The success of peer-to-peer (P2P) lending platforms has proven inconsistent and uneven over time and geography. This paper aims to strengthen our understanding of the market evolution through an analysis of risks on P2P lending platforms, which can be significantly affected by the way the platform design, regulatory structural building, nature of the transaction, and interdependencies between organizational components. We extend the social-technical model and create a systematic framework to map and analyze the financial risks from a hybrid financial and organizational approach. By implementing textual and statistical analysis on a dataset from Renrendai platform, we found that risks are generated not only from the stakeholders but also due to the weaknesses of interdependencies between organizational components and platform design. We also utilize our models to investigate why some P2P platforms such as LendingClub and Upstart (US), Renrendai (China), and Zopa (UK) have succeeded or failed from both finance and IS perspectives, and further propose potential risk-mitigation strategies for P2P lending platforms.
Hinweise
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1 Introduction
Over the course of the last fifteen years, digital technologies have been transforming the financial services industry – eventually triggering the evolution of the Fintech (Hendershott et al., 2017; Liu et al., 2015; Xu & Chau, 2018). “Fintech” refers to business models that use digital technologies to provide financial services (Clarke et al., 2020). Traditionally, banks have dominated the loan market; more recently, however, digital platforms have gradually been transforming the global lending industry, triggering the rise of peer-to-peer (P2P) lending platforms. Such platforms deliver many value propositions. They establish direct connections between lenders who hold funds and are seeking returns with borrowers who require funds, improving the efficiency of fundraising. They also apply online authentication and credit scoring systems, which substantially reduce transaction costs (Lin et al., 2013; Liu et al., 2015; Xu & Chau, 2018). Thus, P2P lending platforms make it easy for fundraisers to reach a large number of investors to raise funds for various purposes (Macchiavello, 2015), enabling individuals and small and medium enterprises (SMEs) to effortlessly access loans and investment (Duarte et al., 2012). Consequently, P2P lending platforms have partly replaced lending services that are traditionally provided by banks (Liu et al., 2015; Xu & Chau, 2018).
Despite the substantial value obtained, the success of P2P lending platforms has been inconsistent and uneven. For example, the Chinese market has recently been subject to a major crisis. In the first six months of 2018 alone, approximately 300 Chinese platforms went out of business, resulting in billions of losses affecting millions of users (Ngai, 2020). Although P2P lending platforms have employed novel approaches to mitigate risks (e.g., big data analytics), an increasing number of investors become aware of the risks associated with P2P lending products. To some extent, the uneven success of P2P lending platforms across the globe is a result of them operating across differing regulatory environments. As such, the platform providers seem to occasionally fail to effectively manage respective risks (Moeini & Rivard, 2019).
Anzeige
Given the great potential of P2P lending platforms by matching borrowers and lenders across the world, the extant literature mainly focuses on the value propositions of P2P lending platforms. There is scant attention paid to potential risks associated with P2P lending. To fill up this research gap, this study adopts an exploratory approach to investigate risks on P2P lending platforms and the antecedents leading to risks. In particular, we develop a systematic theoretical framework that captures risks on P2P lending platforms and proposes risk-mitigation strategies that P2P lending platforms could implement.
In Section 2, we introduce relevant literature on P2P lending platforms. In Section 3, we propose a theoretical framework to categorize risks arising from different types of P2P lending platforms, situated in various regulatory environments. Based on this framework, in Section 4, we provide testable propositions, and employ secondary data from Renrendai (a P2P lending platform in China) to empirically test the propositions. Furthermore, we identify possible strategies to mitigate these risks (Section 5). In the last section, we apply the insights from the theoretical framework on three specific platforms (i.e., LendingClub and Upstart in the US, Renrendai in China, and Zopa in the UK) to understand how risks affect the development of P2P lending platforms.
2 Literature Review
This section draws on the literature on information systems management, finance, and risk management to enhance our understanding of the basic structure of P2P lending platforms and the associated risks.
