Finance: A technological knowledge game?
A bibliometric study of patent data on knowledge sourcing and its impact on innovation in Financial Technology.
Master Thesis
Faculty of Economics and Business
Msc. Business Administration – International management
Author: Cheyenne Seur
Student number: 10868739 Supervisor: Dr. Lori DiVito
2
Statement of Originality
This document is written by Cheyenne Seur, who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Abstract
Firms are subjected to increasing competition, accelerated technological breakthroughs and progressively greater globalization. Powell (1998) states that doing business can be described as “a chronic state of flux, with continual variation in its external environment” (p. 2). The financial sector is no exception and it is a crucial challenge for organizations to adapt (Powell, 1998). As established financial institutions have been repeatedly exposed to economic and regulatory disruptions, new players have entered the financial market (Alt and Pushmann, 2012). FinTechs, non-bank financial institutions that operate on IT based knowledge and IT business models, introduce new products and services that may change finance in the same way Internet has changed the written press (Gelis and Woods, 2014; Deutsche Bank, 2015). In today’s competitive market, in order to obtain and sustain competitive advantage, firms must continuously innovate. Firms who fail to do so will most likely not be able to compete and might not survive (Morris, 2008). Knowledge appears to be the main determinant for innovation. This thesis therefore researches the relation between knowledge sourcing and innovative performance of firms in the Financial Technology industry. This is done by a bibliometic patent citation research. The study furthermore includes the concept of location and questions whether domestic knowledge sourcing or international knowledge sourcing leads to higher levels of innovativeness. This research builds onto widely academically researched concepts of knowledge (e.g. Grant, 1996; Gertler, 2003), innovation (Schumpeter, 1983; Han, Kim, and Srivastava, 1998) and location (Jaffe, Trajtenberg, & Henderson, 1992). This research joins the ongoing conversation and contributes to theory by researching these concepts in the Financial Technology industry. Limitations should however be taken into account and further research is recommended.
4
Acknowledgements
Before you lay my final assignment for my master’s degree in Business Administration. This thesis provided me with the opportunity to showcase what I have learned during my studies at the University of Amsterdam and gave me the opportunity to combine this with a personal interest of mine (technology).
This thesis was written during a very eventful time in my life: with many highs, and sadly a couple lows. I travelled, got a job that I absolutely love and bought my first house. I also broke both my legs, on separate occasions, within a nine-month time frame. Due to all this, writing my thesis took me a couple months longer than expected, but at last it lay before you.
I want to thank my family and friends for supporting me during this time. I want to thank my employer and colleagues for their understanding and for inspiring me to focus my thesis on FinTech. Lastly, I want to thank the University of Amsterdam and its professors for the inspiring lectures during my studies.
5 Index ! Abstract 3! Acknowledgements 4! Index 5! 1. Introduction 7! 1.1 Research focus 7! 1.2 Theoretical relevance 9!
1.3 Structure of the thesis 9!
2. Theoretical foundation 11!
2.1 The Financial sector 11!
2.2 Financial Technology 14!
2.3 Knowledge 16!
2.4 Innovation 21!
2.5 Open Innovation 23!
2.6 Location 25!
2.7 The theory of disruptive innovation 27!
4. Methodology and data 31!
4.1 Patent analysis 31!
4.2 Patent database 32!
4.3 Validity and Reliability 37!
6 5.1 Hypothesis 1 38! 5.2 Hypothesis 2 44! 5.3 Hypothesis 3 46! 6. Discussion 49! 7. Limitations 53!
7. Conclusion and recommendations 55!
8. Sources 58!
7
1. Introduction
1.1 Research focus
International boundaries are blurring and new technologies are simultaneously being introduced. Both factors contribute to the continuously changing environment that firms are competing in today. Traditional managerial and organizational theories become less relevant, as focus is shifting from traditional vertically integrated models to a model of innovation. Focus on innovation has resulted in the fact that a firm’s competitive advantage now lies in its ability to generate, process and apply knowledge and information. This provides opportunity to new players to enter and disrupt an existing market with new business models (Powell, 1998; West, Salter, Vanhaverbeke and Chesbrough, 2014).
“Banking has historically been one of the business sectors most resistant to disruption by technology.” Banks have established themselves as robust businesses, and consumer inertia in banking and financial services is high. This has resulted in an industry with a resilient business model and defensible economics (Dietz, Khanna, Olanrewaju, & Rajgopal, 2016). In the last couple years banks and other financial institutions word-wide have been subjected to this continuously changing environment, and have witnessed the emergence of new players in the financial industry. These new players, non-bank financial institutions (NBFIs) often referred to as FinTechs (abbreviation for Financial Technology) are profoundly changing the financial sector, with the introduction of new digital currencies, networks and wallets (Deutsche Bank, 2015). Since the Netscape IPO in 1995, many have attempted to challenge incumbents (Shinal, 2005). So far, few NBFIs have survived as stand-alone entities and PayPal seems to be the exception to prove the rule: disrupting banks is not easy (Kane, 2002). Some however believe that the rise of financial technology may change the financial sector in the same way Internet changed the written press (Gelis & Woods, 2014). In the
8
period of one year (2013 – 2014), the global investment in the FinTech industry tripled to $12.21 billion (Suh, 2015). At the same time, big and established financial institutions are investing in new technologies to strengthen their own positions (PWC, 2016). As both
FinTechs and established financial institutions are investing in technology, the question arises whether technological knowledge is what is needed to obtain and sustain competitive
advantage in the financial sector. Has finance become a technological knowledge game? The sources of (sustained) competitive advantage have been researched by many scholars, from supporters of the resource-based view as early as the 1960s, to innovation scholars that research modern complex and high-tech industries. Supporters of the
knowledge-based view believe that competitive advantage and the ability to innovate lies in the knowledge a firm possesses (Grant, 1996). According to Kang and Kang (2009), in the age of open innovation, external knowledge is the most important source of technological innovation. The concept of knowledge is however one that has been proven hard to explain, and has been widely researched in the last years but without a clear conclusion or consensus (Grant, 1996). Little is furthermore known about knowledge sourcing in the financial sector, and if and how this contributes to the ability to innovate. A stream within international business research focuses on the impact of location on knowledge sourcing (e.g. Gertler, 2003; Gassmann, Enkel and Chesbrough, 2010). There is no conclusive answer on whether local/domestic knowledge sourcing or international knowledge sourcing contributes more to innovativeness. In order to join to on-going conversations in research, this thesis seeks to further explore the concept of knowledge, questions if external knowledge sourcing
contributes to innovativeness of firms and questions if local/domestic knowledge sourcing or international knowledge sourcing had greater impact on innovativeness of firms in the financial technology industry.
