Markets for Technology and Corporate Groups: Patent Transactions Within and Between Corporations

(1)

Master Thesis

Markets for Technology and Corporate Groups:

Patent Transactions Within and Between Corporations

MSc Business Administration Strategic Innovation Management

Submitted by: Ioannis Kopsacheilis Student number: s3800253

20th_{of June 2020}

Supervisor: Dr. Pere Arqué-Castells Co-assessor: Prof. Dr. Jordi Surroca

Abstract: Markets for technology have flourished and proactive management of IP changed firms’ strategies. Regardless the importance and scale of corporations’ innovative activity, no research exists in this context. This research explores patent transactions in technology markets that are conducted by business units that belong to corporations. Specifically, this paper firstly investigates whether markets for technology increased over time; then, it examines whether corporate groups transact more often and transact more patents within their boundaries; and lastly, which is the role of technological proximity. Results indicate an increase in markets for technology, more frequent trades within group boundaries, a positive relationship between technological proximity and frequency and volume of transactions and lastly greater technological proximity between transacted parties that belong to the same corporation. The findings complement and expand existing literature on markets for technology, corporate groups, and technological proximity.

Keywords: Markets for technology; Corporate groups; Patents; Patent transactions; Technological proximity

(2)

2

1. Introduction

Extensive literature has investigated the emergence and rise of markets for technology

(Caviggioli and Ughetto, 2013). Although innovation was primarily conducted within the firm

in technology intensive industries early in the 20th_{century (Chandler, 1990), firms} significantly changed the way they manage their Intellectual Property (IP) in the last decades. Patent transactions became an important element of competition (Grindley and Teece, 1997). Markets for technology have flourished as indicated by the increase in patent transactions and patenting (Arora, Fosfuri and Gambardella, 2001; Arora and Gambardella, 2010). In this paper, markets for technology will focus solely on patents, and patent transactions will be defined as the sales of patents for money from one firm to another. Specifically, “patent trades” will refer to the frequency of trades, while “patents transacted” to the volume of patents exchanged. Patent transactions will refer to both.

Markets for technology expand firms’ options and change their strategy and competitive behavior as they can manage their IP more flexible (Arora et al., 2001; Monk, 2009). Economic success of companies lies on IP management which is as crucial as the reasearch itself (Green and Scotchmer, 1995; Teece, 1986; Thurow, 1997). Innovation activity is highly concentrated on the business sector and specifically on few large corporations (Eurostat, 2019; OECD, 2019), which are also more innovative than standalones (Belenzon and

Berkovitz, 2010). Despite the long existence and the vast amount of research regarding

technology markets and the strategic importance of IP management (Caviggioli and Ughetto,

2013; Lamoreaux and Sokoloff 1997;Teece, 1986), and despite the increasing importance of

corporations in the development of new technologies (Belenzon and Berkovitz, 2010; Eurostat, 2019; OECD, 2019), there is a lack of research on how corporations behave in the markets for technology. This study will try to fill this gap. The guiding question of this research, is:

How do corporate groups behave in markets for technology, which are the patterns of patent trades and patents transacted within and between corporate groups and which is the

role of technological proximity?

The importance of filling this gap lies on the great scale of innovative activity (Eurostat, 2019; OECD, 2019) and superior innovative performance of corporations (Belenzon and

Berkovitz, 2010). Corporations are more likely to hold and transact a great fraction of patents

(4)

4

contributes to the literatures of markets for technology and strategic management. It firstly contributes to the limited empirical measurements of the size of markets for technology (Arora et al., 2001) by using both the amounts of patent trades and patents transacted as indicators of technology markets size in a new context, that of corporate groups. Additionally, it contributes to the literature of corporate groups as it demonstrates corporations’ behavior in technology markets. Lastly, it shows that technological proximity is not only important for matching transacted parties in markets for technology (Arqué-Castells and Spulber, 2019) but also increases the intensity of patent transactions between them.

This research used a unique dataset constructed by Arqué-Castells, and Spulber (2017) that matched buyers and sellers of patents with their corporate groups by integrating disparate

strands of data from the databases of United States Patent and Trademark Office (USPTO),

Compustat, SDC Platinum and Osiris. Data registered by the USPTO about the transfer of patent rights are suitable especially for research related to IP and markets for technology

(Marco, Myers, Graham, D'Agostino and Apple, 2015; Figueroa and Serrano, 2019). This

integrated dataset allowed me to research for patterns of patent transactions between and

within corporate groups in technology markets. This research used all the patent trades, patents transacted and information regarding assignors and assignees and their corporate groups for the period 1981-2012. The analysis was conducted in three different levels; (year, corporate group, and unique transacted pair) to respond sufficiently to the hypotheses posed.

Using a quantitative analysis, conducted using the statistical software STATA, I examined whether the amount of trades and the number of patents transacted increase over time; whether corporate groups’ patent trades, patents transacted and the bundles of patents are higher within than across their boundaries; whether technological proximity is higher between transacted parties that belong to the same group and whether the number of trades and patents transacted increase with higher technological proximity. I found that both the number of trades and of patents increased over time. More trades take place within corporate groups. Technological proximity has a positive relationship with the number of transactions and is higher for transacted firms that belong to the same corporation.

(5)

5

2. Literature Review

2.1 Patents

In order for an invention to be eligible for patenting it must be novel, useful and have a

commercial application. (USPTO, 2015). Patents grant rights of temporary monopoly to the

invention in return for disclosure (Hall, Jaffe and Trajtenberg, 2001). Patents grant exclusion rather than usage rights (Ziedonis, 2004), for limited time, in a specific geographic area and

subject to a fee (USPTO, 2015)

Patents constitute an isolation mechanism that helps firms capture the value from their innovations. Their effectiveness depends on patent laws, industry, firm, and technological characteristics (James, Leiblein and Lu, 2013; Somaya 2012). Patents are costly because of filling, issuing, maintenance, identification, and legal costs (Somaya, 2003). Inventors must delegate the legal rights of patents to employers through assignments in order for the later to

obtain the legal resource (James et al., 2013). Patents are widely used as a measure of

innovation output as they are considered as a rich source of data for innovation studies (Hall

et al., 2001).

Patents can be used in several ways and the motives of their usage depend on several factors (Giuri et al., 2007). Proactive management of IP has developed significantly because of the increasing importance of technological know-how and the greater protection of IP (Grindley and Teece, 1997). Patents are used for internal exploitation for commercial purposes, licensing out for royalty payments, cross licensing, sales, blocking competitors and for bargaining power (Arora et al., 2001; Grindley and Teece 1997; Hall and Ziedonis 2001; Teece 1986; Ziedonis, 2004). Patents can remain unexploited with option value (Palomeras, 2003).

2.2 Markets for technology

Markets for technology consist of technology that is transacted. Technology can take various forms, like IP (patents), it can constitute part of a product, or it can even not be patented. Technology exchanges can take forms like sales, R&D joint ventures, licensing out, cross-licensing and contracts (Arora et al., 2001). In this paper, markets for technology will focus solely on patents, and patent transactions will be defined as the sales and buys of patents in exchange for money from one firm to another.

