The Relevance of Cluster Economics for the FinTech Industry. Theory and Analysis

Table of Contents

Abstract

List of Figures

List of Tables

List of Abbreviations

1 Introduction

2 Methodology

3 Literature Review and Theoretical Framework
3.1 FinTech
3.1.1 Definition
3.1.2 FinTech Business Segments and Taxonomy
3.2 Economic Clusters
3.2.1 Theoretical Foundations
3.2.2 Porter’s Diamond Model
3.2.3 FinTech Clusters

4 Qualitative Analysis
4.1 Conceptualization of FinTech Clusters
4.2 Application of Porter’s Diamond Model to the FinTech Industry

5 Quantitative Analysis
5.1 Overview of Variables and Data Sample
5.2 Regression Models
5.3 Discussion of Results
5.3.1 Model 1 – Factor Conditions
5.3.2 Model 2 – Demand Conditions
5.3.3 Model 3 – Related and Supporting Industries
5.3.4 Model 4 – Strategy, Structure, and Rivalry

6 Conclusion

7 Appendix

Bibliography

Abstract

While the FinTech industry overall has experienced strong growth in recent years, the geographical presence of FinTechs is not homogeneously distributed. Instead, it can be observed that certain FinTech hotspots are emerging. Within the political arena there is a discussion about how to make business locations more attractive for the accumulation of FinTechs. In this context the theory of economic clusters has to be investigated. What are the underlying factors that contribute to the formation of FinTech clusters? This paper builds on initial studies on this topic and sheds light on various factors that could be decisive for the agglomeration of FinTechs. Using Michael Porter's Diamond Model, various factors were derived, which were then statistically examined using a data sample for the European Union. The results of the study show that besides the availability of positive externalities, such as specialized talent pools, universities, and accelerators, demand factors such as total market size, internet connectivity, and financial literacy are relevant. Furthermore, the results suggest that the presence of competitors and similar firms, as well as the strength of intellectual property protection laws, are related to the size of FinTech clusters. Based on the results, recommendations for policy makers and FinTech entrepreneurs are provided.

List of Figures

Figure 1: Global Transaction Values per FinTech Segment

Figure 2: FinTech Cluster Model

List of Tables

Table 1: FinTech Business Segments

Table 2: Dimensions and Characteristics of FinTech Business Models

Table 3: Overview of FinTech Cluster Research

Table 4: Overview of Hypotheses

Table 5: Overview of Variables

Table 6: Summary Statistics of Data Sample

Table 7: Model 1 Estimates – Factor Conditions

Table 8: Model 2 Estimates – Demand Condition

Table 9: Model 3 Estimates – Supportive and Related Industries

Table 10: Model 4 Estimates – Strategy, Structure, and Rivalry

List of Abbreviations

B2B Business-to-Business

e.g. for example (Latin: exempli gratia)

et al. et alii

ETER European Tertiary Education Register

EU European Union

FCA Financial Conduct Authority

Fig. Figure

HR Human Resources

i.a. inter alia

IPP Intellectual Property Protection

IT Information Technology

p. page

pp. pages

RQ Research Question

VC Venture Capital

WEF World Economic Forum

1 Introduction

The FinTech industry, as a hybrid of financial services and technology, has been experiencing an upswing now for several years. Investment in this new industry is growing strongly. While in 2014 $51.3 billion were invested globally, this figure had already reached $135.7 billion in 2019, an increase of over 260% (KPMG, 2020).

The success of FinTech is multifaceted: innovative products can satisfy previously unmet customer needs, more modern, digital customer interfaces enable a better customer experience, and relatively low regulation, compared to incumbents, gives FinTech a cost and efficiency advantage (Navaretti, Calzolari, Mansilla-Fernandez, & Pozzolo, 2018).

FinTechs are also attracting raising attention from a political perspective, as they entail new regulatory risks, but also contribute to the economic development of regions and cities as drivers of innovation (Cockerton, 2016). As FinTechs seem to agglomerate in certain locations, like the famous FinTech hotspots in London and New York, practitioners and policy makers increasingly develop interest in the cluster effects from which FinTechs benefit. Potential cluster effects often mentioned in the literature are i.a. access to an existing financing infrastructure, availability of talent pools, easy and trustful cooperation with incumbents, and availability of specialized companies from which services can be outsourced (Jutla & Sundararajan, 2016).

