• No results found

When suddenly the environment changes…

N/A
N/A
Protected

Academic year: 2021

Share "When suddenly the environment changes…"

Copied!
52
0
0

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

Hele tekst

(1)

When suddenly the environment changes…

The influence of an economic and financial crisis on the relationship between the outcomes of make-or-buy decisions and financial firm performance

MASTER THESIS Written by Marc Timmerman S1628445 Bergstraat 40a 9717 LT Groningen s1628445@student.rug.nl University of Groningen

Faculty of Economics and Business

Master of Science Business Administration – Strategy & Innovation

1st supervisor Dr. K.J. McCarthy

2nd supervisor Prof. Dr. W.A. Dolfsma

(2)

2

ABSTRACT

This study investigates the influence of an economic and financial crisis on the relationship between innovation output and financial firm performance. Building upon a make-or-buy theory by Graebner et al. (2010) this study proposes positive influences of patents on financial firm performances. An event study consisting of 1771 unique and completed acquisitions in normal and crisis times has been conducted; regression analyses have been used to assess the influence of patents on firm performance. The results indicate that making technological assets and applying for patents has a significant negative influence on shareholder value, but a significant positive influence on market value. An economic and financial crisis strengthens these

relationships even more. This implies that the decision to buy technological assets should be the preferable option for pharmaceutical companies in normal and in crisis situations. However, this decision always depends on the strategy followed by the company, on environmental factors influencing the decision making process and on stakeholder pressure.

(3)

3

PREFACE

The pharmaceutical industry is a fascinating industry that I personally find very interesting. Often the size of the firm affects the innovative performance and the larger the firm grows the harder it becomes to stay innovative. The pharmaceutical industry maybe is the exception to that rule. The industry is dominated by very large players that need to be innovative to get ahead of the competition and need to stay innovative to stay in front of the competition. The subject of this thesis has given me the opportunity to combine my research interest of strategy & innovation with the passion of crunching numbers. Although the process of writing this thesis has not always been simple I am very happy with the final result. I would like to thank my first supervisor, dr. Killian J. McCarthy, for being supportive and critical on my work. It really helped me to come up with a thesis that I can be proud of. I also want to thank dr. Wilfred A. Dolfsma for putting in the effort to grade my thesis.

August, 2012

(4)

4

1. INTRODUCTION

Innovation is in the current era of the knowledge economy often seen as one of the most important means on which firms compete and grow (Mason, Bishop, & Robinson, 2009). In order to win the ongoing battle for the market, firms need to come up with new and innovative ideas. Those ideas will cost precious time and money and those ideas may require specific assets that are not owned by the company. To obtain those assets from targets is the most commonly acknowledged reason for technology acquisitions (Birkinshaw, Bresman & Håkanson, 2000). In essence companies will have to decide whether to make-or-buy. This theory has already been written about in 1915 by Ford & Porter who emphasized the strategic importance of this decision. Making new technologies will create products that are firm specific, but companies need specific assets for achieving that. In the case the company does not have the desired assets a buy decision will come into the picture. Among others; Graebner, Eisenhardt & Roundy (2010) argue that next to the acquiring of desired assets, acquisitions can also enhance market power and initiate strategic renewal. But what happens when the external environment changes in such a way that even big and financially strong companies experience difficulties completing the process of making new technology? Hardly any studies have investigated the influence of a financial crisis on the combination of innovation output and firm performance. This study aims to fill up the gap that has been left open in literature. It will be shown in the literature overview that the decision whether to make or buy technology is a decision of great importance. Internal research is often based on the existing knowledge of an organization and is less likely to produce the sometimes vital breakthrough technologies (Balasubramanian & Lee, 2008; Sørensen & Stuart, 2000).Maybe therefore large firms may choose acquisitions over internal development of technological resources. After all, smaller and younger firms are often a lot more innovative (Graebner et al., 2010).

According to studies by Hall (1992) and Simon & Sullivan (1993) intangible resources such as patents may provide superior returns for shareholders. Another study by Hall, Jaffe &

Trajtenberg (2005) found that patents also have a positive influence on the market value of the firms. But which consequences does an economic and financial crisis have? Does it force

(5)

5 4 that patents have a negative effect on the financial performance of pharmaceutical firms, except for market value. Another result that will be shown is that a financial crisis does not create different relationships between innovation outputs and firm performance. However, the magnitude of the relationships is being strengthened during financial crises. Positive

relationships become even more positive, negative relationships become even more negative. The transaction value of acquisitions drops during a crisis, but patents become more valuable during crises. This study argues that the focus of firms during a crisis situation lies more on protecting their innovation rather than own developing new technology. The results indicate that firms should in most cases choose to buy new technology instead of making it themselves. Developing new technologies and patenting them has a negative relationship with shareholder value, but a positive relationship with market value. This implies that pharmaceutical firms should choose buying technologies over making them themselves to keep shareholders happy and to benefit as a company by creating more market value for the company. More explanations for the results will be given in section 5 on the basis of which conclusions will follow in section 6.

This study contributes to literature by combining innovative output in the form of patents, financial firm performance and the event of economic and financial crises, and by showing the effects of these crises on the financial performance of pharmaceutical firms. Hereby a gap that has been left open by previous studies will be filled up. Some theories will be supported; some will be rejected. Essentially this study will show the effects of a make-or-buy decision made by firms on financial performance during economic and financial crises.

2. LITERATURE REVIEW

2.1 Make-or-buy

(6)

6 (Birkinshaw et al., 2000). It is practically a make-or-buy decision companies have to make. They have to decide whether or not to develop the technology in house, which can be very expensive and time consuming or they can choose to acquire a company that has developed something very similar to what is desired. This make-or-buy decision has strategic influence. It is a topic that has been research quiet often and for a long time. For instance, Ford & Porter (1915) already emphasized the important strategic factors regarding make-or-buy decisions. Furthermore, several other authors (e.g., Cullition, 1942; Ford & Farmer, 1986) have also discussed the strategic implications of a make-or-buy decision. Ford & Farmer (1986) as well as Jauch & Wilson (1979) claim that a make-or-buy decision can shape an organization internally and those decisions will define the relations that organization has with its external environment. Therefore the problem of making or buying can be seen strategically and in the context of environmental threats and opportunities (Ford & Farmer, 1986; Jauch & Wilson, 1979). Graebner et al. (2010) argue that large, established firms may choose acquisition over internal development as a means to build technology resources because smaller, younger firms are often more innovative and those firms are an increasingly important engine of new technological knowledge. Although firms may decide to develop that knowledge internally or acquire a particular technology externally, Cohen & Levinthal (1990) point to a simultaneous external exploration of internal and external knowledge. Cassiman & Veugelers (2006) call it processes that complement each other, as it is nearly impossible to generate all relevant knowledge internally. Of course, this knowledge can be acquired through alliances, but a full acquisition provides access that is exclusive and permanent, thereby providing the acquirer with unique resources which are more competitively valuable (Kale & Puranam, 2004). A make-or-buy decision does not have to be a voluntary. It can be a necessary decision a company has to make. Some research has been done on the drivers of make-or-buy decisions. The decisions depend on several different factors of course, which also differ per unique case, but determining factors for make-or-buy decisions can be: economies of scale and scope, transaction costs, agency costs, management costs, strategic positioning, technology, asset specificity and politics and regulations (Argyres, 1996; Gullition, 1942; Gambino, 1980; Park, Reddy & Sarker, 2000; Williamson, 1975, 1981). Another factor that has been added is capabilities. Jacobides (2008) argues that capabilities should be the main driver of make-or-buy decisions in manufacturing sectors. If an organization possesses the capabilities needed than making can be a viable option, if not than buying is a possibility that is worth considering.

