• No results found

The effect of an initial public Offering on firm innovation : a European perspective

N/A
N/A
Protected

Academic year: 2021

Share "The effect of an initial public Offering on firm innovation : a European perspective"

Copied!
59
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

1

The effect of an Initial Public Offering on firm

innovation

Master thesis Quantitative Finance

July 2018

A European perspective

Name: Boris van Minnen Student number: 11081511

(2)

2 This document is written by Boris van Minnen, 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 reference have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Table of Contents

Abstract ... 6 Acknowledgement ... 6 1. Introduction ... 7 2. Theoretical Framework ... 10

2.1 Initial Public Offerings ... 10

2.1.1 Benefits of going public ... 10

2.1.2 Costs of going public ... 11

2.2 Innovation ... 12

2.3 Managerial behaviour ... 12

2.4 Ownership ... 13

2.5 Literature review ... 14

2.6 Regulatory background: Europe vs. United States ... 16

3. Hypotheses ... 19

4. Data and summary statistics ... 22

4.1 Innovation ... 22

4.2 Patent data ... 23

4.2.1 Patent citations ... 23

4.3 IPO data ... 26

4.4 Financial data and firm characteristics ... 26

4.5 Data consolidation and merging process ... 27

4.5 Summary statistics ... 27

5. Methodology ... 29

5.1 Panel data analysis ... 29

5.1.1 Main sample regression ... 29

5.1.2 Subsample regression ... 30

5.2 Linear vs quadratic vs cubic relationship ... 31

5.3 Differences-in-differences approach ... 32

5.3.1 Propensity score matching ... 32

(4)

4

6.1 Baseline results ... 35

6.1. Impact of an Initial Public Offering on innovation quantity ... 35

6.1.2 Impact of an Initial Public Offering on innovation quality ... 36

6.2 Subsample and cross-sectional analysis ... 38

6.2.1 Innovative countries ... 38

6.2.2 Robustness – Subsample of R&D intensive firms ... 40

6.2.3 Private equity ownership ... 41

6.3 Linear vs quadratic relationship ... 43

6.4 Differences-in-differences ... 45

6.5 Robustness ... 48

6.5.1 General notes on robustness ... 48

6.5.2 Modification of baseline specification ... 49

6.5.3 Subsample of undeveloped equity markets ... 50

7. Conclusion ... 52

7.1 Main conclusion ... 52

7. 2 Limitations and suggestions for future research ... 53

References ... 55

Appendix ... 57

1. Bloomberg Innovation Index 2018 ... 57

2. Country Abbreviations ... 57

3. Variable definitions ... 58

4. Propensity score matching result ... 58

(5)

5

List of tables

Table 2.1 Literature overview ... 18

Table 4.1 Overview of patents and patent citations 24 Table 4.2 Summary statistics 28 Table 6.1 Regression results IPO on patent counts ... 36

Table 6.2 Regression results IPO on patent citations ... 37

Table 6.3 Regression results on innovation for innovative countries ... 39

Table 6.4 Regression results IPO on innovation – R&D intensive sample ... 41

Table 6.5 Regression results on innovation for private equity backed companies ... 42

Table 6.6 Regression results Linear vs Quadratic vs Cubic ... 44

Table 6.7 Firm characteristics treatment and control group ... 46

Table 6.8 Overview of Average Treatment effect results ... 47

Table 6.9 Baseline regression – Robustness ... 49

Table 6.10 Average treatment effect – development of equity markets ... 50

List of figures

Figure 4.1 Average number of citations per patent ... 25

Figure 4.2 Countries represented ... 25

Figure 4.3 IPOs in Europe 1995-2010 ... 26

Figure 5.1 Visual presentation of IPO variable ... 29

Figure 6.1 IPO sample – Mean patents around the IPO year ... 44

Figure 6.2 IPO sample – Mean patent citations around the IPO year ... 44

Figure 6.3 DID plot – Development of number of patents ... 47

(6)

6

Abstract

This thesis analyses the effect of an Initial Public Offering (IPO) on firm innovation using European firm-year data. Innovation is measured with the number of patents (innovation quantity) and the number of citations (innovation quality). The database used is PATSTAT which is supplemented by Orbis. Patents are used from the year 1990 until 2010 and consist of 3,440,240 raw patents from 12,193 unique private and public firms. The research question is addressed using a threefold approach consisting of 1) a fixed-effects panel regression, 2) a quadratic and cubic regression and 3) a differences-in-differences approach with a treatment group matched based on propensity scores. The results imply a positive relationship between going public and firm innovation in terms of quantity and quality. Firms that go public experience an increase in the number of patents and the number of citations granted in the years after an IPO. A number of matching techniques provide sufficient evidence to suggest that firms that go public have on average more patents granted and receive more citations than similar matched private firms. Furthermore, a significant additional positive effect of the impact of IPOs on firm innovation is found for firms that reside in countries that are part of the Bloomberg innovation top ten. In addition, a significant quadratic relationship is found between the amount of years related to the IPO year and the number of citations received per year. This implies an inverted U-shape where the IPO year is the peak. The results found are robust and similar when the regressions are slightly adjusted.

Keywords: Innovation, Initial Public Offerings, Patents, Europe

Acknowledgement

This thesis marks the end of my Master of Science in Quantitative Finance at the University of Amsterdam. I would like to thank my thesis supervisor Derya Güler for her outstanding supervision. She was of great support and always offered critical insights into my work. She allowed me to write my own research while guiding me in the good direction whenever needed.

(7)

7

1. Introduction

In an Initial Public Offering (IPO), as illustrated in its name, a company offers its shares for sale to the public for the first time. There is a wide variety of research available on the effects of an IPO on firm’s performance. Nonetheless, research regarding the impact of an IPO on firm innovation is limited.

This thesis contributes to the existing literature in a number of ways. First, because of a focus on Europe with a corresponding relatively untouched patent database in the field of corporate finance research. Firms in Europe have different characteristics, different equity markets and a different IPO procedure. Second, new analyses are employed to the subject. Not only analysing the differences in innovation between IPO firms and non IPO firms but also within IPO firms.

According to Audrentsch (1995), innovation plays a crucial role in a firm's long-term growth, higher likelihood of survival and in the end, gaining competitive advantage. From a macroeconomic perspective, innovation is a main driver behind a country’s economic growth. Holmstrom (1989) calls attention to the fact that the process of innovation has a high likelihood of failure as it is erratic and takes a lot of time. For these reasons, successful innovation requires precise allocation of (financial) resources combined with strong corporate governance. Because innovation is an activity high in risk and uncertainty, using equity to finance innovation projects could be more appropriate than debt. Indeed, debt holders require a certain level of certainty and predictability for the projects they intend to finance. As a result, an IPO could stimulate innovation by gaining access to equity financing that can be used to finance research and development (R&D) activities.