(6)

6

and Lin, 2006). US firms increased their revenues from licensing their technologies (Dengan, 1998) and the number of transactions (Arora et al., 2001; Marco et al., 2015; Serrano, 2010) and patenting increased significantly (Arora and Gambardella, 2010; Hall and Ham, 1999;

Kortum and Lerner, 1999).

Innovation routines evolved from internal knowledge management to the coordination of internal and external sources in tandem (Pavitt, 2002). Open innovation, referring to intentional inward and outward flows of knowledge, accelerates internal R&D and internal/external commercialization of innovations (Chesbrough; 2006). Competitive advantage depends on firms’ ability to sense and seize external opportunities, technologies and organizational structures and managing IP proactively to support technology transitions (Arora et al., 2001; Iansiti, 1997).

Chien (2010) classified the reasons of patent transactions in monetary and non-monetary ones, while Caviggioli and Ughetto (2013) categorized them on monetary, patent specific, firm specific and managerial, exogenous factors and transaction costs. Transaction costs of technological know-how are contingent on the strength of the appropriability regime, the nature of the technology and the characteristics of the markets for technology that affect firms’ propensity to participate in them (Teece, 1986a). In the presence of heightened transaction costs and in the absence of enforcing power and willingness to bear the cost of patent litigation, firms are more likely to sell their patents (Fosfuri, 2006; Monk, 2009; Teece, 1986). Additionally, firms sell their patents to generate significant cash flows that helps them face financial constraints and increase returns to R&D expenditures (Kulatilaka and Lin 2006). Sellers may engage in patent transactions if they hold superfluous patents, when they exit an industry or product domain and when they do not want to involve in licensing agreements (Monk, 2009). Moreover, firms are increasingly interested in buying patents to use them against infringers, to make them unavailable in other interested firms and avoid litigations (McDonough, 2006). Patent transactions are contingent on the size of the inventor (Giuri et al., 2007), the age, the value, and the generality of the patent and whether it has been traded before (Serrano, 2010). In addition, two important factors that predict the direction of the knowledge transactions are the technological and geographical proximity (Aldieri, 2011; Jaffe, 1986).

2.3 Corporate Groups

(7)

7

belong in corporate groups (Braunerhjelm, 1998) and the greatest percentage of R&D takes

place in multinational corporations (Fors and Svensson, 1994).

Large diversified corporations have advantages related to greater availability of capital, exploitation of economies of scope and scale, and greater spread of risks and uncertainties (Cohen and Klepper, 1996; Cohen and Levin, 1989; Duchin, 2010; Khanna and Yafeh, 2007). Access to more knowledge/R&D resources in the corporate group has a positive effect on the generation of new knowledge (Anderson and Ejermo, 2005). Corporations are more innovative than standalones due to more efficient internal capital markets and lower information asymmetries and transaction costs within groups (Belenzon and Berkovitz, 2010). Corporations exist because CHQs create synergies that prevent loses and create value. (Foss 1997).

Synergies refer to the generated super-additive value or sub-additive costs that result through the interactions in an interrelated portfolio (Tanriverdi, 2005; Tanriverdi and Venkatraman, 2005). Super-additive value or sub-additive costs are created when the combination of two components leads to greater value or lower costs respectively than the sum of the individual values or costs (Farjoun, 1998; Robins and Wiersema 1995).

Transaction cost economics focus on the existence of the firm because of cost minimization and economization through internalization of operations in response to market failures. The need of a governance structure is augmented, when investments in specialized assets are needed to prevent opportunism (Teece, 1986; Williamson, 1979; 1981). Transaction costs of technological know-how are heightened because of incomplete contracts, which enforcement is contingent on the appropriability regime, greater information asymmetries that require disclosing value to buyers, the tacit nature of the knowledge, and, lastly, because of different communication codes, cultures and capabilities of the transacted parties (Teece 1986; 1986a).

Corporate groups have virtues that allow for the avoidance or decrease of transaction costs through the internalization of transactions among their business units (Foss, 1997) According to the theory of corporate culture, organizations’ implicit contracts align incentives and regardless the unforeseen contingencies that may emerge, management will not act opportunistically as firms have plan consistency (Foss, 1999; Kreps, 1990; Von Hayek, 1937). Losses can be prevented within corporations as morally hazardous behaviour and the need for monitoring, enforcement, repetitive negotiations, and explicit contract specifications

are mitigated (Teece, 1986a). Corporate groups benefit from decreased communication costs

and direction advantages of knowledge within the corporate group. It is less costly to delegate

(8)

8

On the other hand, the positive rationale of the existence of the corporate groups is based on the resource-knowledge-based view and the creation of synergies and greater joint productivity through the coordination of knowledge and cooperation (Conner and Prahalad 1996; Foss, 1997; Porter, 1989). Corporations can be conceptualized as a repository of resources and information from which businesses units can pool (Bolton and Dewatripont, 1994; Buchanan, 1965; Osterloh et al., 2002). Corporate groups create synergies because of the parenting function of the CHQs (Porter, 1989).

CHQs practice parenting function and stimulate interactions among the business units through transfer of skills and sharing of activities (Porter, 1989). Value creation is accommodated mainly by knowledge direction and flexibility within groups and the leverage of economies of scope and learning. Knowledge direction refers to the combination of the knowledge endowments that each business unit possesses by making them available to other entities under the same corporate umbrella. This behavior allows cross-learning among business units and is facilitated by the presence of common CHQs by coordinating this activity and performing it more efficiently than if it would be done in the market (Hamel and Prahalad, 1990; Foss, 1997;1999). On the other hand, flexibility refers to the leverage of the unforeseen events that may emerge in the presence of incomplete contracts and provide space for beneficial and unexpected learning. This situation promotes experimentation and knowledge accumulation. In a market setting, the residual rights that arise through incomplete contracts and unpredicted contingencies are more difficult to be captured (Foss, 1997; 1999).

(9)

9

3. Hypotheses Development

R&D was primarily conducted internally through integration of the necessary complementary assets (Chandler, 1990) and the leverage of organizational routines (Nelson and Winter, 1982). While this way facilitated the avoidance of increased transaction costs in the exchange of technological know-how (Teece, 1986a), this dominant paradigm seems to have been altered. Extensive evidence in the literature demonstrates an increase in markets for technology over time measured with the number of patents transferred (Arora et al., 2001; Chesbrough, 2006; Marco et al., 2015; Serrano, 2010), the patenting activity (Hall and Ham, 1999; Kortum and Lerner, 1999) or the money value of the transactions (Arora and Gambardella, 2010; Kulatilaka and Lin, 2006).

Firms manage their patented knowledge proactively and engage in patent transactions horizontally with competing and vertically with non-competing firms to leverage the benefits of participating in markets for technology (Arora et al., 2001). Firms increasingly participate in technology markets both for monetary and non-monetary reasons (Caviggioli and Ughetto, 2013; Chien, 2010). Additionally, innovation routines and practices changed significantly (Pavitt, 2002) and open innovation paradigm emerged with its advantages in knowledge development and commercialization (Chesbrough, 2006a).

