The role of trust in lending-based
crowdfunding investments – a focus on
the lender perspective
Master thesis Business Administration written by Kim Rasmussen (10857982)
Study
MSc Business Administration
Specialization
Entrepreneurship & Innovation
Supervisor
drs. A.C.C. Gruijters
This document is written by Kim Rasmussen, who declares to take full responsibility for the content 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.
1. Introduction to the crowdfunding environment 10
1.1 Background information 10
1.2 Prior research and cause 11
1.2 Research aim 12
2. Literature review 14
2.1 Peer-to-peer lending 14
2.2 The lender investment process 15
2.4 Trust 17
2.4.1 Trust in financial markets 18
2.4.2 Trust in E-commerce 19
2.4.3 Trust development in crowdfunding 20
2.5 Conclusion literature 20
3 Conceptual Model & Research questions 22
3.1 Literature findings – the concepts 22
3.1 Sub-questions and sensitizing concepts 23
4. Research design 24 4.1 Research approach 24 4.2 Data collection 25 4.2.1 Sample 25 4.2.2 Interview process 26 4.3 Data analysis 26 4.3.1 Transcription 27 4.3.2 Open coding 27 4.3.3 Axial coding 27 4.3.4 Selective coding 28 4.4 Evaluation of method 28 4.4.1 Validity 28 4.4.2 Reliability 29 5 Findings 30 5.1 Investment determinants 30 5.1.2 Information transparency 30 5.2 Experience 33 5.3 Relationship 38 5.4 Investment process 41 6 Discussion 42
6.1 Answering the Research Questions 42
6.2 Scientific implications 42
6.3 Limitations & Future Research 43
6.4 The future of crowdfunding 44
7 Recommendations 45
7.1 Customer service 45
7.2 Marketing 46
7.3 Future product extensions 46
8 Conclusion 48
Appendix 2: Interview protocol 54
Appendix 3: Open coding 58
Appendix 4: Axial coding 60
Appendix 5: Memo Merge Interview B, C, D and E 61
Appendix 6: Memo Merge Interview F, G, H and I 62
Lending-based crowdfunding is an exponential growing phenomenon that facilitates borrowing and lending as an alternative financing method in an online market place. The online transaction is characterized by information asymmetry, which can be reduced through the development of trust. Lending-based platform managers have to deal with trust development on a daily basis. This study aims to recommend courses of action to help these managers to increase the trust of their lenders. Exploration of the lender investment process allowed uncovering the trust driving determinants.
This study applied the grounded theory approach, since this best suits the qualitative characteristics of the nascent phenomenon. Data collection was conducted through ten semi-structured interviews with a self-selection sample of lenders of the Dutch lending-based platform ‘Kapitaal op Maat’.
The findings of the research unveiled that trust development seems to be highly influenced by the experience of the lender: investment experience in general, crowdfunding experience and experience with the platform. Theoretically this means that the experience levels of lenders may be indicators for the information needs in the development of trust during project investment. The more experienced the lender is, the less affective information he considers during the investment process. The practical implication is that the platform is the major influencer of the lender experience. Providing information transparently and developing a strong customer relationship seems to facilitate the increase in trust development with lenders.
These findings function as a starting point to better understand the investment process. In addition, the growing use of lending-based crowdfunding stresses the importance for platform managers to integrate customer service and marketing strategies.
Key words: Experience, information asymmetry, investment process, lending-based crowdfunding,
The process of developing my thesis involved the support, cooperation and guidance of many people. Therefore I would like to dedicate special thanks in this section, as without them I could not have finished my study in such a pleasant way.
Firstly, I would like to acknowledge my supervisor drs. A.C.C. Gruijters for the guidance, feedback and patience during supervision sessions. His immense knowledge in the field of qualitative research really helped me and pushed me in the right direction to carry out my thesis.
Secondly, my sincere thanks also goes to the team of ‘Kapitaal op Maat’ with whom I had very pleasant contact. Pim van de Velde, Monique van Holstein and Jeroen Kloeke were always available to answer my questions and assisted me providing with the information I needed and putting me into contact with my interviewees.
Subsequently, as third acknowledgement, I want to thank all the people who devoted some of their valuable time for sharing their thoughts concerning their crowdfunding experience during the interviews I held. The interesting insights they gave me, the openness and the willingness to help me, a total stranger, intrigued me and really motivated to continue my research.
And last, but not least, I would like to thank my family and friends for their continuous support. Without them the completion of my thesis would have been much more difficult. Thank you!
1.1 Background information
Since the latest global financial crisis, businesses and consumers have become more cautious in their financial planning. Financial authorities have caused many failures that resulted in decreasing trust levels in the consumer sentiment. Consequently, alternative finance methods have developed rapidly in the US, UK and continental EU (Wardrop et al., 2015). These new methods, which include crowdfunding, are financial instruments and distribution channels that emerge outside of the traditional financial system and often occur in an online environment.
In general, the changes in the financial climate have caused businesses to be more careful. Therefore the crisis also has affected banks in the way they deal with their daily business. Their precautious behaviour and risk-avoiding attitude makes it difficult for start-ups or high-risk ventures to acquire starting capital. To be more precise, no high-risk loans are supplied even if high potential can be forecasted. Thus, where the bank used to be the preferred stakeholder to approach for financing, now entrepreneurs need to find other methods of financing. As a result, the market requires financing methods that are reliable and trustworthy.
These developments might explain why the transparent rules and regulations that are linked to crowdfunding are experiencing an exponential growth in popularity in many areas in the world. In 2014, 64 million Euros were funded by the crowd in the Netherlands alone, which was an increase of 100 % compared to 2013 (Douw & Koren, 2015). The worldwide prognosis for this year’s total crowdfunding investments amounts to 30 milliard.
The cause of the financial transformation, however, is also rooted in an earlier development: the revolution of the digital environment. Web 2.0 did not only give users new options to communicate and share information, it also enabled companies to reach out to their consumers more easily. Consequently, the consumers (also referred to as the crowd) were gradually taking part in the development of new products and services. Via online communication the crowd has been asked to help in acquiring ideas, feedback, and solutions, which enabled firms to better anticipate the needs of their clients. This new way of interactive value creation was defined as ‘crowdsourcing’ (Howe, 2006). The crowdsourcing movement, enabled by the Web 2.0 and financial changes, explains how crowdfunding came into existence.