3 Stakeholders on P2P Lending Platforms
Previous research has identified three major stakeholders on P2P lending platforms, including the platform intermediary, lenders who hold funds and seek return, and borrowers who require funds. It is worth noting that additional stakeholders may also join the platform (see more information in the findings section).
Anzeige
The platform intermediary is the information illustrator that facilitates the transaction between borrowers and lenders through a variety of mechanisms, such as the pricing mechanism, trust-building mechanism, and the auction mechanism (Einav et al., 2016). The platform receives loan applications from borrowers, gathers investments from lenders, and issues loans to borrowers. Such mechanisms can yield manifold benefits to borrowers and lenders and ensure that lending transactions are successful. A recent study has found that a post-pricing mechanism generated enhanced temporal benefits to lenders and borrowers on the P2P lending platform Prosper (Wei & Lin, 2017). Similarly, another study revealed that the exposure of key and peripheral information positively affects lenders’ trust and that their trust directly relates to the loan performance (Xu & Chau, 2018).
Lenders are the second stakeholder group in the P2P lending process. They take over functions that have traditionally been performed by institutional investors or banks. Compared to institutional investors, however, individual lenders on the P2P lending platforms typically provide smaller investable funds and are characterized by reduced investment expertise (Liu et al., 2015). Their investment behavior may at times be irrational, as they may exhibit herding behavior (Wei & Lin, 2017), or may not have the ability to adequately analyze borrowers’ information to make sound investment decisions (Liu et al., 2019). Interestingly, previous research stresses that lenders on P2P platforms are rather heterogenous. For example, it has been shown that the level of wealth a lender possesses has a negative relationship with their level of risk aversion (Paravisini et al., 2011).
Previous research on the borrowers’ role in the P2P lending process has mainly been concerned with how their characteristics affect the ability to obtain funds or secure a reasonable interest rate from P2P lending platforms (Lin et al., 2013), and how platforms process borrowers’ information to match them with appropriate investors (Burtch et al., 2014). Borrowers’ demographical information, such as gender, race, age, and educational background, has been identified to be major factors determining their success in fundraising/credit scores (Ge et al., 2017).
Regarding the interrelationships between borrowers and platform intermediaries, some extant studies found that female borrowers tend to default less than their male counterparts, raising concerns about discriminative bias in the lending practices (Liu et al., 2019a, b). Another major factor in assessing a borrower’s creditworthiness is their lending history on the platforms (Ngai, 2020). It is common practice for P2P lending and crowdfunding platforms to use the information shared on using social media (Jin et al., 2020), such as Facebook and Twitter (in the US), to rate borrowers’ credit levels based on their information shared on their social networks in multiple scenarios (Burtch et al., 2014), as sharing borrowers’ information across different platforms allow the P2P lending platform to obtain borrowers’ credit worthiness level more accurate.
4 Risks on P2P Lending Platforms
While previous studies on digital platforms mainly focus on risks from financial perspectives (e.g., Moeini & Rivard, 2019; Lim et al., 2011), limited light has been shed on the risks that stem from the P2P platform designs and the regulatory environment in which the platform operates. In this section, we review potential risks related to P2P lending as discussed in the extant literature.
Default risk, which is also named credit risk, is considered to be one of the most prevalent risks in the traditional lending market, as well as in the P2P lending industry (Liu et al., 2019a, b). Due to the information asymmetry and the weakness of information exposure in online lending platforms, lenders cannot anticipate whether borrowers are able to repay money to lenders as agreed (Fu et al., 2021). As high-interest rates set by platforms are known to increase the inability to repay, the borrowing rate directly reflects the default risk: the higher the interest rate, the higher the default risk (Markowitz 1952). The relationship between lenders and borrowers can be characterized by information asymmetry (Liu et al., 2019a, b). Risks arise because lenders typically have limited restricted information about borrowers (Lin et al., 2013). Furthermore, the lending market’s governance of arrangements, systems, and controls may also affect the default risk on P2P lending platforms (Davis & Murphy, 2016). The risk arises when the lack of consistent regulations for credit risk assessment, protection of the rights of lenders and borrowers, and supervision of P2P platforms’ owner (Holmes, 2019).