9
1.2 Theoretical relevance
This research aims to create better understanding on the influence of external knowledge sourcing (and its location) on innovative performance of firm’s in the Financial Technology industry. This research contributes to existing literature in multiple ways. Firstly, this research contributes to existing literature by further exploring the concept of knowledge. According to Grant (1996) the concept of knowledge has been researched widely over the yeas, but without clear conclusion or consensus. This thesis will explore which types of knowledge contribute most to innovativeness in the financial technology industry. It will further build onto the theory of Kang and Kang (2009) and question whether external knowledge contributes to innovation. This thesis further contributes to the widely researched International Business subject of location. There is currently no consensus on the subject of location. Cantwell and Santangelo (1998) focus on international technology networks, while Gertler (2003) and Gassmann, Enkel and Chesbrough (2010) study proximity and believe that being physically close enables the absorptive capacity of firms, giving them access to knowledge.
This thesis lastly contributes to theory by specifically focussing on Financial Technology. So far, little research regarding external knowledge and its impact on innovation in the Financial Technology sector has been conducted.
1.3 Structure of the thesis
In order to provide the reader with an introduction to the playing field Financial Technology, the financial sector and it challenges will be discussed. This is followed by an introduction to the concept of Financial Technology and the firms that operate in this industry. Next the theoretical framework for this thesis will be discussed; the concepts of knowledge and
10
knowledge sourcing will be introduced. This is followed by a discussion on much cited international business literature on (open) innovation and literature on the impact of location. Based on literature, hypothesis for this research will be defined. In the methodology section of this thesis, the research methodology and the research data will be described. Analysis of the data and testing of hypotheses is described in the analysis section of this thesis. Results and limitations of the research will be discussed. This discussion results in conclusions and recommendations for further research.
11
2. Theoretical foundation
This chapter defines the theoretical foundations for this research and defines and describes key concepts in the literature that will be used throughout the research. First the financial sector and its challenges will be discussed. Next theories of the knowledge-based view and (open) innovation theories will be discussed. Traditional international business (IB) literature regarding location will be explained. This will lead to hypotheses that will further guide this research.
2.1 The Financial sector
Firms are subjected to increasing competition, accelerated technological breakthroughs and progressively greater globalization. Powell (1998) states that doing business can be described as “a chronic state of flux, with continual variation in its external environment” (p. 2). The financial sector is no exception and it is a crucial challenge for organizations to adapt (Powell, 1998). Atl and Puschmann (2012) propose that there are four drivers for strong transformation in the financial industry in the coming years.
The first driver they identify is the consequences of the financial crisis. The financial industry, since 2007, has been repeatedly exposed to economic and regulatory disruptions, which forces the industry to change banking (Alt & Puschmann, 2012). Traditionally the financial sector is an oligopoly wherein a small group of players dominates the market. It is a sector with a robust business model, defensible economics and high consumer inertia (Dietz, Khanna, Olanrewaju, & Rajgopal, 2016). Retail banks are highly profitable, secure businesses and are simply too big to fail (Green, 2013; Oracle, 2015). Due to the “chronic state of flux,
12
with continual variation in its external environment” the financial sector has witnessed changes in the last years and disruption at last seems to have reached established MNEs. The financial crisis led to large losses for a number of banks, and for some even to colapse (Alt & Puschmann, 2012).
The second driver Atl and Puschmann (2012) identify is changed consumer behaviour. Consumers, who can often be considered digital natives, expect the use of electronic channels. Consumers want new (digital) products and services avaliable at all time (Oracle, 2015). They are more informed and require more transparency from their financial services provider (Hedley, White, Cormac, dit de la Roche, & Banerjea, 2006). A research conducted by the Federal Reserve (2016) states that the use of mobile banking continues to rise. It was found that more than half (53%) of mobile phone owners with a bank account use mobile banking regularly. 94% of these users use their mobile phones to check their balance or recent transactions, 58% of these users use mobile applications to transfer money. The research furthermore identifies a group (22% of consumers) they name “underbanked”. These consumers have a bank account and in addition use one or more financial services from a NBFI. Customer faith in financial institutions has furthermore declined. The industry fails to inspire trust and fails to offer the personalized experience customers are looking for
(OConnell, 2015). An IBM study showed a 27% gap in the perception retail banks have of customer satisfaction and actual customer satisfaction (Brill, Drury, Lipp, Marshall, & Wagle, 2015).
The third driver is the increase in diffusion of innovative downstream IT. With the general acceptance of smartphones and tablets, software solutions are on the rise. User generated content websites cause a paradigm change in transfer of information and help reshape social interaction. Organizations quickly have to adopt social technologies and the rise of software ecosystems. Banks have heavily invested in IT infrastructure and in electronic
13
channels such as Internet banking and mobile banking to enhance their services for their customers. These systems however focus on operational functionalities for existing banking products, instead of new technologies (Alt & Puschmann, 2012). Consumers want products and services that genuinely enhance their lives (Springford, Vicary-Smith, Pon, Pomeroy, Lipsey, Phillips, Skapinker, Michie, 2011). Bower & Christensen (1995) state that industry leaders are however hardly the first to commercialize new technologies, as they do not initially meet demands of mainstream consumers but appeal to small emerging markets. Leading companies often successfully invest in the technologies that are necessary to retain their customers, but fail to make investments in technology that the customer demands in the future. Basole and Karla (2011) state that to mobile ecosystem is experiencing a wave of transformation due to new business models and the emergence of new players. Which leads to changing roles for existing players.
This is in line with the last driver identified by Alt and Puschmann (2012): the emergence of non-banks that provide innovative IT. These third parties are developing applications that offer more than traditional online banking systems, such as online
investment communities, personal finance applications and wallets. Where large, traditional financial institutions are slowed down by complex development processes, legacy technology and regulations, FinTechs are using agile methods of product development and innovations to quickly respond to and even create costumer need (PWC, 2016). These firms enter the market with IT-based business models instead of traditional banking models (Alt & Puschmann, 2012).
14
2.2 Financial Technology
Financial technology, often abbreviated as FinTech, is a terminology in the financial industry that can be used to describe either:
1. Financial technology in it self (e.g. crypto currencies);
2. A Financial Technology service provider (A NBFI that uses an IT-based business model to offer financial services).
There is no, academically supported, definition for either Financial Technology itself, or a FinTech service provider. Day and Schoemaker (2000) define technology as following: “A set of discipline-based skills that are applied to a particular product or market” (p.2). Financial Technology could therefore be defined as “a set of technological-based skills that are applied to the financial market”. The Wharton FinTech Club uses another definition: “An economic industry composed of companies that use technology to make financial systems more efficient” (McAuley, 2015). Dapp, Slomka, Ag, & Hoffmann (2014) define “FinTech” as digitisation of the financial industry and use the term to describe “Modern technologies for enabling or providing financial services, such as internet-based technologies in the
e-commerce field, mobile payments or early-stage crowd- based financing of start-ups” (p. 5). All definitions mention the concept of Financial Technology in itself, but none mention the companies that deliver these services. The most complete definition appears to be one from FinTech Weekly, which states: “Financial technology, also known as FinTech, is a line of business based on using software to provide financial services. Financial technology
companies are generally startups founded with the purpose of disrupting incumbent financial systems and corporations that rely less on software.” (FinTech Weekly, n.d.).
