since potential buyers might expect higher changes in these financials. The five financial figures that are of relevance in this analysis are: i) Changes in revenue, ii) changes in EBITDA, iii) changes in EBIT, iv) changes in EBITDA margin and v) changes in EBIT margin. The calculation of the first three figures is structured similar. Since the data provide the financial value expressed as last twelve month since announcement date of the add-on transaction, the absolute change is the value of one of those financials at the point of exit !!!! less the value at the point of
acquisition !!. The relative change can be derived as follows:
Xt+1!- Xt
Xt *100
In this study, the EBITDA (EBIT) margin is defined as EBITDA (EBIT) divided by revenues at the same point in time. Hence, the relative change in margin is defined as follows:
margint+1!- margint margint *100
This definition of changes in margins brings limitations to data availability, since margins at entry or exit (or both) might be negative. A change from a negative to a positive margin cannot be defined with the underlying definition. There are methods to reduce this problem, e.g. by increasing all numbers equally until there are only positive numbers left. However, since this method has been previously criticized, this study treats undefined changes in margins as not available. For illustrative reasons, the following regression lists the main variables of interest, but does not list all control variables and fixed effects in detail.
!""! = !!+ !!!""!! + !!∆!"#!!"#!+ !!∆!"#$%&!+ !!∆!"#$!
+ !!∆!"#$%&!!"#$%&!+ !!∆!"#$!!"#$%&!+ !!!"" ∗ ∆!"#!!"#!
+ !!!"" ∗ ∆!"#$%&!+ !!!"" ∗ ∆!"#$!+ !!!"" ∗ ∆!"#$%&!!"#$%&! + !!"!"" ∗ ∆!"#$!!"#$%&!+ !"#$%"&' + !"#$%!!""!#$%
ADDi is a dummy variable that equals one, if the deal acquired at least one add-on
during the holding period. ∆Revenuei, ∆EBITDAi and ∆EBITi represent the changes
in revenues and earnings, respectively. ∆EBITDA!margini and ∆EBIT!margini measure the changes in EBITDA and EBIT margin. All financials are furthermore interacted with the add-on activity dummy. Lastly, the regression contains control and fixed effects (see below).
The key coefficients of interest are β1 and β6 to β10. The first dummy β1indicates whether an inorganic growth strategy is positively or negatively correlated with deal performance. The coefficients of the interaction terms β6 to β10 reveals the additional effect of changes in financials for firms with inorganic growth strategies. For example, if the coefficients of ADDi, ∆Revenuei and ADDi*∆Revenuei are greater than 0, this
means that both, an inorganic growth strategy and an increase in revenues are positively correlated with deal performance. However, since the interaction term is also positive in our example, this would suggest that an increase in revenues would be even more positively correlated with deal performance, if the PE firm use an inorganic
growth strategy. The interaction terms reflect the additional effect of changes in financials on deal performance for deals that made at least one add-on acquisition. Assuming that the second hypothesis is correct, the estimates for all five interaction terms are expected to be positive and significant.
Control variables for this analysis are i) holding period, ii) investment size and iii) club deal. The holding period is defined as the length of the deal measured in months. The transaction dates are not based on the dates that are published in the named data bases in the previous section, but are based on information of the PPMs. Only in cases where the exact month is not stated in the PPMs, the information from the data bases are considered. The investment size is defined as the amount of capital invested by the PE fund. Since the sample includes funds from a variety of different countries, all investments are converted into USD. The historical exchange rates are taken form Compustat on a monthly average interval. The exchange rate for the investment size is the monthly average historical rate from the same month and year in which the PE firm closed the entry deal. All investment figures are measured in millions. The last control variable is club deal. A club deal is defined as an investment deal of a PE firm, where simultaneously at least one other investor is involved in the entry investment. This information is also based on the PPM data.
In order to control for fixed effects, the regressions include a variety of firm and fund characteristics as well as portfolio company and deal characteristics. These fixed effects are i) buyer firm, ii) involved fund, iii) fund county, iv) portfolio company country, v) portfolio company industry, vi) entry year and vii) exit year. The fixed effects buyer firm and fund capture the characteristics that are specific to the PE firm and its fund that is involved in the deal. Since market conditions are different between countries, the country fixed effects for the fund and the portfolio company control for these differences. Furthermore, the industries of the portfolio companies might develop differently, which in turn might effect the deal performance. Hence, the portfolio company industry fixed effect is incorporated additionally. Lastly, it is also controlled for the entry and exit year of the deal. This time fixed effects capture the
influences that emerged during the observed time period and cover market wide changes like the financial crisis etc.