The voluntary crowd contribution toward production processes not only functioned as a diagnostic, more specifically, it also facilitated as an open call for funding from the crowd. With crowdfunding an entrepreneur asks a crowd for the provision of financial resources, either in the form of donations (donation-based), in the exchange for the future product (reward-based) or some form of reward to support initiatives for specific purposes (Belleflamme et al., 2014, p. 588). The reward can be monetary or non-monetary and stands in contrast to traditional fundraising efforts such as securing funds from banks, venture capitalists, and foundations (Gerber et al., 2012). Monetary rewards fall
under the equity-based and lending-based crowdfunding. With equity-based crowdfunding the lenders receive equity or ownership in the company, whereas with lending-based crowdfunding the lenders collect interest on the funding they provided the campaigners with.
According to Douw & Koren (2015) the power of crowdfunding is the direct relationship between the campaign starter and the lenders. The short ties between both stakeholders allow crowdfunding to acquire a strong social character and stress that crowdfunding is more than solely a financial transaction. Apparently crowdfunding responds to the need of human, direct and transparent ways of financing projects, which is not always guaranteed with institutions such as banks.
1.2 Prior research and cause
Prior to the initial research, interviews were held with two platform managers to acquire a better view of how crowdfunding is constructed. Subsequently, their experience enabled me to unveil what concepts are ‘trending’ in their industry and to uncover the difficulties they are dealing with. The first interview was held with Bart Lacroix, the owner of the donation-based platform ‘1% Club’ and the second interview was held with Pim van de Velde, the owner of the lending-based platform ‘Kapitaal op Maat’ (KOM).
In the conversations both platform owners agreed that trust is the fundamental requirement of crowdfunding. According to Bart, the direct contact between borrower and lender facilitates information transparency, which enables the development of trust. Pim even described the word ‘trust’ as the most trending topic within the crowdfunding business since he started his platform 1.5 years ago. According to him the core drivers of trust development in lending-based crowdfunding are information quality, branding and offline endorsement. By this he predominantly refers to trust development with lenders. According to Bart the trust inducing drivers are related to the funders’ association with the project, the information transparency, the platform brand strength and the privacy/security measures.
The main difference between the two platforms is rooted in the transaction that occurs. Bart explained that in donation-based crowdfunding the crowd gains most trust from the borrower. Hence, the borrower starts campaigning for a good cause and looks for help in terms of knowledge and money. In return, the crowd gives knowledge and money with the guarantee that the campaigner makes the promised impact. The role of the platform is to assist in giving tools to make this work. At KOM the borrower requests funding with the purpose to obtain financing for his company that eventually assists in acquiring a higher return on investment. The lenders who fund money also obtain a return on their investment.
Another variable that seems to influence the investment process is the unequal information provision, also referred to as information asymmetry. Furthermore, lending-based crowdfunding is subject to higher risks due to the amounts of invested money. Especially in lending-based crowdfunding, trust development requires more information in order for a transaction to succeed.
1.2 Research aim
The initial research objective of this study is to recommend courses of action to lending-based platform managers in order to increase the trust of their lenders. Conversations with platform owners have pointed out that they are working on the image of being a reliable and trustworthy platform on a daily basis. Since different variables facilitate the development of trust, a better understanding of the trust developing process would not only improve the sustainability of lending-based crowdfunding, but would also increase the trust within the industry.
Given the fact that lending-based crowdfunding is increasing most compared to e.g. donation-based crowdfunding, the purpose of this research is to add theoretical knowledge related to the Dutch lending-based crowdfunding market. As trust is the crucial enabler, evidence concerning its development can aid the efficiency of daily crowdfunding business. Subsequently, researching the lenders’ needs in order to generate trust requires the exploration of the investment process.
Current literature has often elaborated on motivations as influencing factors of the investment process (Block & Moritz, 2014). In addition, research by Mollick (2014) compared the decision-making of the crowd with professional investors such as Venture Capitalists (VCs) and Business Angels (BAs). Greiner & Wang (2010) specifically researched the trust development of lenders and made a distinction between affective and cognitive determinants. However, findings have not resulted in a unique investment process that explains the variables that influence trust. The heterogeneous characteristics (Lin et al., 2014) of the capital providing crowd, in combination with the complexity of financial decision-making (Olsen, 2010), might explain why no golden rules were developed concerning trust.
The combination of the nascent stage of crowdfunding literature, and the practical relevance of improving the lending-based business model in Dutch crowdfunding, explain the relevance of researching trust development. The newness of crowdfunding literature clarifies why researching the phenomenon requires a qualitative approach; therefore this research dominantly applied the strategy of the Grounded Theory (Glaser & Strauss, 1967). Aiming to improve the understanding of trust development, lenders were questioned on their investment process. Consequently, the theme of this thesis is constructed around the central research question:
What is the role of trust in the investment processes of lenders?
And what is the role of the platform in this process?
As a starting point of this research a literature review was conducted to construct an overview of the current knowledge concerning trust development in lending-based crowdfunding and related fields. The literature research resulted in a conceptual model that describes the literature findings and presents the plan of approach for further empirical research. The research design further outlines the
strategy and elaborates how the interviews were constructed and how they were analysed. Then findings, discussion and conclusion follow.
Finally, this thesis aims to develop an understanding of the lender trust development during the investment process in lending-based crowdfunding. Findings could facilitate new insights for literature and assist in developing a better understanding of the lender investment process. Consequently, as indicated previously the practical implications of the findings assist in advising platform owners how to increase trust with their lenders.
2. Literature review
This chapter presents an overview of the lending-based crowdfunding literature. In addition, related topics were added. The literature findings are further discussed in chapter 3, the conceptual model & research questions.
2.1 Peer-to-peer lending
Peer-to-peer or people-2-people (P2P) lending can be defined as a specialized consumer-to-consumer e-commerce model that facilitates borrowing and lending between individuals in online marketplaces (Greiner & Wang, 2010, p.105). Lenders who engage in lending-based platforms obtain interest for the risk they take in the crowdfunding projects. The running time is different per project and often varies from three to six years. The concept of private loans is not a new business model, however, the shift towards transaction in an online marketplace makes the phenomenon innovative.