On the other hand, going public also comes with consequences that could have a negative impact on the innovation process. First, going public increases information asymmetries where managers know more about the true value of a company and the possible outcome of innovation projects than investors do. Alongside, going public could possibly increase agency conflicts when management is disconnected from ownership of the firm. This can lead to problems when ambitions are in conflict. An example of this issue could be, when management would like to pursue long-term R&D projects while current shareholders are more interested in short-term stock price gains. In addition, going public could change manager’s investment horizon due to the mandatory regular disclosure of financial results. Due to pressure from analysts and investors, it is possible that firms focus more on delivering short term results rather than focusing on investing in long term (uncertain) innovation projects. This leads to the research question of this thesis: “What is the effect of an Initial

Public Offering on firm innovation?”, with a focus on the European region. The current literature

about the effect of firms going public and innovation still remains unsatisfactory. Some academics claim that innovation increases after going public (He, Li, & Zhang, 2017), whilst others find that some aspects of innovation decrease (He & Tian, 2013).

(8)

8 While there is a large depiction of United States (U.S.) based analysis, reseach with a European perspective still remains very limited. This leaves this a relatively undiscovered area which misses an accurate analysis. The U.S. and Europe are both different regions with different characteristics that could for example influence the innovation process. The decision to go public could be influenced by different regulations for firms to go public (Ritter J. R., 2003). Next, there is a difference in the development of equity markets between the U.S. and Europe and consequently the difference between private and public firms.

In this thesis the impact of an IPO on firm innovation is analysed through an extensive literature study in combination with various empirical methods. In line with recent literature, patents are used as a proxy to measure the degree of innovativeness of a company. The number of patents is used to measure the innovation quantity and the number of patent citations received to measure the quality of these innovations. Patent data is extracted from the PATSTAT database supplemented by Orbis. These databases are still unexplored and consist of detailed patent information from many European patent offices. The patents are then matched with their Bureau van Dijk (BVID) identifier to provide firm specific financial variables before IPO data is finally matched from Compustat and Orbis.

To test the relationship between IPOs and firm innovation, a panel regression using multiple fixed effects is applied to the number of patents and to the number of patent citations as dependent variables, controlling for firm specific financial variables that could influence the innovation process. In addition, several interaction dummies are included amongst others private equity backing as well as the innovativeness of the home country of a firm. Secondly, the innovation development pattern of firms that decided to go public is visualized and investigated to determine whether the innovation pattern is linear, quadratic or cubic. Thirdly, firms that decide to go public can have different motives and will be at a certain stage in their life. Based on propensity scores, firms that conduct an IPO (treatment group) are matched with a non-treatment group with similar characteristics. The average treatment effect is then calculated and a differences-in-differences approach is used to visualize the development of innovation around the IPO year for both the treatment and non-treatment group.

The results imply a positive relationship between going public and innovation in terms of quantity and quality. Firms that go public experience an increase in the number of patents and the number of citations granted in the years after an IPO. Furthermore, the average treatment effect on innovation of firms that conduct an IPO compared to a similar control group that does not conduct an IPO is positive and significant. This result is in contradiction with findings of Bernstein (2015), which can be explained by the difference in datasets and research methods. Furthermore, a significant additional positive effect of the impact of IPOs on firm innovation is found for firms that reside in innovative countries. No significant differences are found between the impact on innovation of private equity backed IPOs and non-private equity backed IPOs. Finally, a significant quadratic relationship is found for the number of citations received per year and the number of years related to the IPO year.

(9)

9 This implies an inverted U-shape where the IPO year is the peak. The latter can be explained by the increase of attention on a firm around the IPO (EY, 2014).

The subsequent development of this thesis is organized as follows: section two presents a literature review regarding, but not limited to subjects relating to IPOs, corporate innovation and comparable papers. In the third section, a hypothesis derived from the literature review is formulated. The fourth section describes the dataset, variable construction and summary statistics before section five presents the empirical methodology. Section six depicts the empirical results and finally section seven discusses the conclusion, limitations and suggestions for further research.

(10)

10

2. Theoretical Framework

This section gives an overview of the previous academic literature. First, the possible advantages and disadvantages of an IPO on firm innovation are discussed. Next, a number of relevant changing dynamics of a firm are highlighted. Subsequently, a review of available literature on innovation is described. The regulatory landscape of the U.S. and Europe is sketched and the theoretical framework is presented. Lastly, a table is composed to sum up all findings.

2.1 Initial Public Offerings

2.1.1 Benefits of going public

Since the 1960s, IPOs are a widely studied subject in the academic literature. There are several reasons for a firm to go public, such as new financial funds, a way to exit for current shareholders, putting a market price on the company to facilitate mergers and acquisitions or to make shares more liquid.

After going public, firms typically increase in size and have easier access to external financial resources. A firm in post IPO stage would, therefore, have greater financial sources available to fund new innovative activities and conduct more R&D compared to firms that remain private. This is in line with other literature that suggests that the IPO proceeds are used to further finance capital investments (Kim & Weisbach, 2008). Not only access to new capital but a lower debt-to-equity ratio is also beneficial. A high debt-to-equity ratio means that a company has a relatively large portion of debt compared to equity and therefore has a higher chance of default. This influences the interest rate a bank charges to a company due to a different risk profile of the company. Financial debt is often used to finance more predicable projects and often requires a certain level of collateral. This collateral can be for example fixed assets such as a factory or building. Following the Pecking order theory, it is often preferred first to finance a project with debt over financing with equity. Issuing equity is a better fit with financing innovation due to the corresponding risk profile. Concluding, going public could lead to more innovation by giving access to equity financing and receiving better borrowing conditions from a financial institution.

Another advantage of going public is the fact that it is easier to attract, retain and reward valued employees with share option plans (EY, 2014). When a company goes public it usually also get more visible under a broader audience. Also, with being a listed company, comes an element of prestige involved, for current as well as in attracting future employees. This can be because being employed by a listed company often has a higher reputation compared to working for a private company. Overall, based on current literature, going public strengthens the human capital of a company. This could also translate into attracting and retaining talented researchers and inventors. As a result, since an IPO boosts human capital of a company it is likely to have a positive impact on firm innovation.

(11)

11 Additionally, when a firm goes public it receives much attention. The company will be featured in local and financial newspapers and will be followed by retail – and institutional investors (Da, Engelberg, & Gao, 2011). Most firms that are innovation focused and R&D intensive are more business-to-business focused, therefore being in the spotlight can help form strategic alliances or attract new customers that could foster innovation.