Innovation has been seen as a recombinant process of existing or new pieces of knowledge in different combinations (Schumpeter, 1934). R&D productivity declines and inventing is getting harder (Bloom et al., 2017; Griliches, 1994). Jones (2009) referred to this phenomenon with the term “Burden of Knowledge” and showed an increasing difficulty to generate innovations; demonstrated by the increase in the age of the initial invention, the knowledge components that must be combined, the size of the team and the specialization required to generate breakthrough inventions. These findings imply that increased specialization and cooperation is required to generate innovations and firms are less likely to possess all the required resources internally. As there is a higher need for recombination than in the past, when inventing was easier, and there is a greater need to span different technology fields, firms will increasingly transact with other firms in technology markets to get access to the needed technological know-how.

As R&D activity is highly concentrated in large corporate groups (Eurostat, 2019; OECD, 2019) and corporate groups are more innovative than standalones (Belenzon and

Berkovitz, 2010), they are more likely to generate, hold and transact most of the generated

(10)

10

corporations to patent their generated knowledge and subsequently participate in the markets for technology. Therefore, I hypothesize:

Hypothesis 1: The number of patent trades and the number of patents transacted increase over time both within and between corporate groups.

Corporate groups exist because they create synergies that mitigate losses and create greater value (Foss, 1997). Regardless these advantages, firms are not self-sufficient, and they benefit from collaborating with others (Chesbrough, 2006a). Additionally, inventing is getting harder and more knowledge from specialized firms is needed to be combined to create breakthrough innovations (Bloom et al., 2017; Griliches, 1994; Jones, 2009).Consequently, intra-group transactions cannot substitute for inter-group ones but complement each other. However, because of the properties of corporate groups, I expect them to perform more patent trades and to transact more patents internally than outside their boundaries.

On the one side, corporate groups can minimize and economize costs by internalizing costly market transactions due to market inefficiencies. Firms, in the presence of hierarchies can avoid opportunistic behaviour and decrease transaction costs (Teece, 1986a; Williamson, 1979; 1981). This effect is augmented in the case of technological know-how because of incomplete contracts, increased information asymmetries, the tacit nature of knowledge and different cultures and communication codes among the transacted parties. (Teece, 1986; 1986a). Corporate culture is a remedy against incomplete contracts and prevents from opportunistic behaviour (Kreps, 1990). Due to the parenting function of CHQs, sub-additive synergies emerge as communications, negotiations, explicit contract specifications, monitoring, enforcement, delegation of decision rights, evaluation of business units and strategy implementation are performed more efficiently (Demsetz, 1988; Foss, 1997;1999; Teece, 1986a).

(11)

11

contingencies. Corporations can capture greater amounts of residual rights of the generated knowledge during the innovation process that has not been predicted ex ante (Foss, 1997;1999). Flexibility is particularly important in the generation of new knowledge as innovation outcomes are impossible to be predicted accurately and serendipitous outcomes may be generated and be captured easier in the presence of a common CHQ (Fleming, 2001; Foss, 1997).

As corporate groups have more specific and alternative channels, this will affect their behavior in the transactions of their patented knowledge in markets for technology. Corporate groups have several advantages associated with knowledge direction and flexibility in the presence of a joint CHQ. Obviously, more opportunities for knowledge exchanges exist outside group boundaries in comparison to those within group boundaries. Nevertheless, in relative terms out of all patent trades and patents transacted by a given corporate group, I expect more of those to be conducted within its boundaries than across them. Therefore, I hypothesize:

Hypothesis 2a: Corporate groups conduct more patent trades within their boundaries relative to those conducted across them.

Hypothesis 2b: Corporate groups transact a greater number of patents within their boundaries relative to those transacted across them.

The negative rationale of the existence of CHQs implies that corporate groups have the capability to monitor and align incentives. Vertical integration of firms under the umbrella of CHQs can circumvent opportunistic behavior when transactions entail higher asset specificity and there is a need for continuous transactions (Demsetz, 1988). R&D contracts entail great uncertainties that can only be resolved sequentially with the progression of the projects and consequently renegotiations have to take place as contracts cannot entail explicitly all the terms on their outset (Nelson and Winter, 1977; Pisano, 1990). Additionally, contracts are expensive to write and enforce (Pisano, 1990). In the presence of common CHQs, several costs related to negotiations, communication, explicit contract specifications, monitoring, enforcement, evaluation, and delegation of decision rights can be mitigated significantly (Demsetz, 1988; Foss, 1997; 1999; Teece, 1986a).

(12)

12

that belong in different corporate groups are expected to try to reduce the amount of their transactions and negotiate greater mass sales of patents in each transaction. The reason why I expect this to happen is because in inter-corporate group transactions there will be incentives to mitigate the negotiation costs from repeated negotiation rounds and the costs that relate to the formation and specification of several contracts by transacting a greater number of patents in each interaction.

Hypothesis 3: Corporate groups, transact more patents per trade across than within their boundaries.

Firms’ difficulty to access external knowledge is related with the proximate or distant technology bases of the transacted parties (Jaffe, 1988). Technological proximity is defined as the overlapping technology bases between two firms that accommodate the extraction of value from the transacted technology. It is positively related to the matching probability between firms in markets for technology as the gains from intentional inflows and outflows increase in the presence of higher proximity (Arqué-Castells and Spulber, 2019).Knowledge spillovers are more likely if firms operate in similar industries (Teece 1986) or in proximate technological fields (Bloom, Schankerman and Van Reenen, 2013; Jaffe, 1986). Innovators, whose prior knowledge stock has closer technological proximity with the newly developed technology, are more likely to keep it and exploit it in-house as it is more likely to be redeployed (Figueroa and Serrano, 2019). Potential buyers who possess knowledge in technologies more proximate to the newly developed patent are more likely to acquire it (Cassiman and Ueda, 2006).

Interfirm knowledge transfer is facilitated in the more integrated organizational forms (Kogut, 1988) and in the presence of greater absorptive capacity that facilitates its effective identification, acquisition, and commercialization (Cohen and Levinthal, 1990). Knowledge is organizationally bounded, and its transfer, translation, absorption, and usage are more likely with greater technological overlap between the transacted parties (Mowery et al., 1996). Corporations have knowledge coordination advantages (Foss, 1997) and they can redeploy their patents in an alternative use through intra-group transactions, facilitated by the related knowledge bases of the diversified business units (Helfat and Eisenhardt, 2004).

Even though, firms under the same corporation, have low technological proximity and this is not the source of their higher innovativeness relative to standalones (Belenzon and Berkovitz, 2010), knowledge transfer is facilitated by the presence of hierarchies and the

more integrative organizational forms compared to market transactions (Kogut, 1988). In the

(13)

13

because of CHQs’ knowledge coordination function. On the contrary, if different business units that belong to different corporate groups are less technologically proximate, in the absence of a common CHQ, the possibility that patents of interest are identified, transacted, and exploited successfully is less likely. Therefore, technological proximity is expected to be higher between transacted pairs that are not part of the same corporation. Additionally, the probability that two firms match and transact patents increases along with technological proximity has already been tested. However, I extend this analysis on an intensive level; meaning that I am interested in examining whether the amount of patent trades and patents transacted increase with increasing technological proximity. Particularly, I expect this effect to be more profound/stronger as technological proximity increases. Therefore, I hypothesize:

Hypothesis 4a: Transacted pairs of business units that belong to different corporate groups, demonstrate higher technological proximity than transacted pairs of business units

that belong to the same corporate group.

