Changes in the business model of Non-Producing Entities
in ICT patent markets
Master Thesis
Faculty of Economics and business MSc Business Administration
Business & ICT
2
Preface ‘Intellectual Property is the oil of the 21st
century’ – Mark Getty
In this thesis you will find a research on a new type of company: the Non-Producing Entity (NPE). These type of companies do not produce products, but only develop or buy technologies and license these technologies to manufacturers. For a long time, these companies were seen as ‘patent trolls’ that impede innovation. Patent reform laws are drafted to counteract this behavior and subsequently is causing change in the business models of these companies. This research will examine these changes in two NPEs: InterDigital and Rambus.
This thesis is the final assignment in my master of Business & ICT at the University of Groningen. I enjoyed this master very much, and hope to see this master returned.
3
Abstract
4
Table of contents
1. Introduction 6
1.1 Background 6
1.2 Aim of the research and the research question 7
1.3 Method of research 8
1.4 Reading guide 8
2. Theory 9
2.1 How has the technological market changed to create opportunities for NPEs 10
2.1.1 Increasing of technological complexity and convergence and the fragmented market 11
2.1.2 Inefficient patent marketplace 13
2.1.3 Emergence of NPEs 16
2.2 Literature on business models of NPE 17
2.2.1 Different views in literature on NPEs 19
2.2.2 Aspects and strategies of NPEs 20
2.3 Conceptual model on the changing business model of NPE 24
2.3.1 Reasons for changes in patent laws 24
2.3.2 Changes in business models of NPEs 26
3. Methodology 30 3.1 Data collection 30 3.1.1 Time period 30 3.1.2 Collected articles 30 3.2 Data analysis 31 4. Results 35 4.1 Company introduction 35 4.1.1 Rambus 35 4.1.2 InterDigital 35
4.2 Results of counted instances in news articles 36
5
4.2.2 Interactions between NPEs and other NPEs 43
4.2.3 Type of patents 47
5. Discussion 50
6. Conclusion 52
7. References 55
Appendix A – Literary review tables 59
Appendix B – article data 69
6
1. Introduction
1.1
Background
In the past years there has been a significant increase in technological complexity and technological convergence. This is changing innovation in technological markets. Firms are restricted in the knowledge to develop extremely specialized technology in house and turn to external sources for extremely specialized technology (Yanagisawa & Guellec, 2009). This has led to the emergence of specialized knowledge companies which are referred to as Pure-Intellectual Property (IP) firms, or Non-Producing Entities (NPE).
NPEs hardly produce physical products; they primarily develop or trade knowledge and protect this knowledge in patents. Since the start of this development, these companies have had a negative reputation, due to the methods some of these companies have used (McDonough, 2006). These companies would buy cheap, low-quality patents from bankruptcy sales and actively look for companies that could possibly be infringing this patent (these companies would often not even know they were infringing a patent) (Pohlmann & Opitz, 2010). Another method that some patent companies would use is ‘patent ambush’ behavior (Hemphill, 2005). A patent ambush is when a company ‘hides’ their patent and waits until sunk costs have been made to reveal their patent (Hemphill, 2005). One of the first companies to employ this tactic is Rambus, a developer of RAM technologies (used in memory chips) (Pohlmann & Opitz, 2010). Rambus became a member of a Standard Setting Organization (SSO), an organization that brings different parties together that could benefit from a universal standard (Naughton & Wolfram, 2004). Companies that participate in SSO’s could for example be manufacturers of technology, content providers or other companies. Rambus joined JEDEC, an SSO that wanted to establish a universal standard for SD RAM and DDR RAM (Alban, 2004). However, Rambus left the SSO before the standard was completed (Dunvalley & Schallop, 2007). When the standard was eventually completed and the companies producing RAM had incurred sunk costs by buying machinery and other expenses, Rambus came forward with their patent and started suing companies for infringing their patent (Naughton & Wolfram, 2004). These actions led to numerous lawsuits which Rambus won1 and resulted in licensing deals2. This type of behavior, by which
money is made in an unethical manner (hiding patents, using low quality patents, etc.) became known as patent trolling (Geradin, Layne-Farrar & Padilla 2011). Patent trolls were defined as companies that acquire patents without the intention of actually using it to produce a product (McDonough, 2006). This meant that
1, FTC File No. 011 0017, Docket No. 9302; In the Matter of Rambus, Inc., February 17th 2004
2 The Associated Press State & Local Wire, Rambus settles suit with Hitachi, shares soar in after hours trading,
7 most companies that solely produced knowledge, were branded as patent trolls, even if they did not have bad intentions (McDonough, 2006). In 2003, the Federal Trade Commission used the term ‘Non-Producing Entity’ to refer to patent trolls in their report: To Promote Innovation: The Proper Balance of Competition (FTC, 2003). This led to the view that all NPEs are patent trolls, regardless of motive (Gregory, 2006).
However, in recent years, a second, more positive perspective on NPEs can be recognized in the literature (Geradin, Layne-Farrar & Padilla, 2011). This view holds that intermediaries (NPEs) facilitate the market (McDonough, 2006). A more liquid marketplace is achieved by bringing together the different parties. This makes it easier for patent sellers to monetize their patents and decreases search costs for patent buyers. The view of the NPE model has changed from a solely negative view to a view that has positive aspects. Bessen, Ford & Meurer (2012) describe how NPEs claim they have changed their focus from litigation to participation. The companies develop their own knowledge and offer licensing deals to the markets.
At this moment, in the literature there is no clear explanation on what these possible changes entail and how they have originated, creating a gap in the literature on the changes in business models of NPEs. It is not clear if and/or how the business models have changed. The true intentions of the NPEs are unclear: it is not known how these companies aim to achieve profit maximization. This gap in the knowledge could lead to new types of unethical behavior. This problem has occurred before, as can be seen with the Rambus case.
1.2
Aim of the research and the research question
The problem in the literature is that there is a gap in knowledge on (changes in) business models of NPEs that are caused by changes in the technological market. This could lead to new types of unethical behavior. As the firms that are researched in this thesis are engaged in the ICT patent market, the aim of this research is:
To improve our understanding of developments of ICT patent markets and their influence on business models of NPEs
The main research question of this research is:
What are the differences in business models of Non-Producing Entity (NPE) firms over different time frames in ICT patent markets?
8 1) How has the technological market changed to create opportunities for NPEs?
2) What is the business model of NPEs according to literature? 3) What changes on NPEs are envisioned in literature?
On basis of the literature a theory-driven business model is created. This model is used to answer the last sub-question:
4) Are these changes recognizable in NPEs today?