(7)

7 Consequence is that costs will have to be made to adjust, which raises transaction costs. In those cases making is preferred over buying. Furthermore, developing technology in-house allows an objective valuation of real innovation needs and the organization as a whole will develop a unique source of knowledge (West, 2002). But these are not the only advantages; by developing technologies in-house economies of scale can be achieved, organizations can dodge transaction costs and barriers to imitation, such as patents, can be constructed (Contractor & Lorange, 1988). Nevertheless, choosing for a make strategy involves taking risks. The outcome results of the make processes are less predictable, which makes it hard to develop a strategy that is solely based on making technology because for some parts results have to be guessed.

Commercialization will be time consuming when choosing making over buying and most importantly organizations can become stuck in making only one specific technology while the environment can be demanding something totally different (Perrons & Platts, 2004). The strategy to buy instead of making the technology is supposed to have advantages. Kessler & Bierly (2002) claim that buying is more reliable than making and the results from buying are more predictable. Furthermore, buying reduces risk and capacity problems can be dealt with beforehand (West, 2002). However, buying technology implies dealing with other parties which leads to costs concerning negotiating and enforcing contracts (Narula, 2001). But a buying strategy can also be used to complement the internal developed knowledge. External knowledge can be scanned and integrated (Arora & Gambardella, 1990), which is very important for firms in order to survive.

(8)

8 One of the goals of this study is to investigate which decision, making or buying, has a better influence on financial firm performance and if these two decisions create value. Some research has been done on the making strategy, for instance by MacPherson & Boasson (2004) who find that a superior innovation performance, which can be achieved through following a making strategy, is associated with better financial results for the firm. Ceccagnoli (2009) links this outcome with appropriability regimes. He argues that the strong the appropriability regime is at a firm level, achieved through patent protection leads to superior economic performance. He also states that companies that focus more on developing new products will eventually come up with more and more drastic innovations, which is beneficial for the company as a whole

(Ceccagnoli, 2009). This view is supported by Nicholas (2008) who finds that patents have a positive relationship with market value. However, a study by Mahlich (2010) shows that an appropriability regime, measured in patents, does not drive economic performance. So the findings on this topic are a bit mixed. Buying technology will for this study mean acquiring a target company that holds the desired assets. One of the most important things for acquirers is to select a target and to form a view as to how they intend to create value from the purchase (Schweiger, Mitchell, Scott & Brown, 2007). Research on value creation from acquisitions show a mixed picture. For example, several researchers argue that if there is a similarity (e.g., Datta, Pinches, & Narayanan, 1992; Ramaswamy, 1997; Shelton, 1988; Singh & Montgomery, 1987) between acquirer and target or if these companies complement each other (Harrison, Hitt, Hoskisson & Ireland, 1991; Hitt, Harrison, Ireland & Best, 1998; Larson & Finkelstein, 1999; Teece, 1986) value can be created. Gort (1969) explains that the market value will rise due to the acquisition because the market values of both companies are combined. Acquisitions can also help firm performance by saving resources in terms of labour and capital (Calabrese & Erbetaa, 2005). But acquisitions should not be used as a single tool to enhance firm performance (Doig, Ritter, Speckhals & Woolson, 2001). However, other studies show that acquisitions maybe are not that profitable. For instance, several authors argue that 60-80% of all acquisitions fail to create value (Dyer, Kale & Singh, 2004; Marks & Mirvis, 2001). Moreover, less than one quarter of acquisitions achieves their financial objectives (Marks & Mirvis, 2001). An argument that was a bit weakened by Schoenberg (2006) who argues that half of all acquisitions can be seen as a success based on set financial objectives. Jensen (1988) finds that the returns from acquisitions vary closely around zero. Finally, Tuch & Sullivan (2007) find that acquisitions are even value destroying in half of the cases.

(9)

9

H1a: A better appropriability regime (patents) will lead to a better financial performance, while acquisitions create value

H1b: A better appropriability regime (patents) will lead to a worse financial performance, while acquisitions do not create value

2.2 Crisis

While much research has been done on make-or-buy decisions, and on the influence of acquisitions on firm performance, much less research has been done on the influence of a combination of financial crises and the make-or-buy problem on firm performance. Even after some authors have claimed that the make-or-buy decision has to be seen in the context of environmental threats and opportunities (Ford & Farmer, 1986; Jauch & Wilson, 1979). Even examinations of managing firm performance in crises are rare (Daily, Dalton & Cannella, 2003). Maybe there is a general belief that innovation has not much with a financial crisis (Filipetti & Archibugi, 2010) and maybe therefore the make-or-buy decision in times of crises has received not so much attention. The missing literature of the relationship between innovation, economic crises and firm performance is the gap this study aims to fill. This study tries to answer the questions if firm behave differently during a financial crisis and if the performance of these firms differs from “normal periods”. First it is important to define the main characteristics of an

economic and financial crisis, which are: a reduction in available credit, consumer spending is reduced and consumer spending is becoming more price conscious, an increase in

unemployment, a significant move towards establishing cost efficiencies, a reduction in shareholder value and lower return on investments (Stiglitz, 2009). Pearson & Clair (1998) stated that a political and economic crisis, in combination with a legal system that does not work, creates high levels of uncertainty and can affect firm performance. Mitroff, Pauchant and Shrivastava (1988) argue that an environment like that can even threaten the company’s survival. A financial crisis is more or less a major change in the external environment of

(10)

10 opportunities (Haveman, 1992; Meyer, 1982), and firms that recognize these new opportunities can reap significant benefits from it (Wan & Yiu, 2009). But of course, economic and financial crises bring a large set of constraints. The external environment is changing and firms have to react to that development, but financial constraints make that difficult. The decision whether to make or to buy technological assets is a topic that is even more present. Acquisitions can next to providing valuable resources also initiate strategic renewal (Graebner et al., 2010) and that is something which can be needed desperately during an environmental jolt. Acquisitions can also revitalize buyers, preventing them from becoming inert and rigid (Vermeulen, 2005) and a period of crisis can be viewed by firms as an opportunity to reshape resources and capabilities in order to adapt in a better fashion to the changing environment according to Teece, Pisano & Shuen (1997), who describe these phenomena from a dynamic capabilities perspective. The big question is: how do firms react to a financial crisis and how does this affect the performance of the firm? Meyer (1982) argues that a given organization’s behavior during a crisis may be substantially different from its behavior during calmer periods. An example of this different behavior is given by Paunov (2012), who argues that if firms behave differently and cut back on their investments because of a crisis it may be harmful for the performance of the firm and a consequence can be that those firms will not be able to return to their previously followed innovation paths. Wan and Yiu (2009) have found in their study that corporate acquisitions have a positive relationship with firm performance, but this is the only study that combines to a certain extent innovation, firm performance and financial crises. Financial crises have of course their influence on financial performance. Although not much research has been done on these effects, some researchers have tried to explain them. Where Sufian & Habibullah (2010) argue that macroeconomic and financial indicators do not have a significant impact on performance, others argue differently. For instance, Bris, Koskinen & Pons (2004) found that firm

performance in Europe and Latin America did not really get any worse during crises, but the performance of Asian firms did. Yongil & Miller (2004) find that firms, with few exceptions, have a worse financial performance during crises. This finding is also supported by Gonenc & Avbar (2006), who find that the financial performance of firms decreases dramatically during crises.

