MSc Finance
Quantitative Finance Track
Master Thesis
The effect of LBO transactions on firms’ innovation activities
by
Yun Hong
11925973
Student
07.2018
Supervisor: Tolga Caskurlu
Statement of Originality
This document is written by Student Yun Hong 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.
Abstract
This paper examines the impact of the LBO transactions on portfolio companies’ innovation activities. I adopt the difference-in-difference estimation to reveal that, relative to the cancelled LBO transaction group, firms apply a lower number of patents in the short term but the quality of the patents has improved more over a longer timeframe. The treatment effect documents the following facts. First, target firms of successful LBO transactions experience a high debt leverage problem and financial distress, and less the number of patent in the short-term. Second, using the citation-weighted value and market-value of patents, the results show that the innovation activities are improved in the long term.
Table of Contents
1. Introduction ... 4
2. Literature review ... 6
2.1. Private equity and performance ... 7
2.2. Leveraged buyout ... 8
2.3. R&D, patents and innovation activities ... 10
2.4. Hypothesis development ... 12
3. The Sample ... 13
3.1. Identifying Private Equity Transactions ... 13
3.2. Capturing Patent Data... 13
4. Estimation Strategy ... 16
5. Results ... 20
6. Robustness check ... 25
7. Discussion and Conclusion ... 25
1. Introduction
A private equity firm is an investment management company that can provide financial backing and makes investments for small enterprises, such as startups. These companies are called “portfolio companies”. They are acquired by common investment strategies, including leveraged buyout (LBO), venture capital, growth funds, etc. Leveraged buyout is one of the most common strategies that PE like to use. It first emerged as an important phenomenon in the 1980s in the famous case of KKR leveraged buyout collecting a $75 million fee in the RJR takeover. During the leveraged buyout transaction, the portfolio companies are acquired using a high leveraged debt finance in order to pay for the transaction. Their assets usually act as collateral for raising the loan.
There are mainly two views on whether leveraged buyouts can benefit portfolio companies in the future. Some proponents argue that LBOs leads to a change of company structure, which will encourage managers to improve the general performance and pursue better growth opportunities (Liu, 2017). Managers also cut off some unprofitable investments because of the improvement of management efficiency (Smith, 1990; Wright, 1986; Jensen,1989) On the other hand, some papers argue that portfolio companies may suffer more because of financial distress and a lack of cashflow.
Innovation activities is one of the most essential reasons that can affect a company’s market share and its future development. These activities are always related to companies’ management and investment strategies. They can drive companies to be successful and can sustain each company’s unique character in the market. However, it really depends on the size and industry of the firm. In other words, innovation activities can somehow be a measure of the firm’s growth. In this paper, I will use the number of the patents to measure the innovation activities for firms. Patents are widely used in measuring the outcome of PE-backed companies’ innovation activities because patent will be easier to be observable. The most straightforward use of patents involves a simple count of the number of patents that an organization produces, usually coded by time, industry and other characteristics. Lerner et al. (2011) find that firms' level of patenting does not change post-LBO but patents are more frequently cited, which suggests that portfolio firms are producing research that has greater economic impact; however, they are unable to determine whether this is a selection or a causal effect. Besides questions of how transactions affect firm’s innovation activities, there
are a group of studies talking about standing with the right indicators of innovation activities. Using R&D expenditures is popular but there are also some substantial papers putting forward the idea of citation-weighted value of the patent and market-value of the patent.
Ultimately, the nature of the changes in corporate time horizons associated with private equity transactions is an empirical question on the micro-economic level. The research question of how LBO transaction effect the innovation activities in the long run can reflect the growth of the company. Following this research question, we can not only see how LBO transactions affect portfolio companies’ financial structure, but also the aggregate growth of the innovation activities of the portfolio companies. Naturally, the first question answered will be how LBO transaction companies affect the innovation activities. To answer this question, I first examine the relation between the number of patent that is applied after the LBO transaction year. I found that the number of the patent drop 19.6 percent for the portfolio company with successful LBO transactions. The results confirm the negative effect of LBO transactions on the number of patents when compared to the control group. Combined with the result in the probit model, it can indicate that portfolio companies experience the financial distress following the LBO. This result is different from the result found by Liu as that study found improved operational performance in the successful LBO transaction group (Liu, 2017). Secondly, I test the relation between citation-weighted value of patent and the market-value of the patent. The term interactions change to positive for both variables, which shows a positive effect of LBO transactions on the long-term innovation activities. The result suggest there is a positive effect of LBO transactions on the innovation activities in the long-term, which corresponds to the study by Amess (2015). Overall, I was able to show that private equity backed LBO transactions have positive effects on firm’s innovation activities in the long-term. The analysis reveals that the number of patent drop after the LBO transactions but the quality improved more in the future.
In this paper I present evidence about one form of long-run investment, changes in innovative investments around the time of private equity transactions. The contribution of this paper is 1) Most of the paper focus on operational performance but not the innovation activities, which can reflect the growth of the company. 2) This paper uses three variables to test the innovation activities, which can reflect the change in a longer scope. 3) This paper can fit the gap for the study of what effect LBO transactions have on portfolio companies in the
United States market and make the result more recent as most of the studies into this fields are at least a decade older. In this research, I adapt the difference-in-difference methodology as many previous research (Amess, 2015; Liu, 2016; Kogan et al, 2017;). The interaction term captures the post-differential effect on innovation activities of the LBO transactions. For the data construction, I use Capital IQ as the dataset for gathering data of the portfolio companies’ profile from 2000 to 2010, specified in the United States market. I also choose the Seru (2017) pre-structured dataset for gathering information of patents and use the link table from Wharton dataset to link CIK and Permno identifier.
Lastly, the rest of the paper is organized as follows. The following section, section 2, presents a literature review for this study, which compromises of three parts. Subsection 2.1-2.3 contains the literature review for Private equity, Leveraged buyout and indicators of firm innovation. Subsection 2.4 details the hypotheses. Section 3 expands on the dataset and sample. Section 4 presents the methodology. Section 5 shows the results of this study. In Section 6 the robustness check can be found. And finally in Section 7 the conclusion and discussion are presented. This paper also concludes with direction of further research and practical relevance.