Prior lending-based crowdfunding research t focused on P2P lending markets, since this segment is the biggest proportion of the whole crowdfunding market (Wardrop et al., 2015). To illustrate, in Europe the P2P lending market increased with 272% between 2012 and 2014 compared to e.g. reward-based crowdfunding that grew with 127% (Wardrop et al., 2015). Also in the Netherlands the lending-based crowdfunding takes the lead. Since 2005 lending-based crowdfunding made its entry with Zopa UK and in the Netherlands we know ‘Geldvoorelkaar’ as the first online P2P lending platform of the country since 2010.
Stakeholders
The stakeholders of P2P lending are the platform, the borrower and the lender. Lending platforms should be referred to as the brokerage that facilitate the transaction process between people in need of money (borrowers) and people who want to invest money (lenders) (Greiner & Wang, 2010). The platforms are the intermediaries that bring together both the borrowers and lenders, and try to match the expectations of both parties (Bachmann et al., 2011). Bachmann et al. (2011) add that sometimes the lenders are part of a community of similar interests.
B2B or P2P lending
P2P stands for people-to-people or consumer-to-consumer and refers to the direct interaction between humans. The relative ‘new’ term seems to increase in its use and replaces terms such as B2B (business-to-business) or B2C (business-to-consumer). Especially in the online market, there exists a fuzzy area with B2B and B2C, as it has become clear that it is the consumer who decides if a product works, and not the business. Therefore, one could argue if a distinction is really needed for the purpose of this research.
2.2 The lender investment process
Lending-based crowdfunding is based on either an auction process or it is based on a fixed interest rate that the platform calculates. With the auction process the borrower decides upon the highest interest bid the lenders proposed. In case of a fixed interest rate, the intermediary platform decides on the height of the interest rate, which is based on the borrowers’ risk analysis and characteristics (Bachmann et al., 2011). If the lending process leads to a fully funded loan-request, which is linked to a time frame, the loan is granted to the borrower, who will eventually start the repayment process with the lenders (Bachmann et al., 2011). The platform generates revenue via a service fee that is integrated in the funding and interest rates that are paid by both the borrowers and lenders.
Determinants
Online lending-based platforms are different from traditional bank lending systems, where the bank offers services such as guarantees and credit risk evaluations. At online platforms, the lender evaluates the borrower’s capability and motivation to pay back, instead of an intermediary person such as the professional banker. Questionable is to what extend the relative ‘unprofessional’ lender is capable in evaluating the potential of a project compared to the analysis professionals execute. Since little is known about the lender behaviour in crowdfunding, the decision-making process is often compared to the professional financers such as bankers, VCs and BAs (Bachmann et al., 2011; Mollick, 2014; Hornhuf & Schwienbacher, 2014).
When lenders analyse borrower projects they refer to financial, demographic determinants and social capital characteristics. Financial determinants, also named as hard factors, are characteristics alike: credit ratings, detailed information on income and monthly expenses, house-ownership or the debt to income ratio (Bachmann et al., 2011). Greiner & Wang (2010) name this information the ‘economic status’ or cognitive determinants that form the trust antecedent through observations and perceptions of the available information of the borrower. The affective determinants also indicated as social capital or soft factors relate to referral, recommendation, endorsement and feedback from peers. In addition, also visuals, such as photos, fall under social capital. Both Ravina (2007) and Pope & Sydnor (2008) confirm that pictures can positively influence lenders.
As part of the campaigning strategy, borrowers can e.g. influence the lenders with their appearance and content listings. Sometimes platforms also allow members to connect to one another through the use of blogs, networks or communities (Greiner & Wang, 2010). The implication of this information is that social capital does not directly tell us something about how the lender develops trust, however, it gives relevant cues that a lender may consider other information sources for making judgment.
Demographic characteristics such as gender, race and age also fall under one of the determinants lenders take into account. Research of Ravina (2007) illustrates that resemblance
tends to influence the lender is taste based discrimination, which is the subjective propensity of a lender in favour or against a group of people (Pope & Sydnor, 2007). Statistical discrimination occurs when lenders offer higher interest rates to a minority (e.g. old people) caused by the fact that this minority has a statistically proved higher default rate than other borrowers (Phelps, 1972).
Table 1 summarizes the lender determinants that literature has written about so far.
Table 1 Summary P2P lender determinants
Lender
Hard factors (cognition-based) E.g. credit ratings, income information, monthly expenses, house ownership, and debt-to-income ratio.
Soft factors (affective-based) E.g. referral, recommendation, presence of third-party seals, feedback. Demographic factors E.g. Gender, race, and age.
2.3 The professional investment determinants
Since investment process research in crowdfunding is still in its infancy, information about professional investors is included. Their knowledge can form a starting point for researching the investment process of the ‘unprofessional’ investor. Even-though literature has researched the comparisons between the professionals and the lending crowd (Hornuf & Schwienbacher, 2014; Mollick, 2014), Block & Moritz (2014) state that it is still unclear whether the lenders use similar decision criteria as professional investors. The following paragraph elaborates on the investment determinants and characteristics of Venture Capitalists and Business Angels.
VCs and BAs
The fact that the nature of investments is very complex, unitary and dynamic might explain why crowdfunding literature has not yet developed a unique investment process. Pop (2012) explains that an investment process includes the assembly of all decisions, actions and measures adopted and implemented (in the sense of achievement) in the purpose of carrying out the investment projects and their funding arrangements. It may be clear that many variables influence this process. In addition, this chain of steps also depends on the subject of investment (Pop, 2012).
Mollick (2014) compared crowd investors to VCs, and found that the crowd evaluates potential projects on similar determinants, namely: the quality of the product, the team and the likeliness of success. However, since the crowd lacks professional investment experience and knowledge it is important to obtain an overview on what quality signals the professionals take into account.
The evaluating project determinants among professional investors (VCs and BAs) vary through the risks that are accompanied with the potential investment. Nevertheless, according to Mason & Stark (2004) there seem to exist eight overlapping criteria VCs and BAs take into account,
namely: entrepreneur/management team, strategy, operations, product/service, market, financial considerations, investor fit and the business plan.
Financing through banks should be compared to the investment behaviour of VCs and BAs more carefully, since there often exist conflicting interests. To illustrate, bankers evaluate the risks of start-ups through criteria such as the generation of sufficient cash flow and the presence of collateral security. Here, bankers have the difficulty of information asymmetry since business prospects are either unavailable, uneconomic to obtain or difficult to interpret (Mason & Stark, 2004).
