End-user adaptation of cryptocurrencies

MSc Business Administration

Track: Digital Business

Master Thesis

End-user Adaptation of Cryptocurrencies

by

Marco Göbert

11828854

22

of June 2018

15 ECTS, final version

Research period: 05/02/2018 – 22/06/2018

Supervisor/Examiner:

Prof. em. dr. ir. Hans J. Oppelland

i

Statement of Originality

This document is written by Marco Göbert 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.

ii

Table of Content

1.4 Structure of the Thesis ... 4

2 Literature Review... 4

2.1 Definition of Blockchain and its Characteristics ... 5

2.2 Blockchain Network Types ... 8

2.2.1 Public Blockchain ... 9

2.2.2 Consortium Blockchain ... 9

2.2.3 Private Blockchain ... 10

2.3 Blockchain Participants and Roles ... 10

2.3.1 Blockchain user ... 10

2.3.2 Provider ... 10

2.3.3 Certificate authority (CA) ... 11

2.3.4 Blockchain developer ... 11

2.3.5 Blockchain network operator ... 11

2.4 Cryptocurrencies ... 12

2.4.1 Critical assessment ... 12

2.5 Further Current Blockchain Applications ... 15

2.6 Technology Adoption Theories ... 17

2.6.1 Technology acceptance model (TAM) ... 18

2.6.2 Extensions of the technology acceptance model (TAM) ... 19

2.6.3 Unified theory of acceptance and use of technology (UTAUT) ... 21

2.6.4 Diffusion of innovations theory ... 22

2.7 Research Question ... 23 3 Research Concept ... 24 3.1 Conceptual Model ... 24 3.2 Propositions ... 26 4 Method Selection ... 27 4.1 Sample ... 28 4.2 Measures ... 28

iii 5 Data Collection ... 29 6 Research Results ... 30 6.1 Demographics... 30 6.2 Interpretation ... 31 6.3 Interview Results ... 32 7 Conclusions ... 33 7.1 Managerial Implications ... 33 7.2 Limitations ... 34 7.3 Future Research ... 34 8 Bibliography ... 36 9 Appendices ... 44 Appendix A: Demographics ... 44

Appendix B: User Information ... 48

Appendix C: Survey Questions (English) ... 49

Appendix D: Survey Questions (Dutch) ... 53

Appendix E: Survey Questions (German) ... 57

Appendix F: Interview Findings ... 61

1

1 Introduction

1.1 Research Problem

The blockchain has been an increasingly trending topic in recent years and has only lost little of its expectations in the last year (Gartner, 2016; Panetta, 2017). Media coverage of the Bitcoin – one of the most known applications of the blockchain technology (Beck, Avital, Rossi, & Thatcher, 2017) – and its price of nearly US$20,000 in December 2017, caused increased interest by the general public (Beck et al., 2017; Corcoran, 2017).

However, research so far mainly focused on the technological side of blockchain and only investigated few potential applications in a limited amount of industries, such as the financial services industry (Lindeman, Rossi, & Tuunainen, 2016; Risius & Spohrer, 2017).

Risius and Spohrer (2017) therefore pointed out in their work that researchers should focus among other categories on users and society in relation to blockchain technology. As users, they understood individuals using the blockchain technology for transactions. In regards to society, they called for more research being conducted on the societal consequences the blockchain technology implies. In particular, Risius and Spohrer (2017) saw a lack in the literature investigating which blockchain features and designs would influence the adoption and usage of the technology.

The so far narrow technical focus of blockchain technology in recent academic literature emphasises the importance of establishing a better understanding of behavioural aspects of users and their impact on the business world. This thesis will address this currently existing knowledge gap by exploring the factors influencing the adoption of cryptocurrencies which are based on the blockchain technology.

2

1.2 Research Objective

The objective of this Master thesis is to provide a thorough understanding of the blockchain technology and one of its major applications – cryptocurrencies. This research will derive from current literature about technology adaption and acceptance models factors that are relevant for the blockchain technology and to elaborate which of those factors support the end-user adaptation of cryptocurrencies. Additionally, the research aims at exploring which other additional factors may be of relevance for the end-user regarding the adaptation of the technology. Furthermore, the thesis aims at providing recommendations for managers in the business world on how they can positively influence those factors to achieve a faster adaptation of cryptocurrencies or even other blockchain-based services. The findings may also be used as a foundation to widen the scope of potential blockchain applications in other industries.

1.3 Research Method

To achieve the aforementioned research objective, the thesis will first analyse the literature on blockchain technology and its various applications with a focus on the cryptocurrency market. Thereafter, the Technology Acceptance Model (TAM) by Davis (1986), its extension (TAM2) developed by Venkatesh and Davis (2000), and the Unified Theory of Acceptance and Use of Technology (UTAUT) will be evaluated to derive relevant factors for the end-user adaptation of blockchain-based cryptocurrencies.

To further explore whether those and additional factors not previously identified in the literature are relevant for the adaptation, an online survey with self-selection sampling was conducted at one moment in time (cross-sectional). Potential participants for the survey were reached via the personal network of the researcher, online social networks (with focus on blockchain technology and cryptocurrencies), an online research provider, and by approaching

3 others students and former working colleagues to achieve a more heterogeneous database for analysis.

As currently few products and services use the blockchain technology, the survey was limited to the blockchain application in cryptocurrencies (Lorenz et al., 2016).This approach ensured that enough respondents could be reached. Based on the previously derived factors, end-users were asked which factors they perceive as relevant for using cryptocurrencies to its full potential or consider as relevant for using cryptocurrencies in the future. The respondents could also to name additional factors.

Initially, the survey asked participants whether they currently own any cryptocurrency. This approach was selected for two reasons. First, it accounts for the fact that users familiar with cryptocurrencies might have a better experience and understanding of the underlying blockchain technology and, therefore, consider other factors compared to non-users as more relevant. Second, it accounts for the fact that non-users might be prone to indicate a different willingness to use cryptocurrencies compared to the case of actual usage. Therefore, participants were initially asked whether they are currently using cryptocurrencies to allow for a separate analysis of the two groups. Additionally, the survey asked participants whether they plan to use cryptocurrencies in the future to allow for a more detailed analysis.

Even though the intention of non-user to utilise cryptocurrencies in the future might be uncertain at this point, the factors supporting the potential usage can still be relevant and are worthwhile exploring. The derived results can help to provide businesses with recommendations on which factors to focus on, primarily if they seek to increase their user base.

Next to the survey, the researcher interviewed several experts in the area of cryptocurrencies and blockchain technology to explore the view of the professional world and

4 to potentially identify additional factors relevant to the adaptation. Moreover, the research results derived from the survey were discussed with the experts for validation purposes and a critical debate.

1.4 Structure of the Thesis

In the following, this work will analyse the current research on blockchain technology, its different network types and the different parties involved in a blockchain network. Consequently, the subsequent part of the literature review will define and critically assess cryptocurrencies as well as shortly illustrate further blockchain applications. At the end of the section, literature about technology adoption will be evaluated, and the research question will be stated.

The next section focuses on the research concept including the conceptual model and propositions as well as an explanation of the sample and used measures. Thereafter, the method selection and data collection sections follow. The research results section addresses the research findings and interpretation of the latter. In the end, a conclusion is drawn, and limitations as well as aspects for potential future research are presented.