2. Literature review
This part is going to review other relevant studies and develop the hypotheses that I examine in this paper. In order to ensure the literature cited in this paper is from professional and reliable sources, the included related papers are mainly from three sources. Firstly, I’m using two electronic databases, one being Google Scholar and the other one is JSTOR. They were searched for relevant papers using basic search strings. As this paper is mainly concerned with private equity, leveraged buy-out transactions and innovation activities, papers in journals such as the Journal of Finance, Journal of Financial Economics and the Journal of Management are frequently cited. Furthermore, the online library of the University of Amsterdam was also used. There are some related keywords that are being used here, such as “leveraged buyout”, “R&D expenses” and “Patent and innovation activities”. I also found more related literature by scanning reference lists for highly relevant articles. Using this way broadened my thinking and provides me with angles and viewpoints to examine and investigate the subject matter. For example: “value created during failed transactions”, “firms behavior” and “Internal finance”.
2.1. Private equity and performance
Private equity firms have been growing very fast since the 1950s in the United States. So much so that they have played and continue to play an important role in the long-term profit improvement of portfolio companies. Indeed, some research suggests that private equity firms are able to help improve the management activities, financial activities and taxes of a company by providing professional help in these areas (Kaplan and Stromberg, 2009). Furthermore, there is empirical evidence in Kaplan (1989) that for portfolio companies both the operating income and cash flow increased significantly after the public-to-private deals in the 1980s. In his paper, Kaplan (1989) documented a significant increase in operating return following public-to-private deals as well.
This rise in the number of private equity firms over the past decades might be explained by the unwillingness of banks to support smaller businesses. In their paper Stiglitz & Weiss (1981) have used an empirical method to indicate that banks are not willing to support most of the small and medium size firms. This is mainly due to the fact that these small and medium sized firms generally have less history as they often have a relatively short lifespan and because they have fewer assets (Stiglitz & Weiss, 1981). Another explanation of the growth in the number of private equity firms has been examined by Gebhardt et al (2006). It is stated that private equity firms are able to help firms to increase their value in the long-term period because investors have more experience and can bring more information to the table. They can combine management experience, industry information and additional funding in order to increase the innovation activities of portfolio companies (Gebhardt et al, 2006). Particularly interesting in relation to this research there’s empirical evidence which suggests that private equity firms have a beneficial effect on R&D activities, Kortum & Lerner (2001) show, through data collected in more than 20 industries and over more than 60 years that industries with a higher number of patents have more private equity firms involved, referring to the policy in 1979 of the United States. They have also examined on the firm-level and have found out that when there are private equity firms involved firms usually have more cited patents compared to when there are less private equity firms involved (Kortum & Lerner, 2001).
At this point it becomes clear that the literature suggests that the presence of private equity increases the performance of firms in different areas. One of these areas is innovation and research & Development. Capital markets are often plagued with information asymmetries.
Suppliers of financial resources often ration expenditures following the adverse selection problem (Stiglitz & Weiss, 1981), leading to an underinvestment in activities such as R&D (Hubbard, 1998). Research suggests that this is especially true for intangible investments, among which R&D activities, over more tangible investments as their value is more difficult to perceive, at least directly (Brown et al, 2012). The introduction of private equity, such as through an LBO transaction, can alleviate this strain on investments in R&D expenditure. The well-known agency problem (Jensen and Meckling, 1976)is less prevalent with private equity investors as they work with their own money and subsequently work with a longer time horizon, in which the value of R&D is appreciated more. Indeed as In Griliches (1990) it is also stated that R&D expenditures are characterized by small changes only in the short-term period and only after a longer time span will their full effect and perceived benefits become clear.
Conversely there is some disagreement in the literature as some opposing views can be found. For example Tredennick (2001) stated that there are some investors who would only stay within their familiar industries due to knowing the limits of their own expertise and subsequently refuse to invest in a new area, which doesn’t foster the growth of innovation. Furthermore Shand & Bhide (2000) state that some private equity firms only seek short term profits such that they are more likely to eliminate the uncertainty and the risk coming from any investment in innovation without any immediate return.
2.2. Leveraged buyout
Leveraged buyout has been booming over the last 30 years as it is viewed as one of the most popular win-win financial transactions both for the private equity firms as well as the portfolio companies. Leveraged buyout means that the buyout is typically financed with a very high percentage of debt, usually 60 to 90 percent, and where the buying is done by a group of investors (Palepu, 1990; Kaplan and Stromberg, 2009). LBO transactions have been found to always change the financial structure of their target portfolio companies. In the literature one can find various authors that analyses the effect of post LBO transactions on the different measures of company performance utilizing firm-level data. Liu (2017) test the private equity leveraged on portfolio companies’ operating performance after the public-to-private LBO transaction. The result shows that compared to companies who had failed the process, successful LBOs always improve the earnings and cash flow which scaled by total asses size.
Empirical evidence supports these aforementioned views on leveraged buyouts, where several studies show that corporations are likely to experience tax benefits during the post-buyout period as the high level of debt can offset the higher interest rates, leading to an overall positive outcome (Jensen, 1989; Kaplan, 1989b; Liu, 2017). These papers generally found that after the LBO transaction the general performance has improved for most of the companies. Therefore, corporations have more cash flow, which may be of benefit to R&D activities and innovation activities theoretically. Other research shows that when larger organizations, those organized in a number of divisions, are involved the LBO of one or more of these divisions might occur (Williamson, 1985). Generally, such a divisional structure can lead to the reduction of the effectiveness of the internal capital market as a controlling variable (Long & Ravenscraft, 1993; Williamson, 1985) where underinvestment in one or more division follows as a result often hurting company (division) performance (Hill, 1988). By means of a LBO transaction re-investments in an underperforming division lead to an increase in R&D activities and innovativeness (Williamson, 1985). It should be noted that some research shows a more limited increase in profitability following a move from public to private (Guo et al, 2011).