On the contrary to banks, VCs share the success of the business that they invest in, and, thus also the downside in case the business fails. Therefore VCs put their biggest emphasis on the management team, the product/service and the market when analysing the potential of a business (Mason & Stark, 2004). Specifically, within the management ability VCs look at the management skills, quality of management, characteristics of the management team and the management team’s track record. Within the market set they take environmental changes and competition into account and in line with the product or service the VCs like to know more about the opportunities concerning product differentiation.
As VCs attach more towards market risks, BAs attach more towards agency risks. According to Mason & Stark (2004) BAs do so because they lack data to look at the market risks, their funding resources are limited compared to VCs, contracts between BAs and entrepreneurs are more informal and previous market experience of AIs seems to boost their feeling of ability of knowing the market. When linking the investment strategies of VCs and BAs to crowd investing, the major differences, besides the expertise, can be explained through the contractual agreement and security regulation that are involved. According to Hornuf & Schwienbacher (2014) crowd investing relies on more standardized contracts, whereas VCs and BAs rely on more private orderings with the potential start-up investments. In addition, VCs and BAs integrate various securities such as anti-dilution provisions and liquidation preferences, while in crowdfunding these securities are not negotiable.
The following table summarizes the determinants that professionals consider when investing.
Table 2 Summary professional investor determinants
VC/BA
Quality team Management skills, management quality, team characteristics, team track record
Quality product Product differentiation
Market Environmental changes, competition
2.4 Trust
As indicated previously, crowdfunding research has proven that trust takes on the fundamental role of the investment process. Without trust, the whole transaction cannot take place and therefore it is
needed for the survival of lending marketplaces (Greiner & Wang, 2010). As research related to trust development in crowdfunding was not saturated, knowledge about trust development in similar financial environments is included to make comparisons about the investment procedures.
2.4.1 Trust in financial markets
General trust development
Trust plays a key role in human decisions and the necessity of the concept is often only noticed when its absence occurs (Olsen, 2008). An example of such a case is the latest financial crisis, when untrustworthy behaviour provided cogent evidence of the importance of trust in financial markets (Olsen, 2008). Unless the importance of trust in such markets, it is striking that the concept is often excluded from any financial theory or only handled as a probability. Especially since empirical research proved that trust takes on an important role in building a strong client/advisor relationship. Subsequently, investors cannot solely predict the future based on the knowledge they have and therefore they need to settle for ‘second-best’ solutions (Olsen, 2008; Clark-Murphy and Soutar, 2004). Trust is thus complex and therefore requires a working definition that Rousseau et al. (1998) define as “a psychological state comprising the intention to accept vulnerability based upon positive expectations of intentions and behaviour of others.’’
The necessity of trust can be explained through the extrinsic and intrinsic value it adds in the decision-making process. Herewith, extrinsic value reduces transaction costs as it fills a gap to allow agreement and cooperation (Olsen, 2008). On the other side intrinsic value reduces suspicion and animosity, which enables more cooperative behaviour and rapid adaptation to environmental uncertainty. Likewise, to be able to develop this trust, humans undergo a dualistic process in the decision-making mechanism. Hence, they go through both experiential and rational-based processes. Affective trust is the result of the experiential process and cognitive trust arises from the rational decision process. Both processes are also referred to by Greiner & Wang (2010) as trust determinants in the P2P lending segment. According to Olsen (2008) an individual’s basic propensity to trust appears more a function of one’s personality and developmental experience. Therefore decision-makers tend to rely more on affective-based trust in situations where cognitive trust is harder to judge. In addition, professionals appear to put more weight on affective elements in general.
Besides the affective and cognitive processes, trust also comprises some other attributes that drive the mechanism. E.g. trust seems to be fragile and easily broken, it takes a lot of time to create trust compared to losing it and trust-destroying events are usually more salient than trust building events. Here again Olsen (2008) explains that the affective nature of trust is more evident.
Trust as risk factor
According to Olsen (2008) research has shown that trust is an attribute of perceived risk and that perceived risk, as well as expected return, is a primary driver of investment value. In the uncertainty of
the financial world, trust even comprises 90 % of the “risk iceberg” that is below the “waterline” (Olsen, 2008). Moreover, professional investors like financial advisors, corporate managers and regulatory officials are encouraged in their decision-making through the faith they developed rather than the forecast probabilities (Botazzi et al., 2011). Thus, financial decisions are not solely dependent of the financial predictions that are made but they greatly rely on the trust development during relationships. Trust thus functions as a proxy for more objectives measures of risk as investment complexity, time pressure and familiarity. In addition, the lesser the perceived trust, the greater perceived risk.
2.4.2 Trust in E-commerce
Crowdfunding can be defined as a new sort of e-commerce, therefore there might be similarities concerning trust development. Bart et al. (2005) identified drivers of trust for different website categories and different consumer groups. The categories of their research that most relate to crowdfunding are: shopping and e-tailer, finance and community. According to Bart et al. (2005) the dominant drivers of these particular categories are as follows:
- Shopping and e-tailer: security, absence of errors, order fulfilment - Finance: security, absence of errors, brand strength and advice
- Society and community: privacy, absence of errors, community features
In similar research Kim et al. (2008) elaborated on antecedents of trust and risk within the decision-making process. These entities comprise the following:
- Cognitive-based: privacy protection, security protection, system reliability, information quality etc.
- Affect-based: reputation, presence of third-party seals, referrals, recommendation, buyers’ feedback, word-of-mouth
- Experienced-based: familiarity, Internet experience, e-commerce experience - Personality oriented: disposition to trust, shopping style
Their research revealed that cognition-based and affect-based factors increase consumer’s trust as well as decrease a consumer’s perceived risk in completing an e-commerce transaction (Kim et al., 2008). Which is thus very similar to the dualistic decision-making process (cognitive & affective) as indicated by Olsen (2008).
The cognition-based factors Kim et al. (2008) researched were information quality, perceived privacy protection and perceives security protection, which all had strong positive effects on consumer trust. The affect-based factors were third-party seals and positive reputation of selling party, where the authors found that third-party seals do not influence consumer trust, but they did decrease the consumer’s perceived risk.