1.3
Method of research
The first three sub-questions are answered by reviewing literature. The fourth sub question is answered with empirical research. The empirical research consists of two case studies of NPE’s: Rambus and InterDigital. Content analysis of news articles is done on two periods; from 2006-2009 and from 2010 – 2013. Choices for these timeframes are discussed in the methodology section. The input for the content analysis is the LexisNexis database. In these articles instances are counted with the aid of a theoretical model, constructed in sub-question three.
1.4
Reading guide
9
2. Theory
In this section the literature is presented. The questions that are central to the literature research are: (1) How has the technological market changed to create opportunities for Non-Producing Entities (NPEs)? (2) What is the business model of Non-Producing Entities according to literature? and (3) What changes are envisioned in the literature on Non-Producing Entities?
In section 2.1 an overview of the literature on the changing of the technological market is presented, as well as the question how NPEs had an opportunity to enter the inefficient technological market.
Section 2.2 is an overview of research on the business models of NPEs, in which the different views from the literature are presented. Section 2.3 describes research on how the literature predicts the change in the business models of NPEs to achieve a better alignment with the market. Predictions by different authors are collected in a theory-driven model that is used in the empirical research.
Literature research has been conducted to collect literature on the topic of this research. Different databases, like Ebscohost, ScienceDirect, Business source premier and Academic Source Premier have been consulted. The articles that are used are peer-reviewed articles from quality journals. For the first section (2.1) the databases were searched from 1990 until 2013. Different search terms that were used were combinations of: Non-Producing Entity, Non-Practicing Entity, broker, intermediary, Intellectual Property, Patent, patent market, patent troll, technological market, technological complexity, technological change and technological convergence. For the first section, this yielded 153 possibly useful papers of which 13 were selected.
For the second section (2.2), the databases were searched from 2000 until 2013. The following keywords (or combinations of keywords) were used: Non-Producing Entity, Non-Practicing Entity, patent troll, patent shark, patent, standards, business model, strategy, litigation, intermediaries and innovation. This yielded 347 possibly useful papers of which 17 were selected.
10
2.1
How has the technological market changed to create opportunities
for NPEs
In this section the literature is presented that answers the question: How has the technological market developed to facilitate Non-Producing Entities? Due to multiple factors, the NPE business model started to take rise in the second half of the nineties These factors are: the increase of technological complexity and technological convergence (Yanagisawa & Guellec, 2009), the inefficient marketplace (Gans & Stern, 2010; Hagiu & Yoffie, 2013) and fragmented ownership of Intellectual Property (IP) (Hagiu & Yoffie, 2013). The most important literature based on citations is collected in the following papers (full literary-review tables can be found in Appendix A):
Table 1: Overview of literature on changing markets
Year Author Topic Method Findings ID
1994 Arora, & Gambardella
Changing technology of technological change
Literature study and research on how cheap computational power changes industries
‘High tech’ industries have shown the greatest extent of specialization, and an upsurge of network-like arrangements for innovation.
25
2003 Chesbrough, H. W. Open innovation Literature study and empirical study with data of multiple companies (Dupont, AT&T, IBM etc.)
Knowledge has become widespread and these ideas must be used or else they will be lost. These factors create a new logic of open innovation that embraces external ideas and knowledge in conjunction with internal R&D. This change offers new ways to create value.
27
2003 Gans & Stern Startup
commercialization strategy and the implications of these drivers for industrial dynamics
Literature study and empirical study with data on multiple pharmaceutical companies (Pfizer, GSK etc.)
the precise structure and functioning of the market for ideas will depend on subtle factors, including the incentives for reputation-building on the part of established firms, the existence of institutions and brokers that facilitate trade, and the ability to secure exclusive agreements while simultaneously threatening to cooperate with third-parties
28
2004 Ziedonis Fragmented
technology markets
Literature study and empirical data of 67 U.S. semiconductor firms
firms acquire patents more aggressively than otherwise predicted when markets for technological inputs are highly fragmented.
30
2009 Benassi & Di Minin technological environment and patent intermediaries
Case study of 15 NPE companies (i.e. Intellectual Ventures, SRI etc.)
Change in technological and judicial environment aids patent brokers and intermediaries 15 2009 Yanagisawa & Guellec Technological convergence is changing the marketplace
Literary review and empirical analysis of data from trade commissions like FTC, JPO, EC
the activities of IP specialist firms will become more crucial in order to facilitate innovation
3
2010 Gans & Stern Market design in markets for technology
Framework on basis of theory on markets for technology (MfT) based on prior research in the MfT literature
Technologcal markets have characteristics that imply the market is inefficient: Idea
complementarity, user reproducibility and value rivalry
11
2.1.1 Increasing of technological complexity and convergence and the fragmented market
According to Moore’s law, the number of transistors on integrated circuits would double every two years for at least ten years (Moore, 1965). This period turned out to be a lot longer, as it is still in effect to this day. The adding of transistors to chips led to faster chips, which led to cheap computational power. The abundance of cheap computational power has changed technological innovation. Technological innovation was usually the result of research purely based on trial-and-error. This changed with the introduction of cheap computational power, which made it possible to test hypothesis with sophisticated instruments and to simulate processes. This led to more general and abstract knowledge, resulting in more relevant information that is applicable in universal categories. Information became less dependent on context and place. In combination with the development of communications technology, this led to a decrease of cost in intra-firm transactions of technologies. This change is making the production process of new technologies more divisible and thus encourages the division of labor in innovation. Firms will specialize in the stages of the innovation process in which they have a comparative advantage. Specialization will occur mostly in the high tech industries, and lead to a higher technological complexity of products (Arora & Gambardella, 1994). Higher complexity of goods has caused different industries to sharing the same technological bases. This is known as the process of convergence (Rosenberg, 1976). In the consumer goods market the higher complexity of goods and technological convergence is clearly seen. If we take the mobile phone as an example, the change is evident: it started with just a mobile phone to call other people: nowadays it is a phone, music player, camera, web browser, navigation system and has thousands of other applications. Phone makers need access to the knowledge of LCD screens, 3G technology, camera technology, software and other technology. On the other hand, all these technologies are not limited to the phone industry. LCD screens are used in televisions, computer screens, navigation systems, etc. and web browsers are competing not only on computers, but also on telephones, tablets, netbooks and laptops. These are examples of technological bases that are shared among industries (Yanagisawa & Guellec, 2009). The consequence of technological convergence and technological complexity, is that the same technology and intellectual property is spread over different firms and industries.