However, not all research agrees on the beneficial effects of the leveraged buyout on R&D and innovation activities and many papers have an opposing view. They show that the level of capital expenditure as a whole is actually declining after a LBO transaction (Kaplan, 1989a; Liebeskind et al, 1992; Opler & Titman, 1999). These reductions in expenditures usually follows from the cutting of wasteful investment or the profits that follow from successful investments. For example, in Kaplan (1989a), it is shown that a reduction in capital expenditures is mainly due to financial distress and cash constraints following the LBO. What’s more, in Opler & Titman (1999) their results show that companies are not likely to undertake LBOs if they have high capital expenditures to begin with during the initial stages of the pre-LBO transactions. Lichtenberg and Siegel (1990) provide further evidence of this in their research where 43 LBOs are examined during the 1980s where the firms participated in the survey of the Bureau of the Census into research activities prior to and after the LBO transaction. They find that the intensity (read: expenditures) was on average 49% lower after the LBO transaction when compared to the situation pre-LBO (Lichtenberg and Siegel, 1990).
The reasons causing the decline in the level of capital expenditure are mainly twofold. Firstly the decline of capital expenditures is caused by a change in the ownership of the portfolio company, which is usually followed by a significant change in the structure and governance method of said company (Liebeskind et al, 1992). Indeed, this seems to contradict the earlier observation where it was stated that the introduction of private equity would increase income and cash flow following a move to the private sphere (Kaplan (1989b; Kaplan and Stromberg, 2009). It could further be argued that a loss of control can follow when managers fail to be competitive. Instead, private investors (that is: from the private equity firms) are more concerned with ensuring good operating performance as they tend to care more for their own funds that have been invested (Kaplan, 1989a; Muscarella & Vetsuypens, 1990). Long & Ravencraft (1993) state debt has a very close relationship with R&D expenditures because it can show how effective these expenditures have been used. The result of their research shows that post-LBO R&D expenditures are around 40 percent below their pre-LBO level.
2.3. R&D, patents and innovation activities
Research often cites the beneficial effects of a LBO on the R&D activities and innovativeness of a portfolio company. Innovation can be defined rather simplistically as “a new idea, device or method”. However, the literature provides us with more, varying definitions of innovation. According to Joseph Schumpeter’s very first innovation theory, innovation can refer to any new policy that an entrepreneur undertakes to reduce the overall cost of production or to increase the demand for its products (Schumpeter, 1934). This definition suggests that in business it is possible to generate economic profits by introducing successful innovations. Bases on his study other theories and definitions of innovation have been developed. Lynn & Gelb (1996) for example, state that innovation is a process that starts at seeing the potential of one business activity and ends when transferring this potential into products. New product (technology) creation goes hand in hand with patents (Acs et al, 2002; Narin et al, 1987). The relationship between patent based metrics and R&D expenditures has been investigated on many occasions in previous research due to the fact that there is a strong association between them. It is very common in the financial literature to use R&D expenditures as a variable to test how the innovation activities develop (Jaffe & Trajtenberg, 2001; Lichtenberg and Siegel, 1990). Griliches (1990) mentions that the level of R&D expenditures varies widely between big and small firms as well as between different industries. As an example, the
high-technology industry is generally characterized by higher levels of R&D expenditures whereas the retail industry spends less on R&D in general.
Following the lead of previous research, using the number of patents of a portfolio company to measure innovation activities is a valid one, as patents have a long history of being recognized as a measurement of innovation and technology in a firm (Trajternberg, 1990). The number of patents can be taken as a clear signal of how successful a portfolio company is able to utilize the investments in the form of R&D expenditures. Moreover the number of patents a firm has is a more observable metric than the level of R&D expenditure in private equity and private firms. Furthermore, as mentioned before, it is problematic to link R&D expenditures to immediate innovative outcome. Using the number of patents can show the effects of an invention following LBO transactions better, unlike expenditures (Griliches, 1990). The use of patents as a metric can be observed in a number of previous works. In Seru (2014) the impact of the conglomerate form on the scale and novelty of corporate R&D activity productivity is examined using the patent metric. In Lerner, Sorensen & Stromberg (2011), 6,938 patents from 495 firms from 1986 through to 2005 are used to find that the patents of private equity backed firms applied for in the years after the investment are more frequently cited than those applied for before private equity investments. Yet, it can happen that a patent may be applied for unsuccessfully. Also, a patent cannot be cited for every innovation or novelty, depending on the different strategies firms might have (Ham Ziedenis & Hall, 2001). Despite these remarks, the use of the patents as a metric seems well founded and supported in the literature.
Besides solely using the number of patents the citation-weighted value of firm innovation can also be used. The patent citations are like the reference to prior technology, either patents or other scientific literature on which the current patent builds or which it uses. Citations of patents are used to avoid copyright infringements, showing the limits of the patent and protect against possible lawsuits. Furthermore, it can bused by the USPTO examiner or for teaching purposes. There is a substantial amount of literature available that uses this citation-weighted value method for studying micro-economic phenomenon. Aboody & Lev (1998) use citation-weighted patents as a measure of intangible assets. In Hall et al (2001), they use stock of R&D spending, the average yield per R&D dollar and the average citation yield per patent. They found that citations per patent are more important than the patent yield itself and that
when there is one citation per patent increase there is a corresponding 3 percent increase in the market value of the patent. As a whole, therefore, using the citation-value can better test the knowledge transforming power of an innovation as opposed to simply counting the number of patents (Hall et al, 2005). Following the this the market value of patents can show the importance of the related technology. These two indicators, patent citation value and marker value can add more explanatory power for the innovation activities of (portfolio) firms (Korgan et al, 2017).
2.4. Hypothesis development
As can be gathered from the literature review above, the literature on LBO transactions and innovation activities, measured in the number of (citations of) patents is extensive. It has shown that different factors are mentioned as being possible measurements of innovation activities by different authors. These include the number of patents (Trajternberg, 1990), and the citation-weighted value of patents and the market value of firm innovation activities (Kogan et al, 2017). The literature presented two competing views in relation to the effect of LBO transactions on portfolio firm performance, specifically R&D activities and
innovativeness. One the one hand the literature suggest that a move from public to private via LBOs will positively influence a firms innovation, mainly due to re-investment, reducing information asymmetry and a shifting focus from a short to long-term horizon, among others (Gebhardt et al, 2006; Jensen, 1989; Kaplan, 1989b; Kortum & Lerner, 2000; Liu, 2017). Far from reaching consensus, other authors suggest that instead the level of R&D expenditures and innovativeness declines after a LBO transaction, or that investors might have a goal of short-term profit, among other reasons (Shand & Bhide, 2000; Brown et al, 2012; Hubbard, 1998; Kaplan, 1989a; Lichtenberg and Siegel, 1990; Tredennick, 2001). The aim of this paper is to provide an answer to the question of how leveraged buyout transactions affect the R&D expenditures and innovativeness of a portfolio company as measured in the number and citation-weighted value of patents as well as the market value of a firm’s innovation activities.