Opposed to what many think, the concept of financial technology itself is not new. Financial technology has played a key role in the financial sector since the 1950s, with the rise of credit cards and ATMs followed by electronic stock trading, mainframe computing and
15
Internet banking. The difference is however that today FinTech innovations, and the
companies behind them do not only enhance banking services but also replace them. Whereas in the past financial technology was only institutional (business-to-business), it is now
consumer-facing (business-to-consumer) (Desai, 2015). FinTech companies are profoundly changing the financial sector, with the introduction of products directly available to
consumers such as new digital currencies, networks and wallets (Deutsche Bank, 2015). Money20/20, the largest global event focused on payments and financial services innovation, has stated that FinTech companies by nature are borderless, born global firms (Let's Talk Payments, 2015). Knight and Cavusgil (2004) define Born Globals as firms that are early adopters of internationalization. Born globals are “firms that expand into foreign markets and exhibit international business prowess and superior performance, from or near their
founding.” (p. 124). These firms, despite their liability of smallness, scarce financial- and tangible resources leverage innovativeness and capabilities that enable them to achieve foreign market success in a very early stage in their evolution. These firms achieve superior international business success from the application of knowledge-based resources (Autio, Sapienza, & Almeida, 2000; Knight and Cavusgil, 1996; Oviatt and McDougall, 1994; Rennie, 1993). That FinTechs success is increasing is confirmed by the amount of
investments made in the sector. In the period of one year (2013 – 2014), the global investment in the FinTech industry tripled to $12.21 billion, this means investment grew by 201%. In Europe alone, $1.5 billion was invested in FinTech. Most was invested in London-based companies, followed by Amsterdam-based companies and Stockholm-based companies. The same study found that most investments, $1.16 billion, were made in the Payment segment (Suh, 2015). Even though FinTechs are born global firms, that receive large amounts of investments, success of FinTech companies is far from guaranteed. Since 1995 many NBFI’s have attempted to challenge incumbents, so far only a few NBFIs have survived as
stand-16
alone entities and PayPal seems to be the exception to prove the rule: disrupting banks is not easy (Shinal, 2005; Kane, 2002). Some however believe that FinTech may change the financial sector in the same way Internet changed the written press (Gelis & Woods, 2014).
2.3 Knowledge
FinTech is growing and profoundly changing the financial sector. According to Desai (2015) the number of start-up FinTech companies is difficult to determine, but it is estimated that there are approximately 6500 FinTechs word-wide. Why do some FinTechs (e.g. PayPal) survive and why do others fail? Understanding the sources of (sustained) competitive
advantages has been a major area of international business research since the 1960s. According to Ansoff (1965), Wernerfelt (1984) and Barney (1991) sustained competitive advantage can only be gained by implementing a value creating strategy, that is not simultaneously being implemented by a current or a potential competitor and when other firms are unable to duplicate the benefits of the strategy. The resource-based view aims to explain why some firm perform better than others and believes that firms can obtain this advantage by having valuable, rare, inimitable and non-substitutable resources (Barney, 1991). Hammel and Prahalad (1990) believe this lies in the core competencies of the firm and define these as “the collective learning of the organization, especially how to coordinate diverse skills and integrate multiple streams of technology” (p. 82). It is clear that the focus lies in organizational learning and its ability to combine different types of knowledge (Stoelhorst, 2008). The knowledge-based view (a perspective based on the resource-based view) believes that knowledge is the most strategically significant resource to a firm; knowledge is the main determinant of sustained competitive advantage, innovativeness and
17
superior performance. Applying these knowledge-based resources enables firms to leverage innovativeness (Autio, Sapienza, & Almeida, 2000; Knight and Cavusgil, 1996; Oviatt and McDougall, 1994; Rennie, 1993). It is however unknown if a specific type of knowledge leverages more innovativeness than others in the financial technology industry. Therefore the following hypothesis is proposed:
H1: There is a significant relation between the type of knowledge a firm has and the level of
innovativeness.
DeCarolis & Deeds (1999) state that knowledge generated, exploited and transferred throughout the organization yields an advantage that is hard to reproduce in the marketplace. The resource-based view recognizes that transferability of a firm’s resources and capabilities is critical, builds on this and believes that the transferability of knowledge is highly important within firms, but also between different firms. Differences in their knowledge bases cause differences between performances of firms. With external knowledge as on of the main determinants of difference in for performance (and their innovativeness) the following hypothesis is proposed:
Hypothesis 2: Firms relying on high levels of external knowledge sourcing exhibit higher levels of innovativeness than firm inventions relying on low levels of external knowledge sourcing.
Where the resource-based view only sees knowledge as a generic resource; the knowledge-based view is able to distinguish different types of knowledge-based capabilities (Grant, 1996; Alavi and Leidner, 2001). The concept of knowledge itself is one that is hard to
18
define. Scholars like Plato and Popper have touched the concept of knowledge but without clear consensus. The concept of knowledge is complex and difficult to explain. It can however be stated that knowledge can be explicit or tacit. The difference lies in the transferability: explicit knowledge can be easily transferred through communication; tacit knowledge cannot (Grant, 1996). Adbulai (2004) visualizes the difference in an iceberg model.
Figure 1: Explicit and Tacit knowledge (Adbulai, 2004).
Up until the 1960s innovation in technology companies was treated as an unexplained variance in the performance of the firm. An element within technological innovation, tacit knowledge, is nowadays still treated similarly. Interest in tacit knowledge in the technology sector has however grown and its contribution to growth and economic performance of firms has been recognized (Koskinen & Vanharanta, 2002). An answer to why interest in tacit knowledge has grown may lie in the resource-based view of the firm and the increasing competition firms are subjected to (Kay, 1993). In a competitive age, with institutional openness and innovation networks success depends increasingly on the ability to outperform other firms with better and improved products and processes. Tacit knowledge according to Gertler (2003) is the most important basis for innovation based value creation. When every firm has easy access to the codified knowledge in open innovation networks, the creation of unique capabilities depends on the tacit knowledge a firm has and it’s ability to transfer it
19
(Maskell & Malmberg, 1999). In science-based fields, such as the technology sector, tacit knowledge is the largest determent in enhancement of a firm’s core competences and their ability to innovate (Cantwell and Santangelo, 1998).
But what is tacit knowledge? Gertler (2003) defines it by using Michael Polanyi’s phrase “we can know more than we can tell”. He believes that the knowledge of scientists is not limited to clearly articulated rules, statements and algorithms. Within this concept are two things to be considered; the first being the awareness and/or consciousness, as the tacit
knowledge often only exists in the background of consciousness. The second to be considered is the communication difficulties of being unable to express and explain knowledge when it is conscious. Communication such as spoken and written words cannot convey all knowledge and often knowledge can only be communication through the actual performance and the experience of it. In their research on the role of tacit knowledge in innovation processes of small technology companies Koskinen and Vanharanta (2002) also state that tacit knowledge represents knowledge that is based on the experience of individuals.