2.4.3 Trust development in crowdfunding
Clearly, trust plays a major role in the decision-making process in general; moreover trust is an important determinant in the decision-making of investment processes. Impressions of trustworthiness matter in financial transactions as they function as a sort of prediction for investors (Duarte et al. 2012). Thus also in lending-based crowdfunding this rule counts: the more trustworthy a potential investment transaction appears, the higher the probability that the transaction will take place. Subsequently, the accuracy of information that is presented is crucial as it determines how a lender evaluates a project. As mentioned previously, trust determinants in investment procedures are categorized into affective and cognitive mechanisms (Greiner & Wang, 2010; Olsen, 2008). Since trust is developed upon the evaluation of the available information it is crucial that information asymmetries are reduced to boost the efficiency of campaigns. Thus in contrast to offline transactions, where the entrepreneur can intensify investor relationships through personal communication, in online transactions the communication is replaced by pseudo personal forms such as videos or social media messaging (Block et al., 2014).
Particularly in lending-based markets it seems that soft factors (e.g. referral, recommendation) take on an important role in the investment decision. Iyer et al. (2009) found that, similarly to banks, capital providers in lending-based crowdfunding markets primarily rely on cognitive determinants to make investment decisions. Nonetheless, the more risk is attached to a project, the more soft factors (affective determinants) the lenders take into account (Iyer et al., 2009). However, unclear is what particular elements within the soft factors are the generators of trust. In addition, information can be valued differently due to influencing variables like the business model of the platform, the context and the motive of the funders (Lin et al., 2014). Hence, the crowd should be considered heterogeneous and requires dissimilar information to enable its decision-making. Consequently, to overcome information asymmetries literature has proposed various alternatives. Agrawal et al. (2013) and Block et al. (2015) have indicated that the communication of quality signals (e.g. branding, information access, warrantees, patents, trademarks and education/experience of the management team) by the capital-seeking party can help overcome this obstacle. Knowing what quality signals are crucial for the lender decision-making facilitates more sustainable and successful financing through the crowd. Agrawal et al. (2013) explains that the decision-making of backers concerning quality signals is determined by the market design in which the crowdfunding transaction takes place. Besides quality signals, referred to as ‘reputation signalling’, ‘rules and regulations’, ‘crowd due diligence’ and ‘provision point mechanism’ are part of the market design.
2.5 Conclusion literature
Although there is an increasing number of studies on crowdfunding, the findings have not yet led to the development of golden rules. Studies are thus in a conceptual phase and require more quantitative confirmation based on empirical research in order to develop theory (Block & Moritz, 2014)
So far lending-based crowdfunding research focused on the variable ‘information asymmetry’ and how quality signals could overcome this. Subsequently, in aiming to better understand the trust development more extensive research would be needed to further crystallize the lender investment process. The role of the platform is noted as an important influencer in this process, especially for the development of trust. Remarkably, studies focusing on the role of crowdfunding platforms and their optimal business models remain scarce (Block & Moritz, 2014). Consequently, the next chapter will present how this research aims to expand the scientific understandings related to trust development in the investment process.
3 Conceptual Model & Research questions
This chapter presents the main literature findings and explains how I developed my research questions. Furthermore, this section elaborates on the strategy development for executing empirical research that is operationalized through integrating the guidance from sensitizing concepts and sub-questions.
3.1 Literature findings – the concepts
The first concept is the affective and cognitive ‘investment criteria’ that lenders/investors need in order for a transaction to occur. Prior literature dominantly focused on the major quality signals, but did not clarify the exact needs in the investment process or appoint trust indicators. Neither has prior research led to an investment process model or anything similar. Thus, when aiming to better understand the lender investment process further research is required. However, the needs of crowdfunding- and professional investors have been indicated (Greiner & Wang, 2010; Bachmann et al., 2011; Hornhuf & Schwienbacher, 2014; Mollick, 2014; Pop, 2012). Also, criteria of decision-making in e-commerce could apply in lending-based crowdfunding (Bart et al., 2005; Kim et al., 2008).
The second and potentially useful literature finding is that crowdfunding research mentions information asymmetry as the biggest obstacle for lenders. Lenders evaluate a project based on the estimation of the default probability. However, where professional investors are able to execute risk reduction strategies, the crowd just has to deal with the available information (Block & Moritz, 2014). Communication of quality signals by the capital-seeking party can overcome this hurdle (Agrawal et al., 2013; Block et al., 2014). Literature states that the platform is the trust facilitator, as it enables information transfer of both the lender and the borrower in the market place (Greiner & Wang, 2010). Communicating quality signals such as branding, information access and warranties help to overcome information asymmetry obstacles. However, these signals seem rather general. One could question which particular determinants communicated by the platform especially facilitate trust development? And do all sorts of lenders require the same information? As Lin et al. (2014) have indicated the crowd to be rather heterogeneous; are lender characteristics maybe a prerequisite for their information needs?
The third finding relates to research that might be useful to clarify the lenders’ decision-making. First, research found that in both professional and lending-based markets under uncertain cognitive circumstances (poor credit ratings), affective determinants become more important for decision-making (Iyer et al., 2009; Olsen, 2008). The affective determinants, such as experiences that are developed during the relationship between the investor and entrepreneur, then take the lead (Block & Moritz, 2014; Olsen, 2008). This would imply that the affective determinants take on a more significant position within the lender investment decision-making because of the high-risk perception, due to e.g. information asymmetry. However, the source of the affective determinants is unclear. Which affective determinants are counted as trust enablers? Are these only based on the relationship,
or can one find other sources that drive affective determinants? And if the relationship is relevant, does this relationship refer to the borrower, the platform or both?
Up to this point I have a general sense about how investment decision-making is constructed through the use of investment criteria. The major influencers of the decision-making process seem to be relationship development and information provision. The gap is ambiguous: first, there is no lender investment model. And second, the exact trust drivers of this process cannot be appointed. One could argue that this is caused by the complexity of the investment process and the different sorts of lenders. However, one could still expect that there would be some differences or similarities within the lender crowd. Maybe certain lender characteristics require certain quality signals. However, the literature I consulted could not answer these questions. Therefore empirical research is needed to further research. This resulted in the development of the following research questions:
What is the role of trust in the P2P crowdfunding lending process?
And what is the role of the platform in this process?
3.1 Sub-questions and sensitizing concepts
Obtaining trustworthy answers to the abovementioned questions requires research in a sample of the population of lending-based crowdfunding investors. By focusing on their decision-making, I aim to develop an investment process that illustrates the particular determinants responsible for the development of trust. Since one cannot literally ask how their trust is developed, their whole investment process needs to be studied. In addition, analysis should also provide further answers to the research questions. However, to guide the process of answering the research questions, the following sub-questions were developed:
1) What are the steps of the investment process?