12 invention has led to a high degree of fragmentation of Intellectual Property Rights (IPR) in the technological market. In this semiconductor market, even the largest firms do not have more than 30% of the relevant patents (Hagiu & Yoffie, 2013). This results in a market in which most products are based on patents owned by different companies. An example of the fragmentation of the technological market is the diffusion of the 4G/LTE ‘essential’ telecom patents:
Company Percentage Worth
LG 23% $7.9 billion
Qualcomm 21% $7.3 billion
Motorola Mobility 9% $3.3 billion
InterDigital 9% $3.3 billion Nokia 9% $3.1 billion Samsung 9% $3.1 billion ZTE 6% $2.1 billion Nortel 4% $1.35 billion Ericsson 2% $600 million
Research In Motion 1% $400 million
Huawei 1% $200 million
Freescale 1% $200 million
NEC 1% $200 million
Table 2 - diffusion 4G/LTE essential patents (Forbes.com, 20113)
As every single patent-owner has the power to stop a company using their patent (so-called ‘hold-up power’), fragmentation of patents delays the use of innovations (Kultti & Takalo, 2008). A graphical representation of this development can be seen in figure 1.
3 http://www.forbes.com/sites/ericsavitz/2011/09/21/research-in-motion-patents-worth-just-2-5b-analyst-says/
13
2.1.2 Inefficient patent marketplace
In this research, the intellectual property which is discussed, is dominated by patents. According to Brandstetter, Fisman & Foley (2006), over 88% of all licensing fees is generated by licensing patents. In this paper the intellectual property marketplace is the patent marketplace.
The Intellectual Property marketplace is inefficient (Hagiu & Yoffie, 2013). This means that the market does not function in the way it is intended to work. A marketplace emerges when multiple institutions want to trade goods (for money most of the time). According to the Efficient Marketplace Hypothesis (EMH) (Fama, 1970) a market can be a weak, semi-strong or strong efficient marketplace. According to the EMH, in a perfect (efficient) market it is not possible to ‘play’ the market, as every institution has access to the same information. The opposite holds for the weak-form efficient market, where information is not widely accessible and certain parties can take advantage of this information asymmetry. The patent marketplace is an example of a weak-form market. When a marketplace is inefficient, the transfer of goods is inefficient. In the patent marketplace this manifests itself in different problems: high transaction costs, skewed bargaining positions and holdup problems (Lemley, 1997).
In the patent marketplace, it is not possible for an idea seller to play potential buyers against each other, or the other way around. This is because there is no tangible market for patents, where the buyers and sellers can meet and way up different patents or different buyers or sellers. This is known as a bilateral monopoly (Hagiu & Yoffie, 2013). This means that there is elevated bargaining power for either buyers or sellers in the patent market. Gans & Stern (2010) researched why the marketplaces for Intellectual Property, and especially the markets for technology, are inefficient. They found multiple characteristics that lead to an inefficient market. The research of Gans & Sterns (2010) is a combination of Roth’s three basic principles of market design (2007) and three characteristics of ideas that are studied by Gans & Stern (2010).
Roth’s principles of market design are (Roth, 2007):
Market thickness: a market is thick when buyers and sellers have the opportunity to buy or sell their technology to multiple buyers or sellers
14 Market safety: a market is safe if agents do not have incentives for misrepresentation or strategic action that undermine the ability of others to evaluate potential trades (revealing preferences). It must be safe to reveal your preference without risking that other parties may use this information as a strategy to undermine your bid
Characteristics of ideas that influence the formation of an efficient market (Gans & Stern, 2010) are based on literature on Markets for Technology (MfT) combined with the principles of Roth (2007). The three characteristics are:
Idea complementarity: ideas are nearly always only valuable in combination with other ideas; an idea rarely has value on its own. However, in combination with other ideas value is created. It is difficult to value how much each separate idea contributes to the value
Value rivalry: the willingness to pay for ideas can decline with the level of diffusion of that idea. The value of an idea will be less if more people have access to that idea. To give an example of high value rivalry: if different institutions have access to the same financial algorithm to buy stock for a portfolio, the value of this algorithm will go down. Each extra institution that has access to this idea will make it less valuable, as there is less to gain with this algorithm. Art is an example of this: if multiple people look at art, the value (the enjoyability) will not decrease for other people User reproducibility: as an owner of an idea, it is very difficult to appropriate an idea’s full value.
However, showing an idea to a potential customer could be dangerous, as they could copy the idea and produce it at low cost, or sell the idea to other customers (Gans & Stern, 2010)
How these characteristics influence the efficient patent marketplace according to Gans & Stern (2010):
Low Market thickness is caused by idea complementarity: when the value of an idea depends on multiple patents, the potential buyers do not have the incentive to offer a high price for an individual patent. The potential buyers cannot easily weigh different offers, as the technology is scattered among firms and even industries
Market congestion is caused by value rivalry: in a bilateral negotiation the valuation of the seller’s idea is based on mutual secrecy (as explained above: the value of the idea will decrease when multiple people have access to the same idea). On the other hand, the seller’s bargaining power arises from playing different buyers of against each other. The disclosure of the idea to all possible buyers will lower the valuation of the idea and thus sellers are limited in considering multiple offers for a single idea
15 exploitation of the idea. The owner can only sell the idea to one party and subsequently loses all power over the idea
Summarizing Gans & Stern (2010), ideas have characteristics that affect the bargaining power of buyers and sellers that, in their turn, affect the efficiency of the market.
Hagiu & Yoffie (2013) have applied the research of Gans & Stern (2010) to the patent market and have found the following reasons why the patent market is inefficient and illiquid:
1) Patents are very difficult to value:
a. The characteristics that a product must have to be patented are novel, non-obvious and useful. A patent is thus, in definition, unique. This means that there are not any other products that can be compared to a certain patent. Without comparison, valuation is very difficult.
b. Patents are subject to royalty stacking, where a single patent builds on numerous other patents (idea complementarity).
2) Both buyers and sellers face high search costs. Patent sellers are not sure of the full potential of their patent and/or current users/infringers (idea reproducibility). Patent buyers have to research how this potential patent ‘builds’ on other patent to avoid ‘royalty stacking’ (where royalty fees have to be paid for all overlapping patents), or a patent thicket (a patent package where different patent holders combine similar patents). This is very much so the case in fast changing technological markets. These patent thickets usually have multiple owners and increase the holdup risk, the risk that a patent owner will stop a company from commercializing a product due to possible infringement (making unauthorized use) of the patent (Shapiro, 2001). This creates a difficult bargaining position for a buyer.
3) The protection of patents is not perfect. This attracts a lot of litigation and transaction costs. With nearly 50% of the patents being invalidated in court, patent buyers and sellers become cautious and rather settle with the patent seller or buyer, than start a lawsuit. 1,5% of all patents are litigated, of that 1,5%, 95% end in settlement rather than a verdict. A settlement will lead to higher joint profits for both parties because of two reasons: a settlement will prevent the possibility that the patent will be invalidated (and the competition will flood the market) and it avoids litigation costs (Lemley & Shapiro, 2005; Hagio & Yoffie, 2013).
16 hold-up problems, so the value of single patents or patent thickets will be very difficult to determine. This creates high transaction costs and inefficient transaction of goods.