2 Literature Review

First, this section of the thesis will elaborate on the current state of knowledge and research about the blockchain technology the following sections will describe the different blockchain network types as well as participants and roles in a blockchain network. Thereafter, a more detailed look at cryptocurrencies as well as other blockchain applications follows. In the end, the research question will be developed.

5

2.1 Definition of Blockchain and its Characteristics

Before considering different research about blockchain and some of its applications in the business world, it is essential to define and understand the term “blockchain” and its underlying technology in the first place.

The blockchain technology has its origin in a paper published by a person using the pseudonym Nakamoto (2008) in the year 2008. In the paper, a peer-to-peer payment network for electronic cash was theoretically described that was later used as the foundation of the cryptocurrency Bitcoin. Nowadays, the blockchain technology is generally defined as being a distributed system that stores an event log which is linear, consistent, as well as immutable and records the transactions between parties in the network (Risius & Spohrer, 2017). The system encrypts each transaction, and a record can only be changed by performing another transaction and proof-of-work respectively to correct the mistake (Nakamoto, 2008). Each transaction is added to the ledger and is visible to all parties in the network (Nakamoto, 2008). Therefore, transactions can be traced back to its source and records cannot be altered (Nakamoto, 2008; Risius & Spohrer, 2017). Individuals in the network receive an incentive to operate network nodes to perform the proof-of-work required for each transaction; a process that is also referred to as mining (Tschorsch, 2016; Yli-Huumo, Ko, Choi, Park, & Smolander, 2016). Moreover, in modern blockchains so-called “smart contracts” allow for automated execution of transactions if predefined criteria are met (Buterin, 2014). Smart contracts are one example of how blockchain technology finds application in business situations of higher complexity.

Considering the previously outlined definition of a blockchain by Risius and Spohrer (2017), three key characteristics are identifiable. First, a blockchain is a decentralised and distributed system in which the ledger is not stored on a central system, and a majority of network nodes need to reach consensus about transactions to become effective (Kraft, 2016). As a result, it is challenging to hack the network as hackers would need to hold more than 51%

6 of the computing power in the entire network to reach such consensus (Kraft, 2016; Zhang & Wen, 2017). Additionally, the cryptography adds another layer of security (Risius & Spohrer, 2017). On the one side, the cryptography preserves data privacy, which is especially crucial with public blockchains (Xu et al., 2016). On the other side, it also maintains the ownership of coins in the case of cryptocurrencies (Xu et al., 2016). Second, a blockchain is immutable, meaning that the ledger cannot be altered and all transactions can be traced back to its origin (Beck et al., 2017). Third, a blockchain offers finality as each transaction is saved in the decentralised ledger and cannot be undone (Nakamoto, 2008). Only another transaction can correct a mistake (Nakamoto, 2008). Based on these characteristics, the blockchain technology therefore creates and guarantees transparency, a system-wide consensus among the parties, and validity of the entire transaction history (Risius & Spohrer, 2017). As a consequence, the new technology challenges any business that depends on external entities for verification and trust and simultaneously threatens current organisations whose business is to guarantee trust (Beck & Müller-Bloch, 2017). Hence, the blockchain technology is increasingly applied in the financial service industry in which security and trust play a crucial role (Lorenz et al., 2016).

Furthermore, as Iansiti and Lakhani (2017) pointed out, the blockchain technology can be seen as a foundational technology that not only disrupts traditional business models but that can also create new foundations of today’s social and economic systems. As a result, the process of change will be steady and gradual, and it will take decades for an extensive impact to materialise (Iansiti & Lakhani, 2017). In their work, Iansiti and Lakhani (2017) drew a comparison with the transmission control protocol (TCP) and internet protocol (IP). It took the protocols over 30 years to move through all phases of the adoption of foundational technologies and reshape today’s economy. Iansiti and Lakhani (2017) introduced four phases of a technology’s adoption – single use, localized use, substitution, and transformation. They introduced a matrix with the dimensions degree of novelty as well as amount of complexity and

7 coordination to categorise those stages. According to Iansiti and Lakhani (2017), this matrix can help managers to analyse which stage of application the blockchain technology in a particular industry has reached and assess which strategic investments in the company’s capabilities may be required. Additionally, a categorisation helps executives to recognise the required level of collaboration, the regulatory and legislative efforts, and the current challenges (Iansiti & Lakhani, 2017). Figure 1 below illustrates the described matrix.

Figure 1. Adaptation of foundational technologies. Adapted from The Truth About Blockchain (p. 123), by Iansiti & Lakhani, 2017.

In the first quadrant, applications are low in novelty and complexity and therefore gain acceptance first. On the other hand, applications that are high in novelty and complexity form the last stage of adaptation and usually take several decades to develop. However, those applications are also able to transform economies. (Iansiti & Lakhani, 2017)

The first stage – single use – represents applications with a low-coordination and low-novelty characteristic that create improved and highly focused solutions at lower costs. Bitcoin

8 payments also fall into this category, indicating the early stage of blockchain applications and their maturity at the moment. For each different quadrant and stage respectively, Figure 1 contains one example of a further developed blockchain application. The second stage – localization – contains applications that have a high degree of novelty but only require a limited user base to create value. The third stage – substitution – comprises of innovations that have a low level in novelty as they are based on existing single use and localized applications. On the other side, they require a high level of coordination as they aim for an increased public use. The goal of applications in this quadrant is to substitute the approach of doing business. Cryptocurrencies, which build on the single use of Bitcoin payments, strive to be in this category. They seek to replace the traditional currency system but face the challenge that every involved party needs to adopt the new technology. The fourth and last stage of the adaptation of foundational technologies – transformation – includes applications which are entirely new and can change social, political, and economic systems if successful. (Iansiti & Lakhani, 2017) At this point, it should be mentioned that the blockchain technology is not suitable for all types of businesses and should be carefully evaluated beforehand (Beck et al., 2017; Risius & Spohrer, 2017).

2.2 Blockchain Network Types

In addition, a differentiation between public and permissioned blockchain networks can be made. As Cachin (2016) pointed out in his work, in the former anyone can participate. One example of such a network is the cryptocurrency Bitcoin. In permissioned blockchains, however, the members of a network can be controlled, and the identity of parties is usually known (Cachin, 2016). Based on this difference in access to the network, three different types of the blockchain can be derived (Xie et al., 2017).

9

2.2.1 Public Blockchain

A public blockchain is the one described by Nakamoto (2008). As the name already indicates, the transactions of the blockchain are visible to everyone in the network (Xie et al., 2017). Additionally, every entity can participate in the network and can also take part in the consensus process, meaning that every member of the network can operate a node and potentially be involved in reaching consensus (Xu et al., 2016). As the blockchain is entirely decentralised, it is nearly impossible to change any entries, guaranteeing the immutability (Xie et al., 2017). Xie et al. (2017) also pointed out that the large computing power and amount of electricity required to reach consensus also lead to a relatively lower efficiency and higher costs of a public blockchain. On the other hand, the openness of a public blockchain also allows to attract many users and an active community (Xie et al., 2017).