Three tacit knowledge problems are identified by Gertler (2003): the first is that it is unclear how tacit knowledge is produced; second is how firms find it and appropriate tacit knowledge and the third is how is tacit knowledge reproduced or shared. Tacit knowledge, by nature, is difficult to produce and acquire. Tacit knowledge is produced and resides in social relations and is context and history dependent (Hamel, 1991; Badaracco, 1991). In knowledge management literature it is discussed how firms produce knowledge management in the first place. Two dimensions are stated: the first is private (investment in human capital through education and training and obtaining talent) and the second is social (Gertler, 2003).
20
Regarding the production of knowledge Nonaka, Toyama and Konno (2000) believe that the creation process is a spiral existing out of three elements: (1) the SECI process (creation through the conversion of tacit and explicit knowledge), (2) the “ba” which is the shared context for knowledge creation and (3) the assets, outputs, inputs and moderators of the creation process (e.g. a patent). They conclude that organizations can create new knowledge through socialisation, externalisation, combination and internalisation (SECI) taking place on a platform for knowledge conversion (ba) which will lead to a spiral of knowledge creation. In complex industries, such as the financial technology sector knowledge is often complex and tacit, it can thus be sticky and hard to obtain and transfer (Grant, 1996). Transferability is however a key variable of innovation. In Chesbrough, Vanhaverbeke, & West’s (2006) definition of open innovation the emphasis is placed on the inflows and outflows of
knowledge: the transferability. Szulanski (1999) states that knowledge transfer is often seen as an act, but should however been seen as a process. Wolceshyn and Karagianis (2003) describe the concepts of internal knowledge sourcing and external knowledge sourcing. In internal knowledge sourcing creation and acquisition of knowledge takes place within the boundaries of the firm. Members generate and distribute knowledge through activities as research and development, peer learning and communication. When new knowledge is brought in from the outside, external knowledge sourcing occurs. Examples of external knowledge sourcing are acquisition, partnerships, imitation, licensing, and employees obtaining new knowledge from e.g. conferences.
21
2.4 Innovation
Innovation is seen as one of the most important business drivers for growth, and as one of the most important sources for obtaining competitive advantage (Han, Kim, and Srivastava, 1998; McGovern, Court, Quelch and Crawford, 2004; O’Regan, Ghobadian and Sims, 2006). There are many definitions for the concept of innovation. The first economist to draw
attention to the concept of innovation is Joseph Schumpeter. Schumpeter defined innovation as: “Innovation is the commercial or industrial application of something new—a new product, process or method of production; a new market or sources of supply; a new form of
commercial business or financial organization (p. 255)” (Schumpeter & Opie, 1983). The European Commission (2004) has a similar description of innovation: “activities to improve firm performance, including the implementation of a new or significantly improved product, service, distribution process, manufacturing process, marketing method or organization method”. The Organization for Economic Cooperation and Development (OECD)’s definition of innovation is more comprehensive and mentions the scientific, technological, organization, financial and commercial aspects of innovation: “An innovation is the implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organizational method in business practices, workplace organization or external relations. Innovation activities are all scientific, technological, organizational, financial and commercial steps which actually, or are intended to, lead to the implementation of
innovations” (The Organization for Economic Cooperation and Development, 2005). Even though definitions may slightly vary it is clear that the ability to innovate is generally
accepted as a critical success factor to growth and future performance of firms (Carayannis & Provance, 2008). Many firms see innovation as a strong contributor for generating business and profitability, which will lead to high organizational performance and competitiveness. In order to generate sustainable growth, the company requires sustainable innovation (Gupta,
22
2007; Potters, 2009). In order to obtain and keep a competitive advantage, firms must continuously develop new products, processes and services. Innovation may aid in creating new markets, transform industries, but also aids incumbents to stay ahead of challengers. Innovation may even promote the competitiveness of nations on a global scale (Dosi, Freeman, Nelson, Silverberg, & Soete, 1988; Tidd, Bessant, & Pavitt, 2001; Sood & Tellis, 2009). Morris (2008) states that companies that fail to innovate will most likely not be able to compete and maybe not even survive in the continuously changing market environment.
According to Gatignon, Tushman and Smith (2002) innovation is one of the most elusive dimensions of firms to quantitatively comprehend. Much research has been conducted to develop measures of innovation, where the focus lies on inputs, outputs and the
mechanisms to facilitate inputs and outputs (Baruk, 1997; Leenders and Wierenga, 2002). Carayannis and Provance (2008) propose a 3P construct measurement of innovation, which considers the factors “Posture”, “Propensity” and “Performance” and relate this to an
organization’s innovative capacities. In their model the first factor “Posture” is an exogenous factor to the innovation process, that does not consider whether a innovation process or what type of innovation process is in place. Posture is related to the lifecycle of the firm and the industry, and the significance of their strategic activities in one or more markets. The lifecycle of the industry may constrain innovative actions of the firm. The second factor “Propensity” is the ability of the firm to capitalize posture based on cultural acceptance of innovation within the firm. It is believed that a firm may possess sufficient and adequate resources to be
innovative but due cultural constrains have underdeveloped capacity for innovation. The third and last factor is “Performance”. This factor has three levels: output, outcome and impact. Output is a direct result of innovation such as new products, patents and licenses. Outcomes include mid-range results such as increased revenue. Impacts comprehend long-range benefits such as recognition of innovative competence: a status as a top innovator.
23
2.5 Open Innovation
The environment firms operate in has significantly changed since the origins of the resource based view in the 1960s and institutional openness has become increasingly popular in both practice and academia. This openness has resulted in the creation of open innovation networks (Gassmann, Enkel and Chesbrough, 2010). These open innovation networks can server as major sources for spillovers of knowledge (Maskell and Malmberg, 1999). The first open innovation networks were developed by groups of innovation practitioners, whom where mostly active in complex high-tech industries (Gassmann et al. 2010). Chesbrough et al. (2006) define open innovation in the following way: “Open innovation is the use of purposive inflows and outflows of knowledge to accelerate internal innovation, and expand the markets for external use of innovation, respectively’’. Free revealing of knowledge and inventions are the defining characteristics of the open innovation model. These knowledge spillovers can occur by means of compensation (e.g. selling intellectual property, licensing, partnerships) or without compensation (e.g. open source) (von Hippel and von Krogh, 2003; 2006).