Despite that there has been some research on the effects of financial crises on performance; a gap in literature has been left open. Especially the effects of a crisis on the outcomes of a make-or-buy decision have not been investigated. This study aims to fill that gap in literature. The goal of this study is to assess whether or not crises cause differences in financial performance.

Therefore the following hypotheses have been developed:

(11)

11

H2b: During economic and financial crises, a better appropriability regime (patents) will lead to a worse financial performance, while acquisitions do not create value

The objective of this research is to assess whether or not an economic and financial crisis causes differences in the relationship between make-or-buy decisions and financial firm performance. Therefore this study is based on the following research question:

What differences in the relationship between make-or-buy decisions and financial firm performance are caused by economic and financial crises?

The following sections of this study will describe the methodology, the results, the implications and the conclusions.

3. METHODS

The empirical analysis used to explain the influence of the possession of patents on firm

performance in the pharmaceutical industry; based on a shareholders perspective, is conducted with a market-oriented event study approach, thereby building upon a study by Kirchhoff & Schiereck (2011). Their analysis follows the established approach, developed, among others, by Henderson (1990). First, the events that are relevant for this study are identified. Next all independent, dependent, control and interaction variables will be explained. This also includes showing how certain variables have been calculated. Finally, the statistical tests that will be used in this study are explained.

3.1 Data Sample

The data sample for this study has been acquired from the Thomson Financial SDC database. The pharmaceutical industry has been chosen as the industry to conduct this study. First of all; the pharmaceutical industry deals with huge players that need to stay innovative in order to get in front and to stay in front of the competition. And second; the pharmaceutical industry provides an enormous pool of acquisition data with makes it possible to come up with a sample size that can lead to significant outcomes.

The acquisitions involved in this study had to meet the following requirements:

(12)

12 - The acquirers had to be based in Europe or in North-America. Not only because the

majority of pharmaceutical companies is based in these regions, but also because these regions give the most reliable financial data.

- The acquirers had to be listed. All acquirers that were not listed after all, or did not have any financial data available were removed from the data sample. This included all information about the method of payment used to finance the acquisition.

- All acquisitions had to be announced between 01-01-1990 and 31-12-2010. This period has been used because several periods of financial crises have taken place in these 21 years.

- The status of all acquisitions had to be complete.

- The SEDOL-code of the acquirers had to be present. This was needed to acquire additional financial data from Thomson Reuters’ Datastream.

The initial dataset download from the SDC Database included 2,616 acquisitions. After removing all acquirers and acquisitions that did not match the requirements a total of 1,771 acquisitions remained in the final dataset that will be used for the statistical analyses.

3.2 Independent and Dependent Variables

Several independent and dependent variables have been constructed for conducting the statistical analysis of this study. Some of those variables were given by the data that was

acquired, others had to be calculated. All these variables will be explained in this paragraph and a summary of the variables can be found in table 2 at the end of this paragraph.

3.2.1 Independent variables 3.2.1.1 Patents

(13)

13 Trajtenberg, 1990). In this study a distinction has been made between patents published for targets and acquirers. Also a different method of calculating the amount of patents has been used for target and acquirers. The patent data has been acquired from the World Intellectual Property Organization (WIPO, www.wipo.int). The amount of target patents has been calculated by taking the announcement date and adding up all patents the target had published between the announcement date and four years before that date. A distinction has been made between patent rich targets and patent poor targets. Every target that had at least one patent published was considered to be patent rich. Every target that had not any patents published was

considered to be patent poor. The amount of acquirer patents has been calculated by adding up all published patents of the acquirer between 01-01-1990 and 31-12-2010. Here also a

distinction has been made between patent rich acquirers and patent poor acquirers. Every acquirer that had been granted 52 patents or more between 01-01-1990 and 31-12-2010 was considered to be patent rich. Every acquirer that had less than 52 patents granted was

considered to be patent poor.

3.2.2 Dependent variables

3.2.2.1 Cumulative Abnormal Return

A part of this event study is based on cumulative abnormal returns of the acquirers during different event and estimation windows. The data needed for calculating the cumulative abnormal returns has been collected from Thomson Reuters’ Datastream. The cumulative abnormal returns have been calculated by following the “event studies manual” written by L. Dam of the University of Groningen. The cumulative abnormal returns have been calculated as follows: First, the daily returns for the acquirers and the market were calculated:

Acquirer Daily Return = (returnindex – L.returnindex) / L.returnindex

Market Daily Return = (index – L.index) / L.index

The market that has been chosen for assessing the normal returns is the Standard and Poors 500. This stock market exchange has a broad range of companies in it, including several pharmaceutical companies analyzed in this study and this stock market exchange gives a good reflection of how an average stock market in the world performs.

(14)

14

TABLE 1

Overview of Cumulative abnormal returns

Variable Name Event Window Estimation Window

CAR1 -1 , +1 -30 , -60 CAR2 -1 , +1 -30 , -150 CAR3 -2 , +1 -30 , -60 CAR4 -2 , +1 -30 , -150 CAR5 -5 , +1 -30 , -60 CAR6 -5 , +1 -30 , -150 CAR7 -20 , +1 -30 , -60 CAR8 -20 , +1 -30 , -150

After creating the event windows separate regressions have been run on all companies to come up with all alphas and betas. After this step the daily abnormal return was calculated by taking the actual return for each day in the event window and subtracting the predicted return from it. The cumulative abnormal return was calculated as the sum of all daily abnormal returns. To prevent that the outliers of the CARs have an influence on the actual outcomes of the statistical tests performed, winsorized versions of all CAR-variables have been made. The Winsor version of the variable has filtered out the outliers and gives a more consistent variable.

Pakes (1985) found a strong relationship between stock market returns and changes in patent applications. The nature of this relationship is not very clear. This study aims to define the nature of this relationship by measuring the effect of patents on the cumulative abnormal returns of acquirers

3.2.2.2 Return on Assets, Dividend and Earnings per share, and Market Value

To construct a picture that is as complete as possible another four dependent variables have been created to determine what influence the possession of patents has on the financial performance of a firm. All financial information needed to construct these variables has been acquired from Thomson Reuters’ Datastream. First, Return on Assets will be used to measure profitability because it is a widely used measure in innovation studies (Roberts & Amit, 2003; Sher & Yang, 2005) and therefore very well applicable for this study. The Return on Assets has been calculated using the following formula:

Return on Assets = Net Profit (Income) / Total Assets

(15)

15 The dividend per share as well as the earnings per share, and market value of the acquirers did not have to be calculated where Datastream supplied that information completely. Return on Assets as well as Dividend and Earnings per Share are important for shareholders. These are financial indicators used to assess whether a company is performing well or not. Given that shareholders always want maximum return for their investments it is possible to assess which actions an acquirer should avoid to keep the shareholders happy. On this subject of shareholder value a study by Heiens, Leach & McGrath (2007) concluded that intangible assets, excluding goodwill but including patents, have a positive influence on shareholder value. Moreover, evidence from a study by Jensen (1988) suggests that the shareholders in acquired firms derive significant value from acquisitions.

Finally, the variable Market Value has been created to assess whether or not the possession of patents influences the overall value of the acquirers. The market value can be seen as a forward-looking measure of firm performance (Hall, 2000). That is because if stock markets are efficient, the company value equals the sum of discounted future cash flows (Fama, 1970). Market value can therefore be seen as a kind of predictor of how a firm is going to perform in the future. Hall, Jaffe & Trajtenberg (2005) find a strong positive relationship between the number of patent citations the firm receives and the stock market value of the firm. A similar relationship is found by Belenzon (2012) who associates patents with a higher market value. King, Slotegraaf & Kesner (2008) however argue that although acquisitions and thereby the acquisition of

technological resources such as patents, have strategic potential, they often fail to create value.