4.2. Add-on acquisitions and secondary buyouts
The third hypothesis address the relationship of entry channel, add-on activity and deal performance. To test hypotheses three, the regression is set up as follows.
!""! = !!+ !!!""!+ ! !!!""!!"#$,!+ ! !!!"#!+ !!!"#$%!+ ! !!!"#$%&'!+ !!!"#$%&'(! + !!!"#$%&'!+ !!!"" ∗ !"#!+ !!!"" ∗ !"#$%!+ !!"!"" ∗ !"#$%&'!
+ !!!!"" ∗ !"#$%&'(!+ !!"!"" ∗ !"#$%&'!+ !"!!"#$%
+ !"#$%!!""!#$%
The regression includes not only a dummy for ADDi but also a continuous variable
ADDcont,i representing the number of closed add-on acquisitions per deal. SBOi is a
dummy variable which is one if the deal is a SBO deal, and zero otherwise. The classifications for deals to be PBOs or SBOs are taken over form the study by Degeorge et al. (2016). According to their study, an investment of a PE firm into a portfolio company is considered as SBO, if the seller(s) of the portfolio company is another PE firm (or group) that assigns the majority of portfolio company shares to the buying PE firm. Furthermore, all tertiary buyouts etc. are also defined as secondary in this study. This means, that a portfolio company, which is sold for the third, fourth etc. time with PE firms on the sell and buy side, is also defined as secondary. In turn, all other buyouts are categorized as PBO.7
The dummy variable earlyi indicates if a deal was bought early or late. Being bought early or late is in this paper refers to investment timing. The assumption here is that the lifetime of a fund is limited. As already mentioned, funds are usually planned to exist for about 10 to 12 years. In line with the definition in Degeorge et al. (2016), an investment into a portfolio company is defined as bought late, if the fund invests in the company later than 2.5 years after fund foundation If the difference of closing date and fund vintage date is 2.5 years or less, the investment is considered as
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
7 The only exception are deals with a family office on the sell side. These deals are categorized as private
bought early. This study will include these variables since deals that are bought late are assumed to face either selling or buying pressure, which has an impact on their investment decisions (Degeorge et al., 2016).
SBOlatei is a dummy variable that is one if the deal is a SBO deal and bought late. The dummy variables for SBOearlyi and PBOlatei are defined in the same way. All variables are again interacted with add-on activity and the regression contains the same control variables and fixed effects as in the previous regression.
To verify hypothesis III, the estimate of the interaction term ADDi*SBOi should
be significant and positively correlated with deal performance. The variables SBOlatei,
SBOearlyi and PBOlatei will help to get a more detailed view of the effect of add-ons and SBOs.
Additionally, for exploratory analysis, this study also examines the interaction term ADDi*earlyi to gain insight in the performance of buy-and-build strategies, that were
mentioned in the introduction. As already explained, the aim of this strategy is to buy a certain company and grow it over a longer time period by acquiring several additional add-ons. Hence, a positive estimate for ADDi*earlyi would indicate a
superior performance of this specific inorganic growth strategy.
4.3. Add-on acquisitions and holding period
To define short holding period, the sample is divided into two groups. One, that includes deals that are referred as short holding period, and one that includes deals referred as long holding period. Since there is no standardized definition for short holding period, this study uses the commonly used three to five years for an average holding period in PE deals as a reference point. For robustness reasons, this study defines two thresholds that separate both groups. First, we set the threshold for short holding deals to two years. This means, that all deals in this sample that have a holding period of 2 years or less are put into this group. Furthermore, we define as second threshold of 3 years or less, that indicate short holding periods. Hence, the analysis will consider both definitions of short holding period in different regressions.
The structure of the OLS regressions is similar to the one in the previous section. The main difference is, that the variable holding period is in stronger focus. To decrease complexity and avoid repetition in regressions, the investment timing characteristic bought early (late) is grouped with holding period before the analysis, creating four groups of bought late & short holding period, bought early & short holding period, bought late & long holding period and bought early & long holding period. The resulting regression looks as follows.