2.1.2 Costs of going public

There are not only advantages for undertaking an IPO, going public also comes with a cost (Ritter J. , 1987). Academic literature distinguishes direct costs and indirect costs. The most important direct costs are registration fees, legal fees, accounting fees and underwriter and advisory fees.

According to Ritter (1987), an example of indirect costs is the level of under-pricing of the company’s shares. IPO under-pricing in specific is one of the most researched topics within IPOs. Reilly & Hatfield (1969) were one of the first to research this phenomenal, based on a sample of 53 U.S. based firms that went public from 1963 to 1965. They found that the initial day return was on average nineteen percent, later many other types of research confirmed this result. Most studies on under-pricing are centred around the concept of information asymmetry. Rock (1986) proposes the “Winners Curse” hypothesis, in which he assumes that under-pricing is the effect brought my information asymmetry amongst market participants. According to his research, market participants are placed into informed investors and uninformed investors depending on their superior information1 availability. Therefore, informed investors will only participate in “good” IPOs which consequently will have an excess of demand and leaving the bad equity issues with more supply. As a result, issuers under-price IPOs to attract uninformed investors. Managers know more about the true value of a company than investors; therefore a certain level of information asymmetry exists. Hence, insiders in the company have better information available about the outcome and potential successes of R&D projects than investors.

Next, going public could possibly increase agency conflicts when management is disconnected from ownership of the firm. This can lead to problems when ambitions are in conflict. A relevant example could be that management would like to invest in R&D in order to secure long-term growth, however, on the other side, the company is obligated to regularly disclose results of their operations to the public and focusses on short-term results (Michael C. Jensen, 1976). This is confirmed by a research by Azoulay, Graff Zivin, & Manso (2011), they find that fortitude towards failure is crucial to effectively facilitate and maintain innovation. Concluding, when firms go public information asymmetry increases and there is a possibility agency conflicts increase. This could work dissimulating for innovation and thus going public could possibly decrease innovation of a firm.

(12)

12

2.2 Innovation

Previous research emphasizes the importance of innovation for long-term economic growth of companies and their competitive advantage (Hall & Jaffe, 2005; Griep, 2016; Kogan, Papanikolaou, Seru, & Stoffman, 2017). From a macroeconomic perspective, innovation is necessary for the economic growth of a country. Holmstrom (1989) recalls that the innovation process has a high likelihood of failure in that it is erratic and takes a lot of time. Innovation is not part of the daily habitual activities of a firm but it is a stand-alone activity. They find that larger firms are at a competitive disadvantage in doing highly innovative research because there is a cost involved with managing a diverse set of tasks that a smaller company does not have. As stated in the previous paragraph, firms tend to increase in size after doing an IPO. Therefore based on this research going public could harm the innovation process and therefore have a negative effect.

Internally, innovation is generally accomplished by investments in R&D activities such as salaries for researchers and materials used to conduct tests. Robinson (2008) points out that strategic alliances can be formed to foster innovation when internally the options are limited. It is often an advantage to be listed when forming these strategic alliances.

He finds that the amount of R&D expenses contracted to outside organisations grew from $6bn in 1997 to over $10bn in 2003. Especially the biotechnology companies counted on these strategic alliances for their need of capital as a preference to venture capital, IPOs or secondary offerings. In addition to the first paragraph that stated that firms have easier access to capital after doing an IPO, Robinson (2008) suggests that external equity financing is too expensive to finance R&D projects with. As a result, companies rarely use external equity or debt to finance R&D projects.

2.3 Managerial behaviour

As previously mentioned, innovation is created by investments in R&D. Changing status from private to public could change behaviour of managers, based on agency literature these investments could therefore also change. Managerial myopia means that managers are more focused on short-term results in opposite of the company’s results in the long term. Asker, Farre-Mensa, & Ljungqvist (2011) compare investment behaviour of public and private firms. They find that listed firms invest less of their capital and are less responsive to changes in investment opportunities compared to their private peers. Private firms invest approximately ten percent of total assets a year compared to four percent among public firms. Private firms are 3.5 times more responsive to changes in investment opportunities. Investment opportunities were measured by looking at the relationship between the market value and the book value of a company. Private companies do not have a (public) stock price; therefore sales growth is taken as a proxy for investment opportunities. Short-time result pressure was especially present at managers whose company’s stock prices are very sensitive to earnings news. Often in publicly traded firms, manager’s remuneration is linked to the stock price of a firm. Since share prices are based on information that is open to the public, managers can be motivated to share

(13)

13 information that could lead to an increase in share price. Normally, to defeat short-term bias of a manager, public firms take initiatives such as provisions when results are bad due to inappropriate management.

Managerial myopia also supports managers living the quiet life and taking less risk because they are being fired because of missing earnings targets. These anxieties for their career support managers to be risk-averse, which could possibly lead to less investments in R&D.

Looking at the base of the investors, private and public firms are also different. Private firms have fewer investors than their public counterparts; this often allows easy and close contact with their shareholders and often also results in a board position of the main investor. Concluding, managerial myopia has a negative influence on firm innovation. By going public firms could increase managerial myopia, which could consequently have a negative effect on innovation.

Second, Jensen (1986) argues that some managers could have a desire to scale their business as a form of proud. This phenomena is called “empire building”, managers invest disregarding their investment opportunities. The lower investment sensitivity could be explained by this theory.

2.4 Ownership

Firms with a different kind of ownership could also have a different innovation strategy. Information asymmetries and agency conflicts are different between firms before and after conducting an IPO. Before conducting an IPO, ownership is usually less disperse and often consists out of the entrepreneur, relatives and business associates. In the later stage ownership often is extended with external investors such as venture capitalists and private equity. Owners of private companies are often more long-time oriented compared to public firm. Usually, in a private environment management is connected with ownership and therefore have fewer information asymmetries and agency conflicts.

Public companies should have an advantage in innovating because they can spread the risk of failure across a large mass of investors (Aghion, Reenen, & Zingales, 2013). They find that greater institutional ownership is associated with more innovation. This positive result can be explained by the disciplinary effect of institutions on lazy managers or from the level of certainty they are able to offer to their managers concerned about their careers. They confirm the latter hypothesis because they find that probability that a CEO is fired after bad performance is reduced with more institutional ownership.