Moreover, liquidity risk, both the traditional lending market and P2P lending transactions are subject to this type of risk in the platform. Platform providers may create fund pools to conduct private investments, resulting in an inability to repay investors (Tao et al., 2017). In our study, we focus on examining the liquidity risk at the individual loan level.
Although previous research has widely accepted that it is crucial for platforms to manage risks to maintain legitimacy and prevent financial losses, the risks related to P2P lending platforms have not been understood comprehensively. While acknowledging that transactions are subject to risks, there is a lack of a systematic understanding of risks (e.g., risk categories) on P2P lending platforms. In the following section, we take both financial and organizational perspectives to understand how risks are generated and influenced by the interdependencies between key stakeholders and relationships between those risks while providing possible risk mitigation strategies in different regulatory environments.
5 Theoretical Framework
In this section, we introduce a comprehensive theoretical model for Risk Categorization. We first identify 12 major risks and map them to the affected stakeholder groups. Secondly, we build upon Leavitt’s social-technical model to demonstrate how risks are generated on P2P lending platforms.
6 A Framework for Risk Categorization
Our framework is theoretically grounded in research on capturing risk classification and risk management literature in the financial, information systems, and technology and supply chain management industries (Moeini & Rivard, 2019; Lim et al., 2011). It was inspired by recent developments in the P2P platform industry. We extend the existing literature by providing a comprehensive theoretical framework that captures relevant risks on P2P lending platforms.
Our model identifies 12 risk categories that are present on P2P lending platforms. These are (1) platform regulatory risk, (2) platform operation risk, (3) loan investment concentration risk, (4) platform market entry risk, (5) platform information security risk, (6) online credit assessment risk, (7) platform users’ right risk, (8) loan information transparency risk, (9) loan liquidity risk, (10) loan default risk, (11) loan information asymmetry risk, and (12) platform loan interest rate risk. Table 1 provides a detailed definition of each risk. For example, regulation risk occurs when there is a lack of high-quality regulation from regulatory authorities and central banks governing the operational activities and capital flow of P2P lending platforms (Davis & Murphy, 2016). To this end, platform owners may utilize lenders’ capital to generate a private fund pool, conducting illegal financing by exposing false borrowing targets within the platform.
Table 1
Identification and description of risks in an organizational context
Risk types
Description
Financial Regulatory Authority—Borrower
Platform regulatory risk
David and Murphy (2016) stated that regulation risk occurs when there is a lack of high-quality regulation enacted by regulatory authorities and central banks to govern the operational activities and capital flow of P2P lending platforms (Davis & Murphy, 2016). We theorize this concept to specifically explain the risk generated by regulatory institutions of the Peer-to-Peer lending platforms, also the policy execution by P2P lending platforms.
Financial Regulatory Authority—Custodian Bank
Platform market entry risk
We theorize that platform market entry risk occurs when there are no entry requirements and standardized processes set by financial regulatory authorities when starting a P2P lending business. As a result, the platform owners have diminished capability in managing their businesses, ending with significant financial loss. In this case, the custodian bank takes an increased level of pressure to process the fund.
Financial Regulatory Authority—Lenders
Platform users’ right risk
Platform users’ right risk occurs when regulatory authorities have failed to enact policies that protect users’ rights. To be specific, borrowers and lenders can exit the platform anytime once they apply for.
Loan information transparency risk
This risk is when there is a lack of policies introduced by regulatory authorities that ensures all the historical transaction information is managed by and disclosed to both the regulatory party and users.