As mentioned the literature provides two opposing views in relation to the effect of LBO transactions on company R&D and innovation performance. To investigate this effect the following hypotheses have been developed:
Hypotheses 1: The LBO transactions have a negative impact on portfolio companies’
number of patents.
Hypotheses 2: The LBO transactions have a positive impact on citation-weighted of patent
of the portfolio company.
Hypothesis 3 : The LBO transactions have a positive impact on market-value of patent of the
portfolio company.
3. The Sample
To construct the dataset, I identify a comprehensive list of private equity transactions and match the involved firms to U.S. patent records. This section describes this process.
3.1. Identifying Private Equity Transactions
My starting point of sample collection is all the merger and acquisition transactions termed as “Leveraged buyouts” in Capital IQ. Capital IQ covers 11,089 leveraged buyout deals from 2000 to 2010 in the United States. Then I use the following criteria to screen the sample. Firstly, I require the target firms to have a clear transaction status and to have done the transaction in the United States. Then I also required the target firm to have a CIK identifier. Secondly, I drop deals that are classified as “Rumors” or “Pending” and “effective”, due to the ambiguous meaning referred to in the Capital IQ manual. Moreover, I exclude transactions in which the acquiring parties acquire less than 50% of the total shares. In the following step, I display the “deal size”, ”percentage ownership” “revenue” and “R&D” columns, etc. Because my goal for this paper is mainly to focus on the LBO transactions by private equity firms. Unfortunately, because most of the companies go private after the transactions, there is a lot of data which is not eligible for inclusion in the analysis. The final sample consists of 2308 LBOs sponsored by private equities from 2000 to 2010, of which 584 deals fail, and 1724 deals eventually succeed.
3.2. Capturing Patent Data
I restrict our sample to firms with at least one successful patent application after the LBO transactions. I match the firms involved in buyout transactions to their patenting records based on Amit Seru’s constructed dataset from 2016. They use the patent information from
The U.S Patent Office (USPTO) and use the assignee name of each patent to match with companies have CRSP. The number of patents that have been observed and construct after 1926 is 1,801,879. After gathering the information of patents, they examine the relation between citations and the market value of innovation using the number of citations that the patent receives in the future. In Hall at al (2005), he points out the citation can convey information between the investors and assignees along time and space, which can be taken as the “knowledge capital”. Due to the limits of access to databases, using the pre-constructed data can make sure the analysis and result are more reliable and professional for this paper. In Seru (2014), he constructed the market value with the present value of the monopoly profits associated with the good of the patent and the aggregate output. To merge the firms’ profile with Amit’s database, I link the CIK identifier with the link table as constructed by Wharton Database which included the permno identifier for Amit’s dataset. After merging, each company has the number of patents, citation-weighted value and market-valued of the patents. Then I dropped those with duplicates and ambiguous firm names. An observation is only included when I am confident of a match. Finally, I have 910 firms that have LBO transactions and innovation activities with the number of patents of each granted year. During the study, I only used the patent as observation which takes place after the LBO transactions. The dataset also included companies’ locations, industry categories, transaction years and financial indicators.
Table 1. Summary of annual distribution of leveraged buyout and patents
Table 1 reports the annual distribution of LBOs transaction with private equity firms participation, as well as number of the successful LBO and the number of cancelled LBOs. I also combined the number of patent that have been successful applied from 2000 to 2010.
Year Total LBO Successful Cancelled Patents
2000 134 118 16 311 2001 115 97 18 508 2002 164 138 26 477 2003 179 154 25 416 2004 383 189 194 528 2005 199 165 34 365 2006 264 222 42 428 2007 291 235 56 288 2008 172 128 44 117 2009 210 112 98 88 2010 197 166 31 59
Table 2. LBO characteristics
Table 2 reports the mean, median and the standard deviation of characteristics for successful LBOs and Cancelled LBOs. Because most of the cancelled LBOs do not include the financial indicators data in the Capital IQ, the sample number for each variable is less the total number from the previous table.
Panel A: Cancelled LBOs
N Mean S.D
Deal Value 220 813.08 2211.01
Target market cap 238 589.45 1710.24
Target Revenue 282 1684.70 10221.80
Target Debt 281 1020.28 8689.34
Target cash flow 180 1242.37 2830.35
Panel B: Successful
Deal Value 980 935.07 3238.90
Target market cap 715 772.54 2546.84
Target revenue 733 1003.43 3613.03
Target Debt 750 436.38 2124.61
Target cash flow 100 2287.67 4427.89
4. Estimation Strategy
My empirical strategy aims to identify the causal effect of LBOs on firms' innovation outcomes. For this purpose, I constructed variables and combine it with a difference- in differences estimator in order to evaluate the impact of an LBO on portfolio firms.
As I exploit a panel dataset, I can relax the assumption of selection on observables by combining the matching technique with a difference-in-difference estimator (Blundell,2000; Liu, 2007). The control group is the group that doesn’t complete the LBOs transaction. Instead of only comparing the number of patents and R&D expenditures that are grated in two groups, I also focus on the citation-weighted value of the patents and the market-value of the patents. This method allows for the selection of the group of LBO transactions to be based on the expected impact on innovation and on time invariant unobservable characteristics. (Seru, 2014)
The difference-in-differences estimator is obtained by applying weighted least squares to the matched data set. My baseline equation is:
For this equation, the dependent variable ln(I) is the log normal conditional of the outcome of the innovation activities, which is measured by the number of successful patents application per firm in a given year, the citation-weighted value of patent and the market value of the patents. I use the number of patents, the citation-weighted of the patents and the market-value of the firms’ innovation as indicators of observations’ innovation activities.