Kang and Kang (2009) believe that in an age of open innovation, external knowledge is the most important source for technological innovation. Asheim and Coenen (2005) furthermore state that the innovation process of a firm is strongly shaped by its specific knowledge base. Several theories such as Kogut and Zander’s (1992) combinative capabilities theory, Cohen and Levinthal’s (1990) absorptive capacity theory and the dynamic capabilities view (Teece, 2007) state that the ability to combine knowledge from internal and external sources is a necessity for successful innovation. It is theorized by the absorptive capabilities theory that simultaneously investing in internal knowledge and acquiring external knowledge
24
provides firms with the ability to exploit new knowledge commercially (Cohen and Levinthal 1989; 1990).
The concept of open innovation is based on several innovation research streams. In their research on the future of open innovation Grasman, Enkel and Chesbrough (2010) identify several trends in open innovation. The first they identify is born globals. It is indicated that their source of competitive advantage is the leveraging and protection of their intellectual property. By opening their innovation process, SME’s can overcome their liability of smallness (Keupp and Gassmann, 2007; Van De Vrande, Vanhaverbeke, & Gassmann, 2010). A core competency of these rapidly growing SMEs is external technology
commercialization. However, according to Grasman et al. (2010) open innovation is less implemented by SMEs than by MNEs. Another trend is the change in structure. Hagedoorn and Duyster (2002) identify a strong trend in partnerships and alliances. Firms shift from standalone operations to alliances. Value creation can be enhanced by inter-organizational relationships, possibly due to knowledge transfer between the organizations (Enkel, 2010). Grasman et al. (2010) state that due to increasing complexity in technology, even large firms cannot develop new and products alone. They lastly identify a trend intellectual property becoming a tradable good instead of a source of protection for a firm. Which is in contrast with the traditional view of Schumpeter in which patents protect the work of innovators from imitation and thus enable inventors to obtain a (temporary) monopolistic profit. Grasman et al. (2010) state that patents are incentives for entrepreneurs and inventors to invest in innovation. Intellectual property as a tradable good, where patents often document complex technological knowledge, may be an indicator that tacit knowledge is being shared in open innovation networks.
25
2.6 Location
Grasman et al. (2010) state that open innovation can be organized into nine streams. One of which is the spatial perspective. The spatial perspective researches the impact of globalization in innovation. This perspective states that in one hand, being close physically to centres of excellence enables the absorptive capacity for knowledge of firms. On the other hand, globalization offers access to worldwide knowledge. This relates to the traditional International Business theories of agglomeration and clustering. Many knowledge sourcing studies focussed on finding and appropriating tacit knowledge mention the concept
(geographical) location. Scholars have questioned whether participants in the sharing of knowledge must be geographically proximate in order to effectively transfer (Cantwell and Santangelo, 1998; Gertler, 2003; Gassmann, Enkel and Chesbrough, 2010). Cantwell and Santangelo focus on international technology networks and the impact of having firms’ most highly tacit capabilities technologies dispersed internationally. They research why firms have their technologies dispersed internationally and conclude that the major drivers are the interaction between multinational enterprises (MNEs) and the local (national and regional) systems that enables the MNE to tap into local tacit advantages such as highly localized technologies and the company specific (global) strategies that utilize the development of an organizationally complex international network for technological learning. Gassmann et al. (2010) state that even though technology development is easier due to open innovation, physically close to regional centres of excellence enable the absorptive capacity of firms, giving them access to knowledge and talent. This relates to the findings of Gertler (2003) whom wonders how close one must be and what kinds of proximity matters. Another aspect that Gertler takes into account is the difficulty of diffusion of innovations across regional and national boundaries as well as other boundaries such as cultural ones. Jaffe, Trajtenberg and Henderson (1992) research the localization of knowledge spillovers in the form of patent
26
citation research. They believe that knowledge spillovers are quite invisible but that spillovers do leave behind a paper trail in the form of citations. In their research they observed that spillovers where initially geographically localized. It is however observed that geographic localization fades over time and that knowledge spillovers are not confined to closely related regions.
It is however proposed by other scholars such as Casper (2007) that geographic closeness and establishing oneself in a cluster or learning region makes it easier for tacit knowledge to spillover and travel into the firm. Clusters of high-technology firms have become an important source of economic development across the advanced industrial economies, and a central focus of technology policy. The understanding of the dynamics of clusters and networks is however still incomplete.
Besides the positive effects of knowledge spillovers there is also the concern of knowledge leakage. As discussed above, according to the resource based view firms obtain sustained competitive advantage when their strategies are not simultaneously being
implemented by a current or a potential competitor, and when other firms are unable to duplicate the benefits. They need their resources to be valuable, rare, inimitable and non-substitutable (Barney, 1991). So why would firms want to cluster and participate in open innovation networks? According to Mariotti, Piscitello and Elia (2010) multinational corporations tend to only agglomerate with other multinational corporations and not with domestic companies, unless they perceive the knowledge inflows greater and can for example access specialized knowledge they can apply to their own firm. In the financial technology sector it could be proposed that due to the fact that FinTech firms possess superior
27
financial institutions access to specialized knowledge. It is clear that there is no consensus on whether local or international knowledge spillovers leads to a higher level of innovativeness. Due to the born global nature of FinTech firms the following is proposed:
Hypothesis 3: The sourcing of external knowledge internationally leads to higher levels of
innovativeness than firms relying on domestic external knowledge sourcing.
2.7 The theory of disruptive innovation
A specific stream within innovation theory is the theory of disruptive innovation by Bower and Christensen (1995). This theory views innovation-driven growth as the key to success. Leaders of many entrepreneurial companies have used it to obtain a competitive advantage, including technology companies such as Intel and Salesforce. The theory of disruptive innovation differentiates between disruptive innovations and sustaining innovations. Sustaining innovations are similar to what Porter (1985) describes as
differentiation strategy. Sustaining improvements can be either incremental advances or major breakthroughs, but in the end all enable firms to sell more products to established, already profitable customers. Disruption is a process wherein a small company with fewer resources is able to successful challenge established businesses, by successfully targeting a segment that is overlooked by incumbent businesses and by delivering better suited products and
functionalities at a lower price. Incumbent businesses tend to focus on their most demanding, and often most profitable customers and do not notice a smaller company has filled their gaps before they have a foothold. Disrupters are able to create markets where none existed before and find a way to turn non-consumers into consumers. Christensen et al. (2015) identify two
28
ways of disruption: the first being low-end opportunity, where the low end of the market is targeted first before going mainstream. The second is new-market footholds, where the disrupter creates a market where none existed before. FinTech innovations could be either. Many FinTech companies by nature are highly specialized in a specific technology and target this at a small subgroup in the market, after which it incrementally expands into the
mainstream market. Other, such as BitCoin, introduce a new technological concept that established an entirely new market of secure digital currency.
Markides (2006) also defines two types of disruptive innovations: radical (new to the world) product innovations and business-model innovations. He states that these different types of innovations have different competitive effects. They pose radically different challenges to established firms and thus should be treated as distinct phenomena.