Hypothesis 4b: The higher the technological proximity between a transacted pair of business units the more patent trades will be conducted, and the more patents will be

transacted.

4. Sample, Data Sources and Main Variables

4.1 Sample and Data Sources

This research used a unique dataset constructed by Arqué-Castells and Spulber (2017). They created it by integrating data from the USPTO Patent Assignment Dataset and Compustat. They matched “assignees” (buyers) and “assignors” (sellers) in the former, with “GVKEYS” (corporate group identifiers) in the latter. Data registered by the USPTO about the transfer of patent rights are appropriate for research related to IP and markets for technology and innovation (Marco et al., 2015).

(14)

14

groups as well. Standalones are not included in the sample and if either of the firms in a transaction was not matched with a corporate group this record was dropped.

The three datasets provided and used to test my hypothesis were “patent_transactions_bw_wn”, “documentid” and “isseudpat_assignee_gvkey_iddr”. These three datasets were combined and used in three different ways as the levels of analysis required to test my hypotheses were different. Therefore, three samples were constructed as H1 required year level aggregates while H2a, H2b and H3 corporate-group level aggregates and H4a and H4b unique assignor assignee pair level aggregates for within and between groups.

4.2 Variables construction

“Patents_transacted”: The dataset “documentid” contained the variables “rf_id” and “patentid” which constitute unique transaction and patent identifiers, respectively. This sample contained 8,715,825 unique patents and 5,534,135 transaction records. Using that file, I constructed the variable patents_transacted which is the sum of patents transacted in each transaction record.

“Patent_trades”: A variable that sums the total number of transaction records in different levels in each of the three samples constructed.

I measured “patents_transacted” and “patent_trades” in absolute terms and not relative to the number of patents issued or the total number of trades in a given year or by a given corporate group. Serrano (2010) constructed the proportion of patents traded relative to those granted. Their construction in relative terms would not change anything in my analysis and results because of the nature of the comparisons for within and between group trade differences. The transacted patents in a given year or by a given corporate group for both within and between group transactions would be divided by the same number of issued patents for that year or group.

“Between”: The file “patent_transactions_bw_wn” contained information about each transaction, regarding the “year” the transaction executed, the “IDDR” and the “GVKEY” of the “assignors” and the “assignees”. It was comprised of 110,466 observations of transactions between 1978-2013. To identify for every assignment record, whether a transaction took place within or between corporate groups, I constructed the dummy variable “Between” equal to one if the transaction was conducted between assignors and assignees that belong to distinct corporate groups and zero if the transaction was conducted between assignors and assignees that belong to the same corporate group; mentioned as “Within”.

(15)

15

variable estimates the average amount of patents included in a trade for “Between” and “Within” corporate group transactions.

“Technological_proximity”: The last file used, “isseudpat_assignee_gvkey_iddr”, contained the variables “patentid”, “IDDR”, “GVKEY” and “technological_class”. This file contained 1,362,889 observations of patents and their abovementioned information. To examine H4a and H4b, I constructed the measurement of technological proximity accordingly to Jaffe (1986). USPTO, allocates patents in one of the 426 technological classes when they are recorded. (Bloom et al., 2013). In order to determine the measure of technological activity of a firm “i”, I constructed the vector that shows the average share of patents hold in each technological class Ti=(Ti1,Ti2,…,Ti426) where Tij is the number of patents a firm i holds in technological class j. The traditional measurement of closeness according to Jaffe (1986) is TECHij=TʹiTʹj / (TiTʹi)1/2 (TjTʹj)1/2. TECHij can range from zero to one. One means a complete overlap and identical knowledge bases between two firms, while zero denotes that they are totally different and unrelated (Bloom et al., 2013).

“Multiple_trades” and “Multiple_patents”: In order to test H4b I constructed the dummy variables “multiple_trades” and “multiple_patents” which equaled zero if “patent_trades” and “patents_transacted” respectively between a unique assignee-assignor pair was equal to one, and one if these numbers were more. I constructed dummy variables because the distributions of “patent_trades” and “patents_transacted” are highly skewed (Histograms 9, 10, 11 and 12), meaning that many pairs trade only once and a few many times and that many pairs transact one patent and a few many.

4.3 Samples construction

The first steps in the construction of all three samples were the same. Firstly, I merged “documentid” with “patent_transactions_bw_wn” to integrate “patents_transacted” with the other transaction specific data. Consequently, I deleted the duplicates of transaction records, limited the period to 1981-2012 as it is proposed (Arqué-Castells and Spulber, 2017; Marco et al., 2015) and generated “Between”.

“Sample_1”: In order to test H1, I constructed “Sample_1”. After conducting the abovementioned steps, I summed “patent_trades” and “patents_transacted” by “year” and “Between”. I aggregated these variables on “year” level because I wanted to examine for yearly variations in the number of trades and transacted patents within and between corporate groups. Sample_1 contained in total 64 observations. One observation for each year during 1981-2012 for “Within” and one for “Between”, with information on the aggregate amounts of trades and patents (Table_1).

(16)

16

“patents_per_transaction” as explained above. The choice of the corporate group of the assignee over that of the assignor did not have any effect on the analysis and the result, as the same amount of trades and patents were preserved. The choice of aggregation was related to the nature of my hypothesis and the examination of differences in the behavior of corporate groups in the markets for technology regarding the number of trades and patents exchanged within and across their boundaries. Sample_2 included 2,995 observations in total with “patents_transacted”, “patent_trades” and “patents_per_transaction” for each group. From those groups, 697 transacted both within and between while the rest, 1,601, either within or between (Table_6).

“Sample_3”: To test H4a and H4b, I constructed all the unique assignee-assignor pairs that transact in my sample. For each pair I constructed the measurement of technological_proximity and merged the outcome with “patent_transactions_bw_wn”. “Sample_3” included (N=5,852) unique pairs (Table_11) and a value for each of the variables “Between”, “technological_proximity”, “patent_trades”, “patents_transacted”, “multiple_patents” and “multiple_trades” for each pair. Pair level analysis was necessary in order to be able to construct the measurement of “technological_proximity” and examine whether there is a significant difference between transactions conducted “Within” and those “Between”. Additionally, this level of analysis was required to examine the variation in the frequency and volume of transactions as the technological proximity increases between assignors and assignees.

4.4 Descriptive Statistics

This section summarizes the descriptive evidence in the three samples used to test my hypotheses.

Table_1, reports for Sample_1, a sample size N=64 for the period 1981-2012. The total patent_trades conducted in that period were 24,696 with M=717.75 per year. In total 7,655 and 17,041 trades were conducted Between and Within groups respectively. Within transactions group (N=32) was associated with numerically higher patent_trades M=532.53 (SD=501.01) compared to Between transactions group M=239.22 (SD=170.89). The total patents_transacted in that 32-year period was 329,219 with M=10,289.44 per year. From those, 187,596 patents were transacted Within and 141,666 Between. Within transactions group (N=32) was numerically greater M=5,862.35 (SD=6,254.04) than the Between transactions group (N=32) which was lower with M=4,427.06 (SD=3,164.21).