Post is a dummy that equals one if a patent is applied after the LBO transaction year. The
status measures, Treated, is the measure of successful LBO transaction for the firm and equals one if the status is “Successful” and zero otherwise. The interaction term Post*Treated captures the post-differential effect of the LBO transactions. For the fixed facts in the panel data, I included industry (SIC code), company id (CIK identifiers) and the year of the LBO transactions to control for the characteristics of those categories that might affect the dependent variables and eliminate potential omitted variable bias. ε is an error term.
The innovation activity always has a long cycle, which has been affected by exogenous factors and endogenous factors. There are some exogenous factors that are mainly from the macro environment, such as policies. Those factors also affect the whole innovation process and may lead to different results, and have an influence on the final results and profits from the innovations. It is very difficult to use all the control variables. Therefore, I am using four control variables that can be quantized from my data.
The size of the firm. Since the beginning of the innovation activity research, lots of researchers believe that the size of the firm has significant effect on the innovation activities of firms. The famous economist Joseph Schumpeter pointed out that only large firms could afford the change of the important technology because it would make small firms lose their advantages since they are not able to optimize their R&D. In other words, bigger firms can bear the risks from if the innovation failed. Veugeler & Cassiman (1999) state that we can use a “U” curve to describe the relation between the firm size and innovation activities, which means if the firm size has reached a certain high level, it benefits the innovation activity. Stock & Zacherias (2011) stated the opposing conclusion that smaller firms have more incentives to
engage in innovation activities. The main argument for this is that a firm would invest a lot for technology in order to get advanced in the market and the industry. However, after the firm can monopolize the market, it’s motivation of innovation activities is less. Therefore, the size is an important control variable in the whole innovation activity and progress.
Financial performance. There is a lot of research showing that one of the most important related factors of innovation activities is financial health. Usually the innovation activities have a positive correlation with cash flow, revenue and turnover ratio among others. During the leveraged buyout, PE usually use a combination of equity and debt, and the debt will constitute a majority of the purchase. Therefore, the leveraged buyout transaction always changes the capital structure. Beneito (2003) states that when a firm has more surplus capital and high revenue this can motivate the innovation activities. Also, it can shorten the innovation progress and brings more motivations. Galende & De la Fuente (2003) state when a firm is taking higher risk when there it has a higher debt level. Thus, they are less willing to invest in the innovation activity. The cash flow is also an important indicator of financial performance. When the financial performance improves, there is more cash that can be invested in the innovation activities. Therefore, I construct the cash flow variable by using the EBITDA of the latest twelve month minus the capital expenditures, both sets of data coming from the Capital IQ dataset.
Ownership percentage. After the leveraged buyout transactions, the private equity firms own shares of their portfolio companies and this can hold back the “free ride phenomenon when PE firms have the management opportunities”. Chung & Firth (2002) show in their research that it can effectively eliminate and supervise the company’s shareholders only gathering their own profits when PE firms hold more shares. This can also reduce the information asymmetry. Bushee (1998) also states that less portfolio companies cut the R&D for the short-term profits when more shares are owned by the PE firms after leveraged buyout transactions.
Table 3. Variable definitions
This table represent the variable definitions for all the target firms. For examining the financial health, I construct some other variables, such as the Cash flow and Profit %. All the data come from Capital IQ dataset.
Variable Description
Treated = 1 if the LBO is successful, 0 else
Post = 1 if for all year after the year of LBO, 0 else Firm type =1 if the company backed LBO is PE/VC, 0 else Patent count Number of patent applications in current year Asset Asset
Size The market capitalization Cash flow EBITDA- capital expenditure Profit % EBITDA/ Revenue
Citation Citation-weighted value Market Market value
Table 4. Probit model (Dependent variable = Successful )
This table reports the probit regression relating the probability of a deal succeeding for a target to the targets’ financial indicators. Size equals to the market capitalization of the firm. EBITDA/asset presents to the profit percent of the firm. Cash flow equals to EBITDA minus capital expenditure. All the values of variables are log normal transformed . *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.
Independent variable Coefficient (standard error)
ln_asset 0.504** (0.023) ln_debt 1.754** (0.894) ln_Size 0.241* (0.091) ln_(EBITDA/asset) 0.057 (0.040) ln_cash flow -0.227 (0.365) ln_Patent 1.634 (1.08)
Patent citation stock 0.22
(0.486)
Observations 2308
Pseudo R squared 0.6116
Log likelihood -2286.5
5. Results
In this section it is examined whether LBO transaction can affect the innovation activities of firms. Given the full example of LBO transactions with the number of patents, the citation-weighted value and the market value of the patents, the Difference in Difference
methodology is used to test if the LBO transactions have an effect on them.
The results following this analysis are shown in table 5. They confirm the previous findings that LBO transactions decrease the number of patents to be applied for after a LBO
transaction. Following Seru (2014), The result shows the coefficient is negative, which means that the number of patents applied for after a successful LBO transactions is 19.6 percent lower compared to the control group, as can be seen from column (1), at the 99% significance level. What’s more, In column (2), column (3) and column (4) it may be
observed that the coefficient does not change much as it seems stable at around minus 20 percent, whilst also staying very significant. Meaning that even with the control variables added, the results show that the number of patents applied for after successful LBO transactions is nearly 20 percent lower that the control group.
This result is in accordance with Seru (2014) where it is stated that the managers of the firms may have difference incentives about innovation activities. As mentioned in the hypothesis section, the reason may be that after the transactions, the “free rider” problem can be eliminated. Investors may pay attention to the profits more than pre-LBOs.
Therefore, the managers may pursuit a better profit than other inefficient investment. Also, as the result shows in table 4, the debt has been higher after the LBO transactions so the firm may face the financial constraints. It need to cut some investment to acquire more cash flow. This statistics evidence has also been proved in Lichtenberg at nl (1990). In sum, in this part, LBO transactions are not benefit to the innovation activities, with the empirical result of the effects on the number of patent for post LBO transactions showing.
Table 5. The effect of LBOs on the number of patents of the firm
Table 5 presents the main results using the full sample of cancelled LBOs transactions as the control group. A difference-in-difference specification is employed with the innovation activity defined as the number of patents. The unit of the observation is a firm-year. Treated is a binary variable that takes a value one in the treated group and zero for the control group. The results show that successful LBO transactions greatly decrease the number of patents that the portfolio companies applied for after the year LBO transactions took place. All standard errors are clustered at the firm level. Size equals to the market capitalization of the firm. EBITDA/asset presents to the profit percent of the firm. Cash flow equals to EBITDA minus capital expenditure. All the values of variables are log normal transformed . All the standard errors in parentheses. Constants were included in the regressions but are not reported. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.