2.1.3 Emergence of NPEs
The inefficiencies of the market are problematic for small companies and individual inventors. Small companies and individuals, who hold the majority of the patents, are the main victims of the diminishing bargaining power in the market. According to Hagiu, Yoffie and Wagonfeld (2011) individual inventors and small companies own 60% of the patents. The remaining 40% belongs to the big companies. However, 99% of the revenue of these patents is for the big companies. The problem with monetizing these patents has led to patent intermediaries.
17 buyer and seller can easily sell patents. Also, they are brought together more easily. However, according to their empirical research (Benassi et al. 2009), these theories are not comprehensive. They add their own theory to the mix: patent intermediaries actually add an institutional setting to buyers and sellers, but at the same time promote promising IP, in which the intermediary itself has invested. So apart from just bringing buyers and sellers together, they also buy into IP with potential and promote this IP to their customers. They know how potentially valuable a specific patent could be, and they are in the perfect position to promote it (Benassi et al. 2009).
In this section the reason why the market for patents is not efficient is explained. Characteristics of patents in MfT cause impeded bargaining power for buyers and sellers. Furthermore, the reason why patent intermediaries have emerged on the market is explained. They lower the search costs and provide an institutional setting for buyers and sellers. By lowering the transaction costs, the market is made more efficient, and so transferring goods (tangible or not) from one party to another, is executed better and more cheaply. Being in this position between buyers and sellers also offers a great opportunity for patent intermediaries to invest in IP and promote it.
2.2
Literature on business models of NPE
In this section the following question is addressed: What is the business model of Non-Producing Entities according to literature? This is achieved by looking at the definition of NPE and looking at the different views of NPEs in the literature. Also, different aspects and strategies of NPEs are discussed.
The definition of a business model according to Seddon & Lewis (2003) is the following: “A business model is an abstract representation of some aspects of a firm’s strategy; it outlines the essential details one needs to know, to understand how a firm can successfully deliver value to its customers.” The business model thus focuses on how value is made, and how it can transfer this value to customers.
18 suing large companies for infringement, companies that generate ideas for the purpose of patenting and licensing and companies that focus solely on buying patents and enforcing the patent rights (actively looking for infringers) without having any knowledge of the underlying technology (McDonough, 2006).
In the literature there are roughly two streams of thought on NPEs. One stream sees NPEs only as trolls who cause distress and impede innovation. Other authors view NPEs as the opposite: helping small firms and individual inventors to monetize their inventions by making patents more liquid. The following table gives an overview of the key articles, based on citations, on the NPE business model (full literary review tables can be found in Appendix A):
Table 3: Overview of key literature on the NPE business model:
Year Author Topic Method Findings ID
2006 McDonough Alternative view NPEs
Literary review and empirical analysis of lawsuits regarding NPEs
NPEs make the patent market more efficient
13
2008 Bessen & Meurer Failure of patent market
Survey of empirical evidence, research papers
Laws should change to promote patent quality
34
2010 Reitzig, Henkel & Schneider
Assault strategies of NPEs
Analysis of historic data (extensive World Wide Web-based archival research identifying 44 NPEs)
Three different litigation strategies (Filing for Injunction, damage awards or creating switching costs)
9
2010 Pohlmann & Opitz Typology of patent trolls
Case study on 10 companies including EpicRealm, Amazon, EBay etc.
Typology of different NPEs based on patent quality and the activities concerning the patent
5
2012 Bessen, Ford & Meurer
Cost of NPEs Analysis of stock market event studies around patent lawsuit filings
NPEs cost society money, NPEs focus on technological markets
17
2012 Fischer & Henkel Patent acquisitions of NPEs
Analysis of data (a unique dataset of 392 U.S. patents acquired by known NPEs between 1997 and 2006 and a control group of 784 patents)
Patents of NPEs have unique characteristics: NPE acquired patents being, on average, significantly more difficult to substitute and more likely to be upheld in court.
10
2013 Hagiu & Yoffie Different patent intermediaries
Analysis of data on NPEs from patentfreedom.com
In the future, there will only be a few super aggregators left
21
NPE’s are one form of the patent intermediaries that emerged in the 1990s. They soon became known as ‘patent extortionists’ and later as ‘patent trolls’. The term patent troll was popularized by Peter Detkin4, a
lawyer at Intel, who had called Techsearch (a company that Intel was engaged in a law suit with) a patent troll (Sandburg, 2001). The term patent troll was used to refer to every company that held patents they were not practicing. This negative title was earned due to the fact that the business model of a lot of the NPEs during the 1990s was not completely ethical. NPE is a relatively new term; it has been introduced to substitute the derogative term patent troll. In most of the literature, however, the term ‘patent troll’ is still used, as it is better recognized. For clarity, in this thesis, the term patent troll will be substituted with NPE.
4 Interesting fact: Peter Detkin is currently a managing partner for Intellectual Ventures, known as the biggest and
19 The NPE model was based on the exploitation of the ‘loopholes’ in the market (burden of proof lies with alleged infringer, being infringed yields more than issuing licenses etc.). NPEs seek supra-normal returns on patents they own (Reitzig, Henkel & Schneider, 2010). They are usually small firms who are not interested in the knowledge behind the patent, only in the patent itself. They sue ‘infringers’ for royalties. These infringers usually have no idea that they were infringing the patent due to the complexity of the product and the complexity and vagueness of the patent. In the fragmented technological market, sometimes several hundred patents can protect one product. It is very easy to overlook a patent, and when sunk costs are created (costs that cannot be recovered), the patent holder will strike and sue the company for infringement. For these companies, the cost for stopping production or changing to another technology is usually too high, and the same applies for litigation costs. Companies often choose to just pay off the NPE, consequently preventing the possible loss of a court case in which the company could go bankrupt. The reason why the NPE does not offer the license for the technology beforehand, is due to the revenue model of these companies: an infringed patent will generate more income than licensing the technology beforehand (Reitzig, Henkel & Heath, 2007). This incentive to ‘be infringed’ has created ‘patent ambush’ behavior. A patent ambush strategy is a strategy in which the patent holder sees an opportunity to ‘insert’ the patent into the standard that is being developed in standard settings procedures of new technological standards. When the standard is adopted and huge sunk costs are made for this standard, the patent holder can demand royalties or licensing costs for the patent they are using, or according to the patent holder, infringing (Hemphill, 2005).
2.2.1 Different views in literature on NPEs
20 Another argument that aids the negative reputation of NPEs is that they increase the costs for development (when using the patent) with higher than usual royalties and higher search costs for the developer (Yang, 2012). The original view on NPEs is predominantly negative, where NPEs are viewed as a disruptive force and will focus their business model around litigation.