Reaching consensus in such an expensive way is unnecessary in business networks, where all participants are usually known (Xie et al., 2017). Therefore, the following two blockchain network types emerged over time with modified characteristics. Nakamoto (2008) did not describe those two types in his initial work.

2.2.2 Consortium Blockchain

In a consortium blockchain, the visibility of the transactions can or cannot be restricted. Furthermore, only a selection of network nodes is responsible for the consensus determination, meaning that the consensus process is a permissioned one. As several nodes are involved, the blockchain is partially centralized. Compared to a public blockchain, the amount of network nodes is rather small but yet not one. A consortium blockchain requires lower computing power and amount of electricity, making it more efficient and cost-effective. However, one downside of a consortium blockchain is that due to the limited amount of involved network nodes that reach consensus, it is easier to manipulate the records on the ledger. (Xie et al., 2017)

10

2.2.3 Private Blockchain

This type of blockchain networks is the most restricted one (Xie et al., 2017). The blockchain is centralized to one entity and only that same organisation is required to reach consensus, meaning that only one or several network nodes owned by the latter are involved in the consensus determination (Xie et al., 2017). Therefore, the consensus process is entirely controlled by that party and permissioned in nature as well (Xu et al., 2016). Moreover, that entity can also fully control the transaction visibility of the blockchain network (Xu et al., 2016). Xie et al. (2017) pointed out that the even lower computational power and electricity needed makes a private blockchain the most efficient and cost-effective type of networks. Nevertheless, the risk of a ledger manipulation is also the highest for this network type as only a limited amount of network nodes is involved in the consensus process (Xie et al., 2017).

2.3 Blockchain Participants and Roles

This section briefly introduces the different participants and roles in a blockchain network to allow for a better understanding of the technology.

2.3.1 Blockchain user

Usually forming the largest share in a network, the blockchain user performs transactions with other users in the network (IBM, 2018). Blockchain users usually do not provide elementary services to the network but merely use it to perform transactions (IBM, 2018). As a result, they may not be aware of how the blockchain technology operates in the background (IBM, 2018). In a public blockchain, a potential user can simply join the network, whereas a permissioned blockchain requires the user to gain permission to join first (Xie et al., 2017).

2.3.2 Provider

Providers are important for permissioned blockchain networks. They are members of the network who oversee the transactions to maintain the integrity of the ledger. Depending on the

11 policy, they may be excluded to perform transactions on their own. Three different types of providers exist. (IBM, 2018)

First, the initiator who performs tasks to set up the network and its operation policies. In permissioned blockchains, the initiator also selects the members to join the network initially. After those tasks are fulfilled, the initiator typically remains in the network as a normal user. Second, the maintainers who run the network nodes for the consensus process in the blockchain network and are, therefore, integral for the network to function. Maintainers often also are the owner of a certificate authority. The last provider type is the auditor who has the permission of the network to run audit functions. Those include, for instance, analytics, billing, or compliance tracking. Moreover, an auditor can also monitor the permissions granted in a blockchain network. (IBM, 2018)

2.3.3 Certificate authority (CA)

A certificate authority is an entity responsible for issuing and managing certificates that are needed to operate a permissioned blockchain (Wilson & Ateniese, 2015). Those certificates are, for instance, required by blockchain users in such restricted networks (IBM, 2018).

2.3.4 Blockchain developer

Another important member for the blockchain network is the blockchain developer who programs applications and develops the chaincode. The former is required so that blockchain users can perform transactions in the network. Developed applications represent the connection between the user and the blockchain. (IBM, 2018)

2.3.5 Blockchain network operator

The last participant in a blockchain network is the network operator. Individuals in this role have special authority and permissions to run tasks on behalf of other network members. Their responsibility is it to manage and monitor the network. A network operator orders transactions,

12 manages transaction gateways and certificate authorities, and provides other elementary network services. (IBM, 2018)

2.4 Cryptocurrencies

A cryptocurrency is a digital currency that uses cryptography as a security measure to verify and secure transactions (Lindeman et al., 2016; Xu et al., 2016). As previously outlined, cryptocurrencies are based on a public blockchain that allows for monetary transactions in a peer-to-peer network (Xu et al., 2016). As a result, they do not require any centralized and trusted authority like a central bank to transfer or issue new money (Lindeman et al., 2016; Xu et al., 2016). The blockchain logs all transactions that took place in the cryptocurrency system in its public ledger (Xu et al., 2016). In the case of cryptocurrencies, miners are usually rewarded with new coins for their effort to operate a network node to validate transactions in the blockchain network (Tschorsch, 2016). Nevertheless, most users of cryptocurrencies gain access by exchanging fiat money into cryptocurrencies via retail exchanges (Bitpanda, 2018).

Bitcoin is the most known and valued cryptocurrency with a market capitalisation of over US$115 billion (Beck & Müller-Bloch, 2017; CoinMarketCap, 2018b). However, since its creation in 2009, 1,627 additional cryptocurrencies emerged until June 2018 (CoinMarketCap, 2018a; Presthus & O’Malley, 2017). Although high price fluctuations are not unusual for cryptocurrencies, Bitcoin has been the best performing cryptocurrency with the highest market capitalisation most of the time (CoinMarketCap, 2018a; Gil-Pulgar, 2018). Next to Bitcoin, the cryptocurrencies Ethereum and Ripple are currently the second and third largest one respectively in terms of market capitalisation (CoinMarketCap, 2018a).

2.4.1 Critical assessment

Nevertheless, the cryptocurrency market with blockchain as its underlying technology does not only offer positive attributes. The cryptocurrency Bitcoin in particular was recently covered in

13 media and also criticised by the academic world (Beck & Müller-Bloch, 2017). As pointed out by Tschorsch (2016) and Xu et al. (2016), the scalability issue of the Bitcoin network needs to be addressed. The steady increase in the user base to over 23.9 million by the end of the first quarter in 2018 (Statista, 2018) not only caused long waiting times in the network of up to nearly 11,500 minutes on January, 23rd in 2018 (Blockchain.Info, 2018) but also lead to transactions fees as high as US$37.49 on December, 21st _{in 2017 (Bitcoinfees, 2018). These}

long waiting times and high transaction costs made Bitcoin increasingly less attractive as a currency for daily purchases. Therefore, Glaser and his colleagues (2014) argued in their work that due to the missing link between an increased user base and the transaction volume of Bitcoin, new users would use Bitcoin as an asset and not as a currency. After the global cryptocurrency market lost nearly US$500 billion in market value at the beginning of 2018, the global head of investment research at Goldman Sachs, Steve Strongin, even argued that most cryptocurrencies will most likely not survive and lose all of its value (Nishizawa, 2018). Moreover, recent trends showed characteristics of a speculative bubble (Nishizawa, 2018).

Not only did the growth of the cryptocurrency user base cause scalability issues but with it, the energy consumption by the network also increased significantly. As every transaction needs to be validated and, therefore, requires a computation, the energy consumption increased with the larger user base (Tschorsch, 2016). It was recently estimated that the Bitcoin network currently consumes at least as much energy as an average American household in two years to produce one coin (Popper, 2018). Morgan Stanley estimated that the energy consumption of the Bitcoin network alone will equal 130 terawatt hours in 2018, matching with Argentina’s yearly energy needs (Jewkes & Steitz, 2018). As a result, Beck et al. (2017) demanded in their work of future research on new blockchain implementations to increase scalability and improve energy efficiency.