(1) (2) (3) (4)
VARIABLES ln patent ln patent ln patent ln patent
Treated × Post -0.196*** -0.194*** -0.193*** -0.236** (0.0645) (0.0647) (0.0647) (0.0964) Post 0.109*** 0.109*** 0.109*** 0.0970* (0.0362) (0.0362) (0.0362) (0.0532) Treated 0.236 0.236 0.224 0.0624 (0.155) (0.155) (0.156) (0.890) ln_size -0.0318 -0.0449 -0.993 (0.0370) (0.0420) (1.395) ln asset 0.0307 -9.457 (0.0465) (13.57) ln cashflow -0.0224 (0.0227)
Firm fixed effect Yes Yes Yes Yes
Time fixed effect Yes Yes Yes Yes
Industry fixed effect
Yes Yes Yes Yes
Observations 2,841 2,841 2,841 1,480
R-squared 0.677 0.663 0.663 0.696
Next, in order to test the long-term innovation activity development I have examined the effect of LBO transactions on the citation-weighted value of the patents. The results in table 6 reports whether the LBO transactions have effects on the citation-weighted value of the patents. The citation-weighted value of the patents is obtained from Seru’s dataset (Seru, 2016; Kogan et al, 2017). Using the citation-weighted value of the patent as the dependent variable allows for the examination of the forward value of these patents. In column (1), the
result shows that the coefficient of the term of the interaction of Post * Treated is positive and significant at the 99% level. In other words, the citation-weighted value of patents increased by 34.5% compared to the control group, which is those portfolio companies where the LBO transaction has been cancelled. After adding the financial indicators as control variables the result doesn’t change much: the coefficient improved by 0.04 in column (2) and column (3).
Since I am using the pre-constructed dataset by Seru (Seru, 2016; Kogan et al, 2017), it is worth noting that he states that the citation-weighted value of patents is one of the most popular approaches for measuring the innovation output because it can indicate the firm’s own innovation output and future growth. Indeed, as we have seen the results report that the citation-weighted value of the patents improved greatly after successful LBO transactions when compared to the cancelled LBO transactions group. This is in line with my second hypothesis that the innovation activities of a portfolio company improve in the longer term as measured by the citation-weighted value of patents.
This result is consistent with the estimate obtained by Amess et al. (2015). Because of adequate and professional information from both the industry and the inside of the company, firms may have a better quality patents. Also, after the improvement of operational performance, the firm relaxes from its financial constraints so there is more free cash flow for the development the innovation activities.
Table 6. The effect of LBOs on the citation-weighted of patent of the firm
Table 6 presents the main results using the full sample of cancelled LBOs transactions as the control group. A difference-in-difference specification is employed with the innovation activity defined as the citation-weighted value of the patents. The unit of the observation is a firm-year. Treated is a binary variable that takes a value one in the treated group and zero for the control group. The results show that successful LBO transactions greatly increase the citation-weighted value of the patents of the portfolio companies after the year in which the LBO transactions took place. In the panel data, the firm fixed effect, industry fixed effect and time fixed effect are controlled. All standard errors are clustered at the firm level. Size equals to the market capitalization of the firm. EBITDA/asset presents to the profit percent of the firm. Cash flow equals to EBITDA minus capital expenditure. All the values of variables are log normal transformed. Size equals to the market capitalization of the firm. EBITDA/asset presents to the profit percent of the firm. Cash flow equals to EBITDA minus capital expenditure. All the values of variables are log normal transformed .All the standard errors in parentheses. Constants were included in the regressions but are not reported. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.
(1) (2) (3) (4)
VARIABLES l_citation l_citation l_citation l_citation
DiD 0.345*** 0.349*** 0.349*** 0.616*** (0.0786) (0.0895) (0.0896) (0.127) Post -1.484*** -1.485*** -1.485*** -1.507*** (0.05) (0.0501) (0.0502) (0.0700) Treated -0.386* -0.380* -0.375* -1.528 (0.212) (0.214) (0.216) (1.171) ln_size -0.0525 -0.0576 0.397 (0.0512) (0.0582) (1.836) ln_asset 0.0119 4.907 (0.0643) (17.86) ln_value 0.0199 (0.450) ln_cashflow -0.0223 (0.0299)
Firm fixed effect Yes Yes Yes Yes
Time fixed effect Yes Yes Yes Yes
Industry fixed effect Yes Yes Yes Yes
Observations 2,841 2,841 1,480
R-squared 0.794 0.794 0.783
Furthermore, the effect on the market value of the innovation activities is also tested. In table 7, I examine the effect of LBO transactions has on the market value of the patent a portfolio company. In column (1), the result shows the coefficient of the term interaction is positively related to the market value of the patent at the 95% significance level. This is less significant when compared to the number of patents and the citation-weighted value mentioned above. From column (2) through to column (4), more financial indicators were added as control
variables. The coefficients themselves do not change greatly but the significance level is less when compared to column (1). In Column (4), the Post * Treated interaction drops to 0.091 and the significance level is only at the 90% level, where market capitalization, assets, value and cashflow are the control variables.
This result, of a positive coefficient but with a lower significance level than the results in tables (5) and (6), follows the hypothesis above. It was hypothesized that LBO transactions have a positive effect on the market value of patents as opposed to the control group. This is mainly because the long-term performance has been improved. After the LBO transactions, investors from different fields get involved. This can provide a brighter future for the patents and improve their market value. However, the reason the relation may be weaker (and less significant) is because there are a lot of other factors that can affect a firms performance, such as the followed strategy and management style.