The other view in the literature sees potential in NPEs: patent dealers function as market intermediaries, which results in a more efficient market place with more liquid patents. According to some authors, the change of the market that is attributed to NPEs is just a signal of progression and market evolution (McDonough, 2006). Others say that patents held by NPEs improve the competitiveness of the market by helping specialists enter the market (Fischer & Henkel, 2012). This entails specialized patents and thus higher quality patents. Patents held by NPEs can also improve innovation, as rival firms need to innovate in order to keep a competitive advantage (Geradin, Layne-Farrar & Padilla, 2012).
2.2.2 Aspects and strategies of NPEs
In this section, aspects that differentiate NPEs from other IP companies will be addressed firstly. Secondly, different strategies that are described in literature are explained, as these strategies form an integral part of the business model.
Bessen, Ford & Meurer (2012) have studied the literature on NPEs and found that NPEs have the following distinctive aspects in litigation that is not found in specializing operating firms in, for example, the chemical industries:
1) Scale of the litigation: NPEs account for 17% of high-tech patent lawsuits. 2) Many lawsuits have multiple defendants, for greater impact.
3) Nearly all litigation is on software patents. Two studies have found 94% and 90% of the lawsuits encompass software patents.
21 Opitz & Pohlmann (2010) have developed a typology of the different types of IPR enforcing companies. Their typology is based on two variables: patent quality and activities concerning the patent. Patent quality can be either trivial or non-trivial. Trivial patents are patents that either lack quality, or have sufficient quality, but do not affect the technology of the alleged infringer. Activities concerning the patent can be either innovative or not innovative and manufacturing or non-manufacturing. The typology results in the following matrix:
In the figure, the typologies that are relevant for this research are the non-manufacturing types: the patent enforcer, trivial-patent enforcer, patent troll and trivial-patent troll. The patent enforcer wants to be economically compensated for efforts of investment or research. The trivial-patent enforcer is usually a small innovative company or an individual. The IP is not ground-breaking or state-of-the-art. The aim of these companies is to threaten with an injunction, so the company will pay for a license. The companies usually don’t have the funds to engage in a lawsuit or are not aware of the patent quality. The patent troll typologies are studied further and this leads to the following matrix:
22 The two variables of this matrix are ‘License recipient’ and ‘Leverage potential’. License recipient can be an initial licensee, where a patent troll issues a license in advance. The other option is that the patent troll waits for infringement. Leverage potential is based on the height of royalties the patent troll asks. F/RAND stands for Fair, Reasonable and Non-Discriminatory. This is a license obligation that most Standard Setting Organizations will demand, in which is stated that a fair price is asked for a license and discrimination between different licensees is not possible (Layne-Farrar, Padilla & Schmalensee, 2007). Another option is extortive IPR enforcement, where IP holders that do not follow the F/RAND license obligation can ask extortive prices for essential IP. A license supplier is not very common in the literature there is no proof that situations exist. The license extortionist could emerge when an IP holder has a certain piece of essential IP (set by a SSO for example) and can claim high royalties. Royalty claimant is the case of patent ambush, where the patent is hidden from possible licensees. The excessive royalty extortionist is the same as the royalty claimant, except that they will undertake additional actions to gain a bigger settlement/royalty fees by, for example, moving the lawsuit to a country or district that will lead to more favorable judgment (Opitz & Pohlmann, 2010).
Henkel & Reitzig (2008) have looked at the different litigation strategies NPEs can follow. This resulted in three possible strategies according to two variables: (1) probability of patent holding up in court and (2) whether there are negotiation talks between the patent holder and the manufacturer, or not. This gives the following figure:
23
Negotiations Negotiations between holder and manufacturer take place
No negotiations between holder and manufacturer take place Probability of the patent holding up in court Low Injunction
High Switching costs Damage awards
In the situation where negotiations take place between patent holder and manufacturer, the injunction strategy is to file for an injunction. This is a fast action strategy, through which the patent holder can interrupt the manufacturing process of the manufacturer within a couple of days. This can be very damaging for the manufacturer. This is one of the most-used strategies to force the manufacturer into a quick and expensive settlement. The switching costs strategy is to attack a manufacturer with a patent of sufficient quality. This means that the patent holder has sufficient belief that the patent will hold up in court, should this be the case. The important prerequisite of this strategy is that there are no readily available substitutes for the manufacturer to acquire. This means that huge sunk costs (the mentioned switching costs) are created and thus the manufacturer has to use the patent, and thus pay royalties. In the case of the damage awards strategy, there are no negotiations between the manufacturer and patent holder. In this case the manufacturer could be using a technology that may seem trivial and the patent for this technology is overlooked easily. However, this patent can still hold up in court. This strategy is favored above asking for royalty payments in advance because the infringed patent will generate higher revenue, as mentioned earlier (Henkel & Reitzig, 2008; Opitz & Pohlmann, 2010).
In this section, the different views from the literature on the business models of NPEs are displayed. When the NPEs emerged, they had a primarily negative reputation, because of their unethical business model based on litigation. This business model would impede innovation (Tucker, 2011). The business model of NPEs according to this view is based on capturing value in possession of patents and looking for infringement. According to different authors (Reitzig, Henkel & Schneider, 2010; Bessen & Meurer, 2008; Bessen, Ford & Meurer, 2012), the business model is solely based on litigation and has a detrimental effect on innovation. This is based on the fact that an infringed patent has a higher return than offering a license for the same patent beforehand. In later years, however, another view developed. According to other authors (McDonough, 2006; Fischer & Henkel, 2012), NPEs emerge as a sign of evolution of the market. NPEs make patents more liquid and thus help small firms and inventors to monetize their patent. NPEs also help
24 specialized firms to sell their patents, and specialized firms produce higher quality patent. NPEs in Markets for Technology act like an intermediary, where the value is captured in aiding small companies in monetizing patents by finding buyers and sellers. Besides the views on NPEs in the literature, different aspects and strategies of NPEs are described. From the strategies is clearly visible that the business model of NPEs is based on litigation, and less on actually developing new technologies or partnering or collaborating with other technology companies (Henkel & Reitzig, 2008; Bessen, Ford & Meurer, 2012).
2.3
Conceptual model on the changing business model of NPE
In this section the following question is addressed: What changes are envisioned in the literature on Non-Producing Entities? To construct a theory-driven model on the changes in NPE business models, the possible changes of business models are studied in the literature.
2.3.1 Reasons for changes in patent laws
According to Penin (2012) NPEs are changing due to changes in law and more proactive players in the market. The patent law is changing to become more balanced. An example of changing patent laws is the actions of President Obama of the USA, who introduced a proposal to reform the patent laws and formed a task force on high-tech patent issues. Some measures the task force has recommended include: changing the standard to obtain an injunction, demanding more transparency to diminish abusive lawsuits, and tightening of functional claiming5.