2) What criteria do lenders consider in their investment process? 3) What particular criteria contribute to the development of trust?
4) Are there other contributions outside the steps of the investment process that influence decision-making or the development of trust?
5) Are there similarities or differences among the lender characteristics?
The answers to these questions ought to assist me in developing a lender investment process model. Furthermore, analysis of the results facilitated an understanding of trust development and elaborated on how the platform influences trust. The combination of existing literature and empirical research form the base for the establishing findings. The conclusions I have drawn from prior research are referred to as sensitizing concepts, function as a starting point of the explorative research and facilitate an effective start of the study. These concepts are: investment criteria, relationship and information
provision. Despite their deductive character, exploratory characteristics take the lead in my research. The sub-questions I developed illustrate that my research is not forced into existing frameworks or concepts, but is open to gather any relevant data. The sensitizing concepts thus set the context and direction of my study (Bowen, 2006). In addition, according to Bowen (2006), the sensitizing concepts lay a foundation for the analysis of the research data. By using them I am able to advance my research instead of starting from the beginning (Boeije, 2014).
Since the results of this study have major practical relevance, the findings form the base for recommendations to the lending-based platform managers. Initially, these recommendations aim to advise them on how to increase trust with their lenders. However, the explorative character of this research means that findings could also develop other relevant knowledge related to investment decision-making.
4. Research design
This section contains a detailed description and explanation of the applied research strategy and methods used to reach the research objective and answer the research questions. It elaborates on the applied research design, the data collection & sample and the data analysis.
4.1 Research approach
Due to the nascent stage of published research on the lending-based crowdfunding investment process, an exploratory research design is the most suited to develop a better understanding of the process. Such a qualitative design is flexible and adaptable to change, which are advantages for researching the understanding of this quite practical phenomenon. As indicated earlier, the research objective aims to advise platform owners on how to increase trust with their lenders. To illustrate, a quantitative research approach would require existing research findings, as they would serve to measure relationships. As prior research related to lending-based crowdfunding is both complex and dynamic, quantitative research is not appropriate.
The exploratory design applied in this thesis is the grounded theory, which Glaser & Strauss (1967) describe as an approach that aims to discover or generate theory grounded in the data produced from the accounts of social actors. The approach is inductive; in this study it aims to produce theory concerning the investment process. As indicated in Chapter 3 (The Conceptual Model) the findings of the literature review will be used as a starting point for my research. These findings function as sensitizing concepts that set the context and the theoretical background. In addition, these concepts function as a clarification tool during the analytical process. Furthermore, I added sub-questions to guide my research with the aim of answering the main research questions and developing the recommendations. Although this approach integrates some deductive characteristics; the research is pursued objectively and leaves room for signalling any relevant insight. As Bowen (2006) and Boeije
(2014) have indicated, this theoretical starting point can elevate the research and thus develop a deeper understanding of the phenomenon.
The combination of literature findings and empirical research assisted me in answering the research problem and thus added theory; in addition it formed the base for the recommendations for the platform managers. The data were gathered through semi-structured interviews with key investors of a P2P lending-based crowdfunding platform. The data collection and analysis (also referred to as coding) took place simultaneously, which enabled me to develop analytical codes that emerged from the data and to reorganise these data into categories (Saunders et al., 2012). At this stage deduction also took place as interim findings and assumptions of data output were developed, which were consequently compared in next data collection rounds. The sensitizing concepts in the ‘analysis stage’ of the research shape interpretation and sensitize the data, which to a certain extent supports the development of findings and the relevant theory.
4.2 Data collection
As this study focuses on lender decision-making criteria, the population comprises lending-based crowdfunding investors, referred to as lenders. I first interviewed lenders, who are skilled at funding money in lending-based crowdfunding settings, and derived their reasoning from personal experiences, which increases reliability. Subsequently, the primary data collection took place with a variety of lenders from the Dutch lending-based crowdfunding platform ‘Kapitaal Op Maat’ (KOM). This lending-based platform focuses on financial crowdfunding and offers a crowdfunding service for both entrepreneurs (borrowers) and potential investors (lenders).
4.2.1 Sample
Ten semi-structured interviews were held with a group of volunteers of the lender database at the KOM platform. Hence, the research data was integrated in accordance with self-selection sampling, as lenders of the platform were requested to take part in my research (Saunders et al., 2012). Based upon the data outcomes of the first set of four interviews, I awarded some preference criteria for the next subject set. Thus, this thesis also possesses some characteristics of theoretical sampling (Boeije, 2014). This way I was able to investigate potential relationships between the lending subjects.
Both the platform managers and I contacted the subjects. Please refer to Appendix 2 for the e-mail invitation that was sent. Interviews took place between April 2015 and July 2015. The conversations were recorded and lasted between 18 and 51 minutes. Additional notes and memos were recorded in a journal. The subjects were all male, ranging from the age of 28 to 65. Eight lenders were very experienced, one had little experience and one had invested money for the first time. The majority of the participants were experienced in investments. Appendix 1 presents an overview of the interviewed cases.
The interviews were held until the emergent data output and categories tended to converge and a status of saturation was reached (Glaser & Strauss, 1967). After six interviews I felt that I came close to saturation, however, I continued with the interviews to further analyse specific differences among the subjects until I reached ten interviews.
4.2.2 Interview process
The interviews were held by telephone or Skype and were conducted in a semi-structured style with open-ended questions to guarantee free expression of the views and experiences of the subjects. The development of the interview was guided by the initial findings of the literature review; furthermore questions were added upon the data collection. In addition, KOM requested for some questions to be incorporated into the interview.
The aim of the interview was to generate an understanding of how the investment process of a lender is built up. The data output of the interviews formed the base for analysing the investment process and looking for relationships that eventually enabled me to develop recommendations.
The interviews were structured as follows: (1) questions related to the demographics, (2) contextual information, (3) the investment procedure, (4) suggestions and improvements related to the investment process to acquire information on what elements could improve efficiency, (5) risk awareness, (6) future view on crowdfunding use, (7) network and (8) general remarks and additional questions. Please refer to Appendix 2 for the question guide.
The interviews were structured in such a way that the subject was first asked about his own experience. Then, other criteria were named to check if any unconscious factors influenced the decision-making. Thus, subjects were not influenced by pre-determined convictions I developed during the process. The simultaneous data collection and analysis made me update the interview constantly.