Table 7. The effect of LBOs on the market value of patent of the firm
Table 7 presents the main results using the full sample of cancelled LBOs transactions as the control group. A difference-in-difference specification is employed with the innovation activity defined as the market value of the patents. Treated is a binary variable that takes a value one in the treated group and zero for the control group. The results show that successful LBO transactions increase the market value of the patents of the portfolio companies after the year the LBO transactions took place. In the panel data, the firm fixed effect, industry fixed effect and time fixed effect are controlled. Size equals to the market capitalization of the firm. EBITDA/asset presents to the profit percent of the firm. Cash flow equals to EBITDA minus capital expenditure. All the values of variables are log normal transformed .All standard errors are clustered at the firm level. All the standard errors in parentheses. Constants were included in the regressions but are not reported. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.
(1) (2) (3) (4)
VARIABLES L_market value l_marketvalue l_marketvalue l_marketvalue
DiD 0.151** 0.151** 0.150* 0.091* (0.118) (0.119) (0.119) (0.182) Post -0.866** -0.866** -0.866** -0.702** (0.0665) (0.0665) (0.0666) (0.100) Treated -0.0712 -0.0669 -0.0582 -0.839 (0.283) (0.284) (0.286) (1.679) ln_size -0.0647 -0.0742 -0.356 (0.0680) (0.0772) (2.632) ln_asset 0.0222 -2.794 (0.0854) (25.61) ln_value 0.0737 (0.645) ln_cashflow -0.0675 (0.0429) Observations 2,841 2,841 1,480 R-squared 0.646 0.646 0.591
6. Robustness check
In this section some checks of robustness are performed for this study. The check is to ensure the results are robust to sampling methods and the measurement of key variables. First, innovation activities can be very different between industries referring to different level of the demand, strategies and investment selection by investors. As the sample
consists of many companies in various industries it could be that an industry effect is at play. Therefore, I add industry as one of the fixed effects, the coefficients do not shift much as a result. Therefore, the result can be considered robust.
Next, for further showing the robustness, I also use some other control variables. I use the debt divided by EBITDA as the leverage ratio and construct the profit ratio as EBITDA/ profit. After adding new control variables, the results hold, as can be seen from the tables
presented earlier (tables 5, 6 and 7).
Last, I also change the time period to 1995 to 2010. Because usually the citation-weighted and market-value of patents present long term growth, I extend the time period to 1995 and the effect of LBO transactions on innovation activities of firm hold. However, I believe further research should look at this research in an even longer scope to present a better empirical result.
7. Discussion and Conclusion
In this paper the effect of leverage buyout transactions on company performance in terms of R&D and innovative activities was examined empirically, using the
Difference-in-difference method. The innovativeness of a portfolio company was measured by the number of patents, the citation-weighted value and the market value of patents. Using (Citation-weighted value & market value of) patents as a proxy to measure R&D activities and innovation has been widely documented in the literature (Hall et al, 2005; Lerner, Sorensen & Stromberg, 2011; Lev et al, 1998; Kogan, 2017).
This paper contributes to the debate concerning the effects of LBO transactions on the performance of portfolio companies in the field of R&D activities and innovation. The literature is characterised by a lack of consensus on what the effects of LBO transactions on the performance of portfolio companies are, especially in relation to the R&D and
in this literary debate by using the number of patents, the citation-weighted value and the market value of patents as measures of company innovativeness and R&D activities. It was found that after the LBO transaction the number of patents applied for tends to decline compared to our control group, supporting the first hypothesis. The citation-weighted value and the market value of these patents would increase compared to the control group, supporting hypothesis 2 and 3. The decrease in the number of applied patents was expected as it follows the available literature in that a decrease in capital expenditure and R&D activities following a LBO is a well-documented phenomenon (Jensen, 1989; Kaplan, 1989b) and would therefore support the theory here. It is also possible that because investors tend to stay within the industry they know well there’s no need for new patents (Tredennick, 2001). It is also stated that capital expenditures tend to drop after an LBO (Liebeskind et al, 1992; Opler & Titman, 1999) where R&D and innovation often suffers as a result (Lichtenberg & Siegel, 1990). However the results also show that the citation-weighted value as well as the market value of the patents increased significantly (up to 15 and 30 percent respectively). This means that the patents of a portfolio company tend to be valued more after a LBO transaction when compared to those companies who did not go through an LBO successfully. The literature provides a possible explanation for this as the citation-weighted value and market value of patents could represent the aggregate long-term growth of the portfolio company. This result can also be seen as the logical conclusion of the increased performance of a firm following the reduction of wasteful investments and altogether better performance which would increase trust and appreciation in the public sphere.
Some limitations to this empirical analysis should be noted. By means of several other variables used in the analysis an attempt was made to control the effect other variables may have on innovation. However it should be clear that other factors may still be in play. The sample consisted of a large number of businesses, varying in size and industry. It is very possible that different results can be obtained when repeating the research in specific industries. The same can be said for company size as the research already hints at a
difference between smaller and larger firms. Another limitation can be found in the dataset where the time period of the data used is not as long as some previous studies where data was collected over several decades. For example, the decrease in patents applied for could
be reduced initially during the period of upheaval directly following the LBO and later rise when performance optimizations have been realised. Specific research into why the number of patents drop is welcomed. Further research will have to determine the effects of LBO transactions on firms of different sizes and in different industries as well.
It can be concluded that LBO transactions have a negative effect on the innovation activities in the short term. Secondly following the results provided by the citation-weighted value and market value of the patents, LBO transactions have a positive effect on the innovation activities on the firm level in the longer term. In a practical sense, LBO transactions have the capability of improving the overall performance of a firm. Engaging in a LBO to improve cashflow and to change the financial and governance structures might be a successful strategy. Doing so for the sole purpose of improving R&D activities and to improve
innovation should be considered more carefully. A long-term horizon should be employed as benefits from a LBO to innovativeness take time to be realized.
Reference List
Aboody, David, and Baruch Lev. "The value relevance of intangibles: The case of software capitalization." Journal of Accounting research 36 (1998): 161-191.
Acs, Zoltan J., Luc Anselin, and Attila Varga. "Patents and innovation counts as measures of regional production of new knowledge." Research policy 31.7 (2002): 1069-1085.
Amess, Kevin, Joel Stiebale, and Mike Wright. The impact of private equity on firms' innovation
activity. No. 184. DICE Discussion Paper, 2015.
Beneito, Pilar. "Choosing among alternative technological strategies: an empirical analysis of formal sources of innovation." Research policy 32.4 (2003): 693-713.