Christensen’s original theory of disruptive innovation from 1997 was mainly focused on technological innovation, Markides however proposes that technological innovation is fundamentally different from disruptive product innovation and disruptive business model innovation. The first type “radical, new to the word” appears to be similar to Christensen et al.’s (2015) “new-market footholds”. Business model innovation is the discovery and introduction of fundamentally different business models in existing businesses. FinTech companies applying IT-based business models, in a sector where established firms have previously only implemented financial business models could be such an example.
29
3. Research gap
Multiple studies have been conducted on the sourcing of knowledge and the transferability of it (e.g. Gertler, 2003; Nonaka et al., 2000) they are however inconclusive. Open questions regarding knowledge sourcing and transferability remain. Most research on the concept of knowledge has been conducted in knowledge intensive fields; multiple studies focus on the biotechnology industry and some on the information (communication) technology sector. The research of knowledge transfers in these complex, technological, sectors however seems far from complete. No studies that specifically focussed on knowledge transfer in the Financial Technology sector could be found. Another gap is that Grasman et al. (2010) identified that most research on open innovation has focussed on large multinational firms. It has however become clear that medium and small size firms more often participate in open innovation. Most Financial Technology firms are start-ups or medium size firms, such firms have been subjected to research far less than MNEs.
The impact of the location of knowledge sourcing also seems inconclusive. Jaffe, Trajtenberg and Henderson (1992) conclude in their research that geographic localization fades over time and that knowledge spillovers are not confined to closely related regions. It is also stated that clustering however makes it easier for tacit knowledge to spillover. This raises questions: localization fading over time implies that firms seek more external knowledge non-domestically and even internationally and yet clustering is said to be an important source of economic development (Casper, 2007). There is no complete understanding of the impact of the location where knowledge is sourced from and the impact this has on the innovativeness of firms. This thesis aims to join the on-going conversation regarding the concept of
knowledge, its transferability and its effect on innovation. Theoretical analysis identified a gap in research on knowledge sourcing, location and its impact on innovativeness of firms. It
30
has furthermore become clear that much research has been conducted in related fields, but none could be found with a specific focus on firms in the Financial Technology industry.
31
4. Methodology and data
4.1 Patent analysis
In order to test the hypotheses, a bibliometric study and analysis of patent data will be
conducted. Bibliometrics is statistical analysis of written publications. This method is widely known for analysis of books, articles, or other published works. Patents are published works that grant the assignee exclusive ownership of a novel technology for a period of time (Falk & Train, 2016).
According to Acs, Anselin and Varga (2002) economically useful knowledge that leads to innovation plays a large and important role in economic growth and international business. To understand the roles knowledge and innovation play on economic development the measurement of knowledge in- and outputs is critical. They further state that the greatest obstacle to understanding the role of innovation and technological change is the inability to measure it. Knowledge flows are often called invisible, according to Alcacer and Gittelman (2006) knowledge does however leave a paper trail in the form of patents and citations. Jaffe, Trajtenberg and Henderson (1992) also believe that the spill over of knowledge leaves a paper trail behind. They believe this spillover can be researched by conducting a patent study.
Patent citations have been used extensively in the past to measure the diffusion of knowledge across many dimensions, including geographic space. This method is widely used in science and technology (IEEE, 2010). In 2014, the most patents where registered by technology companies. This tells us that when studying knowledge flows within the technology industry, patents can serve as a good measure (Fisher, 2015). Economists refer to citations as
32
technologies are often created by building on, and improving earlier inventions. Relationships between inventions can be traced through citation links between patents (IEEE, 2010).
Research has shown that there is a strong positive relationship between citations and
technological importance (Breitzman & Mogee, 2002). Acs, Anselin and Varga (2002) state that the major aspects of the innovation process can be typically measured, by measures of technological change: inputs into the innovation process (e.g. prior art) and output, (e.g. the number of inventions that have been patented and forward citations).
4.2 Patent database
During this thesis research a patent citation database was build in Excel. A patent is something that grants the assignee (e.g. a company or an individual) exclusive ownership of a novel technology. This is restricted by geographic area (e.g. patents registered with the United States Patent Office only grant the owner exclusive ownership within the United States) and period of time. This economic monopoly, gives patents value. Patented technologies by nature, however differ greatly in quality, and the distribution of values is skewed (Scherer, 1965; Pakes and Schankerman, 1984; Pakes, 1986; Griliches ,1990). When a new patent is registered, the inventor is required to cite relevant previous patents; these citations are called backward citations or prior art. Reporting these technological antecedents is legally important, as it limits the scope of patent claims (Falk & Train, 2016). Citations that a patent receives from other patents are called forward citations. These prior art and forward citations can be seen as flows of knowledge. Studies have shown that the number of forward citations is significantly and positively related to the value of the particular patents (Trajtenberg, 1990; Albert et al, 1991; Harhoff et al., 1999; Hall et al, 2005). Forward citations of patents are utilized by researches as a proxy for patent value (Argyres and Silverman, 2004; Singh,
33
2008), knowledge transfer (Rosenkopf and Almeida, 2003) and for innovation (Ahuja and Lampert, 2001).
The database used in this thesis consists of 220 firm-assigned patents, 3553 backward citations and 2710 forward citations. This includes all firm assigned patents, in selected categories, registered at the European Patent Office ranging in years of application from January 2005 to December 2015. Data was gathered from the European Patent Office and Google Patents and documented in an Excel file. All patents in the categories Payment, Banking, Wealth Management & Capital Market, Insurance, Lending, Internet of Things, Cryptocurrency, Mobile Platform, Cyber Security and Cloud Computing were collected. In the study only firm assigned patents are used. This thesis focuses on overall the impact of knowledge sourcing on innovation in the financial industry, not on the innovativeness of individual inventors. Non-firm assigned inventors furthermore do not have disposal over the same amount of capital as firms, including them in the database would introduce unnecessary uncontrollable increase in the variance of the outcome (Dahlin et al. 2004).
Financial Categories Identifier (IPC) Technology Categories Identifier
Payment G06Q 20/00 IOT (Internet of Things) Keyword
Banking G06Q 40/02 Cryptocurrency Keyword
Wealth Management & Capital Market G06Q 40/06 Mobile Platform Keyword
Insurance G06Q 40/08 CyberSecurity Key word
Lending G06Q 20/24
G06Q 40/25
Cloud Computing Key word Table 1: Financial Technology patent categories
Categories as displayed in table 1 are based on an IP Intelligence Report published by Relecura (2015). The goal of Relecura’s research was to analyse transaction trends and identify key companies, registering, buying, selling and citing patents in the Financial Technology sector. Relecura employed two categories for analysis: Financial and
Technology. Sub categories include the categories mentioned in table 1. In this thesis one category employed by Relecura was omitted: the category Data and Analytics. This
sub-34
category contained a large amount of very broad and generic patents. After initial analysis it was found that many patents in this sub-category were not at all related to Financial
Technology. Due to the time constraints, and the excessive amount of time it would take to filter out the relevant patents, this sub-category was therefore left out. Patents in any of the above mentioned categories that were found to be not relevant for this thesis (e.g. patents that focussed specifically on biotechnology or aerospace) were manually filtered out. Patent data for 220 patents was collected and documented in Excel. Please find an overview of collected patent data in Appendix A. In order to structure the data an overall patent count was kept, all patents were furthermore assigned a unique identifier. In order to avoid duplicates, error alerts were created. The original patent registration number, filing date and publication data were registered in the database. In the event a duplicate was created, the researcher would be notified. After initial registration of the patents in the database, all citations (prior art and forward) were registered corresponding to the patents. In total there were 3553 prior art and 2710 forward citations. For each citation information was collected (e.g. name, firm,
location). Please refer to appendix A table 2 for more information.