14 The Chinese government is currently trying to reduce electricity consumption and financial risks connected to cryptocurrency consensus calculations and trading (Wildau, 2018). About three quarters of the calculations needed to find consensus in the Bitcoin network are conducted in China, where local companies specialised on those calculations take advantage of low electricity costs in regions with coal and hydroelectric power (Wildau, 2018). However, the Chinese government does not plan to shut down those companies immediately, but rather enforce policies on tax collection, land and electricity use, and environmental regulations (Wildau, 2018). Nevertheless, because of those increased regulations, some Bitcoin miners plan to move to other regions with cheap hydroelectric power and climatic advantages such as Norway or Sweden (Jewkes & Steitz, 2018; Rapier, 2018).

The above-illustrated issues show that the application of the blockchain technology in the cryptocurrency market still faces some major issues and is not mature yet. Nonetheless, new consensus algorithms used in other cryptocurrencies than Bitcoin aim at being more efficient and hence reduce the energy needed in the mining process (Xie et al., 2017). For instance, the cryptocurrencies Ethereum and Ripple try to overcome the scalability issue with newly developed consensus algorithms, leading to less usage of energy and a faster processing time per transaction (Xie et al., 2017). Developers of the Bitcoin network also tried to improve the chaincode to overcome the scalability issue but had limited success so far (Tschorsch, 2016).

On the other hand, Iansiti and Lakhani (2017) pointed out that hacks on cryptocurrency exchanges or wallet software reported by the media are not caused by a weakness in the blockchain but rather by programming issues or errors in the linking of separate systems. For instance, at the end of 2017, a programming mistake by the wallet software provider Parity caused over 500,000 units of the cryptocurrency Ethereum with a market value of US$150

15 million at that time to be no longer accessible by the respective owners (Kyriasoglou, 2018; Lardinois, 2017).

The often proclaimed anonymity of cryptocurrencies also has limited validity. Although no direct personal information is shown in cryptocurrency networks, all transactions are stored on the public ledger, allowing anyone to analyse and review them (Moser, Bohme, & Breuker, 2013). Therefore, the network does not provide anonymity but rather pseudonymity (Moser et al., 2013). Research already demonstrated that it is possible to reveal the identity of Bitcoin users based on the publicly accessible transaction history (Moser et al., 2013). Additionally and as previously pointed out, most users gain access to cryptocurrency networks via retail exchanges for which they have to register with personal information (Moser et al., 2013).

In addition, Beck et al. (2017) pointed out the criticism that the blockchain technology in general may limit the freedom to make decisions as people will face a record of transactions that cannot be altered or erased.

2.5 Further Current Blockchain Applications

Research so far mainly focused on the technological component of the blockchain technology (Lindeman et al., 2016) and less on its different potential applications and involved parties.

Bitcoin was the first application of the blockchain technology (Iansiti & Lakhani, 2017; Nakamoto, 2008) and, therefore, blockchain is often only seen as the technology behind Bitcoin and other cryptocurrencies (Beck et al., 2017). However, by now the technology has been applied in other settings as well. The following section provides some examples.

The first example is from the food industry in which the food quality and traceability play an essential role (Aung & Chang, 2014; Zhong, Xu, & Wang, 2017). The application of the blockchain in this sector would mean that all information about a product is stored on a

16 decentralised ledger to which all parties (farmer, wholesaler, retailer, and consumers) could have access to know when the food was produced and where it is coming from. In case of a food scandal, it would be much easier to identify affected items. (Koonce, 2017)

Another example is how blockchain technology could help to resolve the issue of lost identification documents. Especially in times of increased amounts of refugees, this application could help to identify persons more easily (UNHCR, 2018). As an example, the state of Illinois in the US has announced a pilot project which allows Illinois citizens to access their birth certificate online via the blockchain (Barrena, 2017). Furthermore, Ølnes (2016) pointed out in his work that the blockchain technology could potentially be used for storing all sorts of permanent or relatively permanent public documents besides certificates. The example from Illinois can also be seen as a pioneer project of a transformative application of blockchain technology in the area of public identity systems (Iansiti & Lakhani, 2017).

The third example is about how the shipping company MAERSK in cooperation with the technology company IBM is trying to use the blockchain technology to make its supply chain safer and more efficient. The goal of the two companies is to use the blockchain technology to make the unchangeable record of supply chain transactions available to any relevant partner in real time. As a large amount of paperwork that needs to be processed is involved in shipping goods, the joint venture aims at digitalising and automating the filing of paperwork. Not only would the project increase transparency in the supply chain but it would also reduce time and costs. Other companies such as General Motors have also shown interest in this project as it could potentially change the foundations of global trade. (Milne, 2018; White, 2018)

The last example is about a more complex application of the blockchain technology. As Mainelli & Smith (2015) reported in their explorative research, incorporating a shared ledger

17 into trusted mediator systems can back services such as the ones of insurance companies. The insurance company Allianz tested the use of blockchain for smart contracts in a pilot project in cooperation with Nephila Capital Limited for executing natural catastrophe swaps (Allianz Risk Transfer AG, 2016). Those swaps are a financial instrument that hand over a specific risk set from the insurer to an investor (Allianz Risk Transfer AG, 2016). The project also illustrated how the transactions and settlement between investors and insurers could be simplified and accelerated by smart contracts (Allianz Risk Transfer AG, 2016). In such a case, the policy conditions of the insurance are coded into a smart contract, which is stored on the blockchain (Buterin, 2014). At the moment an insurable event takes place and is reported by a trusted party, the smart contract will be automatically executed, leading to the processing of the claim without any additional manual intervention (Allianz Risk Transfer AG, 2016; Buterin, 2014). As highlighted by Iansiti and Lakhani (2017), fully self-executing smart contracts fall into the category of transformation and require a high level of collaboration among involved parties. They pointed out that it can take decades until a technology can transform an entire industry. That observation can also explain why the Allianz project only has been a pilot yet. Furthermore, Lorenz et al. (2016) also pointed out in their work that the insurance industry lacks behind the finance one in terms of the adoption rate of the blockchain technology.

2.6 Technology Adoption Theories

This section will analyse the literature about the technology acceptance model (TAM) and its extensions, the unified theory of acceptance and use of technology (UTAUT), and the diffusion of innovations theory, followed by the research question. The analysis will be the foundation for the conceptual model.

18

2.6.1 Technology acceptance model (TAM)

The technology acceptance model (TAM) by Davis (1986) can be seen as the most applied model to understand and predict the underlying motivational factors that cause users to accept a technology (Yi, Jackson, Park, & Probst, 2006). The TAM is an adaptation of the theory of reasoned action (TRA) (Davis, Bagozzi, & Warshaw, 1989).

Davis (1986) presented in his work a model that depicted two factors influencing a user’s attitude toward using a system. The latter is, in turn, a major determinant of actual system use (Davis, 1986). The two factors influencing the attitude are perceived usefulness and perceived ease of use (Davis, 1986). In his model, perceived ease of use also has a causal effect on perceived usefulness, causing the latter to be a mediator between perceived ease of use and attitude toward using a system (Davis, 1986). Davis (1986) argued that external factors are directly influencing perceived ease of use and perceived usefulness. He defined perceived usefulness as the degree an individual thinks that using a specific system would improve job performance (Davis, 1989). On the other hand, perceived ease of use represents to which extend an individual believes that using a specific system would be free of effort (Davis, 1989).