Another type of ownership is private equity ownership. Private equity funds are usually structured as a limited partnership with an established lifecycle. So-called General Partners (GPs) manage these funds and investors in the fund are named Limited Partners (LPs). Life of these funds is normally ten years but it possible to extend to fourteen years. The private equity firms will make financial investments in firms and tend to exit in two to five years (Phalippou & Gottschalg, 2009). Venture capital investors (VCs) and Private Equity investors (PEs) can exit their investments with the following five options:, secondary buyouts, initial public offerings (IPOs), trade sales, buy-backs and

(14)

14 write-offs (Wright & Robbie, 1998). There is an ongoing discussing whether PE firms add and create value for the firms they acquire. They are frequently blamed for focussing too much on short-term results instead of long-term results. Also, since they usually acquire companies with a high amount of debt that increases the risk of default, they are often more focused on short-term cost-cutting than long-term value creation. Previous research on the impact of private equity on innovation from the European Central Bank finds a slight positive impact on innovation (Popov, 2009). However, this research is conducted at the moment of the leveraged buy-out and can therefore possibly have a certain selection bias and therefore the outcome of this research is discussable.

Lerner, Sorensen & Strömberg (2011) investigate 495 private equity-backed Leveraged Buy Outs in the United States between 1986 and 2005. They apply a Poisson and a regular ordinary least squares (OLS) regression. In line with this thesis, they use patents and patent citations to measure innovation. They find differences in promoting innovation between public and private firms. They find in the long term no evidence that leveraged buyouts (LBOs) are associated with a decrease in innovation. They do find that the quality of innovation, measured in the number of citations increased with private equity ownership. Their research does not distinguish different industries, which could lead to different results since industries differ in their innovation pattern. Since their research is centred around LBOs, it only sheds light on part of the relationship in innovation between public and private firms.

2.5 Literature review

Ferreira, Manso, & Silva (2014) model the impact of public and private ownership structures on firm’s inducement to invest. They find that it is optimal to go public when a firm would like to exploit concepts/ideas and that it is ideal for firms to be private to investigate new ideas. Private firms are less transparent to investors than public firms. This is because they are often not obligated to disclose operational results on a regular basis. They claim that investors in private firms are more tolerant of failures and more motivated to invest in innovative projects. Private firms are more likely to choose projects that are complex and untested. Divestitures, M&A, and changes in management or organization structure are easier when private. On the other side, public firms tend to choose more conservative projects and projects which are understandable for the market. Prices of publicly traded stocks react fast to good news; this gives an incentive to managers to choose conventional projects and cash in early. This results in a favour for private companies when it comes to innovation.

Firms that are traded on public equity markets are often covered by financial analysts. He & Tian (2013) examine the effect of financial analyst coverage on firm innovation. To assess how analyst coverage affects innovation, they estimate a number of different OLS regressions with fixed effects on a panel dataset. They include a dummy variable for analyst coverage that equals one if a firm receives analyst coverage in that specific year, they include a vector of firm specific control variables. This approach is similar to the baseline regression in this thesis. A positive aspect of coverage by financial analysts is that it decreases information asymmetry. Financial analysts are often

(15)

15 accused of putting a high pressure on managers and stimulating managerial myopia. Their baseline results confirm this and show that firms covered by a larger number of analysts generate fewer patents and patents with fewer citations. Their research is relevant for this thesis because if firms decide to go public, they are very likely to be covered by financial analysts. This could therefore lead to a decrease in firm innovation due to pressure on managers for short-term results.

Directly comparable with the research question, Bernstein (2015) measures the impact of going public on firm innovation in the United States using both patents and number of citations. His research focuses on innovative firms (that have at least one patent) that filed for an initial registration statement with the U.S. Securities and Exchange Commission (SEC) in attempt to go public and compares firms that completed or pulled back from their filing. This approach is chosen to avoid a selection bias; all the firms from his sample intended to go public. In this thesis, the selection bias is addressed by selecting a control group based on propensity scores. In his research, he uses the NASDAQ fluctuations as an instrumental variable for IPO completion. He finds that the innovation declines for the firms that went public compared to those that remained private. In specific, this paper finds that quality of internally created innovation declines and they focus more on accumulative innovations. As a possible fundament behind his results, he finds that the drivers behind the patents, the so-called inventors, experience an outflow in number of skilled inventors and a decline in productivity of the remaining inventors. Next, he finds that firms that went public experience a large increase in acquisition activity. These acquisitions are often associated with the purchase of new technologies, patents acquired are of higher quality2 than internally created technologies. As a result, the innovation strategy of firms seems to change; they generate more incremental innovation internally as they also rely on acquiring technologies externally. Bernstein (2015) concludes innovation measured by the number of patents is a linear relationship that is downsloping. However, it is not tested whether this relationship is quadratic. Next, this research also ignores the ownership of the companies. Private equity - or institutional backed firms have different characteristics and could, therefore, have a different innovation pattern.

In line with Bernstein (2015), Johnson (2015) investigates the impact of IPOs on innovation for firms in the pharmaceutical and medical supplies industries that went public between 1979 and 2002. He uses the number of patents published and not the number of patent citations. The findings of the statistical analysis conclude that the IPO act as a brake on innovation in the years going to and immediately following the IPO. They look at the relationship between going public and innovation from a cubic and linear perspective. Their linear regression yields into a positive relationship between going public and innovation, while their more robust cubic relationship indicates a negative relationship between going public and innovation. This non-linear relationship can be explained by the fact that innovation itself is not a linear process.

(16)

16 He, Li, & Zhang (2017) from the international monetary fund approach the question from a Chinese market perspective. They investigate whether the stock market boosts firm innovation using a differences-in-differences approach to mitigate endogeneity. They use a propensity score matching algorithm to construct the treatment and non-treatment groups, similar as in this thesis. Their results show a significant positive causal effect of IPOs on firms patent output in terms of number of patents and in number of citations. Next, they find that the impact on innovation is different across firm’s financial constraints, ownership structures and corporate governance. Especially firms that have a governance structure aligned with the interest of managers see a larger increase in innovation after going public. They believe that the reason why firms increase quality and quantity of innovation after an IPO is dependent on their financial constraints. Finally, by leveraging the Chinese inventor database, they find that firms that conduct an IPO not only have a higher retention rate of inventors, they also attract more external inventors. This is possible also contributing to increasing innovation activity.

In contrast with the findings of Bernstein (2015) and (He, Li, & Zhang, 2017), Acharya & Xu (2014) compared a sample of U.S. privately held firms with public firms. They discover that public firms spend more money on R&D and have a more excellent patent portfolio than their private peers in case of external finance independence. However, public firms in internal finance dependent sectors are not better at innovating than their private peers. As a result, the impact of a public listing on innovation depends on the need for external capital. This need for external capital is different in each industry.