Since London, as the “global FinTech capital” (Bain, 2020), could, according to some experts, be weakened by the Brexit, there is the opportunity for other regions to strengthen its own FinTech clusters (Sohns & Wójcik, 2020). The European Commission for instance, has already published several action plans explaining measures to make the European Union (EU) more FinTech-friendly and promote innovation in this new sector (European Commission, 2018). Other regions are also striving to attract promising FinTechs and strengthen the business location for this relatively new industry (Maddock, 2019; Mueller & Piwowar, 2019).

Although the relevance of economic clusters in the FinTech industry is of great importance for various stakeholders such as entrepreneurs, investors, and politicians, research on this topic remains rare. Therefore, the thesis aims to give a current overview of the scientific literature on FinTech and the relevance of cluster economics, and to provide evidence for this phenomenon. The contribution to the existing literature is provided by answering the following research question (RQ):

“What are the underlying factors that contribute to the formation of FinTech clusters?”

In addition to the introduction, the thesis consists of four further chapters. Chapter 2 discusses the methodology of the thesis, explaining the research approach in greater depth. Chapter 3 discusses the existing scientific literature on the topic of FinTech and economic clusters. In this chapter, the theoretical background of FinTech including definition and taxonomy is established and fundamental basics of economic cluster theory is provided. In Chapter 4, Porter's Diamond Model is applied to analyze the relevance of economic clusters for the FinTech industry qualitatively. In this context, testable hypothesis for the subsequent empirical analysis are extracted. Chapter 5 deals with the conducted statistical analysis that is performed to test the hypotheses from the qualitative analysis with a data set of the EU. After explaining the variables, the underlying data, and the regression models, the results of the analysis are critically discussed. The thesis concludes with Chapter 6, which summarizes the work and its most important findings, highlights limitations, and proposes recommendations. The recommendations address both policy makers who want to strengthen FinTech clusters as well as FinTech entrepreneurs who deal with the choice of a suitable location for their business.

2 Methodology

In the following, the methodological approach is described. The aim is not only to show what the main steps in the research were, but also to critically assess the extent to which the chosen methodologies best serve to answer the research question.

The research approach can be divided into three components: (1) Literature review, (2) qualitative analysis, and (3) quantitative analysis. Each component was carried out in such a way that on the one hand the achievement of the research objective is ensured, and on the other hand that the scope of this thesis is adhered to with scientific diligence.

The first component, the (1) literature review, aims to provide a theoretical basis which is a prerequisite for the subsequent analyses. Both textbooks, mostly published by Springer Verlag, and peer-reviewed scientific papers were used as resources to ensure a high level of scientific quality and objectivity.

In the second component, (2) qualitative analysis, Porter's Diamond Model (see Chapter 3.2.2) was used as an analysis tool. The goal of the analysis was to clarify in which dimensions cluster economics are relevant for the FinTech industry by deriving testable hypotheses on the underlying factors for cluster formation. The Diamond Model enjoys great popularity among scientists and practitioners due to its structured and clear design. Various studies use the model for the analysis of economic clusters (Amiri Aghdaie, Seidi, & Riasi, 2012; Chung, 2016; Eickelpasch, Lejpras, & Stephan, 2011; Mann & Byun, 2011). In the qualitative analysis, a meta perspective was adopted to analyze the general factors of FinTech clusters. In contrast to this, the analysis of a case study, which is also frequently carried out in this context, would be an alternative option (Gnirck & Visser, 2016; Haas & Bierbaumer, 2016; Linder, 2018). While case studies have the advantage of providing in-depth details and can yield very concrete knowledge, a high-level analysis is more appropriate for the purpose of this paper, as it gives less weight to case-specific factors and provides a more objective basis for the subsequent quantitative macro-analysis.

The third component, the (3) quantitative analysis, forms the evidence part of the work by statistically testing the hypotheses made using a data sample of the EU. For the analysis, the data sample was independently compiled from several sources. For each hypothesis, a variable was identified, for which appropriate data were collected and evaluated using the statistics software JMP by SAS. The data sources differ depending on the variable (see Chapter 5.1), but the main sources were Eurostat and Crunchbase, renowned resources used in several studies (e.g. Garlick, 2015; Haddad & Hornuf, 2019).

The data sample is cross-sectional and thus reflects the most current status of the data. This has the advantage that hypotheses can be approved or disproved relatively easily and a large number of variables are available. The disadvantage is that no temporal dimension is included and thus potentially endogeneity and biased data are present. Especially the aspect of endogeneity should be mentioned as various variables in the data set can be influenced by reverse causality. For example, the number of resident venture capital funds may influence the number of FinTechs, but it is also possible that the number of FinTechs may influence the number of venture capital (VC) funds. Therefore, it should already be noted that the quantitative study is not intended to prove causality, but only to discover correlations between variables. Whether, and to what extent, causality could be present will be briefly examined in the discussion of the results (Chapter 5.3). The limitations of cross-sectional data could be circumvented by using longitudinal panel data, but at the time of the analysis, a sufficient data set is not available. Since the analysis of a severely restricted number of variables is only of limited utility, it was decided to carry out the analysis with the present cross-sectional data. If longitudinal panel data become available, it is recommended to repeat the analysis and perform appropriate endogeneity tests.