In a later publication, Davis, Bagozzi, and Warshaw (1989) added the variable behavioural intention to use between the variables attitude toward using and actual system use. Although in conflict with the underlying TRA, they argued that other researches show a direct connection between a belief, to which also perceived usefulness belongs, and behavioural intention to use (Davis et al., 1989). The reasoning behind that connection is that individuals in organisations develop intentions towards a behaviour they think will improve their job performance (Davis et al., 1989). Those intentions would develop no matter how the person feels about the behaviour as such (Davis et al., 1989). Venkatesh and Davis (2000) claimed that within ten years the TAM had become a robust, proven, and economical model for

19 predicting the acceptance of new technologies. Figure 2 depicts the described technology acceptance model.

Figure 2. Technology acceptance model (TAM). Adapted from User Acceptance of Computer Technology: A Comparison of Two Theoretical Models (p. 985), by Davis, Bagozzi, &

Warshaw, 1989.

2.6.2 Extensions of the technology acceptance model (TAM)

Pavlou (2003) looked in his work at the acceptance of electronic commerce by users. Not only did he set the technology acceptance model by Davis (1986) into relation with the end-user in the B2C e-commerce market but also extended the existing TAM by the variables trust and perceived risk (Pavlou, 2003). Pavlou (2003) pointed out that the TAM initially focused on the usage of technology in the workplace, but research started to apply it for the understanding of website usage to understand the consumer perspective. He claimed that trust is required in all interactions, especially in uncertain environments. Pavlou (2003) argued that trust positively influences perceived usefulness as users become vulnerable to the provider to ensure that they get the expected transaction. Additionally, he claimed in his work that trust also positively influences the perceived ease of use as it would reduce the necessity by the user to control and monitor the situation. In turn, the transaction would be facilitated and connected with less effort (Pavlou, 2003).

20 Venkatesh and Davis (2000) extended the initial TAM by factors of social influence and cognitive instrumental processes to better explain perceived usefulness and usage intentions. Named TAM2, the extension added two independent variables for the social influence (image and subjective norm) and three independent variables for cognitive instrumental processes (job relevance, output quality, and result demonstrability). In addition, the variables intention to use and usage behaviour replaced the original variable construct of attitude toward using, behavioural intention to use, and actual system use.

Venkatesh and Davis (2000) defined the variable image as the degree to which an individual perceives that the use of a technology is enhancing the status in his or her social system. Image was modelled to have a direct positive influence on perceived usefulness. The second factor of social influence – subjective norm – describes to which extend an individual perceives relevant other’s beliefs that he or she should or should not perform a certain behaviour (Venkatesh & Davis, 2000). Besides also having a direct positive effect on perceived usefulness, subjective norm was depicted to be a direct determinant of intention to use and to have a positive effect on the variable image (Venkatesh & Davis, 2000). The relationship between subjective norm and intention to use is subject to the same argumentation made by Davis, Bagozzi, and Warshaw (1989) for the initial TAM. Venkatesh and Davis (2000) argued that subjective norm would influence image as an action by an individual would increase the standing of the latter in the social group if important members of the individual’s social group think that he or she should perform that action. In addition, Venkatesh and Davis (2000) added two moderators. The first – experience – moderates the relation between subjective norm and intention to use and between subjective norm and perceived usefulness (Venkatesh & Davis, 2000). For both cases, Venkatesh and Davis (2000) claimed that an increase in experience would mitigate the direct effect of subjective norm on intention to use and perceived usefulness respectively (Venkatesh & Davis, 2000). The second moderator – voluntariness – also

21 moderates the relation between the variable subjective norm and intention to use. If the technology use is perceived voluntary, no significant direct effect will occur (Venkatesh & Davis, 2000).

All three independent variables for cognitive instrumental processes – job relevance, output quality, and result demonstrability – are direct determinants of the variable perceived usefulness with a positive effect (Venkatesh & Davis, 2000). Venkatesh and Davis (2000) defined job relevance as how far the technology is capable of performing and whether it matches the individual’s job goals. Under output quality they understood how well the technology does perform the tasks it is expected to fulfil. In regards to result demonstrability, they argued that the level to which a person can attribute improvements in job performance to the use of a technology will directly influence perceived usefulness. The better this attribution is possible, the higher the awareness of a technology’s usefulness would be.

2.6.3 Unified theory of acceptance and use of technology (UTAUT)

In 2003, Venkatesh, Moris, Davis, and Davis developed the unified theory of acceptance and use of technology (UTAUT) to measure user acceptance and technology usage (Venkatesh, Morris, Davis, & Davis, 2003). The theory integrated eight models and their extensions of user acceptance theory (Venkatesh et al., 2003). The considered models were from the theory of reasoned action, the motivational model, the theory of planned behaviour, the technology acceptance model, the innovation diffusion theory, the model of PC utilisation, the social cognitive theory, and a model combining the theory of planned behaviour and the technology acceptance model.

The formulated UTAUT has four fundamental determinants for the intention and usage of a technology and up to four moderators for the main relationships (Venkatesh et al., 2003). The first determinant – performance expectancy – evaluates to which extend an individual

22 thinks that the technology will improve job performance (Venkatesh et al., 2003). Effort expectancy, which is the second determinant, represents the ease associated with the use of a system (Venkatesh et al., 2003). The third determinant – social influence – reflects to which degree an individual perceives relevant others’ beliefs that he or she should use the new system (Venkatesh et al., 2003). The last determinant – facilitating conditions – defines to which extend an individual thinks that technical and organisational structures are in place to support the system use (Venkatesh et al., 2003).

In the developed UTAUT model the variables gender, age, voluntariness to use, and experience moderate the relationship between the four fundamental determinants and the intention and usage of a technology (Venkatesh et al., 2003). Venkatesh et al. (2003) pointed out in their work that managers could use the UTAUT as a tool to determine the chance of success for newly introduced technologies. Furthermore, the theory could help to comprehend the factors driving the acceptance and to adopt the business strategy to reach users with a lower intention to use the new technology accordingly (Venkatesh et al., 2003).

However, Bagozzi (2007) criticised the Unified Theory of Acceptance and Use of Technology. Although being an approach with good intentions, he claimed that the UTAUT with its 49 independent variables to predict intentions and behaviour would be a model in the technology adoption theory that causes chaos.

2.6.4 Diffusion of innovations theory

Presthus and O'Malley (2017) ran a small study in 2017 that built on the diffusion of innovation theory by Rogers to investigate the barriers and motivations of end-users to use bitcoin as a digital currency. Rogers (2003) defined diffusion as the process by which an innovation is communicated among the participants of a social system over time. He argued that the main factors driving the diffusion of new ideas are the innovation itself, the used communication

23 channels, the passed time, and the social system in which the spread of the new idea takes place (Rogers, 2003).