3.3 Control Variables 3.3.1 Diversification

Diversification has been created as a control variable because research has suggested a negative relationship between diversification and R&D intensity (Baysinger & Hoskisson, 1989;

Hoskisson & Hitt, 1988). Based on this predicted relationship diversification can be controlled for in this study. The extent of diversification has been measured through the SIC codes of the targets and acquirers. A distinction has been made between diversified and undiversified acquisitions. Every acquisition for which the difference between the SIC code of the target and the acquirer is more than 100 the acquisition is considered to be diversified. If the difference between the SIC codes is less than 100 the acquisition is considered to be undiversified.

3.3.2 Transaction Value

(16)

16 transaction value for all acquisitions, but those missing variables have been replaced by the mean value of the sample. SPSS had done this otherwise automatically. There has been made a distinction between Expensive, Average and Cheap valued deals. Expensive deals are deals with a transaction value over $ 372 million. Average deals are deals with a transaction value between $ 372 million and $ 34,63 million and Cheap deals are deals with a transaction value that is less than $ 34,63 million.

3.3.3 Acquirer Employees

Firm size effects innovation, and the innovative capacity of an organization. This relationship is pretty strong as has been shown by Hitt, Hoskisson & Ireland, 1990. Which way the relationship goes is a contested subject. Schumpeter (1961) hypothesized that larger firms are more

innovative because their Research and Development programs are efficient and sustained. This vision is contested by Kamien & Schwartz (1982), who argued that the relationship between the size of the firm and innovation is a non-linear inverse-U relationship. For this study the size of the firm has been defined as the amount of employees each acquirer had. A number that has been given by Thomson Reuters’ SDC database. Also in this case there were some missing values; these values have also been replaced by the mean of the total sample. A distinction has been made between Large, Medium and Small firms. Large firms had over 12,700 employees. Medium sized firms had between 12,700 and 651 employees. Small firms had fewer than 651 employees.

3.3.4 Control variables based on patents

To give some extra control for this study and to give some extra information about certain ratios that can be useful for assessing if the possession of patents has an influence on firm

performance. First there is the price per patent, which has been calculated as follows:

Price per patent = Transaction Value / Target Patents

Second the ratio of acquirer employees and the amount of acquired patents, which has been calculated as follows:

Acquirer Employees per patent = Acquirer Employees / Target Patents

Next the ratio of acquirer patents per target patent. This ratio can show how many patents are bought relatively to the amount of patents that already was possessed. The ratio has been calculated as follows:

(17)

17 Finally the ratio of acquirer employees per acquirer patent shows how many employees were “used” for one patent. Of course, not every employee of the company works on patents and several employees are responsible for multiple patents, but this ratio gives an indication of how innovative per employee the acquirers are. The ratio has been calculated as follows:

Acquirer Employees per Acquirer Patents = Acquirer Employees / Acquirer Patents

3.4 Interaction/Selection Variables

The interaction variables constructed for this study are only used as selection variables. The primary variable on which every interaction variable is based is Target Patent. The interaction variables have been constructed by taking the patent rich targets and assessing by every other variable (crisis, diversification, large firms, medium sized firms, small firms, expensive deals, average priced deals, cheap deals, cash deals, other deals, stock deals and international deals) if it is considered to be both a 1. So the interaction variable will be used if, for example, the acquisition contains a patent rich target and a large acquirer.

3.4.1 Crisis

(18)

18

TABLE 2

Overview of all Variables

Variable characteristics Variable name

Independent

Variable Target Patents tar_pat

Independent

Variable Acquirer Patents acq_pat

Dependent Variable Cumulative Abnormal Return car1

car1_winsor car2 car2_winsor car3 car3_winsor car4 car4_winsor car5 car5_winsor car6 car6_winsor car7 car7_winsor car8 car8_winsor

Dependent Variable Return on Assets ROA

Dependent Variable Dividend per Share DPS

Dependent Variable Earnings per Share EPS

Dependent Variable Market Value MV

Control Variable Diversification divers_sic

Control Variable Transaction_value trans_val

Control Variable Acquirer Employees (Firm Size) acq_empl

Control Variable Price per Patent price_per_patent

Control Variable Acquirer Employees per patent empl_per_patent

Control Variable Acquirer Patents per Target Patents acq_pat_per_tar_pat

Control Variable Acquirer Employees per Acquirer

Patents empl_per_acq_pat

Interaction Variable Target Patent * Crisis IPRC

Interaction Variable Target Patent * Diversification IPRD

Interaction Variable Target Patent * Large Firm IPRL

Interaction Variable Target Patent * Medium Sized Firm IPRM

Interaction Variable Target Patent * Small Firm IPRSM

Interaction Variable Target Patent * Expensive Deal IPRE

Interaction Variable Target Patent * Average priced Deal IPRA

Interaction Variable Target Patent * Cheap Deal IPRCH

Interaction Variable Target Patent * Method of payment: Cash IPRCA

Interaction Variable Target Patent * Method of payment:

Other IPRO

Interaction Variable Target Patent * Method of payment:

Stock IPRST

Interaction Variable Target Patent * International Deal IPRI

Selection Variable Crisis CRISIS

3.5 Statistical Tests

(19)

19 it possible to see what the influence is of the independent variables on the dependent variables and also shows the degree of change in the dependent variables when the independent and control variables are changed by 1%.

4. RESULTS

TABLE 3

Descriptive Statistics for All Variables

N Minimum Maximum Mean Std. Deviation

tar_pat 1771 0 10809 153,34 730,745 acq_pat 1771 0 19706 2065,82 4858,560 divers_sic 1771 0 7474 1191,01 1989,586 trans_val($mln) 1771 0,03 89167,72 500,8235 3372,62174 acq_empl 1771 1 122000 15161,64 27225,522 price_per_patent ($mln) 1771 0 8300 35,28753026 286,8036796 empl_per_patent 1771 0 122000 1461,868173 7048,988737 acq_pat_per_tar_pat 1771 0 19706 227,1306798 1353,784285 empl_per_acq_pat 1771 0 12757 323,2328637 1369,355066 car1_winsor (%) 1771 -0,187536000 0,0666483000 -0,0255220725 0,0791111530 car2_winsor (%) 1771 -1,829619000 0,0544525000 -0,3740921690 0,7020544454 car3_winsor (%) 1771 -0,1447679000 0,0880923000 -0,0145023714 0,0713349015 car4_winsor (%) 1771 -0,3038978000 0,0687987000 -0,0583380253 0,1313313003 car5_winsor (%) 1771 -0,1311090000 0,1211980000 -0,0019893586 0,0854428805 return_on_assets (%) 1771 -208 2 -0,2798725881 5,327141247 dividend_per_share ($) 1771 0 127,36 1,565711061 8,133551998 earnings_per_share ($) 1771 0 304,14 4,452618510 18,09693188 market_value ($mln) 1771 0,06 290445,1000 17242,71073 41899,48552 4.1 Descriptive Statistics

(20)

20 the winsorized versions gave any chance of significance and only five of these winsorized

versions actually gave significance. Therefore only those five versions of CAR measures have been included. When looking at the performance measures we find that over the whole period of 21 years the average cumulative abnormal returns as well as the average return on assets have been negative. Based on these outcomes it can be said that the overall influence of an acquisition on the stock market price and the return on assets has been negative. On the other hand,

acquisitions are beneficial for dividends and earnings per share.