In 2003 the Federal Trade Commission (FTC) published a report6 on problems in the patent law and
how it impeded innovation. One of the main conclusions was that ‘questionable’ patents can harm innovation. Specifically: In Industries with Incremental Innovation, Questionable Patents Can Increase “Defensive Patenting” and Licensing Complications. They also recommended different reforms that would improve the patent system. Some of the recommendations:
- Enact legislation to create intervening or prior user rights to protect parties from infringement allegations that rely on certain patent claims first introduced in a continuing or other similar application
- Enact legislation to create a new administrative procedure to allow post-grant review of and opposition to patents
5 http://www.whitehouse.gov/the-press-office/2013/06/04/fact-sheet-white-house-task-force-high-tech-patent-issues 6 To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy, A Report by the Federal
25 This report has led to different so-called ‘patent reform acts’. In 2005, 2007 and 2009 different patent reform acts were proposed, and in 2011 it was finally passed in the senate. Through these reform acts, different laws have been created that affect the sustainability of the business models of NPEs that relied heavily on litigation strategies (Davis, 2007; Myers, 2006).
A big turning point in the patent legislation is the case MercExchange, LLC v. EBay, Inc in 2005. In the case MercExchange, LLC v. EBay, Inc, EBay was sued for infringing a patent of MercExchange which describes the ‘buy now’ functionality in online auctions. This case is famous as the Supreme Court did not automatically grant MercExchange an injunction (which was the usual course of business), thereby taking away an important litigation tool for patent holders (Davis, 2007). The Supreme Court stated that in case of patent infringement, an injunction should not be awarded automatically. Instead, the court should test the patent infringement according to a four-factor test prior to awarding an injunction (Myers, 2006).
26
2.3.2 Changes in business models of NPEs
In this section the different changes that are predicted in literature are researched. These predictions are usually done using mathematical models or theories and are usually not supported with empirical research, as can be seen in table 4.
Table 4: Overview of literature on the change of NPEs
Year Author Topic Method Findings ID
2012 Cheng, Lan, Liu & Liu
Interactions NPEs-PEs Mathematical game theory model on patent market
Changing interactions between NPEs and PEs and between NPEs and other NPEs
37
2012 Cheng, Lan & Liu Patent market dynamics and NPEs
Analysis of data from different sources (USPTO, Westlaw)
The changing business models of NPEs with and without R&D capabilities
19
2012 Penin Influence of NPEs on R&D investments
Mathematical game theory framework based on literature
Patens trolls impede R&D investment, Pure IP brokers do not. Policy is changing to stop trolling
23
2012 Fischer & Henkel Patent acquisitions of NPEs
Analysis of 392 U.S. patents acquired by known NPEs between 1997 and 2006
Patents of NPEs have unique characteristics
10
2012 Rey & Salant Abuse of IP Standard horizontal differentiation mathematical framework based on literature
NPEs will issue a socially sub optimal amount of licenses to control technology
7
2012 Bessen, Ford & Meurer
Cost of NPEs Analysis of multiple data sources (Patent Freedom database, Center for research in security prices US stock database)
NPEs will focus on software patents due to fuzzy boundaries
17
Research has been conducted on what changes could be happening in the patent market and how NPEs are changing. According to Bessen, Ford & Meurer (2012) NPEs convince the world they have changed. They are a new type of company with a business model that is not yet fully understood, but will be beneficial for society. They create, as they call it, a capital market for technology by buying and selling patents and licenses (McDonough, 2006). This will realign the incentives of the entities of the market by bringing the buyers and sellers together in a market they create. This will result in a more liquid patent marketplace with transparent patents (Bessen, Ford & Meurer, 2012).
27 language. These fuzzy boundaries are mostly associated with software patents. Even technological companies cannot easily understand what they exactly claim. This results in a rise in litigation. Bessen, Ford & Meurer (2012) predict that the new business model for NPEs is not just licensing patents. It is licensing software patents, as the result of the fuzzy boundaries.
Cheng, Lan & Liu (2012) have studied two groups of NPEs: the biggest NPEs, called the Giant Troll (GT) and the most aggressive (litigious) NPEs, called the Aggressive Troll (AT). A NPE is a GT if it is in the top 20 in terms of patent quantity and is an AT if it is in the top 20 in terms of patent litigation activities. They also distinguish between firms with or without Research & Development (R&D) capabilities. The first finding is that between the GT and the AT there is hardly any overlap. This means that NPEs can have very different business models. They pursue either the route of litigation as the main business income (in which the threat of litigation is their aim, instead of just means to an end), or the licensing of patents, where contracts are signed and knowledge is exchanged without the interference of the court.
Most NPEs are small companies without R&D capabilities. However, more than half of the GTs that were studied had significant R&D capabilities. The research shows that GTs with R&D capabilities are clearly less litigious than GTs without research capabilities (Cheng, Lan & Liu, 2012).
They also found a strong indication that R&D-based giant NPEs deal with a selected few manufacturers, and engage in parnerships, collaborations and joint ventures. This “close and exclusive” relationship between Practicing Entities (PEs) and NPEs implies that PEs might work with NPEs to control certain critical technologies. This could be explained by research from Rey & Salant (2012). They found that an Intellectual Property (IP) holder can choose to issue less or more licenses than is socially optimal. An IP holder will issue too many licenses when variety is particularly valuable. However, when downstream products (products that are sold to customers) are close substitutes, profits will disappear (as they can just choose a substitute) and the IP holder will issue too few licenses, or charge too high a license fee for the licenses (Rey & Salant, 2012).
Another finding is that the relation between NPEs and PEs will change from litigation to “cooperative licensing and assignment” for NPEs with R&D capabilities. This means that litigation will not be the aim anymore, but more of an exception. The aim of the new business model is to engage in contracts with PEs and develop a fruitful relationship that is not created by threats (Cheng, Lan & Liu. 2012).
28 like universities and research institutes, to commercialize their technology (for example: licensing the technology from them and then issuing their own licenses). Another option for NPEs is to guide other NPEs to study directions that are aligned with the market needs. NPEs could enable technology push/market pull movement. This means that instead of pushing certain types of technology, NPEs can ‘listen’ to the market needs and focus research capabilities on those needs (Cheng, Lan, Liu & Liu, 2012). These findings from the literature are combined in a model shown in figure 5.
Relationship between NPEs and PEs
Close and exclusive relationships between NPEs
and PEs
Cooperative licensing and assignment
Litigation
Type of patents
Focus on licensing based on software patents Focus on
technological patents
Relationship between NPEs and other NPEs
No relation
Helping to commercialize patents of other NPEs (e.g.
universities)
Direct research of other NPEs to align to market needs (e.g.
universities)
Chan, Lan & Liu (2012)
Chan, Lan, Liu & Liu (2012)
Bessen, Ford & Meurer (2012)
29 In figure 5, the first box identifies the possible changes in the relationship between NPEs and PEs. A change can be seen from a business model based on litigation to a business model that is dominated by cooperative licensing and assignment. This means that the focus of NPEs will be to supply them with licenses instead of trying to find companies that may be infringing their technology. Another change is the cooperation with PEs. At the moment, NPEs target a lot of companies with their lawsuits. This will change in close and exclusive relationships, comprising of collaborations, joint ventures and partnerships.