Presthus and O'Malley (2017) concluded in their work that users accept Bitcoin because of technological curiosity and not because of financial incentives. On the other hand, non-users would stay away due to security and currency value concerns (Presthus & O’Malley, 2017). With the emergence of other large cryptocurrencies, however, the Bitcoin is no longer the only currency of interest (CoinMarketCap, 2018a). Therefore, this research is focusing on cryptocurrencies in general.

2.7 Research Question

Similar to Pavlou’s work (2003), this research will not focus on investigating an underlying technology but rather one of its applications. Pavlou (2003) focused in his work on the acceptance of electronic commerce rather than of the internet by users. Correspondingly, this work mainly concentrates on cryptocurrencies and not on the blockchain technology as such. In addition, the focus will not be on cryptocurrency services such as retail exchanges but rather on the end-user and his or her relation to cryptocurrencies. As Lorenz et al. (2016) pointed out, cryptocurrencies are the most spread application of blockchain technology in the end-user market so far.

The literature review has shown a lack of academic research with respect to the usage and adaptation of blockchain technology in the end-user cryptocurrency market. Although frequently used to investigate the adoption of new technologies, technology adoption theories have not been thoroughly applied to cryptocurrencies yet. Furthermore, as the different examples about blockchain applications illustrated, there is a lack of experience on how to address the end-user market and knowledge about the blockchain technology in the business world. Based on those observations, the following research question was developed:

24 Which factors are supporting the adaptation of cryptocurrencies in the end-user market?

3 Research Concept

Based on the previously outlined literature and the factors identified as relevant for the adaptation of the blockchain technology, the conceptual model was developed. This research mainly used the original Technology Acceptance Model (TAM) developed by Davis (1986) and the extension of it (TAM2) as a foundation to develop propositions. As Venkatesh and Davis (2000) pointed out in their work, the TAM has shown to be a robust and proven model to predict the acceptance of new technologies. Therefore, the TAM serves as a foundation for this research. Furthermore, the work of Pavlou (2003) influenced the research concept with its focus on consumers and the factors of trust and risk.

3.1 Conceptual Model

For the conceptual model, nearly all variables of the TAM2 were adopted. However, as this research is concerned with the end-user adoption of cryptocurrencies, the variable job relevance from the TAM2 was not transferred. As the user of cryptocurrencies is not seeking to achieve specific job-related goals, the researcher considered this variable as inadequate.Moreover, the relationship between subjective norm and intention to use was not investigated because the end-user is not part of an organisation in which he or she develops intentions towards a behaviour of which he or she thinks will improve job performance. In addition, the original TAM by Davis (1986) as well as the underlying theory of reasoned action did not include this relation (Davis et al., 1989). The moderator voluntariness was not included as no external party such as an organisation is forcing the end-user to own and use cryptocurrencies. The entire usage of cryptocurrencies is assumed to be voluntary. As in the original TAM by Davis (1986), a direct positive effect of perceived ease of use on perceived usefulness is theorised because an easier to use technology will create a higher usefulness for the user.

25 In an adjusted form, the researcher integrated the variables trust and risk presented by Pavlou (2003) into the modified TAM2. The variable trust is modelled to have an indirect effect on the intention to use through the variables perceived usefulness, perceived ease of use, and perceived risk only. Furthermore, the researcher added the independent variables wealth, which is measured via the income, and age to account for additional factors that may influence the adoption of cryptocurrencies. The variable wealth was included as it necessary for an individual to own some capital to exchange it into another currency. Hence, the user can use some of his or her income and potential savings to exchange it into cryptocurrencies. It is theorised that wealth will have a positive direct effect on perceived usefulness as people with capital value cryptocurrencies as an alternative currency more. Age is theorised to have a direct effect on perceived ease of use. This relation is based on the findings by Czaja et al. (2006) who stated that younger adults are more likely to use technology in general. Figure 3 depicts the conceptual model.

26 In the model, the variable experience acts as a moderator between the variables subjective norm and perceived usefulness. The variable perceived usefulness mediates the relationship between perceived ease of use and intention to use. The variables subjective norm, image, output quality, result demonstrability, wealth, age, and trust do not directly influence intention to use and usage behaviour; they are rather indirectly affecting those variables through perceived usefulness and perceived ease of use respectively. The variable usage behaviour is the dependent variable in the developed conceptual model.

3.2 Propositions

In the following, the propositions corresponding to the conceptual model are outlined: Proposition 1: Subjective norm has a positive direct effect on perceived usefulness.

Proposition 2: The positive relationship between subjective norm and perceived usefulness is moderated by experience, so that this relationship is stronger for lower values of experience. Proposition 3: Subjective norm has a positive direct effect on image.

Proposition 4: Image has a direct positive effect on perceived usefulness.

Proposition 5: Output Quality has a direct positive effect on perceived usefulness.

Proposition 6: Result Demonstrability has a direct positive effect on perceived usefulness. Proposition 7: People’s wealth has a direct positive effect on perceived usefulness.

Proposition 8: People’s age has a direct positive effect on perceived ease of use. Proposition 9: Trust has a direct positive effect on perceived usefulness.

Proposition 10: Trust has a direct positive effect on perceived ease of use. Proposition 11: Trust has a direct negative effect on perceived risk.

27 Proposition 12: Risk has a direct positive effect on intention to use.

Proposition 13: Perceived ease of use has a direct positive effect on intention to use.

Proposition 14: The positive relationship between perceived ease of use and intention to use is mediated by perceived usefulness.

Proposition 14a: There is a positive relationship between perceived ease of use and perceived usefulness.

Proposition 14b: There is a positive relationship between perceived usefulness on intention to use.

Proposition 15: Intention to use has a direct positive effect on usage behaviour.

4 Method Selection

The literature review showed that the academia mainly focused on the technological aspect of blockchain technology and a limited amount of applications (Lindeman et al., 2016; Risius & Spohrer, 2017). Moreover, there is a lack of academic research covering the usage adaptation of cryptocurrencies by the end-user. As a result, this research is more of exploratory nature, trying to identify the factors supporting the adoption of cryptocurrencies (Saunders, 2011).

This research follows an inductive and deductive approach. By assessing the existing literature, a conceptual model could be derived. The latter was then used as a foundation for this exploratory research. With a mixed method approach, this work aims at triangulating findings by applying quantitative and qualitative methods (Iii, 2013). On the one side, a survey was carried out. On the other, five interviews with people who have experience in the cryptocurrency market or blockchain sector were conducted. This research design has the advantage of attaining higher external validity compared to an experiment (Saunders, 2011).

28 However, no causal relationships can be inferred (Saunders, 2011). The cross-sectional data will allow the comparison of different variables at the same moment (Saunders, 2011).

4.1 Sample

The population for this research are users and non-users of cryptocurrencies. However, due to limited information on the user segment, the actual sampling frame is mostly unknown. Therefore, non-probability sampling techniques were be applied (Oates, 2005; Saunders, 2011).

For the interviews, individuals with experience in the cryptocurrency market or blockchain technology respectively were approached. By covering the underlying technology (i.e. blockchain) as well as the application (i.e. cryptocurrency) a broader understanding could be reached.

4.2 Measures

The survey included an extended section of questions on demographics to compensate for the fact that the sample frame is mostly unknown. Users of cryptocurrencies were also asked additional questions about which currencies they own, which cryptocurrency services they have used so far, and what their objective behind owning cryptocurrencies is.