Users’ information security risk
Users’ information security risk occurs when there is a lack of rules imposed by regulatory authorities to guarantee that platform owners have strong protection regarding user personal information. If the risk management mechanism is not sufficiently consolidated, P2P lending platforms may suffer cybersecurity issues, such as data breaches and malware that are exploited by criminal gangs. Hackers can easily access borrowers' personal information, and fraudsters may utilize capital from one lender to pay another lender, resulting in a default on the loan repayment to the original lender.
Loan investment concentration risk
This risk occurs when lender’s investments are not diversified on the lending platforms, so the lender takes on a large volume of risk on a single loan if the borrower defaults. This incident happens frequently because of the lack of a diversification system on the platforms also investment knowledge provided by platform. It also occurs due to the existence of platform information asymmetry risk, as lenders overestimate the repayment capability of borrowers leading to an instance where the lender invests too much capital.
Borrowers—Lenders
Loan liquidity risk
This relates to the difficulty for the platform provider to repay lenders when the amount of loan exceeds the P2P lending platform’s solvency. Therefore, lenders are not able to withdraw cash and will not receive repayment. This risk is measured by using funding time and funding quantity of each loan.
Loan default risk
Loan default risk occurs when the borrower is not able to repay the lender with the guaranteed amount of loan on time once the loan is matured, meaning the borrower will break the contract (Du et al., 2020). This is due to the weakness of the credit scoring system to assess borrower’s creditworthiness.
Loan information asymmetry risk
Information asymmetry risk occurs when borrowers reveal insufficient personal information, so the lenders and the platform are unable to assess their credit scores accurately (Lin et al., 2013). This affects lenders’ investment decisions. Furthermore, if there is insufficient information revealed by borrowers and lenders, the platform will be unable to measure lenders’ risk-aversion levels and offer appropriate loans to lenders.
Platform operation risk
Platform operation risk originates from technical and management issues, such as the weaknesses of software operation and platform management team (Lim et al., 2011), and can be affected by the economic environment. In P2P lending platforms, it originates from the weakness of the online credit assessment system, the internal risk management mechanism, and in some cases the low quality of data processing. This can lead to platform information asymmetry between lenders and borrowers, making borrowers’ credit scores inaccurate.
Others
Loan interest rate risk
We theorize this concept in this paper as interest rate risk for online lending platform particularly. It occurs when regulatory authorities have no restrictions on the loan interest rate in P2P lending platform, then platform owners can set interest rate as higher as they expect to earn profit. Borrowers will have higher chance to default, and lenders might invest irrationally.
7 Mapping Risks to Major Stakeholders
What follows is an explanation of some of the identified risks and the stakeholders affected by them. We introduce these stakeholders and use the model (see Table 1) to explain how some of the risks illustrated in Table 1 can be mapped to the key stakeholders involved in the P2P lending process.
In addition to the three stakeholders that we mentioned in the literature review section, we also involve another two stakeholders namely the regulatory authority and custodian bank. These two stakeholders are playing a key role in maintaining sustainable internal and external structures in the P2P lending market. Financial regulatory authority refers to the financial institutions or central banks that enact policies and regulations that control and govern activities and capital flow of P2P lending platforms across different countries. Furthermore, custodian banks are financial institutions separate from the P2P lending platforms that manage customer funds to prevent them from being used by platform providers (Liu et al., 2019a, b).
On most P2P lending platforms, lenders are taking most of the risks. As being the fund supply side, lenders absorb the loan default risk created by borrowers and platform intermediaries, as these two parties may choose to leave the platform environment. They also absorb the loan liquidity risk generated by the platform if they are not able to withdraw cash once their loan has matured. Further, they are vulnerable to the loan investment concentration risk generated by platform intermediaries if the platform provider does not establish an effective internal mechanism to diversify a lender’s loans. In addition, they are susceptible to the loan information asymmetry risk if borrowers’ information is not adequately gathered and revealed, which may misdirect their investment decisions due to the inherent weakness of the credit scoring system. This can also increase the loan default risk. Platform operation risk harms the lenders directly when the platform is not able to accurately assess borrower’s credit levels due to the weakness of the internal risk assessment mechanism, and when they are offered a low-quality, unsuitable loan product due to the weakness of the software system and the internal governance system. They may also confront information security risks if the platform providers have failed to protect user data, leading to information leakage, in turn generating a financial loss.