4.3 Data analysis
As indicated, the data collection and analysis took place simultaneously (Glaser & Strauss, 1967). After transcription of the interviews the first data output and findings were analysed, restructured and coded. The strategy for analysis of the interview transcriptions comprised the comparison of the coding schemes. First, open coding was applied where I reorganized the data into categories. Secondly, I recognized relationships between the categories, which can be referred to as axial coding, and in the final stage I applied selective coding (Saunders et al., 2012). In addition, during the data collection, the findings of the data output and analysis phase I wrote memos that functioned as a hinge between my thinking and writing process. In addition, I also looked for clarification within the meaning of the sensitizing concepts. The sub-questions assisted me in answering my research questions. The process combination of memo writing, information abstraction and moving towards
selective coding enabled the development of theoretical findings. The following paragraphs explain the coding process further.
4.3.1 Transcription
The first stage of data analysis was transcribing the interview output. After every interview I directly transcribed the results so that any recurrent themes or extra needed understanding could be integrated. I used the program ‘Evernote’ for transcribing, in which I also recorded the interviews. The transcription was very time consuming, however, very valuable as it facilitated the creation of an overview and representation of all the data. Hence the data were prepared for further stages of the analysis, namely the coding. I organised the transcriptions per subject in the ‘Evernote’ program and printed them for the first phase of coding.
4.3.2 Open coding
After the transcription of an interview I first coded the printed version. This way I explored the research field without any boundaries and could make the first steps in categorizing the concepts. According to Corbin & Strauss (1998) coding involves finding the right word that conceptually describes what researchers think is discussed in the data. The codes I attached to data thus do not refer to a specific word, but to the essence of the data. The first part of Appendix 3 shows an example of the open coding phase of ‘subject I’; the second part depicts the overview of all the open codes that I attached to the interview transcripts. This made it easy to return to particular codes of the open coding of the transcribed interviews at a later stage.
After coding the first four interviews I wrote a sub-conclusion memo about my ideas, expectations and reflections towards the next interviews (Boeije, 2014) and started to prioritize and subcategorize codes. This can be referred to as a first move towards axial coding.
4.3.3 Axial coding
In the phase of axial coding, I prioritized and further categorized codes, which enabled me to abstract the most relevant categories. As indicated by Boeije (2014) the axial coding phase functions as a bridge between open and selective coding. Other activities were related to comparing fragments of codes, ordering codes, testing provisional ideas and expectations and further memo writing.
The reorganization of the axial coding was documented in MS Excel. This enabled me to obtain an overview of all interviews (subject B – K) and compare the codes to find overlaps, differences or relationships. The document is labelled with the axial codes in the left vertical column and with subjects in the upper horizontal row.
Appendix 4 provides an overview of the relevant axial codes including findings and implications of the data comparison. After every interview I wrote a memo to maintain my thinking process. After the first four interviews I merged the written memos into one document, which led to
the summary memo of the first interviews showed that I did not yet possess sufficient data to show valuable results or relationships. Therefore, at that stage I continued to collect data through interviewing and added questions to the protocol. Please refer to Appendix 5 for details.
The next set of interviews however enabled me to extract reoccurring topics and develop assumptions. At this stage I noticed several important aspects: (1) the investment process seems to be dominantly based upon cognitive determinants. (2) The affective determinants seem to come into place due to the interaction with the platform before and during the investment process. Here, I noticed a link with the literature of Olsen (2008) that explains that in financial investments, affective determinants are needed to develop trust prior to decision-making. However, in crowdfunding the relationship with the intermediary, i.e. the platform, seems to be more relevant than the relationship with project initiator, i.e. the borrower. (3) Thus the platform is an important influencer of the investment process. Therefore, I added a few questions that specified the role of the platform more in-depth. This is an example of how I continuously updated the study. Please refer to Appendix 6 for the memo details that describe the above-mentioned process.
4.3.4 Selective coding
The phase where the most important and relevant categories are selected, assuming that these assist in answering the research question, is referred to as ‘selective coding’. After ten interviews I was able to develop an understanding of the investment process that lenders undergo in the P2P lending-based crowdfunding environment. In addition, the data output saturated the three main topics and thus my data gathering was sufficient to answer my research questions. The topics that returned most frequently to influence the trust development in the investment process were: ‘investment criteria’, ‘experience’ and ‘platform relationship’. The memos show my thinking related to these reoccurring topics (Appendix 7). The last section of Appendix 3 depicts the coding scheme.
4.4 Evaluation of method
The following section evaluates the quality of the data through looking at the validity and reliability of the methods I applied for the collection and analysis of data (Saunders et al., 2012).
4.4.1 Validity
The internal validity of my study relates to what extent my findings can be based on the causal conclusion and how bias was minimised (Saunders et al., 2012). In the case of this study, this reflects the extent to which I was able to access the interviewees’ experiences and perceptions.
The data collection is based upon the lenders’ personal experience and perceptions. Furthermore, the applied interview strategy tried to be as objective as possible. Therefore I believe that the findings accurately represent the views of the lenders, which adds to the credibility of my research. Although probing questions might have been suggestive, they were never posed immediately, but only after interviewees’ finished or completed their responses on a certain topic. Therefore I do not
consider the internal validity as a limitation of this research, but as very credible. In terms of the external validity this may not apply.
External validity is concerned with whether the findings are generalizable (Saunders et al., 2012). Hence, the extent to which my research findings are viable in other research environments. When referring to qualitative research, the intention is not to generalize a larger population, but to gather more in-depth information and explore an area without drawing general conclusions (Saunders, et al., 2012). This study supports this view, as the sample I used for data collection is small and not representative for the whole lender population. Consequently, if this research would be conducted with a larger number of subjects the findings may achieve additional support. In addition, adding quantitative research methods, such as surveys, would make it possible to increase both the internal and external validity of my research. Another indicator that would increase the external validity would be if this study would be conducted at a variety of lending-based platforms. Therefore my results and assumptions might not be representative or generalizable for the entire lending-based crowdfunding industry, but dominantly apply to the KOM platform.