!""! = !!+ !!!""!+ ! !!!"#!+ !!!ℎ!"#!+ ! !!!"#$&!ℎ!"#!+ !!!"#$%&!ℎ!"#! + !!!"#$%&!"#$!+ !!!"" ∗ !!"!+ !!!"" ∗ !ℎ!"#!
+ !!!"" ∗ !"#$&!ℎ!"#!+ !!"!"" ∗ !"#$%&!ℎ!"#! + !!!!"" ∗ !"#$%&!"#$!+ !"!!"#$% + !"#$%!!""!#$%
The dummy variable shorti, indicates if the deal holding period was short. The
variable late&shorti represent deals that were bought late and held short. The
variables early&shorti and early&longi are defined in the same way. All variables are again interacted with add-on activity. Fixed effects are similar to the previous regressions.
This study expects that a short holding period is positively correlated with deal performance. The variable of major interest is the interaction term of add-on activity and short holding period. As mentioned in section 2, this study expects the estimate of this interaction term to be negative. The additional variables help to get a more detailed picture of the concrete relationship.
4.4. Proportion of add-on acquisitions
The last hypothesis is primarily concerned with the proportion of add-on acquisitions relative to the portfolio company. As argued in section 2, it is reasonable to expect that a higher proportion of add-on acquisitions results in greater deal performance. An obvious approach to measure the proportion of add-on is to set the company value of the acquired company in relation to the portfolio company value at the time of acquisition. However, this approach needs an extensive data set that includes, among other things, a reasonable measurement for the portfolio company at
several points in time during the holding period. As the data set for this analysis is not extensive enough to provide this kind of data quality, a second best approach was used. This approach measures the proportion of add-ons through three financials (revenue, EBITDA and EBIT) and is performed as follows: For revenues, the accumulated amount of revenues of all add-on acquisitions is related to the amount of revenues of the portfolio at entry investment. The proportion of add-ons measured by EBITDA and EBIT is calculated in the same way. To ensure validity of the results, the analysis only considers deals that have the relevant financial data needed for the entry deal and all add-on acquisitions. In addition, the sample only considers deals that used an inorganic growth strategy. This strongly limits the sample up to eleven observations. Not all control variables and fixed effects are hence included into the regression.
!""! = !!+ !!%!"#"$%"!+ !!%!"#$%&!+ !!%!"#$!+ !!ℎ!"#$%!!!"#$%&!
+!!!"#$%!!"#$!+ !!!"#$!!"#$!
To verify the last hypothesis, the estimates of β1 to β3 should be significant and positive. Since the sample only includes deals with inorganic growth strategies, positive coefficients would suggest that a higher proportion of add-on acquisitions is beneficial with respect to deal performance for inorganic growth strategies.
5. Empirical results
The hypotheses are tackled by analyzing four different tables. The tables are in general structured in the same sequence as the hypotheses. Table 3 shows deal performance of buyouts and changes in financials of the underlying portfolio firm. Table 4 takes special attention to the relationship of deal performance, the entry channel and investment timing, whereas table 5 emphasizes changes in deal performance and the length of holding period. Finally, table 6 analyzes the effect of add-on proportion on deal performance. All regressions include standard errors clustered by entry years of the portfolio companies and by the related PE firms. Furthermore, the regressions are controlled for fixed effects that cover firm and fund
characteristics as well as portfolio company and deal characteristics. Given the specification, control variables are holding period, investment size (in million) and a dummy for club deal.
5.1. Add-on activity
The first hypothesis that focuses in general on the relationship of deal performance and inorganic growth strategies, is subject to table 3, 4 and 5. All these three tables include a dummy variable for add-on activity and hence indicate the effect of add-on acquisitions on deal performance. The add-on dummy is present throughout multiple tables since all hypotheses are related to add-on activity.
The first regression specification in table 3 shows the effect of add-on activity on deal performance including control variables, but neglecting fixed effects. The effect is positive, but statistically insignificant. When the fixed effects (specification 2) are included, the effect is significant at a 10% level. Interpreting this value, on average deals that make at least one add-on transaction have investment returns that are about 14 percent points higher than deals that do not use any inorganic growth strategy. However, even if the effect in the subsequent five specifications is also positive and similar in magnitude, only the add-on variable in the last specification is significant.