3 Literature Review and Theoretical Framework

In this section the two underlying thematic areas are introduced. First, the relatively new topic "FinTech" with its multifaceted definitions, segments and taxonomies is explained. Then, the theoretical foundations of economic clusters are presented including the theoretical framework, the Diamond Model by M. Porter, and the state of research on FinTech clusters.

3.1 FinTech

3.1.1 Definition

From its origins in 1972, when “FinTech” was firstly used by Abraham L. Bettinger (1972, p. 62), former Vice President of Manufacturers Hanover Trust, as “an acronym which stands for financial technology, combining bank expertise with modern management science techniques and the computer”, the term has gained significant popularity within the financial realm. But, even though the term “FinTech” is used at an increasing rate, scholars and practitioner’s alike sill have not reached a consensus about a clear definition of the term. Semantically, the term “FinTech” is the abbreviation of “Financial Technology” and thus represents a combination of “Financial Services” and “Information Technology” (Cambridge Dictionary, 2020). However, when it comes to defining the term, opinions diverge.

Basically, a distinction can be made between two different interpretations of the term: FinTech as (1) a new type of business model, and as (2) a whole business sector or industry. Supporters of the first interpretation of FinTech as a new form of business models focus on the innovative creation and delivery of financial services and products to the customer. For example, Leong, Tan, Xiao, Tan, & Sun (2017) define FinTech as the "design and delivery of financial products and services through technology". Mention (2019) adds that FinTech “can be used to describe any innovation that relates to how businesses seek to improve the process, delivery, and use of financial services”. Advocates of the second interpretation, on the other hand, place more emphasis on the economic perspective and speak of a new industry that exists separately or as a sub-sector from the traditional financial industry. Čižinská, Krabec, & Venegas (2016), for example, see FinTech as an “economic industry composed of companies that use technology to make financial services more efficient”, while Shim & Shin (2016) define FinTech as an “emerging financial services sector that includes third-party payment, MMF (money market fund), insurance products, risk management, authentication, and peer-to-peer (P2P) lending”.

For the purpose of this thesis, a rather broad definition of FinTech, tied to the second type of interpretation, is most suitable due to the following two reasons: Firstly, the thesis discusses a holistic view of the FinTech industry, in which macro- and micro-economic factors, such as competitive landscape, are analyzed. With a few exceptions, factors on a business model level are not in the scope of this thesis. Secondly, in the empirical research part data from Crunchbase.com are used. Crunchbase Inc. (2020) also assigns the term "FinTech" to the second type of definition category, as they refer to FinTech as a sector which "includes everything from mobile banking technology to investment apps to cryptocurrency". This is important because the definition determines which companies are declared as a FinTech and subsequently can be found in the analyzed database. It is noted that the choice of the definition may have limiting effects on the empirical analysis in Chapter 5, as Crunchbase Inc. does not clearly specify which attributes are decisive for them to declare companies as FinTech. However, as Crunchbase is a renowned and often cited data source, the validity of the declaration will be assumed.

3.1.2 FinTech Business Segments and Taxonomy

In the literature there is no consensus about how to segment the business areas that FinTechs are active in. This is partially due to the fact that the industry is highly dynamic and new business models emerge constantly. In general, it can be observed that FinTechs position themselves along the entire value chain of banks and other financial companies and offer innovative products and services to customers, usually specializing in one area of the value chain (Navaretti et al., 2018).

A comprehensive segmentation of the business areas of FinTech is provided by Dorfleitner, Hornuf, Schmitt, & Weber (2017). They generally divide the business segments of FinTech in four categories: (1) Financing / Funding, (2) Wealth Management, (3) Payment Services, and (4) other FinTechs. Table 1 presents a more detailed representation of the segments and sub-segments.

Abbildung in dieser Leseprobe nicht enthalten

Table 1 : FinTech Business Segments

Source: Own illustration, following Dorfleitner et al. (2017), p. 11.