4.4.2 Reliability
Reliability concerns the precision of applying the methods of data collection. If reliable data collection methods are used, repetition of the research will lead to similar outcomes in case the phenomenon that is measured stays unchanged (Boeije, 2014). The flexibility of a semi-structured interview is therefore appropriate to qualitative research, as it is adaptable to different situations. However, this results in the possibility of different outcomes upon repetition of the research. Therefore quantitative research is needed to confirm the reliability of this research.
5 Findings
This section presents the main findings that evolved during the coding of the data. Especially the phase of selective coding permitted the abstraction of the main categories that seem to influence the development of trust during the investment process of lenders. The variables are: investment determinants, experience and platform relationship.
5.1 Investment determinants
Analysis of the interview data has unveiled that cognitive determinants take the lead in the investment process. Sometimes affective determinants are followed directly through e.g. platform reputation, and sometimes the evaluation of affective determinants takes place at another stage of the investment process, e.g. prior to investment or through communication with the platform. Only in two cases the investment process started with affective determinants followed by cognitive determinant evaluation.
The cognitive determinants mostly comprised a combination of one or more of the following determinants: risk rating, interest, running time, project potential, background of the entrepreneur and state of the company. However, the development of affective determinants seemed to be more complex. First, lenders looked at the background presentation of the borrower (cognitive information) that gave them a particular feeling that enabled their decision-making. And second, interaction with the platform developed affective decision-making. For example lenders posed questions through a feedback form on the project web page or contacted the platform if clarification was needed. Sometimes affective determinants were also derived from the image the platform had developed through interacting experiences. Please refer to table 3 for the quote overview.
5.1.2 Information transparency
An influencing factor that goes beyond the cognitive and affective determinants concerns information transparency. As the parts of the decision-making process are dominantly based on information, the transparency influences the flow of information processing. The fact that cognitive determinants take a leading role in the actual investment decision-making is not strange. These antecedents are the information that is directly available for evaluating the potential of the investment. This information directly overcomes information asymmetry through the use of qualified information. The affective antecedents are mostly retrieved through indirect interaction, which is derived from the image the platform had created through feedback or prior experiences of lenders.
The transparent way platforms communicate and share information concerning their business policies and the project presentation is very important for lenders. As most lenders are distrustful traditional institutions such as banks, they require full transparency from the new financial intermediary, in this case the platform. The interview subjects dominantly argued that the transparency of information is the enabler of crowdfunding. According to the lenders the responsibility of participation, including risk, rests completely in their own hands. This is in contrast to traditional
financing methods, where the investor or client often makes an institution responsible for making bad decisions. However, in return for taking full responsibility for their investment risk, the lenders explained that they require clear and transparent information supply (Information reporting, table 3). Consequently, all information concerning business policies and project presentation is needed. Additional information updates, such as the presentation of new projects or news flashes including project proceedings also facilitate this transparent image. In case there are mistakes in the information reporting this will directly damage the trust levels of the lender and thus also change the lender’s platform image (Information reporting, table 3).
To conclude the first finding section, I have now gained an understanding of the cognitive and affective determinants that generate trust, and the fact that information transparency satisfies the lenders’ requirements for decision-making. However, this finding did not unveil sufficient information concerning relationships within the trust development in the investment process.
Understanding how the construction of cognitive and/or affective determinants develops helps to create a better understanding of the trust development during investing. Subsequently, I compared data to abstract more valuable information concerning possible relationships between variables. This resulted in the assumption that the experience of the lender influences the specific information needed in the investment process. The next paragraph will further elaborate on this.
Selective code
Category Factor Quote Investment
determinants
Cognitive determinants
Risk/Rating “Actually I look at the combination of risk, return and the background of the entrepreneur or company.” – Subject D Interest “I only look at the interest and the running time, because I
trust the analysis KOM performs.” – Subject H
Running time “Sometimes the running time of projects is shorter, e.g. 36 months. In such a case I accept less interest, however the running time should be well evaluated in relation to the total lended amount.” – Subject F
Project potential “I evaluate the project in such a way that it should show market potential according to the knowledge I have gathered from e.g. the newspapers I have red. (…) If a project has no chance of survival, then I will not take part. I do not invest based on my emotions.” – Subject I
“The project must be related to Internet. (…) Blokker, V&D and Miss Etam are examples of companies that were too late with switching to the online segment.” – Subject J Background
entrepreneur
“I carefully read the background story of the entrepreneur and look if his story inspires me.” –Subject J
State of the company (track record)
“I want to know what the purpose of the needed money and I want to know the state of the company, thus in which phase is the company currently situated.” – Subject D “I want to know if it is an existing company and not a start-up, for example a family company. In the green keeper industry for example, people are often performing in their branch for more than 40 years. These people know exactly what they are doing.” – Subject I
Affective determinants
Feeling information borrower
“If I tear down my investment process, I first refer to my gut feeling. If the feeling is good, then I pursue the process by further looking at more objective criteria.”- Subject E
Selective code
Category Factor Quote
Interaction platform: - Feedback form - Posing questions
“If there is a mistake in the provided information or if I have a question then I post the inquiry on the platform or I e-mail the platform manager. Depending on their answer I make my decision.” – Subject F
“If there is a technical aspect in the project information that I don’t understand, I’ll contact the platform. Mostly, I don’t ask questions related to the story of the
entrepreneur.” – Subject K Transparency Information
reporting
“The fact that KOM declines 75% of its requests is a very strong characteristic for a start-up. There are other platforms accepting every request that eventually will result in major misfortunes. KOM does that differently, which is a good thing.” – Subject H
“If there exist mistakes in the reporting of the project analysis of the platform this definitely damages my trust and lowers the chance of participating in the respective project.” - Subject E
Communication “What I think is an advantage with KOM is that the lenders receive an e-mail upon the launch of a new project, in contrast to e.g. Geldvoorelkaar who don’t participate in this service.” – Subject K
5.2 Experience
Comparing the lender investment criteria unveiled that the level of personal investment experience, crowdfunding experience and the experience with a platform seem to influence the investment process. The more experienced the lender, the less the needs for affective determinants during investment. Hence, the lender probably integrates the evaluation of affective determinants based on previous experiences. This assumption could also be interpreted from the opposite direction as this would imply that the less experienced a lender is, the more he looks for trust confirmation in affective determinants prior to investment. This finding is new when referring to the sensitizing concepts I presented in the conceptual model.
To compare the experience with the investment determinants I integrated all data in table 4. A distinction was made in ‘very experienced’, ‘experienced’ and ‘not experienced’ to classify the lender experience.