Previous research is therefore contradictory; some papers find a positive effect of going public and equity markets on innovation while other papers find a negative effect. Aghion (2005) investigates the relationship between competition and innovation. From his panel data sample, he finds strong evidence for an inverted U-shape in the relationship between innovation and competition. Using a nonlinear estimator, he finds industries distributed across either decreasing or increasing sections of the U-shape. It is possible to test for a U-shape with a quadratic regression estimator, a similar approach is applied in this thesis to test the innovation development around the IPO year. The differences are explained by the level of competitiveness in each sector, while some firms operate in a “neck-and-neck” sector, on the other hand, firms are operating in sectors where innovations are made by laggard firms with already low-profit margins.

2.6 Regulatory background: Europe vs. United States

Previous academic research is mostly focused on U.S. based firms using the National Bureau of Economic Research (NBER) patent database. This collection of data includes information on every patent granted by the United States Patent and Trademark Office (USPTO) from 1976 to 2010. This database has successfully been exploited by a handful of papers used in the literature review (Kogan, Papanikolaou, Seru, & Stoffman, 2017; Strömberg, 2011; Bernstein, 2015). In line with the research Bukanski & Enes (2017) who analyse the impact of mergers and acquisitions (M&A) on innovation in

(17)

17 Europe, this thesis has a focus on Europe. They base their research on a sample of 1,419 European listed firms and find a positive impact of M&A on innovation quality for the acquirer. The result of their research is even more evident for a subsample of the most innovative countries in Europe based on the Bloomberg Index 2017. M&A has a supportive role in innovation activities rather than being the main innovation driver. From their sample, firms mostly acquired smaller targets with complementary technologies which allow them to internally produce innovations of higher quality.

European and U.S. markets are in many ways different (Ritter J. R., 2003), the process of going public is different with different corresponding regulations. Next, registering innovations in the form of patents and collecting citations is different in Europe and the U.S. Po-Hsuan Hsu (2014) conducts research on a sample of 32 developed and emerging countries and applies a fixed effects identification strategy. He discovers that the level of development of the credit- and equity markets affect technological innovation. More developed equity markets compared to credit markets translates to relatively more R&D investments. Their research suggests that firms based in countries with more developed credit markets have a stronger base to internally generate innovation through R&D investments. As a result, results are likely to differ between countries in the sample of this thesis and between Europe and the U.S. Po-Hsuan Hsu (2014) concludes that equity markets in Europe (excluding the United Kingdom and Sweden) are not as developed as in the U.S. The amount of IPOs also differs in the U.S. and Europe. The level of development of equity markets is strongly correlated with the level of analyst coverage (He & Tian, 2013). Firms in the U.S are therefore expected to be more in the spotlight compared to their European counterparts. When firms are more in the spotlight, the pressure on managers for short-term results is likely to increase.

Another difference between Europe and the U.S. concerning public firms is the frequency of mandatory disclosure of operating financial results. In Europe, it is mandatory to disclose results every six months while in the U.S. it is every quarter. This could impact the level of present managerial myopia thus the difference between public and private firms can be explained.

Concluding this literature review, since the existing literature in the U.S. and Europe remains unsatisfactory, the impact of an IPO on firm’s innovation remains an empirical question. Different researches with various methods applied to testing the subject have different results. This thesis is contributing to the current literature because 1) this research explores a different region with different characteristics and a different equity market and IPO process, 2) the corresponding PATSTAT database is still unexplored and consists of detailed patent information from many European patent offices, 3) methods explore not only the differences between firms that went public with those that remained private but also investigates the changes within firms that decided to go public and test whether the relationship is linear or quadratic.

(18)

18

Table 2.1 Literature overview

This table presents an overview of related literature about the impact of an IPO on innovation or the impact of private/public ownership on innovation. In the third column the measurement variable is presented which acts as a proxy for innovation. The measures vary from innovation outputs: patent counts and/or patent citations to innovation input: R&D investments. Further details on these measurement variables can be found in chapter 4. The fourth column presents which status of a firm is in favour when it comes to innovation, either private or public. The fifth column states the focus region of the region and in the sixth column, a brief conclusion relevant for our research can be found. Authors Year Measurement variable Public/Private Region Conclusion

He, Li, & Zhang 2017 Patents & Citations Public China

Based on Chinese firms, results indicate a significant positive causal effect of IPOs on firms patent output in terms of number of patents and in number of citations.

Bernstein 2015 Patents & Citations Private U.S.

Innovation declines for the firms that went public compared to those that stayed private

Johnson 2015 Patents Private U.S.

Linear regression yields into a positive relationship between going public and innovation, while their more robust cubic relationship indicates a negative relationship

Ferreira, Manso, & Silva 2014 R&D Private U.S.

It is optimal to go public when a firm would like to exploit

concepts/ideas and that it is ideal for firms to be private to investigate new ideas

Acharya & Xu 2014 Patents & Citations

Depending on external finance

dependence U.S.

Public firms spend more money on research and development and have a more excellent patent portfolio than their private peers in case of external finance independence. However public firms in internal finance dependent sectors are not better at innovating than their private peers.

Po-Hsuan Hsu 2014 Patents & Citations

Depending on the level of development

of equity markets Multiple

The level of development of the credit- and equity markets affects technological innovation. More developed equity markets compared to credit markets translates to relatively more R&D investments.

He & Tian 2013 Patents & Citations Private U.S.

Because if firms decide to go public, they are very likely to be covered by financial analysts. This could therefore lead to a decrease in firm innovation

Lerner, Sorensen &

Strömberg 2011 Patents & Citations No evidence found U.S.

They find no evidence that LBOs are associated with a decrease in innovation. They do find that the quality of innovation, measured in the number of citations increased with private equity ownership.

(19)

19

3. Hypotheses

The literature gave an overview of the latest research on IPOs and innovation. The results presented are contradictory, although most research based on patents show a negative relationship between going public and firm innovation. In this section the hypotheses are presented. In this research, a distinction is made between innovation quantity and quality. Innovation quantity is measured by the number of patents of a firm. Innovation quality is measured by the number of citations of a patent. It is important to measure these differences because not all innovations are of similar value, some are minor while others can be groundbreaking (Hall & Jaffe, 2005).

The first hypothesis addresses the innovation quantity, measured by the number of patents granted. For this hypothesis, two alternative hypotheses are constructed because of the contradictory results of existing literature.

H1 Null: Conducting an IPO has no impact on the number of patents granted

Based on an increase in information asymmetries and possible agency problems between management and ownership that could occur when going public, it is expected that going public has a negative impact on innovation. Next, going public forces a company to publish financial results on a regular basis, usually every three or six months. This could lead to short-term result pressure for managers by investing in short-term projects instead of long-term projects. In line with the results of U.S focused research of Bernstein (2015) the following first alternative hypothesis is set up.