Table 4 drafts a similar picture about the relevance of add-on activity. All four estimates are positive, but only two estimates are significant at a 5% level.
Nonetheless, table 5 displays again that all estimates are positive and this time, four of the six specifications are significant at the 5% level. Overall, the results support the findings of Hammer et al. (2016), and Nikoskelainen and Wright (2007), even if not all estimates are significant. The findings speak for a positive relation of inorganic growth and deal performance and hence confirm the first hypothesis.
As mentioned before, for exploratory analysis performance of buy-and-build strategies was examined. This study assumes that the PE management of a target company, that was bought early and acquired at least one add-on company, was following a buy-and-build strategy. One would expect that this strategy turns out to
be successful since the PE management does not face any selling pressure and hence has time to carefully select its add-ons and make better investment decisions. However, as can be seen in table 4, the estimate for this interaction term reveals the opposite effect, significant on a 5% level. Following the previous argumentation, buy-and-build strategies tend to turn out as value destroying strategies.
Table 4 reveals another conspicuous aspect about add-on acquisitions. As already indicated by figure 1, the number of add-on acquisitions is not significantly correlated to deal performance. In addition, both estimates of the continuous variable for add-on activity are almost zero. A reason for this might be that the underlying data for add-ons are not sufficiently documented regarding the size of add-add-ons, meaning that there is no differentiation between a deal that acquired relatively large add-ons and a deal that added the same amount of companies, but relatively small ones. These differences in add-on sizes are subject to hypothesis 6 and will be discussed in section 5.5.
Results on add-on activity hence suggest that, in general, add-ons relate positively to deal performance, which supports hypothesis I. However, the positive effect does not depend on the number of add-ons. Furthermore, buy-and-build strategies that do not face selling pressure, do not seem to enhance deal performance.
5.2. Changes in financials
The second hypothesis predicts that increases in financials of the portfolio firm have a stronger positive impact on deal performance, when the managing PE firm pursues an inorganic growth strategy. The assumption here is that inorganic growth is assumed to lead to higher increases in revenues and earnings through e.g. cross selling potentials and higher potential for efficiency improvements (cost reductions) across the platform company and the added firms. Hence, this study assumes that improvements in financials function as a signal for potential buyers, which is important since an add-on strategy leads to greater uncertainty about the firm performance for outsiders.
Table 3 shows the effects of changes in the three main financial figures revenue, EBITDA, EBIT and also the changes in EBITDA margin and EBIT margin. Evaluating the isolated effect of theses changes, without considering add-on activity, the regressions do not depict a clear picture. Improvements in revenues are positively correlated with deal performance at a 10% significance level. This finding is in line with Acharya et al. (2013), who also find a similar but even stronger positive correlation. However, the coefficient for changes in revenues turns negative, but not significant, when including changes in EBITDA and EBIT.
For changes in EBITDA only one of the three estimates is positively and significantly correlated with deal performance. The relationship of changes in EBIT and deal performance that are presented in specification 5 and 6 also remain vague. Both estimates are positive, but only one is significant at the 10% level.
Overall, the findings hence do not mirror the results of e.g. Acharya et al. (2013). On the other hand, all effects that are significant support that each of the changes is positively correlated with deal performance.
For hypothesis II, the estimates of interest are the interaction terms of the main financials with add-on activity. The interaction term of add-on activity and changes in revenue, shown in specification 4 and 6 of table 3, suggests a positive correlation with deal performance at a 5% and 10% significant level, respectively.
The effect of changes in EBIT for deals with inorganic growth strategies are also in line with the second hypothesis, although results are not consistent over all specifications. Changes in EBITDA for deals that acquired at least one add-on, however, show an opposite picture. All three estimates are negative, whereby one is significant at the 5% level. Lastly, the interaction terms of add-on activity with EBITDA margin and EBIT margin are both insignificant.
In total, the findings regarding the impact of changes in financials on deal performance are mixed. While findings for revenues and EBIT (even if only partly significant) support hypothesis II, results form the interaction of EBITDA contradict the hypothesis. Due to this inconsistency, this study confirms the second hypothesis only with limitations for revenue and EBIT.