8 Extended Social-Technical Risk Relationship Model
Building upon the risk categorization, we further analyze risks on P2P lending platforms from a view of constant variation in a social-technical circumstance. Specifically, we incorporate the factors of social-technical model because risks in P2P lending platforms are generated not only by the uncertainty of users and platform owners’ quality, but more importantly by the constant variation in Fintech P2P lending markets (e.g., from regulations, users’ characters, platform internal structures, and relationships between stakeholders). The way of analyzing risks from previous literature is mainly from a financial perspective that emphasizes independent factors, for instance, investigating how the inefficiency of risk management systems generates risks, which is no longer comprehensive enough. However, it is necessary to understand how these risks are generated from a macro perspective.
The original Leavitt’s open system social-technical model classifies the key dimensions of organizational change. It has been extensively utilized and extended in the information system literature. Therefore, we refer to the key factors of the social-technical model to analyze risks on P2P lending platforms in Fig. 1. Leavitt’s model includes four components which are the actor, task, technology, and structure (Leavitt, 1964). Actor refers to all the stakeholders such as users, managers, and designers. Structure refers to project organization and institutional arrangements. Technology originally refers to development tools, methods, internal technical mechanisms, and platforms. Task originally refers to what goals that software developer aims to achieve. The change of one component will have an impact on others, which is internal disequilibrium (Lyytinen et al., 1998; Leavitt, 1964). The disequilibrium between the four components can generate risks on P2P lending platforms.
Fig. 1
Relationships between risk categories
×
8.1 Technology-Actor interdependencies
In P2P lending platforms, risks occur due to the weaknesses of Technology-Actor interdependencies instead of the weaknesses of Technology only, because the involvement of Actor determines the characters of the risk taker. The risks are generated by improper matching of people with technology, or by introducing untried technologies that cannot meet the demands of each Actor. In P2P lending platforms, the occurrence of loan investment concentration is because the internal loan recommendation systems are unable to promote matchable loans to lenders according to lenders’ capability of taking risks. Besides, the interdependencies between platform software engineers, the platform owner, and fund management systems generate the platform operation risk. Therefore, we include that users’ information security risk, loan credit assessment risk, loan investment concentration risk and platform operation risk are generated by the interdependencies between these two factors.
From the organizational perspective, platform owners should regularly set up-to-date goals for software engineers to keep optimizing risk management systems to better allocate funds in the platform for the users, as internal credit assessment standards for borrowers should be modified constantly. The failure of keeping the risk management systems optimized decreases the liquidity of the platform in the sense that lenders who have sufficient money cannot be matched with borrowers who have higher creditworthiness.
8.2 Structure-Actor Interdependencies
Structure-Actor interdependencies originally explain the relationship between actors and the structure of organizations. In this study, we assume that the mismatch between Actor and Structure in P2P lending generates major risks. We summarize the loan interest rate risk, loan information asymmetry risk, and platform users’ right risk and loan are generated due to the weakness of Structure-Actor interdependencies. Particularly, how efficiently and effectively the regulatory authority constructs the supervision systems and enacts efficient policies in the P2P lending market determines whether users’ rights are being protected and the quality of support from custodian banks to manage users’ funds in each platform. Further, loan liquidity risk is generated because of the lack of enactment of regulatory policies and an efficient supervision check that requires platforms to prepare sufficient risk reverse funds for lenders. These are playing a key role in making P2P lending platforms operate smoothly in the long run.