Although the segmentation is often referred to, there is no financial information, such as market sizes, for those specific areas. To get an impression of the diverging relevance of the individual segments, it is useful to review Statista's segmentation including transaction values and forecasts. Although Statista's Digital Market Outlook (2020) uses a different classification than Dorfleitner et al. (2017), the dominant role of digital payments with more than 80% of the global transaction value has been evident in the past and is expected in the future, as shown in Figure 1.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1 : Global Transaction Values per FinTech Segment

Source: Own illustration, following Statista (2020), p. 6.

After segmenting the FinTech industry into different businesses, it is outlined what FinTechs actually do and which characteristics constitute their business models. Providing a consistent taxonomy of FinTech business models is important for understanding the basic aspects of FinTech businesses and distinguishing them from traditional players within the financial industry.

The term “taxonomy” can be defined as a “set of dimensions each consisting of a set of mutually exclusive and collective exhaustive characteristics” (Nickerson, Varshney, & Muntermann, 2013). In the scientific literature, several taxonomies for FinTech exist, which are summarized in Appendix 1. In the following, the taxonomy of Eickhoff, Muntermann, & Weinrich (2017) is briefly presented. This taxonomy provides a coherent basis for the further development of the thesis for three reasons. First, it covers a wide range of FinTech business models, since it has no restrictions, e.g. regarding the customer group (compare e.g. Gimpel, Rau, & Roeglinger, 2017). Secondly, the selected dimensions and characteristics were empirically validated. Third, for validation data from Crunchbase was used, which ensures a sound fit for the empirical analysis later in this thesis.

The taxonomy consists of six dimensions, namely (1) Dominant Technology Component, (2) Value Proposition, (3) Delivery Channel, (4) Customers, (5) Revenue Stream, and (6) Product/Service Offering. Each dimension can have several characteristics which, in combination, ultimately form the FinTech business model. A detailed illustration of the characteristics is given in Table 3. In order to validate the taxonomy, ten archetypes were identified based on Crunchbase data, which reflect the most relevant FinTech business models. Appendix 2 illustrates these ten archetypes.

An example archetype is the "Payment Service", which is worth noting because of the already mentioned relevance of FinTech in the payment area. The dominant technology component here is "Transaction Processing System" and the value proposition is "Convenience / Usability", addressing the complicated and sometimes insecure credit card payments – online and offline. The delivery channel is primarily via App and is mainly served to business customers. According to Eickhoff et al. (2017), the revenue stream is mostly "unknown". This could be due to the fact that various models exist for generating revenue, ranging from a monthly user fee to a fixed or variable fee per transaction. Lastly, the product offered is not surprisingly a "payment service".

Abbildung in dieser Leseprobe nicht enthalten

Table 2 : Dimensions and Characteristics of FinTech Business Models

Source: Own illustration, following Eickhoff et al., (2017), p. 10.

3.2 Economic Clusters

3.2.1 Theoretical Foundations

In the following sub-chapter, the theoretical background of economic clusters is established. Besides explaining the origin of the theory and the economic forces underlying the formation of clusters, the key components of such, and the resulting advantages on a company and regional level are elaborated.

The origin of the theory of economic clusters can be traced back to Alfred Marshall's work "Principles of Economics" (1890), in which he speaks of "industrial districts", which are accumulations of companies with similar demands. The term "cluster" became well established after Michael Porter (1990) used it in his work "The Competitive Advantage of Nations" and defined a cluster as “a geographically proximate group of interconnected companies and associated institutions in a particular field, linked by commonalities and complementarities”. Even though there are multiple definitions in the literature, Porter’s definition is the most widely-known one among scholars and practitioners (Martin & Sunley, 2003). Moreover, the geographical scope of clusters is multifaceted and can range “from a region, a state, or even a single city nearby to a group of neighboring countries” (Porter, 2000).

The concept of economic clusters enjoys high relevance in both politics and science around the globe (Li, Webster, Cai, & Muller, 2019). Famous examples such as Silicon Value, Hollywood, or Wall Street illustrate the theory in glamorous manner. At first glance, though, the popularity of cluster economics may be surprising as ongoing globalization seems to reduce the relevance of geographical proximity in the business world (Cairncross, 1997; Gray, 1998; O’Brien & Royal Institute of International Affairs, 1992). However, Porter (1998) rejects this argument as follows:

“In a global economy – which boasts rapid transportation, high speed communications and accessible markets – one would expect location to diminish in importance. But the opposite is true. The enduring competitive advantages in a global economy are often heavily localized, arising from concentrations of highly specialized skills and knowledge, institutions, rivalry, related businesses, and sophisticated customers” (Porter, 1998).