By nature, patented technologies greatly differ in quality (Pakes & Schankerman, 1984). The quality of the patents is not a research subject in this thesis, as patents are only used to visualize knowledge flows. It should however be noted that the knowledge acquired by knowledge sourcing thus may depend on the quality of the patents. Therefore a short analysis on the quality of patents was conducted, this analysis can be found in Apendix B.
In order to test the hypotheses, data in the database needed to be prepared for analysis. Prior art and forward citations were originally registered in the database as absolute numbers, for purpose of analysis these variables for categorized into categorical variables “low”,
“medium” and “high”. The low class includes the bottom 33% of the distribution, the medium class includes the middle 33% of the distribution and the high class includes the top 33% of
35
the distribution. An additional variable was furthermore created to measure whether there is a domestic or international knowledge spillover. To create this variable (IntDomCitation) a formula was made to observe and compare the location of the patent to the location of the citation. If the location of the patent and the location of the citation matches, it was coded as 0. This indicates a domestic knowledge transfer. In the case the location of the patent and the citation did not match, it was coded as 1. This indicates an international knowledge spillover.
In this thesis innovative capacity of single patents is not measured as this says little about innovation in the financial technology industry. Innovativeness is measured on a firm level, as firms are the players in the financial technology industry. In order to measure innovative capacity of firms in the sample the 3P framework of Carayannis and Provance (2008) is used. This framework uses the factors Posture, Propensity and Performance. The first factor “Posture” is an exogenous factor that is related to lifecycle of the firm and the industry. In their analysis Carayannis and Provance (2008) used the variable “ONEW” as the innovative metric for Posture. This metric included the newness of the industry and the newness of the firm. Opposed to Carayannis and Provance’s (2008) research, this study only contains firms that operate in the Financial Technology industry. Therefore newness of the industry is equal for all firms in the sample. As many firms in the FinTech industry are start-up firms, and incumbents are often established MNE’s, there is a large amount of variance in the years of existence. As Posture is an exogenous factor, this is not included in the overall metric for innovativeness. The second factor “Propensity” is measured by Carayannis and Provance (2008) using multiple metrics. As detailed information regarding internal workings of innovation processes and culture within organizations in the sample could be obtained, this variable is only measured by external indications of innovative culture: inclusion of
innovation in the mission statement. A mission statement is defined as: “A written declaration of an organization's core purpose and focus that normally remains unchanged over time.”
36
(Business Dictionary, n.d.). The mission statement describes the core purpose and
corresponding culture of a firm. The inclusion of innovation in a firm’s mission statement is thus an indicator of innovative culture. The third factor is performance. This is measured among three levels. Output is measured with the number of total number of patents registered. In order to process the data in the model, the patents were divided in to 10 different classes. The bottom 10% of the distribution was assigned a score of 10. The top 10% of the
distribution was assigned a score of 100. Outcome is measured by the increased revenue in %. Impact is measured by whether the firm is mentioned in Forbes “The World's Most Innovative Companies” top 100 or the FinTech top 100. The propensity and performance data were then combined into a metric variable between 0-100, where 0 stands for no innovative capacity and 100 stands for perfect innovative capacity. For more information on innovation metrics, please refer to appendix C.
The innovation metric is based on Carayannis and Provance (2008) 3P model. However, since only one metric can be used for Propensity the model used in this thesis differs from the original model. Therefore the reliability of the metric is tested in reliability analysis wherein the Cronbach's alpha is analysed. Cronbach's alpha is a much used statistical method for determining scale reliability and is often used to determine reliability of likert scales. SPSS can however also perform a reliability analysis using Cronbach's alpha on interval data. A Cronbach's alpha of .662 was found based on the four-item scale. Traditional statistics state that a Cronbach's alpha of .7 is considered good. In this case the Cronbach's alpha is slightly lower than .7. The corrected item total correlations show that none of the items would substantially positively affect the reliability if they were removed from the scale. Researchers in the field of innovation are still looking for a reliable model to test innovation. As developing a model for testing innovation is not the purpose of this thesis, no further research was done into improving the reliability by adding items. For the purpose of this
37
thesis, a Cronbach's alpha of .662 was found acceptable. Please refer to Appendix C for more information.
4.3 Validity and Reliability
Validity is defined by Blumberg, Coopers and Schindler (2005) as the extent to which a test measures what we actually wish it to measure. Reliability on the other hand refers to a
measurement that generates consistent results. Out of the two, Blumberg et al. (2005) consider validity to be most essential. Reliability is mainly determined by the degree to which a study minimizes bias. In this thesis, data was continuously and rigorously checked for errors and inconsistencies to minimize the bias. The original data was collected from reliable sources such as the European Patent Office, Google Patents and published annual reports of
companies. Validity refers to the generalizability of the study. Generalizability for this study may be difficult. As it is not exactly known how many firms operate in the financial
technology sector, it could not be determined what the minimum representative sample size should be. Data in this thesis is furthermore based on patent data. It is likely that not all firms in the financial technology industry document their knowledge in the form of patents. Patent registration is costly, the PCT-procedure of application, European and national phase and the costs of a patent attorney can very from €50.000 – €100.000. On top of this annual costs of € 4.000 - € 5.000 are to be paid to remain registered (Netherlands Enterprise Agency, n.d.). Firms who do not register patents are thus automatically excluded from this dataset. In order to ensure the highest possible amount of generalizability all Financial Technology related patents registered at the European Patent Office between 2005 and 2015 are in the dataset instead of a sample.
38
5. Analysis and results
In this thesis several hypothesis will be tested. Hypothesis testing is the use of statistical tests to determine the probability of the hypothesis being true. When testing a hypothesis we assume the following: The null hypothesis (H0) is assumed to be true. The H0 states that there is no effect and uses the assumption that there is no significant difference. The alternative hypothesis (H1) is a statement that directly contradicts this.
Before testing the hypothesis, several preliminary steps were taken to ensure high data quality. This chapter described the alternative hypotheses and the corresponding methods of analysis that will be used to test the hypotheses. The hypothesis will be tested using the statistical application SPSS and results will be reported.
5.1 Hypothesis 1
H1: There is a significant relation between the type of knowledge a firm has and the level of innovativeness.