8.3 Structure-Task Interdependencies
Structure-Task interdependencies illustrate the correspondence between tasks and structure in organizations, particularly in P2P lending platforms. In this study, we assume that both internal and external structures of P2P lending platforms affect the success of the platform. Regarding the internal structure of the P2P lending platform, the way of loan information is exposed to users, and fund management systems involving platform owners and the custodian bank affect the overall revenue and security level of P2P lending platforms. For the external structure, whether the financial regulation authorities implement efficient supervision systems and establish strict P2P lending market entry threshold for platform owner in the market also influence the revenue and security level of each P2P lending platform. From an organizational perspective, we conclude that loan information transparency risk, platform market entry risk, and platform regulatory risk are generated due to the weakness of interdependencies between these two factors.
8.4 Task-Actor Interdependencies
Task-Actor interdependencies originally indicate the association between tasks and actors in organizations. We assume that this concept is a correspondence level between risk management mechanisms and the annual revenue and security level of P2P lending platforms. As most of the extant research has shed light on the usage of technologies affecting the annual revenue and security level in the P2P lending platform, which may not comprehensively explain the reason why some P2P lending platform succussed at a macro level.
8.5 Risk Relationships
All 12 risks are strongly interrelated. To this end, Fig. 1 illustrates the relationships between different types of risks that exist across P2P lending platforms. We find the existence of both positive and negative relationships between the risks inherent across the platforms.
Platform regulatory risk occurs when regulatory institutions are unable to enact suitable and applicable policies to control and support the development of platforms. Platform regulatory risk also generates loan default risk when there are no specified entry requirements for P2P lending platform owners to enter the market, there will be a lack of funds available in the platform due to the low capability of collecting and managing funds. This risky situation will increase the occurrence of loan liquidity risk and lead to financial loss to lenders. Platform operation risk generates loan liquidity risk when the operating system does not manage funds efficiently before and after the loan is funded. In addition, loan information transparency risk occurs when there is a lack of policies to constrain platform owners to keep the internal operational process transparent to regulatory institutions, thus, platform users’ rights will not be guaranteed. For instance, lenders will not be guaranteed to withdraw money when the loan is matured. Additionally, the users’ information security risk occurs since the platform's internal design is not robust enough to protect users’ personal information so hackers might easily access and steal users’ information, which generates a financial loss for platform users.
Furthermore, loan interest rate risk occurs when platform regulatory risk is high because there are no certain policies to limit the maximum interest rate so platforms might set it much higher than it is supposed to be, and borrowers might not be able to repay lenders. Therefore, loan interest rate risk is positively associated with loan default risk in the platform, which generates financial loss to lenders.
The first catalog in Fig. 1 is related to the platform service design, which generates loan credit assessment risk, users’ information security risk, loan investment concentration risk, and platform operation risk. Firstly, if a P2P lending platform has not built a robust credit assessment system by applying big data techniques, therefore the credit score they generate will not be accurate. In this situation, borrowers' credit level might be set higher than it has, and loan default risk will be high.
The quality of supervision infrastructure systems determines the generation of loan information asymmetry risk, wherein lenders may lack sufficient information about borrowers if the platform design team fails to implement an intelligent system for supervising and gathering borrowers’ information. This situation also indicates that a high investment concentration risk significantly elevates the default risk for lenders if borrowers are unable to fulfill their loan repayments, and limited investment knowledge is provided by the platform. In such circumstances, lenders may typically overestimate the loan quality and allocate a large portion of their funds to it, thereby, increasing the loan default risk. We further conclude that there is a positive correlation between loan credit assessment risk and loan default risk, as the inaccuracy in credit evaluation can obstruct the mutual disclosure of information between borrowers and lenders.
9 Propositions and Justification
In a platform, achieving annual plans is not only according to the task assigned but also on how to coordinate and adapt with the market structure. The relationship between Structure and Task, from an organizational perspective, is a determining factor in efficiently sustainably achieving and resolving a company’s goals or issues in accordance with a long-term pattern. There are some approaches to strengthening the relationship between these two factors. For instance, optimizing risk management mechanisms to provide risk premiums with lenders and supervising the procedure of scrutinization borrowers ‘ creditworthiness, and creating services to encourage borrowers to repay on time (Duarte et al., 2012) to maintain the credit quality of borrowers who join the P2P lending platforms. However, the most effective element of consolidating the relationship is the enactment of high-quality regulations.