(17)

17

patents_transacted M=120.72 (SD=733.8) in comparison to the Between group M=98.31 (SD=402.48). Lastly, Within group had lower patents_per_transaction M=11.62 (SD=31.92) compared to Between group M=15.65 (SD=36.2).

Table_11 reports that Sample_3 had N=5,852 observations of unique assignee-assignor pairs that transact. The transacted pairs in my sample that belonged to the same corporate group (N=2,797) were associated with numerically higher patent_trades M=16.44 (SD=110.01), patents_transacted M=47.07 (SD=339.05) and higher technological_proximity M=.45 (SD=.38). By comparison, the transacted pairs that belonged to different corporate groups (N=3,055) were associated with lower patent_trades M=13.92 (SD=73.54), patents_transacted M=43.90 (SD=228.31) and lower technological_proximity M=.33 (SD=.3).

5. Methodology and Results

5.1 Methodology

To test H1, Sample_1 was used. Following Marco et. al. (2015), I reported changes of patent_trades and patents_transacted over time with line-graphs. Το verify whether these trends were statistically significant I regressed the continuous variables patent_trades and patents_transacted on the continuous variable year for both Within and Between group categories. These regressions used as a descriptive tool.

To test H2a, H2b and H3 Sample_2 was used while for H4a Sample_3 was used. The group samples of Within and Between for patent_trades, patents_transacted, patents_per_transaction and technological_proximity, needed to be compared to understand whether there are statistically significant differences between them. The samples of Within and Between groups were independent from each other and all four variables were interval type. To compare the means of these independent samples I used independent samples t-tests of unequal variances (Table_L) and Wilcoxon rank sum tests. Both a parametric and a non-parametric test were used in each hypothesis as the sample sizes were greater than 30 but the normality condition was not satisfied (Histograms 1, 2, 3, 4, 5, 6, 7, 8 and Table_SW) (Keller, 2012; Weinberg and Abramowitz, 2002). To verify the validity of these results, simple linear regressions were used as a descriptive tool. I regressed the continuous variables patent_trades, patents_transacted, patents_per_transaction and technological_proximity on the binary variable Between.

(18)

18

follows “it approximates how much more likely (or unlikely) it is for the outcome to be present among those with x=1 than among those with x=0” (Hosmer, Lemeshow and Sturdivant, 2013, p. 50).

5.2 Hypothesis 1: Patent trades and patents transacted over time

H1 stated that the number of trades and the number of patents transacted increase over time both Within and Between groups.

The line-graphs in Figure_1 demonstrate an upward trend over time for patent_trades both Within and Between. While in 1981, patent_trades for Within and Between were 19 and 52 respectively, over time these numbers increased significantly. This increasing trend picked for Between in 2003 when they were conducted 593 patent_trades. For Within, the highest amount of trades conducted in a single year was in 2011 and accounted for 1697. The linear regressions, Table_2 and Table_3, report that the models are highly significant (p=.000) and that 82% and 69% of the variations of patent_trades in Within and Between groups respectively is explained by year. For every additional year, the expected number of patent_trades increase by B=48.26, p=.000 for Within and by B=15.08, p=.000 for Between. These results show that the number of trades increased significantly over time both in Within and Between group categories.

The line-graphs in Figure_2 demonstrate an upward trend over time for patents_transacted both Within and Between. In 1981 patents_transacted were 780 Between and 383 Within corporate groups. This amount picked for Within in 2003 with 26972 and for Between in 2010 with 12863 patents transacted in total. Table_4 and Table_5 report that both regressions are statistically significant with p=.000; and 43% and 55% of the variations of patents_transacted are explained by year. The results show positive and significant effects of year on patents_transacted, B=435.5, p=.000 and B=249.08, p=.000, for Within and Between respectively. These results show that the number of patents transacted within and across corporate group boundaries increased over time.

The results from the analysis of hypothesis 1 demonstrate a significant increase both in patent_trades and patents_transacted over time for both Within and Between corporate groups. The figures presented and the results of the regressions demonstrate an increase in the markets for technology as corporate groups increasingly engage in transactions more frequently and with greater volume of patents over time. Hypothesis 1 is supported.

5.3 Hypotheses 2a and 2b: Patent trades and patents transacted within and

between corporate groups

(19)

19

patents is expected to be transacted within group boundaries relative to those transacted across them.

As reported by Table_7, both the independent samples t-test and the Wilcoxon rank-sum test indicated significant and similar results for patent_trades. Within transactions group was associated with higher patent_trades M=10.97 compared to the Between transactions group M=5.31. The independent samples t-test was associated with a statistically significant effect, p=.000. On the other hand, Wilcoxon rank-sum test indicated that the rank sum of Within transactions group was significantly higher, p=.006, than the rank sum of Between transactions group. The simple linear regression, Table_8, suggests that Between explained only 0.5% of the variance (R2_{=.005, p=.000). Within group transactions reported a} patent_trades score that is significantly higher B= -5.65, p=.000 than that of across group boundaries trades. This difference between the means of the two groups is identical to the result reported by the conducted independent samples t-test.

Therefore, I can reject the null hypothesis and infer that patent_trades conducted by corporate groups within their boundaries are more than the patent_trades conducted by corporate groups across their boundaries. The conducted regression, t-test and the Wilcoxon rank sum test indicated the same result congruent to H2a which is supported. These results indicate that corporate groups transact more frequently within their boundaries than across them in technology markets.

Regarding patents_transacted Table_7, reports the results of the independent samples t-test and the Wilcoxon rank sum t-test. Both t-tests indicated non-significant and similar results. Descriptive evidence demonstrated that Within group had higher patents_transacted M=120.72 than Between group M=98.31. Nevertheless, the independent samples t-test demonstrated that this difference was non-significant, p=.148>.05. Similarly, while the rank sum of patents_transacted of Within transactions group was higher than the rank sum of Between transactions group, the Wilcoxon rank-sum test showed an insignificant difference between the two groups as well, p=.245>.05. Even though, simple linear regression, Table_9, reported that patents_transacted are on average higher Within than Between (B= -22.41), this difference between the two groups was not statistically significant (p=.305>.05).This difference between the means of the two groups is identical to the result reported by the conducted independent samples t-test.

(20)

20

5.4 Hypothesis 3: Patents per transaction within and between corporate groups

Hypothesis 3 stated that corporate groups transact more patents per trade in transactions conducted across their boundaries than within them.

As reported by Table_7, Within transactions group was associated with lower patents_per_transaction M=11.62 compared to the Between transactions group M=15.65. The independent samples t-test was associated with a statistically significant effect, p=.000. On the other hand, Wilcoxon rank sum test indicated a statistically significant difference as well, p=.000, between the rank-sums of Within transactions group and Between transactions group. Table_10, of linear regression, reports that there is on average 4.03 difference between Within and Between meaning that between group transactions reported a patents_per_transaction score that is significantly higher B=4.03, p=.000 than that of within group boundaries transactions. This difference between the means of the two groups is identical to the result reported by the conducted independent samples t-test.

Thus, I can reject the null hypothesis that indicates that the mean patents_per_transaction of Within and Between group transactions are equal. The data provide strong evidence to infer that the number of patents that are transacted per trade is lower in the trades that take place within corporate group boundaries than in the trades that are conducted across their boundaries. Corporate groups transact a smaller bundle of patents in each transaction within their boundaries than across their group boundaries. All three conducted tests indicated the same result congruent to H3 which is supported.