In the literature there are a variety of mostly micro-economic approaches to explain the formation of clusters. In the current publication by E. Feser & Bergman (2020), five concepts that explain the emergence of clusters are highlighted: (1) local externalities, (2) innovative environment, (3) cooperative competition, (4) rivalry, and (5) path dependence.

The first concept goes back to Marshall (1890), who noted that agglomeration of firms results in positive local externalities as they benefit from labor pooling, knowledge spillovers, partnerships, shared access to R&D infrastructure, and entice supplier and service firms to locate there as well. These externalities lower the market entry barriers and help the co-located companies to reduce costs and stay competitive on the global market as their productivity increases while risks are mitigated (Pike, Rodriguez-Pose, & Tomaney, 2008).

Next, clusters can be seen as an innovative environment in which firms’ knowledge exchange serves as a mechanism to innovate. According to Roelandt & den Hertog (1999), innovation is a dynamic social process which depends on trading tacit, non-codified knowledge. This information exchange is mostly successful when done between individuals, not only businesses, benefiting the formation of economic clusters (Lundvall, 1999).

Further, cooperative competition is regarded as one major force driving cluster formation. In areas such as “lobbying, foreign market research, joint export promotion, trade fairs, and specialized infrastructure investments” (E. Feser & Bergman, 2020) cooperation between competing firms is beneficial as it reduces costs and helps to build trust. Often, traditional, social, or cultural ties govern the relationships between competitors.

Porter (1990) emphasizes the role of rivalry in cluster establishment. From a neo-classical perspective, competing firms must ensure low-cost levels, innovative product development, and productivity increase. This phenomenon is enforced when companies co-locate in a specific region, as they do not only compete for customers, but also for funding, human resources (HR), media coverage, and political support. The co-location additionally helps to better monitor the performance of competing firms and thus increase their motivation further.

Lastly, path dependence can be an important factor for cluster emergence. E. Feser & Bergman (2020) state that “path dependence refers to the general notion that technological choices […] can assume a dominant lead over alternatives and be self-reinforcing […]” (p. 13). The decision, whether accidental or not, to establish a company in a certain location can be decisive for the development of an economic cluster, even in the absence of the other listed reasons. The so-called first-mover advantage can therefore play a critical role in the formation of clusters, as it can be determinant for the co-location of further companies (Martin, 2010).

After having discussed the most commonly noted underlying concepts of economic cluster formation, the main components of clusters will be established in the following. These components are also recalled in the conceptualization of FinTech clusters for the qualitative analysis (Chapter 4.1).

According to Feser (2004) clusters generally consist of three components: (1) Trading sectors, (2) related sectors, and (3) supporting institutions. The first component, “trading sectors”, describes the buyer-supplier relationship between several companies along their value chain. Key players in this category include production companies, capital and consumer good suppliers, and research and development service providers. The trading sector is generally more important in industry clusters than in tech clusters. Geographical proximity can reduce friction along the value chain (vertical trade relations), e.g. in the supply of materials from the manufacturer to the production plant (St. John & Pouder, 2006). While vertical trade relationships can exist in tech clusters, e.g. in the supply of laboratory materials, these play a rather minor role, while the following two components gain importance. The second component, "related sectors", covers companies that use similar technologies, want to attract the same talent, or pursue similar business strategies. The third component, "supporting institutions", is made up of players who provide complementary services to the companies located in the cluster. Typical institutions in this category are educational institutions, such as universities, training centers, autonomous R&D facilities, as well as government agencies that are responsible for regulatory or development efforts for the cluster.

Due to the mentioned advantages that emerge when clusters evolve, the region in which the cluster is based in benefits from economic prosperity. It appears that these benefits are greater than those resulting from increased mobility and economic deregulation (Benner, 2012). Wennberg & Lindqvist (2010) found that companies located in clusters not only have a higher probability of survival, but also pay higher salaries, generate more jobs, and increase overall tax revenues. In the long term, clusters create a stronger middle class, which, due to higher living standards, strengthens the housing market and other regional economies (Bailard Real Estate, 2017; Beriatos & Gospodini, 2004; Mommaas, 2004). All the mentioned advantages for both the companies and the regional economy itself lead Porter (1998) to the conclusion that a cluster’s “value as a whole is greater than the sum of its parts”. Due to the emergence of several highly successful clusters and the associated positive effects for the economy, clusters are increasingly recognized as a market-based policy tool to foster regional economies and competitiveness (Held, 1996).

Economic clusters can be found in a wide spectrum of industries. However, the reasons why, where, and how clusters form can differ from sector to sector due to the varying business models and demands. Appendix 3 gives an overview of commonly analyzed clusters and their main attributes.

[...]