To measure the items of the variables subjective norm, image, output quality, result demonstrability, perceived usefulness, perceived of use, and intention to use a validated seven point Likert scale (1 = strongly disagree, 7 = strongly agree) with a Cronbach’s alpha ranging from 0.8 to 0.98 by Venkatesh and Davis (2000) was used. For the items of trust, a validated seven point Likert scale (1 = strongly disagree, 7 = strongly agree) with an averaging Cronbach’s alpha of 0.9 by Pavlou (2003) was used. A seven point Likert scale (1 = very low, 7 = very high) measured the variable experience. For each item of perceived risk, a different validated seven point Likert scale with an averaging Cronbach’s alpha of 0.9 by Pavlou (2003)

29 was used. Item 1 and 3 used a Likert scale with the elements ranging from 1 = significant risk to 7 = significant opportunity. Elements for item 2 ranged from 1 = very negative situation to 7 = very positive situation. For item 4, the elements ranging from 1 = high potential for loss to 7 = high potential for gain were used.

By using validated measurement scales from previous researchers, a higher reliability can be achieved. The researcher modified the question for each item to match the context of cryptocurrencies and blockchain technology. He also included questions about the current state of knowledge about cryptocurrencies and blockchain technology by the participant. For those questions, a five point Likert scale (1 = strongly disagree, 5 = strongly agree) was used. Moreover, a seven point Likert scale (1 = not at all important, 7 = extremely important) measured the importance of ten different items. The same measurement scale was used to measure the importance of additional factors that the respondent indicated as being important to him or her.

The survey was created in English and translated into Dutch and German as the researcher assumed that most respondents would come from the United Kingdom, the Netherlands, and Germany. A complete list of all questions with the corresponding measurement scales in all three languages can be found in Appendix C-E respectively.

5 Data Collection

Before collecting responses via the survey, a pre-test with six students covering all languages was conducted. As the participants match the sample and only minor changes were made, they remained in the data set for further analysis. Appendix G depicts all the changes made after the pre-test.

Afterwards, the survey was distributed in the personal network of the researcher, to 67 online social networks groups with focus on blockchain technology and cryptocurrencies, and

30 to others students and former working colleagues. In the personal network, the survey was directly distributed to 46 individuals. Thirty-five persons completed the survey, leading to a completion rate of 85%. To improve the response rate, one reminder was sent out to the personal network. Moreover, an online research provider (crowdsourcing platform Amazon Mechanical Turk) was used to collect responses.

In total, 369 responses were collected in Qualitrics. To test for careless responses, a response time analysis was conducted. Moreover, an instructed response item was included in the survey. The later led to the exclusion of one response. Another response was excluded as only a minority of questions were answered. As a result, 367 responses were used for the analysis.

For the interviews, five individuals with experience in the cryptocurrency market and experience with blockchain technology were approached. Two of the interviewees had a particular expertise in cryptocurrencies. One interviewee had extended knowledge about the blockchain and smart contract market. Appendix F includes a short introduction to each interviewee.

6 Research Results

This section will first present and then analyse the collected survey data. Thereafter, the data will be interpreted and each proposition considered. In the end, the major findings derived from the interviews are described. Tableau 10.5 was used for the visualization and interpretation of the survey results.

6.1 Demographics

In total, responses from 32 countries were collected. However, the majority with over 27% were coming from the United Kingdom, followed by the Netherlands and Germany.

31 Overall, 69.7% of the respondents were male. By considering the different user types (user, non-user, potential user), a more detailed image can be derived: Of the users, the majority were male (86%) whereas for the non-user the opposite was the case. Here, 54% of the respondents were female. Overall, 48% of the respondents were cryptocurrency user. 29% of the individuals reported that they intend to own cryptocurrencies in the future, and 23% indicated that they are non-users.

The majority of answers were coming from individuals in the lower income segment, followed by users with an average of 20,000-45,000 Euro annual gross income. The later range is reflecting the average salary in the United Kingdom, the Netherlands, and Germany. Most of the respondents were employed for wages (51.2%). The second largest group in the data set were students with 23.8%.

Appendix A includes the most relevant figures about the demographics of the participants. Appendix B depicts more in detail the key characteristics of cryptocurrency user.

6.2 Interpretation

The following table depicts which findings for each proposition could be derived from the data set.

Table 1. Proposition summary.

Nr. Proposition Findings

P1 Subjective norm has a positive direct effect on perceived usefulness.

Data suggests a positive relationship.

P2 The positive relationship between subjective norm and perceived usefulness is moderated by

experience, so that this relationship is stronger for lower values of experience.

Data suggests a contrary moderation.

P3 Subjective norm has a positive direct effect on image.

32 P4 Image has a direct positive effect

on perceived usefulness.

Data suggests a positive relationship. P5 Output Quality has a direct positive

effect on perceived usefulness.

Data suggests a positive relationship. P6 Result Demonstrability has a direct

positive effect on perceived usefulness.

No clear support by data.

P7 People’s wealth has a direct positive effect on perceived usefulness.

No clear support as respondents mostly had low income level.

P8 People’s age has a direct positive effect on perceived ease of use.

No clear support as also older respondents indicate a high perceived ease of use. P9 Trust has a direct positive effect on

perceived usefulness.

Data suggests a positive relationship. P10 Trust has a direct positive effect on

perceived ease of use.

Data suggests a positive relationship. P11 Trust has a direct negative effect on

perceived risk.

Data suggests a positive effect from trust on perceived risk.

P12 Risk has a direct positive effect on intention to use.

No clear support by data. P13 Perceived ease of use has a direct

positive effect on intention to use.

No clear support by data. P14a There is a positive relationship

between perceived ease of use and perceived usefulness.

Data suggests a positive relationship.

P14b There is a positive relationship between perceived usefulness on intention to use.

Data suggests a positive relationship.

P15 Intention to use has a direct positive effect on usage behaviour.

Data suggests a positive relationship.

Survey participants had the possibility to name additional factors they perceive as important. However, no additional factor with a meaningful amount of responses was recorded.

6.3 Interview Results

Interviewee 2-5 confirmed that according to their experience most individuals in the cryptocurrency and blockchain sphere are male.

33 Interviewee 1,2, and 5 also pointed the young age of those individuals out. Although agreeing with them, interviewee 3 and 4 highlighted that in the corporate environment those individuals can be much older. Interviewee 1,3, and 5 agreed that the industry in which blockchain technology is currently applied the most is the financial one. Interviewee 4 argued that there is no clear industry taking the lead but rather technology companies being the pioneers.

Interviewees 1-4 all agreed that individuals are currently still very cautious with the blockchain technology and its applications. Interviewee 3 and 5 argued that most users do not have any knowledge about how the blockchain technology in the background actually works. Interviewee 1 stated that most companies would be eager to learn how the technology works. Interviewee 4 argued that, as blockchain applications becoming more widespread and people with a lower technical aversion would use them, the knowledge about the blockchain technology would actually decrease over time.

Interviewee 5 agreed that the social environment can have a large impact on whether an individual feels that he or she has to use cryptocurrencies. Interviewee 4, however, argued that in the business world individuals would be less influenced by others as such but more by the fear of missing out. They both agreed that the application of a blockchain is seen as more risky by the end-user compared to the underlying technology.