H1.1A: Conducting an IPO has a negative and significant impact on the number of patents granted

Europe and the United States show differences as well as similarities, therefore it is possible that results might differ. Next, as presented in the literature review, going public also comes with advantages such as attracting new funds and creating a strong brand name around the company. New funds can be used to invest in R&D projects that eventually yield in innovative output in the form of patents and patent citations. Going public comes with a certain status that can attract talented employees that are able to coordinate and produce innovations. He, Li, & Zhang (2017) finds that innovation increases after going public. As a result, the following the following second alternative hypothesis is constructed.

H1.2A: Conducting an IPO has a positive and significant impact on the number of patents granted

The second hypothesis addresses the quality of innovations, measured by the number of citations. Bernstein (2015) found in his U.S. based research that the number of patent citations increased significantly after going public. Also based on Chinese data, He, Li, & Zhang (2017) confirm this result, therefore the following null and alternative hypothesis are set-up.

H2 Null: Conducting an IPO has no impact on the number of citations received

(20)

20 Inspired by Bukanski & Enes (2017) who research the impact of M&A on innovation in Europe, a hypothesis based on the Bloomberg Innovation Index (2018) is set up. The Innovation Index ranking is based on a number of variables, amongst others: R&D intensity, Manufacturing value-added, Productivity, High-tech density, Tertiary efficiency, Research concentration and Patent activity (Bloomberg, 2018). The most innovative European countries in the top ten are Sweden, Germany, Switzerland, Finland, Denmark and France. The (business) environment of a company and laws and support a company gets to support it innovativeness is therefore likely to have a positive impact on the number of patents and number of citations of a company. For example, some countries have tax advantages for companies that pursue innovative projects. As a result, the following null and alternative hypotheses are tested.

H3 Null: Conducting an IPO does not have a positive and significant impact on innovation in

innovative countries

H3A: Conducting an IPO has a positive and significant impact on innovation in innovative countries

Different companies from different countries and industries all have their own innovation strategies. While some companies are heavily focused on innovation and spend a high percentage of their sales on Research & Development, others pursue innovation in a different way. Previous literature concludes that R&D expenses are the main driver behind innovation, it is therefore self-evident that companies with a relatively high R&D expenses have a more active innovation ecosystem. Hypotheses four is constructed to test if the baseline hypothesis holds even though the dataset is changed.

H4 Null: Conducting an IPO does not have a positive and significant impact on innovation output for

R&D intensive firms

H4A: Conducting an IPO has a positive and significant impact on innovation output for R&D

intensive firms

The fifth hypothesis addresses the (additional) impact of private equity backing of an IPO on innovation. Previous research on the impact of private equity ownership in an IPO setting is mostly limited to the under-pricing topic and not related to firm innovation. In the past, there has been a number of research conducted on innovation in an LBO setting. For example, Lerner, Sorensen & Strömberg (2011) find that the quality of innovation, measured in the number of citations increased with private equity ownership. Private equity firms are frequently blamed for focussing too much on short-term results instead of long-term results. Previous research on the impact of private equity on innovation from the European Central Bank finds a slight positive impact on innovation (Popov, 2009).

H5 Null: There is no significant difference between the impact of private equity backed and

non-private equity backed IPOs on firm innovation

H5A: There is a significant difference between the impact of private equity backed and non-private

(21)

21 Hypothesis six addresses the development of innovation for firms that decided to go public. Aghion (2005) investigates the relationship between competition and innovation and find a quadratic relationship between competition and innovation. Looking at the original dataset extracted from Orbis, firms do not publish patents every year. This also comes forward in the high standard deviation of the firm-year observations in the number of patents, number of patent citations and the highest cited patent. Some firm characteristics could very much stimulate innovation after an IPO, while some firm characteristics can discourage the innovation process. Next to firm characteristics, there could also be external factors influencing innovation and can sometimes stimulate a lot and sometimes work against innovation. As a result, the following hypotheses are tested on the number of patents and on the number of patent citations. Johnson (2005) finds in his analysis on the innovation pattern around the IPO year that a cubic relationship fits the data better than a linear estimator. As a result, it is interesting to see how innovation develops around the IPO.

H6 Null: There does not exist a significant quadratic relationship between an IPO and firm

innovation

(22)

22

4. Data and summary statistics

4.1 Innovation

Measuring innovation is challenging and imperfect. In previous academic literature, innovation is measured based on input and/or output variables. The input variable is investments in R&D and the output variable is often the count of patents and the number of patent citations. Patent-based metrics are considered to be the most meaningful proxy for innovation. Patent data is objective data and is available for many years. In opposite to R&D expenditures, this is an accounting variable that is easy to manipulate. In addition, in many European countries it is not mandatory to disclosure and to justify R&D expenditures. This could, therefore, lead to incorrect results in analysing the impact of IPOs on innovation. Also, many private firms are not obligated to disclose operational results on a regular basis.

Following recent literature on innovation, this thesis employs patents as a measure of innovation. Arundel & Kabla (1998) calculate sales-weighted propensity scores for nineteen different industries in Europe to estimate what percentage of innovations a patent application is made. The average propensity score for product innovations is 35.9 percent and 24.8 percent for process innovations. The four sectors that surpassed the 50 percent score are pharmaceuticals, chemicals, machinery and precision instruments. For the latter, this means that more than half of the innovations are patented. They find that larger firms on average apply for more patents for their inventions compared to smaller firms. Also, some industries have higher propensity scores due to the competitiveness of the industry. Concluding, previous literature points out that patents are not a flawless proxy for innovation, patents do not capture the full innovation. For example, some innovations are not patented but protected by trade secrets. However, using patents as proxy for innovation is broadly accepted by previous literature and therefor the most suitable choice to address the research question in this thesis.

It is also important to highlight the differences between patents and patent citations and their characteristics. The baseline measure is the count of patents, which is the number of patents granted to a company. Nonetheless, the number of patents does not differentiate between additional features on patents (Bernstein, 2015) or with groundbreaking innovations (Griliches, 1990). It does give information about how active a firm is in patenting and how this pattern possibly varies over time or around a specific moment, such as an IPO.

By including patent citations in this thesis it is possible to capture the importance or quality of each patent. Patent citations itself are also not completely waterproof, not all citations are registered and not all citations are of equal value. Kogan, Papanikolaou, Seru, & Stoffman (2017) present in their research that equity markets positively react to approval of patents which in the course of time receive many citations. Griep (2016) analysed firms across 27 countries from 1973 to 2013. He finds that in ten countries stock price reactions are positive and significantly correlated with the traditional

(23)

23 measure of quality, the number of patent citations. Concluding, based on previous research the number of patents granted and the number of citations is a reliable proxy for firm innovation.