5.5 Hypotheses 4a and 4b: Technological proximity

H4a stated that technological proximity was expected to be higher between transacted pairs that belong to different corporate groups relative to those that belong to the same corporation.

(21)

21

The data provide strong evidence to infer that technological_proximity is higher between transacted parties that belong to the same corporation. Assignee-assignor pairs that belong to the same corporation are technologically closer than assignee-assignor pairs that belong to different corporations. These finding are opposite to hypothesis H4a which is not supported.

H4 stated that more trades are going to be conducted and more patents are going to be transacted when the technological proximity increases.

Table_14 reports that technological_proximity significantly predicted the multiple_trades, B=.56, p=.000. Table_14 reports that the exponential of the coefficient B or stated differently the odds ratio equals OR= 1.757. This result indicates that the probability of more than one patent trade between a transacted assignee-assignor is higher for higher technological_proximity. Stated differently, for an increase of 1% in the measurement of the technological proximity the odds of more than one patent trades is higher by 75.7% (1.757*100-100). The overall explanatory power of the model LR Chi2=42.50.

Table_15, reports that technological_proximity significantly predicted multiple_patents, B=.21, p=.038. Table_15 reports that odds ratio equals OR=1.228. This result indicates that the probability of more than one patent to be transacted between a transacted assignee-assignor pair is higher for higher technological_proximity. Stated differently, for an increase of 1% in the measurement of the technological proximity the odds of more than one patent to be transacted is higher by 22.8%. The overall explanatory power of the model LR Chi2_=4.36.

These result are congruent with H4b. As technological proximity increases between a transacted pair more patent trades and more patents are transacted. This means that in the markets for technology higher knowledge overlap between transacted parties can stimulate more frequent trades and transactions of greater volume. H4b is supported.

6. Discussion, Conclusion, Limitations and Future Research

6.1 Discussion and Conclusion

Despite the long existence of and the vast amount of research on markets for technology (Caviggioli and Ughetto, 2013; Lamoreaux and Sokoloff 1997) and despite the increased innovation activity and performance of corporations (Belenzon and Berkovitz,

2010; OECD, 2019) no extensive research regarding patent transactions between and within

corporate groups exists. The purpose of this study was to address the following research question: How do corporate groups behave in markets for technology, which are the patterns of patent transaction within and between corporate groups and how technological proximity affects this behavior?

(22)

22

both benefits, monetary and non-monetary ones, and the need to do it because of the “Burden

of Knowledge” (Bloom et al., 2017; Caviggioli and Ughetto, 2013; Chien, 2010; Jones,

2009). This study documents that corporate groups conduct more trades related to patents and transact a greater number of patents over time both within and between group transactions. These increasing patterns are congruent with H1 and previous findings that demonstrate an expansion of markets for technology over the years (Arora et al., 2001; Arora and Gambardella, 2010; Chesbrough, 2006; Kulatilaka and Lin, 2006; Marco et al., 2015; Serrano, 2010). These findings complement and expand existing literature and demonstrate the expansion of markets for technology measured both with the number of trades and the number of patents transacted. Additionally, this increase is demonstrated in the context of corporate groups which constitute a novel contribution made possible by the unique dataset provided by Arqué-Castells and Spulber (2017).

Corporate groups exist because they create sub-additive and super-additive synergies,

meaning that they can economize transaction costs and create greater value (Foss, 1997; Porter, 1989). This ability is related to the presence of hierarchies and incomplete contracts that exploit the functions of knowledge direction and flexibility (Foss, 1997; 1999).Based on this rationale, I hypothesized that corporate groups would engage in more trades and would transact more patents within than across their group boundaries. I found that corporate groups conduct significantly more patent trades within their group boundaries relative to those conducted across them. Nevertheless, the number of the patents transacted by corporate groups within their group boundaries is not significantly higher relative to the number of patents transacted across them. These results imply that even though business units transact more frequently with business units under the same corporate umbrella than with others that are not, they do not transact a greater volume of patents with them. The result of hypothesis 3 gives a better insight in the abovementioned findings by demonstrating that corporate groups transact significantly greater bundles of patents per transaction in their inter-group trades relative to intra-group ones.

These results probably demonstrate that corporate groups deploy and transact their readily patented knowledge inside their group boundaries more often thanks to the functions of knowledge direction and flexibility. Smaller bundles of patents and greater frequency were

made possible probably because of the decreased costs related to negotiations,

(23)

23

open innovation (Chesbrough 2006; Jones 2009; Bloom et al. 2017). Corporate groups are more innovative than standalones (Belenzon and Berkovitz, 2010) probably because of the frequent and continuous deployment of readily available knowledge and the redeployment of serendipitous generated knowledge among the subsidiaries that are under the mandate of a common CHQ, but not the amount of transferred knowledge measured by patents. This evidence, shows the benefits of firms belonging to corporate groups that can facilitate recombinant innovation through frequent transactions but at the same time shows the complementarity between within and between group trades of patents that probably indirectly shows the complementarity between internal and external R&D and knowledge sourcing.

Lastly, knowledge flows and patent transactions in the markets for technology are

more likely between firms that possess overlapping technology bases (Arqué-Castells and

Spulber, 2019; Jaffe, 1986). Structural and cognitive barriers can hinder knowledge transfer (Cohen and Levinthal, 1990; Kogut, 1998; Mowery et al., 1996). In the presence of common CHQs I expected transacted pairs of firms that belong to the same corporation to demonstrate lower technological proximity than the technological proximity of transacted pairs of firms that belong to different corporations. This research demonstrated an opposite result to that predicted, and higher technological proximity among transacted firms that belong to the same corporate group. This result probably denotes that even though business units in a corporate group are diversified, they are still technologically proximate to facilitate knowledge exchanges (Helfat and Eisenhardt, 2004) and in order CHQs to exercise knowledge direction and flexibility functions successfully (Foss, 1997). To avoid strengthening competition (Caviggioli and Ughetto,2013) firms are possibly less likely to transact patents with technologically proximate firms that belong to different corporations. Lastly, this finding is possibly explained by the fact that firms are benefited by collaborating with moderately technologically proximate partners to tap into diverse knowledge that increases innovativeness (Sampson, 2007).

Finally, this study shows that technological proximity not only contributes to the

match between firms in markets for technology (Arqué-Castells and Spulber, 2019), and to

the internalization and extraction of value from it (Cohen and Levinthal, 1990) but also that it has a positive relationship with the number of trades and patents transacted. These findings show that increased knowledge overlap between two transacted parties not only facilitates identification, sharing and transfer of knowledge but also it can intensify and stimulate more frequent trades and a greater volume of patents to be transacted between them.

(24)

24

boundaries. It demonstrates that technological proximity could be a useful managerial tool to increase the frequency and volume of transactions.