Blundell, Richard, and Monica Costa Dias. "Evaluation methods for non‐experimental data." Fiscal
studies 21.4 (2000): 427-468.
Brown, Reilly. "Analysis of investments & management of portfolios." (2012).
Bushee, Brian J. "The influence of institutional investors on myopic R&D investment behavior."
Accounting review (1998): 305-333.
Chung, Richard, Michael Firth, and Jeong-Bon Kim. "Institutional monitoring and opportunistic earnings management." Journal of corporate finance 8.1 (2002): 29-48.
Galende, Jesús, and Juan Manuel de la Fuente. "Internal factors determining a firm’s innovative behaviour." Research Policy 32.5 (2003): 715-736.
Gebhardt, Georg, and Klaus M. Schmidt. "Conditional allocation of control rights in venture capital finance." (2006).
Griliches, Zvi. "Patent statistics as economic indicators: a survey." R&D and productivity: the
econometric evidence. University of Chicago Press, 1990. 287-343.
Guo, Shourun, Edith S. Hotchkiss, and Weihong Song. "Do buyouts (still) create value?." The Journal
of Finance 66.2 (2011): 479-517.
Hall, Bronwyn H., Adam B. Jaffe, and Manuel Trajtenberg. The NBER patent citation data file:
Lessons, insights and methodological tools. No. w8498. National Bureau of Economic Research,
2001.
Ham Ziedonis, Rosemarie, and Bronwyn H. Hall. "The effects of strengthening patent rights on firms engaged in cumulative innovation: insights from the semiconductor industry." Entrepreneurial Inputs
and Outcomes: New Studies of Entrepreneurship in the United States. Emerald Group Publishing
Limited, 2001. 133-187.
Hill, Charles WL. "Internal capital market controls and financial performance in multidivisional firms."
The Journal of Industrial Economics (1988): 67-83.
Hubbarrd R. Glenn. “Capital market imperfections and investment.” Journal of Economic Literature 36 (1998): 193-227
Jensen, Michael. “Eclipse of the public corporation” Harvard Business Review 67, (1989): 61-74
Jensen, Michael C., and William H. Meckling. "Theory of the firm: Managerial behavior, agency costs and ownership structure." Journal of financial economics 3.4 (1976): 305-360.
Kaplan, Steven. "Management buyouts: Evidence on taxes as a source of value." The journal of
finance 44.3 (1989b): 611-632.
Kaplan, Steven. "The effects of management buyouts on operating performance and value." Journal
of financial economics 24.2 (1989a): 217-254.
Kaplan, Steven N., and Per Stromberg. "Leveraged buyouts and private equity." Journal of Economic
Perspectives 23.1 (2009): 121-46.
Kogan, Leonid, et al. "Technological innovation, resource allocation, and growth." The Quarterly
Journal of Economics 132.2 (2017): 665-712.
Kortum, Samuel, and Josh Lerner. "Does venture capital spur innovation?." Entrepreneurial inputs
and outcomes: New studies of entrepreneurship in the United States. Emerald Group Publishing
Limited, 2001. 1-44.
Lerner, Josh, Morten Sorensen, and Per Strömberg. "Private equity and long‐run investment: The case of innovation." The Journal of Finance 66.2 (2011): 445-477.
Lichtenberg, Frank R., and Donald Spiegel. The effects of leveraged buyouts on productivity and releated aspects of firm behavior. Journal of financial Economics 27 (1990) 165-194
Liebeskind, Julia, Margarethe Wiersema, and Gary Hansen. "LBOs, corporate restructuring, and the incentive-intensity hypothesis." Financial Management (1992): 73-88.
Liu, Ping. "Horse Picker or Right Jockey? An Examination of Private Equity Value Creation Through the Lens of Withdrawn Leveraged Buyouts." New York Times (1988).
Long, William F., and David J. Ravenscraft. "LBOs, debt and R&D intensity." Strategic management
Lynn, Michael, and Betsy D. Gelb. "Identifying innovative national markets for technical consumer goods." International Marketing Review 13.6 (1996): 43-57.
Muscarella, Chris J., and Michael R. Vetsuypens. "Efficiency and organizational structure: A study of reverse LBOs." The Journal of Finance 45.5 (1990): 1389-1413.
Narin, Francis, Elliot Noma, and Ross Perry. "Patents as indicators of corporate technological strength." Research policy 16.2-4 (1987): 143-155.
Opler, Tim, et al. "The determinants and implications of corporate cash holdings." Journal of financial
economics 52.1 (1999): 3-46.
Palepu, Krishna G. "Consequences of leveraged buyouts." Journal of Financial Economics 27.1 (1990): 247-262.
Schumpeter, Joseph A. "Change and the Entrepreneur." Essays of JA Schumpeter (1934). Seru, Amit. "Firm boundaries matter: Evidence from conglomerates and R&D activity." Journal of
Financial Economics 111.2 (2014): 381-405.
Shand, Ric, and Shashanka Bhide. "Sources of economic growth: Regional dimensions of reforms."
Economic and Political Weekly (2000): 3747-3757.
Smith, Abbie J. "Corporate ownership structure and performance: The case of management buyouts."
Journal of financial Economics 27.1 (1990): 143-164.
Stiglitz, Joseph E., and Andrew Weiss. "Credit rationing in markets with imperfect information." The
American economic review 71.3 (1981): 393-410.
Stock, Ruth Maria, and Nicolas Andy Zacharias. "Patterns and performance outcomes of innovation orientation." Journal of the Academy of Marketing Science 39.6 (2011): 870-888.
Trajtenberg, Manuel. "A penny for your quotes: patent citations and the value of innovations." The
Rand Journal of Economics (1990): 172-187.
Tredennick, Nick, and Brion Shimamoto. "An engineer's view of venture capitalists." IEEE spectrum 38.9 (2001): 67-70.
Veugelers, Reinhilde, and Bruno Cassiman. "Make and buy in innovation strategies: evidence from Belgian manufacturing firms." Research policy 28.1 (1999): 63-80.
Williamson, Oliver Eaton. The Economic Institutions of Capitalism: Firms, markets, relational
Contracting. Free Press, 1985.
Wright, Mike. "The make‐buy decision and managing markets: the case of management buy‐outs."