4.1.1 Descriptive Statistics in a Crisis Situation

TABLE 4

Descriptive Statistics for Variables in a Crisis Situation

N Minimum Maximum Mean Std. Deviation

tar_pat 790 0 10809 151,41 785,274 acq_pat 790 0 19706 2062,16 4848,800 divers_sic 790 0 7474 1174,53 1992,732 trans_val($mln) 790 0,04 67285,70 466,7880 2878,93150 acq_empl 790 1 122000 15449,50 28285,866 price_per_patent ($mln) 790 0 8300 38,96986652 328,6545460 empl_per_patent 790 0 101800 1417,748825 6506,101650 acq_pat_per_tar_pat 790 0 19706 208,58110101 1375,135709 empl_per_acq_pat 790 0 12757 328,6900020 1395,023661 car1_winsor (%) 790 -0,187536000 0,0666483000 -0,0210679200 0,0745083356 car2_winsor (%) 790 -1,829619000 0,0544525000 -0,3332272100 0,6729195506 car3_winsor (%) 790 -0,1447679000 0,0880923000 -0,0141175315 0,0689400707 car4_winsor (%) 790 -0,3038978000 0,0687987000 -0,0532834348 0,1271573634 car5_winsor (%) 790 -0,1311090000 0,1211980000 -0,0057696075 0,0827511654 return_on_assets (%) 790 -35,20145631 1,625642032 -0,1329759965 1,479480772 dividend_per_share ($) 790 0 87,69 1,8560 8,65850 earnings_per_share ($) 790 0 203,26 5,0220 18,89589 market_value ($mln) 790 0,06 290445,1000 17720,3311 42990,90094

(21)

21 percentage for return on assets and the dividend and earnings per share a lower in a crisis situation. However, the company with the highest market value of the sample did the acquisition during a crisis period. When comparing the means it can be said that the number of granted patents for targets as well as acquirers is somewhat lower. The transaction value is lower on average, but companies pay more for a single patent. All cumulative abnormal returns are lower during a crisis period as well as the return on assets. However, the dividend per share, earnings per share and the market value are all higher on average compared to the overall data sample.

4.2 Correlations

Table 1 in the Appendix shows the correlations for all numeric variables involved in this study. The results show that Target Patents are negatively significant correlated with diversification and positively significant correlated with the transaction value of the acquisition and the market value of the acquirer. It shows that the acquisition of a patent rich target will cost acquirers more, because the transaction value should be higher, but the acquisition will be beneficial for the acquirer because the market value goes up. The results also show that Acquirer Patents are negatively significant correlated with car1_winsor, car2_winsor and car4_winsor. However Acquirer Patents are positively significant correlated with dividends per share, earnings per share and the market value of the acquirers. It shows that the more patents an acquirer already possesses at the time of acquisition, the worse it is for the abnormal returns a shareholder derives. The possession of patents is, however, also beneficial for shareholders because it increases the paid dividend per share and the earnings per share. It is also beneficial for the acquirers because it increases market value.

4.2.1 Correlations in a Crisis situation

When comparing Table 1 in the Appendix with Table 2 in the Appendix the following can be said. Of the 153 observations that have been made by this correlation analysis; 35 of them have a less negative relationship than in a “normal situation”. 31 of these observations show a relationship that is more negative, again 31 other observations show a less positive relationship, 54

observations shows a relationship that is more positive than in a normal situation and 2 observations stayed exactly the same. What stands out is that in 89 of the 153 cases the relationship between variables improves during a crisis.

4.3 Testing hypotheses

(22)

22 review. So first, the cumulative abnormal returns, followed by return on assets, dividend per share, earnings per share and market value. Only the models in which variables have shown a significance level of 95% or more have been included. There is one exception; model 32 has been included because this model assesses the influence of a financial crisis. To assess whether or not the outcomes of the total datasets stay the same; twelve additional regressions have been executed using samples created by the interaction variables that were introduced in the Methods section. Finally a regression analysis will be performed to assess whether or not a financial crisis has any influence. Only acquisitions that were announced in a crisis period are included in that analysis. The total sample size is 790. These regressions will be performed at the end of each dependent variable under investigation.

4.3.1 Cumulative Abnormal Returns

The assessment of the influence of patents on CARs can be seen in models 1 to 32. Five different versions of winsorized CARs gave significance in the linear regression tests. Car1winsor,

Car2winsor, Car3winsor, Car4winsor and Car5winsor have been used as dependent variables. When looking at Car1winsor, the event window was 3 days; the estimation window was 30 days. Target patents do not have a significant influence, however, acquirer patents do: β = -0,148 (ρ < 0,01). This means that the cumulative abnormal return decreases by 14,8% for every 1% rise in acquirer patents. The next models include interaction variables. Model 2 (IPRCR): β = -0,224 (ρ < 0,05). Model 3 (IPRL): β = -0,205 (ρ < 0,05). Model 7 (IPRO): β = -0,340 (ρ < 0,05) and Model 8 (IPRI): β = -0,222 (ρ < 0,05). These significant betas are achieved for acquirer patents. All other models also gave a negative relationship in the regression, but were not significant at a 95% level or better.

Car2winsor had an event window of 3 days and an estimation window of 120 days. In this case target patents also did not have any significant influence on the cumulative abnormal return. Acquirer patents did have a significantly negative influence: β = -0,207 (ρ < 0,01). The next results will again concern samples created by interaction variables. Only acquirer patents are significant. Model 11 (IPRCR): β = -0,246 (ρ < 0,05). Model 12 (IPRL): β = -0,294 (ρ < 0,01). Model 14 (IPRE): β = -0,308 (ρ < 0,01). Model 16 (IPRST): β = -0,608 (ρ < 0,05) and Model 17 (IPRI): β = -0,352 (ρ < 0,01). Also in this case all other models gave a negative regression, but they were not significant at a level of 95% or better.

(23)

23 The next dependent CAR variable, Car4winsor, had an event window of 4 days and an estimation window of 120 days. The model that was executed for the total data sample (model 21) did not give any significance for target patents, but the model did give significance for acquirer patents: β = -0,179 (ρ < 0,01). The following models include samples given by interaction variables. Significance only was found for acquirer patents. Model 22 (IPRCR): β = -0,262 (ρ < 0,05). Model 23 (IPRL): β = 0,227 (ρ < 0,05). Model 26 (IPRE): β = 0,248 (ρ < 0,05). Model 28 (IPRST): β = -0,598 (ρ < 0,05) and Model 29 (IPRI): β = -0,305 (ρ < 0,01). Also in this case it can be said that all other regression analyses that were not mentioned had a negative influence on the dependent variable.

Finally, for Car5winsor can be said that this variable had an event window of 7 days and an estimation window of 30 days. Only the sample created by interaction variable IPRD resulted in being significant for acquirer patents: β = 0,256 (ρ < 0,05). Interesting to see is that the outcome is positive.

Overall it can be said that the influence of target patents seems nonexistent. Something else can be said about the influence of acquirer patents. For all models, except one, the influence was significantly negative. Proving the argumentation of Pakes (1985) and giving direction to the strong relationship that was described in that study. The effects of these findings will be described in paragraph 4.3.11.

4.3.2 Cumulative abnormal return in a crisis situation

In order to assess whether a financial crisis has any influence on the regressions concerning cumulative abnormal returns five separate regression have been run including only acquisitions that were announced during financial crises. The only significance was found for acquirer patents. The results are presented in Model 9: β = -0,269 (ρ < 0,01). Model 18: β = -0,266 (ρ < 0,01). Model 20: β = -0,196 (ρ < 0,01). Model 30: β = -0,256 (ρ < 0,01) and Model 32: β = -0,131 (ρ < 0,10). These results indicate a strong negative relationship between acquirer patents and cumulative abnormal returns, certainly because only one model does not have significance at a level of 99%. Which kind of consequences these results have for the crisis hypotheses will become clear at the end of this chapter.