Under the H0 it is assumed that there is no significant difference between the different types of knowledge. Therefore the type of knowledge all have the same effect on the level of innovativeness. The mean level of innovativeness should be similar for all types of
innovations. In order to test the hypothesis and to conclude whether the H0 can be rejected, two variables are analysed. The first is the independent variable “Invention Type”. This represents the knowledge of the firm and the type of knowledge. “Invention Type” is a categorical variable, with two categories: “Financial” and “Technology” The dependent
39
variable is the innovation rate, which is an interval variable between 0-100. This variable reflects the firms level of innovativeness.
Variable type Variable name Nature of variable Level
Independent Invention Type Categorical 2 independent groups
Dependent Innovation Rate Interval 1-100
Table 2: Variables H1
In the Excel database invention types were listed by their names “Financial” and
“Technology”. In order to use the categorical variable of Invention Type in SPSS, it was recoded into a new variable. “Financial” was recoded as 0 and “Technology” was recorded as 1. Labels with the original names were given to the recoded values.
As the independent variable is categorical with two categories, and the dependent variable is an interval variable, a two independent sample t-test will be performed in order to determine whether the means of two independent groups differ.
An assumption of a two independent sample t-test is that the two populations must be independent from each other. This means that the observations from the first must not have any effect on the observations of the second. As the patents that are measured are assigned to either the Financial or Technology category they are considered independent. It was
furthermore found that the dependent variable is approximately normally distributed for each group.
The first step in data analysis is checking the descriptive statistics of the data, this is done to detect errors and missing values. No errors or missing values were found. Please refer to appendix D for more information and output. A two independent sample t-test was then performed.
InventionType N Mean Std. Deviation Std. Error Mean
Innovation Rate Financial 101 27.101 18.1555 1.8065
40
Table 3: Group Statistics
The group statistic, as displayed in table 3, shows that the mean innovation rate (31.940) for invention type technology is higher than the mean innovation rate for invention type financial (27.101).
As the two groups must be independent from each other, Levene's test for equality of variances is performed. This tests the homogeneity of variances assumption. The test assumes that the variance is the same in both samples. The standard deviation is the square root of the variance. Studying this variable gives us an indication of the variance. Initial observation of the group statistics table, as displayed in table 3, shows that the standard deviations of the two groups (18.1555 for Financial and 16.2853 for Technology) are relatively close. This indicates little variance. Levene's test for equality of variances, as displayed in table 4, reports an F of 1.610 and a p value of .206, which is not statistically significant as .206 > α .05. This means that equal variances can be assumed. Therefore the T value and P value for the first row “Equal variances assumed”, displayed in table 4: Independent samples test can be further interpreted.
Levene's Test for Equality of
Variances t-test for Equality of Means
F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper Innovation Rate Equal variances assumed 1.610 .206 -2.083 218 .038 -4.8395 2.3228 -9.4175 -.2615 Equal variances not assumed -2.065 202.990 .040 -4.8395 2.3436 -9.4603 -.2187
41
Table 4: Independent Samples Test
The P value is .038, which is < α .05. Due to this the H0 can be rejected. Data supports the claim at the 95% significance that there is a relationship between type of knowledge and level of innovativeness. The data furthermore show that the level of innovativeness was
significantly higher (31.940) for technology knowledge than financial knowledge (27.101). In order to further analyse if a specific type of knowledge has more impact on the level of innovativeness, invention type sub-categories are analysed.
Variable type Variable name Nature of variable Level
Independent InventionSubType Categorical 10 independent groups
Dependent Innovation Rate Interval 1-100
Table 5: Variables H1
In the database invention sub types were listed by their names “Payment”, “Banking”, “Wealth Management & Capital Market”, “Insurance”, “Lending”, “Internet of Things”, “Cryptocurrency”, “Mobile Platform”, “CyberSecurity” and “Cloud Computing”.
In order to use the categorical variable of Inventions Sub Type in the analysis, it was recoded into new variables. The variables were recoded as displayed in table 6.
Original value Recoded value
Payment 1
Banking 2
Wealth Management & Capital Market 3
Insurance 4 Lending 5 Internet of Things 6 Cryptocurrency 7 Mobile Platform 8 CyberSecurity 9 Cloud Computing 10
Table 6: invention sub types recoded
As the independent variable is categorical with 10 categories, and the dependent variable is an interval variable, a one-way ANOVA test will be performed. One-way ANOVA
42
can be used to determine if there are any significant differences between the means of the 10 groups. The original source of patent information (European Patent Office and Google Patents) classify patents by class. Classes may however not be fully exclusive. It is possible that a patent belongs to more than one type (e.g. a patent registered in the internet of things category, may also contain knowledge regarding a cloud technology). It can thus not be stated that the categories are either fully independent or fully dependent. As independence is a requirement for a one-way ANOVA, the Levene's test for equality of variances will be performed to determine whether there is independence.
In order to detect missing values or errors, the data was analysed by checking the descriptive. No errors or missing values were found. Please refer to appendix D for more information and output.
One of the assumptions for an ANOVA is the homogeneity of the variances of the residuals. In order to test this Levene test was conducted. Levene's test for equality of variances reports a P value of .066, which is not statistically significant as .066 > α .05. The assumption has thus been met. It can be assumed that the variables are independent.
Levene Statistic df1 df2 Sig.
1.934 7 191 .066
Table 7: Test of Homogeneity of Variances
As equal variances can be assumed the data from the ANOVA is further analyzed. The data, as displayed in table 8, shows that there is no statistically significant effect of Invention Sub Type on the innovation rate F (7, 191) = 2.052 and P = .051 which is slightly higher than α .05. In case of significance one would also expect the F value to be lower. Therefore further results are not interpreted.
43
Sum of Squares df Mean Square F Sig.
Between Groups 4080.143 7 582.878 2.052 .051
Within Groups 54266.601 191 284.118
Total 58346.744 198
Table 8: ANOVA
In this section hypothesis 1: There is a significant relation between the type of knowledge a
firm has and the level of innovativeness was analysed using a two independent sample t-test and a one-way ANOVA. Data supports the claim at the 95% significance that there is a
relationship between type of knowledge and level of innovativeness. It was furthermore found that knowledge documented in “Technology” category patents leads to a higher mean of innovativeness (31.940) than “Financial” knowledge (27.101). In order to clarify which specific types of knowledge lead to higher levels of innovativeness, the relationship between The variables “InventionSubType” and “Innovation Rate” was analysed in an ANOVA. It was found that there is no statistically significant relationship P = .051 > α 0.5. Therefore it could not be determined which specific type of knowledge leads to higher levels of innovativeness. The H0 could not be rejected by a difference of .001. If a significance level of 90% would have been assumed, the H0 could be rejected. It is suspected that this slight difference is due to the nature of the data, and the uncertainty whether the variables in “InventionSubType” are truly independent. This is however not statistically supported.