6.2 Limitations

The first limitation of this study is related to the nature of the data used to test my hypotheses. The exclusion of all the standalones and the inclusion only of business units that belong to and transact with corporate groups makes the analysis biased against transactions that occur beyond group boundaries (Arqué-Castells, and Spulber, 2017). Most of the transactions occur beyond group boundaries and most of the transacted parties are standalones. Nevertheless, as R&D activity is highly concentrated in a few corporations and they are more innovative (Belenzon and Berkovitz, 2010; Eurostat, 2019; OECD; 2019); trades with other corporate groups are a good indicator of their behavior in markets for technology. Another limitation of the study is related to the measurement of technological proximity. As indicated by Bloom et al. (2013) Jaffe’s measurement captures only the spillovers that occur into the same technology field and not between different ones. This assumption of knowledge transfer depends on the different levels of aggregation of technology fields. This study used the 426 technology classes as indicated by the USPTO to get a more fine-grained classification and a more precise construction of the measurement. Another limitation is the measurements of multiple_trades and multiple_patents. Both are measured with dummy variables instead of numerical continuous variables. The reason of this choice of measurements was the highly skewed histograms. More fine-grained measurements could reveal interesting insights. Lastly, this study does not infer any causality.

6.3 Future Research

(25)

25

7. Bibliography

Aldieri, L. (2011). Technological and geographical proximity effects on knowledge spillovers: evidence from the US patent citations. Economics of Innovation and New Technology, 20(6), 597-607.

Anderson, M., & Ejermo, O. (2005). How does accessibility to knowledge sources affect the innovativeness of corporations? evidence from Sweden. The annals of regional

science, 39(4), 741-765.

Arora, A., & Fosfuri, A. (2003). Licensing the market for technology. Journal of Economic Behavior & Organization, 52(2), 277-295.

Arora, A., Fosfuri, A., & Gambardella, A. (2001). Markets for technology and their implications for corporate strategy. Industrial and corporate change, 10(2), 419-451. Arora, A., & Gambardella, A. (2010). The market for technology. In Handbook of the Economics of Innovation (Vol. 1, pp. 641-678). North-Holland.

Arqué-Castells, P., & Spulber, D. F. (2017). Matching assignees and assignors in the USPTO Patent Assignement Dataset to Compustat firms. Northwestern University.

Arqué-Castells, P., & Spulber, D. F. (2019). Firm Matching in the Market for Technology: Harnessing Creative Destruction. Northwestern Law & Econ Research Paper, (18-14). Belenzon, S., & Berkovitz, T. (2010). Innovation in business groups. Management Science, 56(3), 519-535.

Bloom, N., Jones, C. I., Van Reenen, J., & Webb, M. (2017). Are ideas getting harder to find? (No. w23782). National Bureau of Economic Research.

Bloom, N., Schankerman, M., & Van Reenen, J. (2013). Identifying technology spillovers and product market rivalry. Econometrica, 81(4), 1347-1393.

Bolton, P., & Dewatripont, M. (1994). The firm as a communication network. The Quarterly Journal of Economics, 109(4), 809-839.

Braunerhjelm, P. (1998). Varför leder inte ökade FoU-satsningar till mer högteknologisk export? Industriens utredningsinstitut (IUI).

Buchanan, J. M. (1965). An economic theory of clubs. Economica, 32(125), 1-14. Cassiman, B., & Ueda, M. (2006). Optimal project rejection and new firm start-ups. Management Science, 52(2), 262-275.

(26)

26

Chandler, A. D. (1990). Strategy and structure: Chapters in the history of the industrial enterprise (Vol. 120). MIT press.

Chesbrough, H. (2006). Emerging secondary markets for intellectual property: US and Japan comparisons. National Center for Industrial Property Information and Training (INPIT), Tokyo.

Chesbrough, H. (2006a). Open innovation: a new paradigm for understanding industrial innovation. Open innovation: Researching a new paradigm, 400, 0-19.

Chien, C. V. (2010). From arms race to marketplace: the complex patent ecosystem and its implications for the patent system. Hastings Lj, 62, 297.

Cohen, W. M., & Levin, R. C. (1989). Empirical studies of innovation and market structure. Handbook of industrial organization, 2, 1059-1107.

Cohen, W. M., & Klepper, S. (1996). Firm size and the nature of innovation within industries: the case of process and product R&D. The review of Economics and Statistics, 232-243. Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative science quarterly, 35(1), 128-152.

Conner, K. R., & Prahalad, C. K. (1996). A resource-based theory of the firm: Knowledge versus opportunism. Organization science, 7(5), 477-501.

Demsetz, H. (1988). The theory of the firm revisited. Journal of Law, Economics, & Organization, 4(1), 141-161.

Dengan, S. A. (1998), The Licensing Payoff from U.S. R&D, Journal of The Licensing Executives Society International, 33(4), 1-8.

Duchin, R. (2010). Cash holdings and corporate diversification. The Journal of Finance, 65(3), 955-992.

Eurostat. (2019). Intramural R&D expenditure (GERD) by sectors of performance. Retrieved 2020, January 10, from

http://appsso.eurostat.ec.europa.eu/nui/show.do?wai=true&dataset=rd_e_gerdtot Farjoun, M. (1998). The independent and joint effects of the skill and physical bases of relatedness in diversification. Strategic Management Journal, 19(7), 611-630.

Figueroa, N., & Serrano, C. J. (2019). Patent trading flows of small and large firms. Research Policy, 48(7), 1601-1616.

(27)

27

Fors, G., & Svensson, R. (1994). R&D in Swedish Multinational Corporations (No. 406). IUI Working Paper.

Fosfuri, A. (2006). The licensing dilemma: understanding the determinants of the rate of technology licensing. Strategic Management Journal, 27(12), 1141-1158.

Foss, N. J. (1997). On the rationales of corporate headquarters. Industrial and Corporate Change, 6(2), 313-338.

Foss, N. J. (1999). The use of knowledge in firms. Journal of Institutional and Theoretical Economics (Jite)/zeitschrift für die gesamte staatswissenschaft, 458-486.

Giuri, P., Mariani, M., Brusoni, S., Crespi, G., Francoz, D., Gambardella, A., ... & Hoisl, K. (2007). Inventors and invention processes in Europe: Results from the PatVal-EU

survey. Research policy, 36(8), 1107-1127.

Green, J. R., & Scotchmer, S. (1995). On the division of profit in sequential innovation. The RAND Journal of Economics, 20-33.

Griliches, Z. (1994). Productivity, R&D, and the Data Constraint. The American Economic Review, 84(1), 1-23.

Grindley, P. C., & Teece, D. J. (1997). Managing intellectual capital: licensing and cross-licensing in semiconductors and electronics. California management review, 39(2), 8-41. Hall, B. H., & Ham, R. M. (1999). The patent paradox revisited: determinants of patenting in the US semiconductor industry, 1980-94 (No. w7062). National Bureau of Economic

Research.

Hall, B. H., & Ziedonis, R. H. (2001). The patent paradox revisited: an empirical study of patenting in the US semiconductor industry, 1979-1995. rand Journal of Economics, 101-128. Hall, B. H., Jaffe, A. B., & Trajtenberg, M. (2001). The NBER patent citation data file: Lessons, insights and methodological tools (No. w8498). National Bureau of Economic Research.

Hamel, G., & Prahalad, C. K. (1990). The core competence of the corporation. Harvard business review, 68(3), 79-91.

Hargadon, A. (2003). How breakthroughs happen: The surprising truth about how companies innovate. Harvard Business Press.