Compared to the traditional lending market where financial regulations are developed maturely, the P2P lending market is still in an under-development regulatory environment. Under these circumstances, the exactness and rationality of regulations implemented by the financial regulatory authorities for platform owners determine how the strength of relationships between Structure and Task, which in turn, help platforms address potential problems easily, and maintain the relationship between Actor and Task, in terms of timely loan repayment by borrowers. We posit that the absence of policies and regulations from financial authorities weakens the interdependencies between Structure and Task, which are positively associated with the interdependencies between Task and Actor. As a result, this may lead to an increase in loan default risk and a higher likelihood of user defaults on the platforms. Therefore, we propose:
Proposition 1: The platform regulatory risk increases the loan default risk in the P2P lending platforms.
On an organizational level, accomplishing Task within a platform not only following and implementing regulations and policies enacted by financial regulatory authorities but also ensures information transparency. In P2P lending platforms, historical transactions of loans indicate borrowers’ creditworthiness and the quality of loans, as lenders can observe the speed of funding and whether each of the borrower’s loans was successfully funded. However, due to the operational configurations of most P2P lending platforms and strategies borrowers employ for information disclosure, lenders often lack access to authentic and complete loan information. This gives rise to loan information transparency risk, which in turn undermines the relationship between Structure and Task.
Additionally, the loan information transparency risk leads to a situation where lenders, particularly those lacking adequate financial literacy on investment are less likely to diversify their portfolios. They may invest larger sums into a loan if the platform does not fully disclose negative information about the loan. This scenario heightens the loan investment concentration risk and weakens the relationship between Technology and Actor, and making it less probable for users to use platform services to mitigate risk by themselves. Therefore, we assume that the absence of complete information disclosure for each loan, as required by financial regulatory authorities, weakens the interdependencies between Structure and Task, which directly affect the interdependencies between Technology and Actor. Therefore, we propose:
Proposition 2: The loan information transparency risk increases the loan investment concentration risk in P2P lending platforms.
In P2P lending platforms, credit assessment systems leverage machine learning techniques to evaluate the creditworthiness of borrowers based on their individual economic status and produce a credit score for each borrower. However, some P2P lending platforms only evaluate borrowers’ hard information through an online survey (e.g., income status, educational background, and mortgage) but neglect borrowers’ soft information such as textual comments, motivation for applying for loans, and previous trading behavior in the platform. Prior studies suggest that soft information can sometimes better provide more insightful understanding of an individual’s character, similarly, which can help accurately identify borrowers’ creditworthiness in the P2P platforms (Agarwal & Hauswald, 2010). Hence, the credit scores generated by the platform sometimes can result in biased and failure to represent borrowers’ creditworthiness level accurately, which increases information asymmetry risk in the platforms between borrowers and lenders.
From an organizational standpoint, the relationship between Structure and Actor indicates how the external and internal structure of the P2P lending market cater to stakeholders as the Actor. The absence of supervisory systems provided by regulatory authorities to facilitate platforms in setting specific requirements for borrowers to disclose their credit information before joining the platforms can weaken the interdependencies between Structure and Actor. This reduction can lead to lenders making irrational investment decisions when the credit score is overvalued by the credit assessment systems, generating loan information asymmetry risks. This also weakens the relationship between Technology and Actor, as lenders are less likely to use platform services to diversify their investment. In such a situation, P2P lending platforms need to devise an investment diversification system to help users to diversify their loans. Therefore, we propose:
Proposition 3: The loan information asymmetry risk increases the loan investment concentration risk inP2P lending platforms.