4.2 Patent data

Most previous research on innovation is based on the NBER patent database and as a result focused on the U.S. This thesis exploits a different, relatively untouched, different patent database focused on Europe. This patent dataset is similar to Bukanski & Enes (2017) who investigated the impact of M&A on innovation in Europe. The data is collected from the most prominent patent database in Europe, the European Patent Office (EPO). This data is available from the PATSTAT database supplemented by Orbis. The PATSTAT database consists of more than 74 million individual records and contains information about amongst others patent applications, e.g., inventors and owners, technology fields, titles and abstracts, publication instances, and citations. Orbis has matched each unique patent with its owner and corresponding Bureau van Dijk identification number (BVID). Since the size of this database, the full patent database is only accessible and editable with SQL.

This thesis uses patent data between 1990 and 2010; this period is chosen to maximize the quality of the dataset. Financial information is not available in the Bureau van Dijk database before the years 1990 and only limited available before 1995. The end of the range, 2010 has been chosen to capture the full range of citations. After a patent is published it can take several years to be recognized and to capture citations. The sample in this thesis consists of 12,193 different firms and a total of 3,440,240 patents by private and public companies published in Europe. This thesis uses the application year. A possibility exists that application year differs from the grant year, application year captures the level of innovation better (Hall & Jaffe, 2005), only patents that are eventually granted are included in the final dataset.

4.2.1 Patent citations

A recent paper from Lerner & Seru (2017) gives useful insights into the use of patent data in corporate finance research. They point out several methods to use to measure patent value and bring light to pitfalls of this data that can lead to incorrect results. The dataset from Orbis contains variable, patent citations. This is the sum of all the citations following the application date of the patent, this variable is updated every year. The result is that recently published patents have a shorter period to be cited than older patents. This issue is illustrated in figure 4.1, as a result, the number of average citations per patent decreases over time. According to Lerner & Seru (2017), this issue can be solved in several ways to make this variable useful for research, for this thesis two options were considered. The first way is to estimate the distribution of the number of citations and project forward citations based on the historical distribution. The second way to solve this issue is to collect citations for a limited number of years after a patent is granted. In line with the research of Strömberg (2011) and Bukanski & Enes (2017) this thesis uses a corrected version of the number of citations. These are the citations captured in the five years following the publish year of the patent. According to their

(24)

24 research, the peak in patent citations is reached after five years after publishing. An example of this is a patent published in the year 2000 by the Dutch company Philips. This patent has 50 citations received from 2000 until today (2018). Since we are interested in the citations in the first five year after granting, the PATSTAT file of 2005 is consulted and 40 citations are found, thus 40 citations are registered in the new dataset. Looking at the total dataset, the method from the example is applied on all the patents from our first dataset. The result is that the total number of original citations 953,679 is corrected to 328,529 citations. Figure 4.1 illustrates how the number of corrected citations compares to the original number of citations. This issue is solved because over the full estimation period the patent citations remain stable. The downside of this correction is that two third of the citations is not taken into account in this dataset. Some innovations might be patented early on and exploited later and big breakthrough innovations made early on, might not get the full credit for how important they are as they are.

Table 4.1 Overview of patents and patent citations

Table 4.1 lists the number of patents and the total number of (corrected) citations per year. The number of citations is the sum of all the citations of patents published in the corresponding year. The original number of patent citations is extracted from the Orbis database based on PATSTAT and are all the citations received until today. The third column presents the number of corrected citations is the number of citations five years after publication date following the approach of Strömberg (2011) and Bukanski & Enes (2017).

Number of patents Number of citations Number of citations (corrected) 1990 139,518 41,579 8,057 1991 136,784 41,639 8,498 1992 138,985 47,208 10,204 1993 146,015 54,131 13,799 1994 153,018 59,345 16,155 1995 163,887 61,768 16,996 1996 173,584 66,682 18,906 1997 184,812 67,507 19,541 1998 189,503 64,203 17,958 1999 202,353 61,677 18,130 2000 214,484 57,198 16,557 2001 205,488 53,286 16,890 2002 191,908 46,380 16,245 2003 185,803 45,499 18,576 2004 178,727 41,803 18,784 2005 167,017 37,261 18,915 2006 158,230 32,983 19,379 2007 146,000 26,112 17,521 2008 139,092 20,651 15,524 2009 120,468 15,818 12,968 2010 104,564 10,949 8,926 Total 3,440,240 953,679 328,529

(25)

25

Figure 4.1 Average number of citations per patent

Figure 4.1 displays the average number of citations per patent per year from 1990 until 2010. The original number of patent citations is extracted from the Orbis database based on PATSTAT and these are all the citations received until today. More recent patents have fewer citations than patents which are published earlier. The second line represents the average number of corrected citations. The number of corrected citations is the number of citations five years after publication date, following the approach of Strömberg (2011) and Bukanski & Enes (2017).

Figure 4.2 Countries represented

Figure 4.2 displays the dispersion of patents amongst the different countries in the dataset. The patents are not equally divided over the different countries in Europe; Germany is the largest patent holder of our sample with 25 percent of all patents. A description of the abbreviations used can be found in appendix two. 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 A ve rage n u m b er of ci tat ion s p er p at en t Year Original Corrected DE, 25% IT, 12% GB, 11% FR, 9% ES, 8% NL, 6% SE, 3% CH, 3% Other, 24%

(26)

26

4.3 IPO data

IPO data collected is collected from the Bureau van Dijk Amadeus database and supplemented by Orbis. The main variable of interest is IPO year, other variables of interest are ownership and the percentage of equity sold. Private equity ownership is not always available in the databases, as a result, hand collected or checked from the IPO prospectuses. The final sample consists of companies that published at least one patent between the years 1990 and 2010. Figure 4.3 presents all the IPOs in Europe from 1995 until 2010. Remarkable is the high peak in the years 1999 and 2000, this peak can be explained by the tech bubble.

Figure 4.3 IPOs in Europe 1995-2010

Figure 4.3 presents the number of IPOs per year from 1995 until 2010 in the European Union (EY, 2012). The focus of this thesis is on companies that have at least one patent, as a result, not all IPOs in Europe are used in the analysis.

4.4 Financial data and firm characteristics

The patent dataset is supplemented by firm-specific (financial) data. This financial data is useful to get insight into the background of the companies that pursue innovation as well that they act as control variables. Firm data is collected from the Bureau van Dijk Amadeus database and contains firm-year observations. The sample consists of public and private companies, considered to be “Very Large” firms by Amadeus. The Firms are identified by their industry group. According to innovation literature, R&D expenses are the main input measure for innovation. However, in many European countries it is not mandatory to disclosure and to justify R&D expenditures. In the sample from Bureau van Dijk, this is confirmed since there are many missing R&D expenses in the dataset. Other variables collected from the database are: Sales, Total Assets, Fixed Assets, Total Equity, Total Debt, EBITDA, EBIT, Net Income and Industry. Financial ratios are not extracted from the database but are self-constructed using the untouched raw variables from the financial statements.