Helfat, C. E., & Eisenhardt, K. M. (2004). Inter‐temporal economies of scope, organizational modularity, and the dynamics of diversification. Strategic Management Journal, 25(13), 1217-1232.

(28)

28

Iansiti, M. (1997). Technology integration: making critical choices in a dynamic world. Harvard Business school press.

Jaffe, A. B. (1986). Technological opportunity and spillovers of R&D: evidence from firms' patents, profits and market value.

Jaffe, A.B., 1988. Demand and supply influences in R & D intensity and productivity growth. Rev. Econ. Stat. 70 (3), 431-437.

James, S. D., Leiblein, M. J., & Lu, S. (2013). How firms capture value from their innovations. Journal of management, 39(5), 1123-1155.

Jones, B. F. (2009). The burden of knowledge and the “death of the renaissance man”: Is innovation getting harder?. The Review of Economic Studies, 76(1), 283-317.

Katz, M. L., & Shapiro, C. (1985). On the licensing of innovations. The RAND Journal of Economics, 504-520.

Keller, Gerald. Managerial Statistics. International Edition. Cengage Learning, 2012. Khanna, T., & Yafeh, Y. (2007). Business groups in emerging markets: Paragons or parasites? Journal of Economic literature, 45(2), 331-372.

Kogut, B. (1988). Joint ventures: Theoretical and empirical perspectives. Strategic management journal, 9(4), 319-332.

Kortum, S., & Lerner, J. (1999). What is behind the recent surge in patenting? Research

policy, 28(1), 1-22.

Kreps, D. M. (1990). Corporate culture and economic theory. Perspectives on positive political economy, 90, 109-10.

Kulatilaka, N., & Lin, L. (2006). Impact of licensing on investment and financing of technology development. Management Science, 52(12), 1824-1837.

Lamoreaux, N. R., & Sokoloff, K. L. (1997). Location and technological change in the American glass industry during the late nineteenth and early twentieth centuries (No. w5938). National Bureau of Economic Research.

Marco, A. C., Myers, A., Graham, S. J., D'Agostino, P., & Apple, K. (2015). The USPTO patent assignment dataset: Descriptions and analysis.

McDonough III, J. F. (2006). The myth of the patent troll: an alternative view of the function of patent dealers in an idea economy. Emory LJ, 56, 189-228.

(29)

29

Monk, A. H. (2009). The emerging market for intellectual property: drivers, restrainers, and implications. Journal of Economic Geography, 9(4), 469-491.

Mowery, D. C., Oxley, J. E., & Silverman, B. S. (1996). Strategic alliances and interfirm knowledge transfer. Strategic management journal, 17(S2), 77-91.

Nelson, R. R., & Winter, S. G. (1977). In search of a useful theory of innovation.

In Innovation, economic change and technology policies (pp. 215-245). Birkhäuser, Basel. Nelson, R.R. and S.G. Winter, 1982, An evolutionary theory of economic change (Harvard University Press, Cambridge, MA).

OECD. (2019). Business R&D structure and dynamics and the impact of public support for business R&D. Retrieved 2020, January 10, from http://www.oecd.org/sti/microberd.htm Osterloh, M., Frost, J., & Frey, B. S. (2002). The dynamics of motivation in new

organizational forms. International Journal of the Economics of Business, 9(1), 61-77. Pavitt, K. (2002). Innovating routines in the business firm: what corporate tasks should they be accomplishing? Industrial and Corporate Change, 11(1), 117-133.

Palomeras, N. (2003). Sleeping patents: any reason to wake up. IESE Research Papers No D/506.

Pisano, G. P. (1990). The R&D boundaries of the firm: an empirical analysis. Administrative science quarterly, 153-176.

Robins, J., & Wiersema, M. F. (1995). A resource‐based approach to the multibusiness firm: Empirical analysis of portfolio interrelationships and corporate financial

performance. Strategic management journal, 16(4), 277-299.

Porter, M. E. (1989). From competitive advantage to corporate strategy. In Readings in strategic management (p. 234-255). Palgrave, London.

Sampson, R. C. (2007). R&D alliances and firm performance: The impact of technological diversity and alliance organization on innovation. Academy of management journal, 50(2), 364-386.

Serrano, C. J. (2010). The dynamics of the transfer and renewal of patents. The RAND Journal of Economics, 41(4), 686-708.

Schumpeter, J. (1934). The theory of economic development Harvard University Press. Cambridge, MA.

(30)

30

Somaya, D. (2012). Patent strategy and management: An integrative review and research agenda. Journal of management, 38(4), 1084-1114.

Tanriverdi, H. (2005). Information technology relatedness, knowledge management capability, and performance of multibusiness firms. MIS quarterly, 311-334.

Tanriverdi, H., & Venkatraman, N. (2005). Knowledge relatedness and the performance of multibusiness firms. Strategic management journal, 26(2), 97-119.

Teece, D. J. (1986). Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research policy, 15(6), 285-305.

Teece, D. J. (1986a). Transactions cost economics and the multinational enterprise An Assessment. Journal of Economic Behavior & Organization, 7(1), 21-45.

Thurow, L. C. (1997). Needed: a new system of intellectual property rights. Harvard Business Review, 75, 94-107.

USPTO, (2015), General Information Concerning Patents: Assignments and Licenses. Retrieved March 2020, from https://www.uspto.gov/patents-getting-started/general-information-concerning-patents#heading-27

Von Hayek, F. A. (1937). Economics and knowledge. Economica, 4(13), 33-54.

Weinberg, S. L., & Abramowitz, S. K. (2002). Data analysis for the behavioral sciences

using SPSS. Cambridge University Press.

Williamson, O. E. (1979). Transaction-cost economics: the governance of contractual relations. The journal of Law and Economics, 22(2), 233-261.

Williamson, O. E. (1981). The modern corporation: origins, evolution, attributes. Journal of economic literature, 19(4), 1537-1568.

(31)

31

8. Appendix

Table_1

Between Within Total

Variables patent_trades patents_transacted patent_trades patents_transacted patent_trades patents_transacted

mean 239.22 4,427.06 532.53 5,862.38 717.75 10,289.44 sum 7,655 141,666 17,041 187,596 24696 329,262 min 33 780 19 349 66 1,163 max 593 12,863 1,697 26,972 1,968 29,936 st. dev. 170.89 3,164.21 501.01 6,254.04 6,554.90 7,926.48 variance 29,201.8 10,012,221 251,012 39,113,036 429,667 62,829,151 N 32 32 32 32 32 32

*Table_1 reports year level descriptive statistics for Sample_1

Table_2

Variables Coefficient SEM

year 48.26*** 4.18 Intercept -95,815.79*** 8,340.24 N 32 Model F statistics 133.46*** Model R2 _0.82 Adjusted R2 _0.81

Table_2 reports the results of linear regression for Within corporate group transactions

Dependent Variable: patent_trades

Independent Variable: year

(32)

32

Table_3

Table_3 reports the results of linear regression for Between corporate group transactions

Dependent Variable: patent_trades

***p<.001

Table_4

Table_4 reports the results of linear regression for Within corporate group transactions

Dependent Variable: patents_transacted

Markets for Technology and Corporate Groups: Patent Transactions Within and Between Corporations

Master Thesis