4.3.3 Return on Assets

(24)

24 34): β = 0,254 (ρ < 0,05). A significant positive relationship has been found, but thereby not supporting the results of Hitt et al., (1991) and Artz et al., (2010).

4.3.4 Return on Assets in a crisis situation

The statistical assessment of the influence on the return on assets in a financial crisis can be found in Model 35. No significance was found.

4.3.5 Dividend per Share and Earnings per Share

The regression model (36) describing the full dataset and using dividend per share as a dependent variable gives a negative regression for acquirer patents: β = -0,397 (ρ < 0,01). Target patents are not significant. The next models are all samples caused by interaction variables. Significance was only found for acquirer patents. Model 37 (IPRCR): β = -0,458 (ρ < 0,01). Model 38 (IPRD): β = -0,566 (ρ < 0,01). Model 39 (IPRL): β = -0,383 (ρ < 0,01). Model 41 (IPRE): β = 0,464 (ρ < 0,01). Model 42 (IPRA): β = 0,262 (ρ < 0,05). Model 43 (IPRCH): β = -1,058 (ρ < 0,01). Model 44 (IPRCA): β = -0,496 (ρ < 0,01) and Model 45 (IPRI): β = -0,437 (ρ < 0,01). In eight of twelve cases were different samples have been used the outcome is a negative regression.

The outcome of the regression where Earnings per Share is the dependent variable and the full dataset is used can be found in Model 47. Significance has only been found for acquirer patents: β = -0,328 (ρ < 0,01). Regarding the other models; significance has been found for target patents only in Model 51 (IPRM): β = -0,141 (ρ < 0,05) and significance has been found for acquirer patents in multiple models: β = -0,328 (ρ < 0,01). For the following models can be said that significance has been found for acquirer patents. Model 48 (IPRCR): β = -0,345 (ρ < 0,01). Model 49 (IPRD): β = 0,496 (ρ < 0,01). Model 50 (IPRL): β = 0,309 (ρ < 0,01). Model 52 (IPRE): β = -0,382 (ρ < 0,01). Model 54 (IPRCH): β = -0,936 (ρ < 0,01). Model 55 (IPRCA): β = -0,405 (ρ < 0,01) and Model 56 (IPRI): β = -0,347 (ρ < 0,01). A very strong negative relation has been found for patents and acquisitions on shareholder value. Thereby the outcomes of Hall (1992), Heiens et al., (2007) and Simon & Sullivan (1993) have not been supported. The effects of these results will be given in parapraph 4.3.11.

4.3.6 Dividend per Share and Earnings per Share in a crisis situation

(25)

25 Significance has been found for acquirer patents: β = -0,344 (ρ < 0,01). The effect of these

outcomes on the hypotheses developed for crisis situations will be described later in this section.

4.3.9 Market Value

In contrast to the previous dependent variables target patents have on their own a significant influence on this dependent variable: Model 58: β = 0,078 (ρ < 0,01). In combination with the other (control) variables target patents (Model 59: β = 0,039 (ρ < 0,01)) as well as acquirer patents (Model 59: β = -0,235 (ρ < 0,01)) have a significant influence. Because in the following models target patents as well as acquirer patents are significant a distinction will be made. First the models will be presented with target patent significance, followed by the models with acquirer patent significance. Target patents models with significance: Model 60 (IPRCR): β = 0,097 (ρ < 0,01). Model 63 (IPRM): β = 0,137 (ρ < 0,05). Model 64 (IPRE): β = 0,072 (ρ < 0,05). Model 69 (IPRST): β = -0,119 (ρ < 0,01) and Model 70 (IPRI): β = 0,134 (ρ < 0,01). Acquirer patents models with significance: Model 60 (IPRCR): β = 0,235 (ρ < 0,01). Model 61 (IPRD): β = 0,130 (ρ < 0,05). Model 62 (IPRL): β = 0,233 (ρ < 0,01). Model 64 (IPRE): β = 0,194 (ρ < 0,01). Model 65 (IPRA): β = 0,328 (ρ < 0,01). Model 67 (IPRCA): β = 0,360 (ρ < 0,01). Model 68 (IPRO): β = 0,314 (ρ < 0,01). Model 69 (IPRST): β = -0,244 (ρ < 0,05) and Model 70 (IPRI): β = 0,278 (ρ < 0,01). One thing that stands out the most is that the relationships in these regressions, except for Model 69, are positive. It is a finding that is in strong contrast with the other dependent

variables of this dataset and by which the outcomes of the studies conducted by Hall et al., (2005) and Belenzon (2012) are supported. The effects of these findings will be explained in paragraph 4.3.11.

4.3.10 Market Value in a crisis situation

Model 71 describes the regression analysis for market value in a crisis situation. In this model target patents β = 0,085 (ρ < 0,01) as well as acquirer patents β = 0,278 (ρ < 0,01) are significant at a 99% level. So also in a crisis situation the possession and acquiring of patents has a positive influence on the market value of the firm. The effect of this finding on the crisis hypotheses will be described next.

4.3.11 The overall effects of patents on financial performance

(26)

26 cannot be accepted because of these results. In contrast hypothesis H1b can be accepted because patents do cause a worse financial performance and the descriptive statistics have shown that the average acquisition in this dataset does not create value.

4.4 The influence of a crisis situation

Finally the influence of a crisis on the financial performance measures will be assessed. As the outcomes of comparing the descriptive statistics already predicted there are some differences. Pharmaceutical companies spend less on average on acquiring but pay more for a single patent. However, the means for dividend per share, earnings per share and market value are somewhat higher.

When looking at the effects of a financial crisis on financial performance it can be said that for the cumulative abnormal returns, the dividend per share and the earnings per share a crisis aggravates the negative effects on financial performance. However, for market value a crisis is better, because a crisis makes the relationship with firm performance more positive. Thereby Wan & Yiu (2009) are to a certain extent supported in their findings. Unfortunately, no

significance was found for return on assets. All and all it can be said that the overall influence of patents on financial performance during crises has been negative. The descriptive statistics have shown that also in crises acquisitions do not create value. Therefore hypothesis H2a cannot be accepted. However, hypothesis H2b can be accepted.

The following chapter will give explanations for the phenomena that have been found in this chapter.

5. DISCUSSION

After the statistical assessment of the influences of economic and financial crises on the financial results of a make-or-buy decision several findings will be presented in this chapter. Every key finding will be treated separately in a new paragraph.

(27)

27 cannot. By not using several parts of the target organization firms essentially throw some value away. Another explanation can be that the expectations for value creation of the acquisition were higher than the value that was created in the end. Dealing with those expectations is tricky, especially when the results of the process are lower than the results that were initially

predicted. Shareholders may derive less value than expected, which in turn can cause shareholders to get rid of their stocks and that is harmful for firms.