7 Conclusions

7.1 Managerial Implications

One of the implications for businesses that can be derived from this research is that companies should focus more on establishing trust. The survey as well as the majority of interviewees indicated that users of cryptocurrencies tend to be cautious and assess the application of the blockchain as more risky than the technology itself.

34

7.2 Limitations

The research faces several limitations. First, due to the fact that the sampling was non-probability based, the representativeness and generalizability of the results cannot be entirely guaranteed. Second, as the survey was only carried out once (cross-sectional), the reliability of the results cannot be completely ensured. This issue was addressed by using validated measuring scales. As a survey is used as data collection tool, the research faces the problem of self-reported data only and the common method bias. These issues were addressed by conducting additional interviews with experts of blockchain technology and cryptocurrencies. Regarding the generalizability of this research, it is questionable whether the results can be applied to other countries than the United Kingdom, the Netherlands, and Germany. As the majority of respondents came from those countries, it is more likely that the findings also only apply here. Moreover, the findings may be transferred to sectors very similar to the cryptocurrency market. Blockchain applications in the insurance sector could attract similar customer for who the same findings hold true in the future.

Additionally, as interviewee 2 pointed out, the characteristics of cryptocurrency users changed extremely in the past. It cannot be ruled out that the findings presented in this research hold true in the future. Nevertheless, as the technology becomes more mature, it can be argued that also the user base will become more stable.

7.3 Future Research

As this research was more of exploratory nature with different propositions presented, future research should collect data in a way that allows for testing hypothesis statistically.

As done by Venkatesh (2000), future research could focus on identifying factors that influence the variable perceived ease of use. As perceived ease of use is a key variable in the

35 technology acceptance model, identifying additional factors influencing the latter in the area of blockchain applications may lead to worthwhile results (Venkatesh, 2000).

36

8 Bibliography

Allianz Risk Transfer AG. (2016). Blockchain technology successfully piloted by Allianz Risk Transfer and Nephila for catastrophe swap. Retrieved June 1, 2018, from

https://www.agcs.allianz.com/about-us/news/blockchain-technology-successfully-piloted-by-allianz-risk-transfer-and-nephila-for-catastrophe-swap-/

Aung, M. M., & Chang, Y. S. (2014). Traceability in a food supply chain: Safety and quality

perspectives. Food Control, 39, 172–184.

https://doi.org/10.1016/J.FOODCONT.2013.11.007

Bagozzi, R. P. (2007). The Legacy of the Technology Acceptance Model and a Proposal for a Paradigm Shift. The Legacy of the Technology Acceptance Model and a Proposal for a

Paradigm Shift, 8(4), 244–254. Retrieved from http://aisel.aisnet.org/cgi/viewcontent.cgi?article=1406&context=jais

Barrena, P. (2017). Blockchain-like ID may mean end of paper birth certificates | New

Scientist. Retrieved January 26, 2018, from

https://www.newscientist.com/article/mg23531454-500-blockchaininspired-project-means-you-are-who-you-say-you-are/

Beck, R., Avital, M., Rossi, M., & Thatcher, J. B. (2017). Blockchain Technology in Business and Information Systems Research. Business & Information Systems Engineering, 59(6), 381–384. https://doi.org/10.1007/s12599-017-0505-1

Beck, R., & Müller-Bloch, C. (2017). Blockchain as Radical Innovation: A Framework for Engaging with Distributed Ledgers. In Proceedings of the 50th Hawaii International

Conference on System Sciences. Retrieved from http://scholarspace.manoa.hawaii.edu/bitstream/10125/41815/1/paper0666.pdf

37 Bitcoinfees. (2018). Bitcoin Transaction Fees. Retrieved January 26, 2018, from

https://bitcoinfees.info/

Bitpanda. (2018). About – Bitpanda – Buy and Sell Cryptocurrencies. Retrieved June 1, 2018, from https://www.bitpanda.com/en/about

Blockchain.Info. (2018). Average Confirmation Time - Blockchain. Retrieved January 26, 2018, from https://blockchain.info/de/charts/avg-confirmation-time?timespan=180days Buterin, V. (2014). A Next Generation Smart Contract & Decentralized Application Platform.

Retrieved from http://www.the-blockchain.com/docs/Ethereum_white_paper- a_next_generation_smart_contract_and_decentralized_application_platform-vitalik-buterin.pdf

Cachin, C. (2016). Architecture of the Hyperledger Blockchain Fabric. Retrieved from https://pdfs.semanticscholar.org/f852/c5f3fe649f8a17ded391df0796677a59927f.pdf CoinMarketCap. (2018a). All Cryptocurrencies | CoinMarketCap. Retrieved June 19, 2018,

from https://coinmarketcap.com/all/views/all/

CoinMarketCap. (2018b). Bitcoin | CoinMarketCap. Retrieved June 19, 2018, from https://coinmarketcap.com/currencies/bitcoin/

Corcoran, K. (2017). Bitcoin is climbing on the last day of 2017. Retrieved January 26, 2018, from https://www.businessinsider.nl/bitcoin-price-value-increasing-on-final-day-of-2017-2017-12/?international=true&r=UK

Czaja, S. J., Charness, N., Fisk, A. D., Hertzog, C., Nair, S. N., Rogers, W. A., & Sharit, J. (2006). Factors predicting the use of technology: findings from the Center for Research and Education on Aging and Technology Enhancement (CREATE). Psychology and Aging, 21(2), 333–352.

38 Davis, F. D. (1986). A technology acceptance model for empirically testing new end-user

information systems: Theory and results. Management, (April), 291. https://doi.org/oclc/56932490

Davis, F. D. (1989). Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology. MIS Quarterly, 13(3), 319–340.

Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User Acceptance of Computer Technology: A Comparison of Two Theoretical Models. Source: Management Science, 35(8), 982–1003. Retrieved from http://www.jstor.org/stable/2632151

Gartner. (2016). Gartner’s 2016 Hype Cycle for Emerging Technologies Identifies Three Key Trends That Organizations Must Track to Gain Competitive Advantage. Retrieved January 26, 2018, from https://www.gartner.com/newsroom/id/3412017

Gil-Pulgar, J. (2018). Bitcoin Remains the Best Performing Currency: Up 700% Since January 1, 2017 - Bitcoinist.com. Retrieved June 19, 2018, from http://bitcoinist.com/bitcoin-price-up-700-since-january-1-2017/

Glaser, F., Zimmermann, K., Haferkorn, M., Weber, M. C., & Siering, M. (2014). Bitcoin-asset or currency? Revealing users’ hidden intentions. In Twenty Second European Conference on Information Systems, Tel Aviv 2014. Retrieved from https://www.researchgate.net/profile/Florian_Glaser2/publication/286338705_Bitcoin_-_Asset_or_currency_Revealing_users%27_hidden_intentions/links/5a1bbbb2aca272df0 80f2f07/Bitcoin-Asset-or-currency-Revealing-users-hidden-intentions.pdf

Iansiti, M., & Lakhani, K. R. (2017). The Truth About Blockchain. Harvard Business Review, 95(1), 118–127.