In the second box the changes in the relationship between NPEs and other NPEs are identified. At the moment there is no real interaction between NPEs and other NPEs. However, NPEs can work together to encourage the competition in the market. NPEs can help other NPEs, for example universities or research institutes, to commercialize their IP. They can also help NPEs by aligning the research with the market needs, which can also help the commercialization.
30
3. Methodology
In this research two companies are studied using the theory-driven model that is constructed in the literature study. The two companies are NPEs with R&D capabilities, namely Rambus Incorporated (hereafter: Rambus) and InterDigital. The different research techniques that are used are described in this chapter.
3.1
Data collection
In this section the two aspects of the data collection are discussed: the chosen time period and the collected articles. To collect data on the firms, the name of the firm has been used as search query (Rambus & InterDigital) in the LexisNexis database.
3.1.1 Time period
For both firms, data has been collected from 01-01-2006 until 01-07-2013. For every year, from 01-01 until 31-12, the relevant newspaper articles from major US and world news were collected. 2006 is chosen as starting date, as it follows the verdict of the EBay case (October 2005), in which the Supreme Court caused a turning point in IP litigation as is described in 2.3.1. In the results section the data will be presented in two different ways: data collected per year and in two timeframes. The choice for displaying the data in different ways has been made to remove the bias that could arise from an insufficient amount of articles on particular topics. By presenting the data in two four-year timeframes, this will give a clear view of the change on the different topics.
3.1.2 Collected articles
For Rambus, a total of 168 news articles haven been collected. For InterDigital 154 news articles were collected. In total, of these 322 news articles, 198 were used in the analysis. The articles that were not used, comprised of ‘double articles’ (same news article, but from another source) and articles that did not have relevant content for this study. To determine whether a news article was relevant or not, different criteria were used:
- The article mentions a license agreement - The article mentions patent sale
31 - The article mentions infringement
- The article mentions new technology
- The article mentions collaboration/joint ventures - The article mentions Standard Setting Organizations
The list of collected articles and a table with the collected and used articles per company and per year can be found in appendix B.
3.2
Data analysis
In section 2.3, the theory driven model has been proposed as a research tool for this research. In this model codes are added that are used to analyze the collected news articles. The model is divided into three parts and the codes that are assigned can be found below.
In figure 6, the first part of the model has been assigned codes for the different possibilities that can occur. News articles that describe a relationship between NPEs and PEs that is dominated by litigation will be assigned code 1A. According to literature, the aim of business models of NPEs was litigation (Reitzig, Henkel & Schneider, 2010; Bessen, Ford & Meurer, 2012; Opitz & Pohlmann, 2010; Henkel & Reitzig, 2008). News articles that describe a relationship between NPEs and PEs that mention licensing will be assigned code 1B. A patent sale is also grouped as 1B. If the relationship between NPEs and PEs is close and exclusive, it will be assigned code 1C. This also entails collaborations and joint ventures between NPEs and PEs (Chan, Lan & Liu, 2012). Two hypotheses are derived from this model:
Relationship between NPEs and PEs
1C
Close and exclusive relationships between NPEs
and PEs
1B
Cooperative licensing and assignment
1A
Litigation
32 Hypothesis 1: The relationship between NPEs and PEs shifts from litigation to cooperative
licensing and assignment.
Hypothesis 2: The relationship between NPEs and PEs shifts from litigation to a relationship with a close and exclusive nature.
In figure 7 the second part of the model is shown. In the literature, there is no mention of a structural relationship between NPEs and other NPEs. This can be attributed to the fact that the business model of NPEs is based on litigation and mostly on infringement (McDonough, 2006). NPEs, however, cannot be sued for infringement, as the basis for an infringement suit is that a product is produced using IP. NPEs do not produce products, and thus cannot be sued for infringement (Mello, 2006).
If there is a clear mention of no relation between the researched NPEs and other NPEs, code 2A will be assigned, but the prediction is that this will not often be the case, as there will not be a lot of news articles that will mention a lack of interaction between the researched NPEs and other NPEs. To be comprehensive, it is added to the model.
If there is a mention of NPEs helping to commercialize patents of other NPEs, it will be assigned code 2B. This could be a partnership, collaboration or even a joint venture with a university, research center or commercial NPE. If there is a mention of an NPE directing research of other NPEs, it will be assigned code 2C. This can happen in multiple forms, for instance leading or participating in a technology forum, platform, special interest group or congress. Joining a Standard Setting Organization is also a form of directing the research of NPEs to one standard. Two hypotheses are derived from this model:
Relationship between NPEs and other NPEs
2A
No relation
2B
Helping to commercialize patents of other NPEs (e.g. universities)
2C
Direct research of other NPEs to align to market needs (e.g. universities)
33 Hypothesis 3: NPEs will intensify their partnerships with other NPEs to commercialize
patents.
Hypothesis 4: NPEs will intensify their cooperation with other NPEs to direct research with market needs.
In figure 8 the third and final part of the model is shown and the codes are assigned. This part of the model is about the type of patents that NPEs will focus on. According to Bessen, Ford & Meurer (2012), NPEs will generally focus on software patents instead of technological patents. They attribute this to the ‘fuzzy boundaries’ software patents have. ‘Fuzzy boundaries’ means that these types of patents are vague, can be interpreted in multiple ways and even technological companies cannot easily find them or understand what is claimed in the patent (Bessen, Ford & Meurer, 2012).
In order to research the types of patents, a definition of the term ‘software patent’ must be given first. Literature shows that software patents are hard to describe and there is no clear consensus on a definition of a ‘software patent’ (Graham & Mowery, 2003; Bessen & Hunt, 2004; Layne-Farrar, 2006; Hall & MacGarvey, 2010; Graham & Vishnubhakat, 2013). Graham & Mowery (2003) constructed their definition of software patents based on the patent classifications of the United States Patent and Trademark Office (USPTO). Bessen & Hunt (2004), however, argue that this system does not distinguish whether the underlying technology is software or something else. Bessen & Hunt (2004) have constructed a way of identifying software patents called ‘Bessen/Hunt technique’. The Bessen/Hunt technique consists of keyword searches with three parameters:
1) the description of the patent must contain one of the following: ‘software’, ‘computer’ or ‘program’ 2) the description of the patent cannot contain the following words: ‘antigen’, ‘antigenic’ or
‘chromatography’
3) the title of the patent cannot contain the words: ‘chip’, ‘semiconductor’, ‘bus’, ‘circuit’ or ‘circuitry’
Type of patents
3B
Focus on licensing based on software patents
3A
Focus on technological
patents
34 This method, however, has been criticized by Hahn & Wallsten (2003, they criticized the first working paper). Hall & MacGarvey (2010) have combined the Graham & Mowery (2003) and Bessen & Hunt (2004) methods and added their own method (the Hall & MacGarvey method, which also consists of using classifications of the USPTO) to improve this method. This literature review shows that there is no universally accepted method to identify software patents. Bessen & Hunt (2004) thus also claim: ‘researchers must construct their own definitions’. Following this recommendation in this research, patents will be as classified as software patents when they meet the following criteria:
- the name or description mentions it as software/client
- the ‘novel’ requirements of the patent is software related, the software is not used solely as support The different technologies will be tested according to these rules and specified as a software patent, or a technological patent. If a news article mentions a technological patent, it will be assigned code 3A. If a software patent is mentioned, it will be assigned code 3B. This leads to the final hypothesis:
Hypothesis 5: NPEs will change their focus from technological patents in general to software patents.