Another interesting result from the statistical tests is that during crises firms get fewer patents granted compared to the average period. Targets get granted about two patents less in crises, acquirers get granted about three patents less. In addition to that finding, patents become worth more during crises. The average price that has been paid for a patent over the whole 21 years was around $ 35 million. In crises this price goes up to about $ 38 million per patent. It seems that firms in this sample see a making strategy as a bit more risky during crises. This finding supports West (2002) who considers that a buying strategy involves less risk than a making strategy. It also seems that firms consider crises as a period of protecting rather than innovation because of the fewer patents granted and the rise in patent value. Reasoning might be that firms, in times of crises, choose to make use of outside knowledge more than during normal periods. The environment is changing fast and within the organization changes have to be made to create a fit with the new environment again because demands have changed (Perrons & Platts, 2004). Internal processes may not be capable of making such changes. Therefore the environment has to be scanned for external assets which can be acquired and integrated (Arora & Gambardella, 1990) which can lead to strategic renewal (Graebner et al., 2010). Buying can thus be seen as a better option during crises.

(28)

28 crisis may threaten the survival of companies (Mitroff, Pauchant & Shrivastava, 1988) and these companies are not willing to take the same risks as in normal periods because the consequences can be a lot more severe. However, the threat of not surviving can also be a trigger for

companies to handle the process of acquiring more carefully, thereby causing to lose less value compared to normal periods. Firms seem to get more careful during crises. They pay less for their transactions ($ 34 million less in this study), they award more value to patents and all and all they take more care for the acquisition process in trying to lose as less value as possible.

One of the other findings of this study is that the market value of companies is higher on average during financial crises. This can be explained by again looking at the price paid per patent. Firms tend to grant a higher value to patents during crises, therefore the overall value of the company can go up. In addition, although the overall transaction value of acquisitions drops during crises more real value in the form of patents is acquired which may explain the increase in market value. Another explanation might be that the integration process of the acquisition has been handled with more care which causes less value to get lost in the overall process.

The following key findings of this study will focus on the relationships patents have with financial performance. The results show that during the overall (average) period that has been under investigation in this study; the influence of patents on financial firm performance has been significantly negative. This implies that firms should not make all their technologies on their own, because in the end this making strategy has a negative effect on performance of these firms. But why have these negative relationships, except for market value, emerged? After all, several authors (Belenzon, 2012; Hall et al., 2005; Heiens et al., 2007; Jensen, 1988) have claimed that patents have a positive relationship with financial performance. The results from this study therefore do not only contradict heavily with results from previous studies, but they are therefore also very interesting. It goes to show that even when firms are at the top

innovatively are not creating benefits for their shareholders. A reason might be that the overall costs for developing innovations highly exceeds to benefits that are reaped from that innovation. However, this seems highly unlikely because firms should go bankrupt very quickly if this were true. Another explanation might be that patents are primarily seen as means of protection rather than a reward for manufacturing excellence. Additionally a point can be made for the

commercialization time of the innovations. It is a time consuming process (Perrons & Platts, 2004) and to get the commercialization completely right the very first time it will cost a lot of effort and time. However, new innovative products often have to be presented to the market as quickly as possible. It can therefore be possible that firms have to compromise on the

(29)

29 can cause the derived benefits to be less and can eventually lead to a worse financial

performance. The argument of patents being a protective means can also be applied to this reasoning. Because commercialization has to be fast and firms do not want copied versions of their product on the market within a week, patents are used to protect the innovation from being imitated. In the process the value of the patent is pretty high, causing the market value of the firm to go up and because commercialization has not been as perfect as it could have been still less benefits are derived from the process.

Finally the influence of a make-or-buy decision during an economic and financial crisis on the financial performance will be discussed. At first glance the results of the crisis sample do not seem that different from those of the overall sample. However, when looking closely at the results one might see that the magnitude of the relationships becomes bigger during financial crises. The effect of patents on cumulative abnormal returns decreases from an effect of –17,8% per 1% rise in granted patents to an effect of –22,4%. For dividends per share the effect

decreases from –39,7% to –40,7% and for earnings per share the effect decreases from –32,8% to –34,4%. As already has been mentioned before the relationship between patents and market value are positive. This relationship holds during crises and experiences the same effect as the other financial performance measures; however this effect works the other way around. Target patents go from a positive effect of 7,8% to 8,5% so the buying option helps improving

shareholder value. Acquirer patents go from a positive effect of 23,5% to 27,8% which

practically means that for a huge increase in market value, firms should choose for the making strategy. Interesting to see in these results is that the relationship between patents and financial performance strengthens during a financial crisis. The decrease for cumulative abnormal

returns, dividend per share and earnings per share can be explained using the same reasoning that has been used above. Quick commercialization can lead to missed value because the process was not optimized. In addition to that reasoning; an economic and financial crisis causes less credit to be available, price conscious consumer spending and a reduction in consumer spending (Stiglitz, 2009). These are all factors that influence sales and that have an effect on the financial performance. When looking at the effect of patents on market value also the same reasoning as has been used above can be applied. During financial crises patents become even more

important as means of protection. There will be less innovative activity, as this study has shown, and that causes a focus that lies more on protecting the innovations that are already out there. Therefore patents become even worth more during crises and because of this higher value the market value of companies owning patents will also rise with each extra patent they possess.

(30)

30 shareholders to lose value and companies to gain value. And third, these effects will only become stronger during financial crises. So which effects do these outcomes have for a make-or-buy decision? The results from this study suggest that companies should choose buying technologies over developing those technologies themselves. Although the possession of patents is beneficial for the market value of the company; the shareholder value will drop when making technologies and after all shareholders are the owners of the company. Buying technologies enables not only the acquisition of desired resources, but it also enhances market power and it may initiate strategic renewal (Graebner et al., 2010). The latter can be extremely useful when dealing with financial crises. And also for these situations the buying decision is preferred, because the effects of patents on financial performance only become stronger.

6. CONCLUSION 6.1 Key findings and contributions

This study has argued that an economic and financial crisis has an influence on the relationship between acquisitions including innovation output, measured as the number of granted patents, and the financial performance of organizations. The literature section of this study has

introduced three main angles to look at the problem. First, the concept of a make-or-buy

decision has been introduced building on a study by Graebner et al., (2010) who argue that firms can not only acquire valuable assets from a target but an acquisition can also increase market power and initiate strategic renewal. The latter concept can especially be relevant during an economic and financial crisis. That was the second angle of this study, the subject that makes this study contribute to literature by filling in the gap that has been left open by previous studies. This study combines the concepts of patents, acquisitions and economic and financial crises from a make-or-buy perspective, something that has not been done before. The study has been conducted as an event study using pharmaceutical firms which has been explained in section 3. Section 4 gives the results from the data analysis and these results turned out to be somewhat mixed but interesting. Firms pay less for transactions, but pay more money per patent during a crisis which indicates that the value of patents becomes higher during crises. Patents turned out to have a significant strong negative relationship with cumulative abnormal returns, thereby giving direction to the relationship already described by Pakes (1985).

(31)

31 This study was based on the following research question:

What differences in the relationship between make-or-buy decisions and financial firm performance are caused by economic and financial crises?

The answer to this question is that relationship in essence does not change. However, the magnitude of the relationship does change. The negative relationship between patents and financial performance became even more negative and the positive relationship between patents and market value become even more positive. When assessing the effects of the results of this study it can be said that the buying option should be preferred by organizations. First of all the buying option should be preferred because of the negative relationship between

innovation output and financial performance. Although market value rises through an increase of patents, shareholders lose value with every extra patent and after all, they are the owners of the company. Second, as Graebner et al., (2010) already pointed out, acquisitions cannot only help in acquiring desired resources, but they can also enhance market power and initiate strategic renewal. Especially strategic renewal is a tool for dealing with financial crises. So because acquisitions do not only, more or less, prevent shareholders from losing value; they can also help dealing with financial crises and that is why the buying option should be preferred by firms.