0 100 200 300 400 500 600 700 N u m b er of I P O s Year

(27)

27

4.5 Data consolidation and merging process

In the first phase, patents are collected from the Orbis database. Due to a data limit of Orbis, it was only possible to download 80.000 patents per time. It was mandatory to do this by hand and after downloading the patents were merged together using a loop STATA and checked for duplicates. Following up, corrected citations have been created using SQL in PATSTAT. A code is used to automate the process to collect the citations five years after the publishing year of each patent. This process was very time consuming because for each patent the loop has to consult the specific file to count exactly five years of citations after the publication.

In the second phase, the total list of patents and (corrected) patent citations is transformed into firm-year observations. As a result, each firm has the total number of patents granted and received (corrected) citations per year sorted in a dataset.

In the third phase, firm-specific financial data and if applicable IPO data is merged in the main file. Financial data is merged using Bureau van Dijk ID numbers and the IPO data using International Securities Identification Numbers (ISIN). The IPO data is combined from Compustat and from Orbis and had to be supplemented with handpicked data from the IPO prospectuses.

4.5 Summary statistics

The final sample consists of 27,140 firm-year observation which consists of 12,193 unique European firms. The firms in the sample are private and public. The number of patents and the number of citations has been transformed to the logarithm + one patent due to the right-skewed distribution and missing firm-year observations are automatically transformed to zero. Only ten percent of all patents receive a citation and the total number of citations is 28 percent of the total number of patents. The average of the highest cited patent is 4.46 times and the highest cited patent is cited 638 times originally and 442 corrected for five years after the application year. The standard deviation of the number of patents is relatively large (47.84), this implicates the number of patents per firm-year observation is varying substantially. This can be explained by the innovation intensity of some industries and that the number of patents contains all patents in the sample. The latter is important because there is no distinguishment between small add-on innovations and breakthrough innovations. The company size is measured with the logarithm of total assets, the sample mostly consists of very large firms with an average asset size of 1.85bn euro. In order to indicate the intensity of R&D expenses spend, a variable R&D-to-total Assets is constructed. Since R&D disclosure is not obligated in all countries of Europe and the fact that this accounting measure is vulnerable to manipulation, one has to be careful with interpretation. The variable Cash-to-Total Assets is constructed to measure by dividing the amount of cash & cash equivalents on the balance sheet by the total assets. The availability of cash could indicate ability and flexibility to invest in innovative projects according to innovation literature. Next, the variable Fixed-to-total assets gives insight in the relationship of fixed assets to total assets. Fixed assets include buildings, plant, machinery and equipment and gives an

(28)

28 indication of the company’s asset structure and flexibility. On average, 37 percent of total assets are fixed assets. Firms are varying from very asset-intensive businesses to asset-light businesses. Asset-intensive business could have up to 96 percent of their total assets being fixed, an example is a production company with a lot of factories. Firms with only little assets could be for example technology firms that mostly have their capital in intangibles and virtually have zero fixed assets. All dependent and independent variables are winsorized at the 1st and 99th percentile to correct for outliers.

Relevant characters for this research about IPOs are amongst other the age of a firm when they went public if a firm is part of the treatment group and if the firm has a private equity investor. Firm age is calculated by subtracting the year of incorporation from the year a company goes public. On average firms go public thirty years after they are incorporated, however, the median is fourteen years. This is comparable to the firms in the U.S. based sample of Bernstein (2015) which has an average age of twelve years before filing to go public.

Table 4.2 Summary statistics

This table presents the summary statistics of the European firms in the sample. These are firm year observations from 1990-2010.

Innovation

N

Mean

Median

SD

Min

Max

Patents 27,140 8.20 2 47.84 1 1920

Citations 27,140 1.36 0 12.17 0 1128

Citations (corrected) 27,140 0.98 0 7.91 0 638

Citations best corrected patent 27,140 4.47 1 12.59 0 442

Financial information

Log (Total Assets) 26,564 18.02 17.72 1.98 14.024 24.27

R&D/Total Assets 1,784 0.07 0.03 0.13 0.000 1.41

EBITDA/Total Assets 23,948 0.11 0.10 0.13 -0.45 0.52

Cash/Total Assets 26,001 0.10 0.05 0.13 0.000 0.67

Fixed-to-Total Assets ratio 26,554 0.37 0.34 0.23 0.007 0.96

Employees 22,889 1,313.84 225 4,642.06 3 38,530

IPO characteristics

Firm age 2,104 30.18 14 37.77 0 157

IPO treatment in data time 27,140 0.05 0 0 0 1

Private equity backed 22,608 0.02 0 0 0 1

It is relevant to compare the summary statistics of this thesis with datasets from similar studies on innovation. A direct comparison can be made with Bukanski & Enes (2017) who look at the relationship between M&A and innovation in Europe. The firms in their sample have on average more patents granted per year and a higher amount of citations. This can be explained by the fact that their research is only applied to public firms and that this research covers private and public firms.

Comparing the summary statistics with the research of Bernstein (2015) and other American studies using patent data, there are differences in the summary statistics. On average the firms in the sample of this thesis have fewer patents granted each year and earn fewer citations. This can be explained by regulatory differences between the granting of citations in Europe and the U.S.

Referenties

GERELATEERDE DOCUMENTEN

The fact that the results of table VIII don’t show a relation between risk and the value premium effect on a firm level, is more in line with Lakonishok, Shleifer

The independent variables are measured by; Management ownership (MANOWN), Board ownership (BOARDOWN) inside ownership (INSIDEOWN ) as the absolute difference in percentage

Even though the importance of the resistance barrier was significantly different for small, medium and large firms it should be noted that the mean score indicates a low importance

Approach/Design/Methodology/Purpose – Building on signalling and legitimacy theory and drawing from literature on corporate transparency and information disclosure, this study uses

Multiple determinants for corrupt firm behaviour are considered, including: country-level corruption, length of applications, gender of the owner, gender of the

following characteristics significantly influence R&D expenditures: age, tenure, type of education, stock ownership, nationality and gender. Below we will discuss

Model 1 includes all control variables, model 2 shows the regression results including the independent variable EMISSION REDUCTION, model 3 involves the moderating variable

Early childhood education rediscovered, New York, Holt, Rinehart & Winston.. The research process in