35
4. Results
In the results section, the companies will be introduced first. Both of the researched companies are American technological companies with R&D capabilities. The results consist of content analysis of news articles from 01-01-2006 until 01-07-2013. The quantitative data is presented in graphs and clarified with excerpts from the news articles.
4.1
Company introduction
4.1.1 Rambus
Rambus is an American technology company that licenses patents. Rambus is known as a Non-Producing Entity, or NPE. They only buy or develop intellectual property (IP) to license or sell to other companies. Rambus is the first company associated with a so-called ‘patent ambush’ (Pohlmann & Opitz, 2010). A patent ambush strategy is a strategy in which the patent holder sees an opportunity to ‘insert’ the patent into the standard that is being developed in standard settings procedures. When the standard is adopted and huge sunk costs are made for this standard, the patent holder can demand royalties or licensing costs for the patent they are using, or according to the patent holder, infringing (Hemphill, 2005).
From 1991 to 1996 Rambus was a member, or at least a participating guest, in the Joint Electron Device Engineering Council (JEDEC). In 1993 the JEDEC adopted and published a standard for SDRAM, a vital part of electronics. In 1996, after Rambus had left JEDEC, JEDEC adopted DDR-SDRAM. However, the standard only reached its final phases in 1999 or 2000, at least three years after Rambus left JEDEC. This is important due to the fact that JEDEC members had the duty to disclose any patents they had regarding the standard during the final stages of the setting of the standard, the formalization of the standard (Stern, 2001; Alban, 2004). In 1999, long after Rambus left JEDEC, Rambus sues Infineon Technologies (and several other companies) for infringement of four patents in Infineon’s JEDEC-compliant DRAM products (Naughton & Wolfram, 2004). This leads to several lawsuits and counter-lawsuits. Some companies are quick to settle, for instance Hitachi. Rambus settles the lawsuits with the European Commission for monopolizing of the market, and receives over 50 million a year in SDRAM royalties (Rey & Salant, 2012).
4.1.2 InterDigital
36 for $11 million. They stayed on the course of wireless and telecom research and became a big player in 2G, 3G and 4G patents as well as the 802 Wi-Fi standard. At this moment 50% of the 3G technology is reliant on InterDigital patents7. They have filed multiple big lawsuits against Huawei, Nokia, ZTE, Samsung and
others, which lasted years . In 2011 the company tried to sell itself, but lacking adequate offers (they were looking for at least 3 billion) decided against it in 20128. However, they did sell 1700 patents to Intel in
2012 for $375 million9.
4.2
Results of counted instances in news articles
In this section, the counted instances are represented in graphs. To clarify the different results, excerpts from the researched news articles are shown.
4.2.1 Relationship between NPEs and PEs
In figure 9 the results of the content analysis from 01-01-2006 until 01-07-2013 are shown with respect to the relationship between NPEs and PEs.
7 http://www.law360.com/articles/128960/itc-jams-interdigital-s-section-337-claims-v-nokia 8 http://www.electronista.com/articles/12/01/23/interdigital.decides.against.sale.to.others/ 9 http://newsroom.intel.com/community/intel_newsroom/blog/2012/06/18/interdigital-agrees-to-375-million-patent-transaction-with-intel 0 5 10 15 20 2006 2007 2008 2009 2010 2011 2012 2013
counted instances on relations between NPEs
and PEs
1A 1B 1C
Chart 1 - Interactions between NPEs and PEs
1A - Litigation
37 2006-2009
From 2006 until 2009 the relationship between the researched companies and Practicing Entities were dominated by litigation (1A). Cooperative licensing and assignment (1B) and Close and exclusive relationships (1C) were not common, as can be seen in chart 1.
From 2006-2009 litigation dominates the news articles of the researched companies. InterDigital was involved in different lawsuits with LG, Samsung, Nokia and the ITC. Rambus was engaged in lawsuits with Hynix Semiconductor, Micron Technology, Samsung, NVidia and the FTC. At the same time, there were also a few deals that focused on licensing based on cooperation. Rambus had signed Toshiba and Advanced Micro Devices for long-term patent deals. However, these licensing deals did emerge under the threat of litigation. In this period, however, there are already different indications that there is a shift leading to less litigation, as the patent licensing business on its own is not sustainable. In a news article10 of May 2006, the
following comment is made by Rambus:
“Rambus, the DRAM interface IP supplier, is looking for customers prepared to take a total package of patents as well as product licences backed by engineering services. The firm may also be branching out into flash memory interfaces. "Patent licensing on its own is not a long-term sustainable business model, we want to add value to our customers," Sharon Holt, senior vice-president for worldwide sales, licensing and marketing, told EW. Rambus has been seen as a firm which licenses products to DRAM manufacturers. Now it sees system companies, which want to use its interface technologies such as XDR, Advanced Backplane, FlexIO and RDRAM, as potential customers.”
In this excerpt, Rambus itself indicates that the licensing model is not sustainable anymore. The verdict in the case MercExchange, LLC v. EBay (Davis, 2007) impacted the litigation strategies of NPEs: threatening ‘infringing’ companies with injunctions was not possible anymore, as injunctions were no longer granted automatically. After years of earning royalties through fast-acting litigation strategies, a new business model had to be contemplated.
Rambus indicates they want to move to a more profound relationship with other companies, with added support and package deals. The need for change is also confirmed after losing a ruling to the FTC as described in the following excerpt11:
“Rambus Inc. was dealt what could turn out to be a major challenge to its intellectual-property licensing business model last week, when the Federal Trade Commission ruled that the company had unlawfully monopolized the markets for computer memory technologies incorporated into industry standards for DRAM chips. […]Given the company's historically contentious relationship with chip makers, some industry observers believe chip companies could use the FTC decision as leverage to force Rambus to renegotiate terms of its royalty arrangements.”