6.2 Limitations

This study has some limitations that have to be taken into account. First of all the use of patents as a measure of innovative output. As has been argued in section 3, patents have some

limitations as a measure of innovative output. Not every company that develops technological assets will patent these assets. Not every technological asset a company develops is patentable. And patents differ greatly in economic value (Cohen & Levin, 1989; Griliches, 1990; Trajtenberg, 1990). Although it is very difficult to treat all patents as equals, patents have been used to measure innovative outputs, because patents are directly related to innovativeness (Walker, 1995) and because patents represent an external validation of technological novelty (Griliches, 1990). Besides it is an enormous job and nearly impossible to effectively figure out what the value of each patent is and to figure out which technological assets have not been patented and to give a value to those assets. However, this study can be a starting point for a more extensive study in which the effects of a financial crisis on the value of individual patents can be

(32)

32 possibility to get all the financial data that was needed for this study, but it also gives a

geographical scope that may be a bit too narrow for recent times. Pharmaceutical companies from Asia are beginning to make a bigger impact on the global pharmaceutical industry as a whole. The problem was that of these companies not all financial information that was required for this study was available. However, it is a problem that can be solved in the future. Asian pharmaceutical companies will get more attention in the future when they get bigger and more information will become available. That will be the time to conduct this event study on a global basis.

6.3 Managerial Implications

The implications of the findings of this study are presented in this paragraph. Based on the theories that have been tested and the outcomes of the data several implications for managers can be identified. First, the results of the data imply that firms should choose buying

technologies over manufacturing technologies themselves. Both options cause the market value of the companies to go up, but manufacturing technology and the application for patents will cause negative effects for shareholder returns. Managers will have to find a balance between keeping their shareholders happy and still improving their market value. Manufacturing technologies has a positive effect on market value that is higher than buying technologies, but manufacturing technologies itself has a strong negative influence on shareholder value. These effects are even getting stronger when firms have to deal with financial crises. And although the market value of the company is getting better through the development of new technologies, managers should focus more on protecting their innovations than on developing new

technologies. Especially during a financial crisis where every shareholder is even more

(33)

33

REFERENCES

Agarwal, R., & Helfat, C.E. (2009). Strategic Renewal of Organizations. Organization Science,

20(2), 281-293

Amburgey, T.L., & Miner, A.S. (1992). Strategic Momentum: The effects of Repetitive, Positional, and Contextual Momentum on Merger Activity. Strategic Management Journal, 13(5), 335-348

Argyres, N. (1996). Evidence of the Role of Firm Capabilities in Vertical Integration Decisions.

Strategic Management Journal, 17(2), 129-150

Arora, A., & Gambardella, A. (1990). Complementarity and External Linkages: The Strategies of the Large Firms in Biotechnology. Journal of Industrial Economics, 38(4), 361-379

Artz, K.W., Norman, P.M., Hatfield, D.E., & Cardinal, L.B. (2010). A Longitudinal Study of the Impact of R&D, Patents, and Product Innovation on Firm Performance. Journal of Product

Innovation Management, 27(5), 725-740

Audia, P.G., Locke, E.A., & Smith, K.G. (2000). The Paradox of Success: An Archival and a

Laboratory Study of Strategic Persistence Following Radical Environmental Change. Academy of

Management Journal, 43(5), 837-853

Balasubramanian, N., & Lee, J. (2008). Firm Age and Innovation. Industrial & Corporate Change,

17(5), 1019-1047

Baysinger, B., & Hoskisson, R.E. (1989). Diversification Strategy and R&D Intensity in Multiproduct Firms. Academy of Management Journal, 32(2), 310-332

Belenzon, S. (2012). Cumulative Innovation and Market Value: Evidence from Patent Citations.

Economic Journal, 122(559), 265-285

Birkinshaw, J., Bresman, H., & Håkanson, L. (2000). Managing the Post-acquisition Integration Process: How the Human Integration and Task Integration Processes Interact to Foster Value Creation. Journal of Management Studies, 37(3), 395-425

Bris, A., Koskinen, Y., & Pons, V. (2004). Corporate Financial Policies and Performance around Currency Crisis. Journal of Business, 77(4), 749-795

(34)

34 Capron, L., & Mitchell, W. (2009). Selection Capability: How Capability Gaps and Internal Social Frictions Affect Internal and External Strategic Renewal. Organization Science, 20(2), 294-312

Cassiman, B., & Veugelers, R. (2006). In Search of Complementarity in Innovation Strategy: Internal R&D and External Knowledge Acquisition. Management Science, 52(1), 68-82

Ceccagnoli, M. (2009). Appropriability, Preemption, and Firm Performance. Strategic

Management Journal, 30(1), 81-98

Chattopadhyay, P., Glick, W.H., & Huber, G.P. (2001). Organizational Actions in Response to Threats and Opportunities. Academy of Management Journal, 44(5), 937-955

Claessens, S., & Djankoc, S. (2000). Corporate Performance in the East Asian Financial Crisis.

World Bank Research Observer, 15(1), 23-46

Cohen, W., & Levin, R. (1989). Empirical studies of innovation and market structure. In

Handbook of Industrial Organization (II). Schmalensee, R., & Willig, R.D. (Eds.). Amsterdam:

Elsevier, 1059-1107

Cohen, W.M., & Levinthal, D.A. (1990) Absorptive Capacity: A New Perspective on Learning and Innovation. Administrative Science Quarterly, 35(1), 128-152

Contractor, F.J., & Lorange, P.P. (1988). Competition vs. Cooperation: A Benefit/Cost Framework for Choosing Between Fully-Owned Investments and Cooperative Relationships. Management

International Review (MIR), 28(4), 5-18

Cyert, R.M., & March, J.G. (1963). A behavioral theory of the firm. Englewood Cliffs, NJ: Prentice-Hall.

Daily, C.M., Dalton, D.R., & Cannella Jr., A.A. (2003). Corporate Governance: Decades of Dialogue and Data. Academy of Management Review, 28(3), 371-382

Danzon, P.M., Epstein, A., & Nicholson, S. (2007). Mergers and Acquisitions in the Pharmaceutical and Biotech Industries. Managerial & Decision Economics, 28(4/5), 307-328

Datta, D.K., Pinches, G.E., & Narayanan, V.K. (1992). Factors Influencing Wealth Creation from Mergers and Acquisitions: A Meta-Analysis. Strategic Management Journal, 13(1), 67-84

Referenties

GERELATEERDE DOCUMENTEN

AXES SYSTEM ARCHITECTURE - SEARCHING BASED ON AUDIO-VISUAL CONTENT Considering that an archive may grow over time, we define our system such that new content (videos

The approach we take in designing POINTER and LOITER, our serious games for interpersonal skills training in the domain of law enforcement, expands the possibilities of

Hence, model 3 is run for both variables separately (this is shown in appendix 1; model 3.1 and model 3.2). higher intellectual or executive) and the other job

Ondanks het feit dat clandestiene bladen veel minder talrijk waren dan de gecensureerde pers, hadden ze toch de mogelijkheid om de Belgische bevolking van nieuws te voorzien die

The next step in the methodology, after having found evidence of long-run cointegration relationships between stock market development, economic growth and investment, is the

However, police members who experienced stress because of lack of resources and police stressors also showed a higher professional efficacy, which are feelings of

Stefan Kuhlmann is full professor of Science, Technology and Society at the University of Twente and chairing the Department Science, Technology, and Policy Studies (STePS). Earlier

In short, where section 2.5 discussed the possible effects of the small scale of operations, the low visibility, the involvement in local communities and the lack