The lack of precise and systematic regulations enacted by financial regulatory institutions on maintaining the entry standards for P2P lending platform owners results in the low quality of the platform’s fund operational services. For instance, the overall capital flow in the platform cannot operate smoothly when the owners create a private pool fund and use the pooled funds to conduct external investments and are not managed to return the fund invested by lenders back to the platform on time due to the ongoing external investments. In this situation, there are insufficient funds in the platform, and borrowers cannot receive the loans from the platform when the loans have been successfully funded. Additionally, lenders cannot withdraw their principal and interests on time when loan borrowers cannot receive the loans from the platform as agreed upon when the repayment is due. This generates the platform operation risk and weakens the relationship between Technology and Actor, indicating that fund operational systems are not efficient in securing users’ funds. This decreases the overall liquidity of the P2P lending platform and leads to the occurrence of loan liquidity risk.
At the organizational level, the weakened relationship between the efficiency and effectiveness of the operational system of the platform weakens the interdependencies between the factors of Technology and Actor, directly affecting the interdependencies between Structure and Actor. This indicates that, in P2P lending platforms, users' profit cannot be supported and guaranteed due to the ineffectiveness of the internal structure of the platforms. Therefore, we propose:
Proposition 4: The platform market entry risk increases the loan liquidity risk in P2P lending platforms.
10 Empirical Validation of Propositions
10.1 Data and Research Context
To empirically test the propositions, we collected a large dataset from one of the popular Chinese P2P lending platforms (i.e., Renrendai). This dataset spans from October 2010 to October 2016 and includes 7,519,488 trading transaction observations generated between lenders and borrowers. In particular, there are 102,736 loans recorded. It includes detailed transactional information such as borrowers' personal information and investors' purchasing information.
Additionally, there are several important events and services launched for P2P lending platforms, and they can affect the risks level in the platforms, therefore, we include these events and launched policies as important variables in our study. Regarding the regulations enacted for P2P lending platforms in Chinese market, China Banking Regulatory Commission announced that there are four key requirements for P2P lending platforms that need to be met on 21st April 2014. First, they made clarification of the mediating feature of the platform. Second, they stated that the platform itself shall not require lenders to pay collaterals. Third, they stated that the collected fund shall not be put into the pool fund created by the platform owner. Fourth, the illegal absorption of public funds shall not be allowed. In the same year, they also clarified the market entry requirements for P2P lending platform owners on 22nd August 2014.
Regarding platform services launched in Renrendai, the 'Preferred Financial Planning' launched on 7th December 2012, which is a smart bidding tool that helps users take precedence over manual bidding for loans. For instance, if the user chooses a certain time range of bidding, such as six months, the system will automatically help them to bid after the selection. In addition, Renrendai also launched a mobile APP for their users on 29th June 2014. In this App, users can check the company's dynamic information and manage their own account status, and invest in the scattered bid quickly. This App eliminates the geographical restrictions of traditional finance and makes the transaction more convenient. Furthermore, in terms of users’ account security, Renrendai’s mobile APP sets the automatic exit of an account after 24 h of logging in, which effectively protects users' privacy and fund security. After two years, Renrendai officially launched fund depository cooperation with China Minsheng Bank On 29th February, 2016. This partnership helps manage all users' funds flow in a comprehensive way, which avoids the occurrence of the illegal capital pool created by platform owners.
The real borrowers’ credit score is calculated by the platform based on the personal information they exposed on the platform. For instance, the income level, educational background, age, whether they own a car or house, and description of applying for loans. This information will be processed by the platform immediately once borrowers fill out the survey online. However, this type of hard information may not be enough to generate borrowers’ credit scores as it may not clarify the intention of borrowing the amount of money. For instance, the borrowers can borrow money from the current platform to repay another loan in another platform which is extremely risky for the current platform. In this case, to better measure the risk level of borrowers on Renrendai, we utilized text analysis techniques to evaluate the textual description of each loan that describes the reason for applying for the loan as borrowers’ soft information.
10.2 Methods
The three main risks of interest are loan default risks, loan concentration risk and loan liquidity risk. First, to measure loan default risk \({{\varvec{y}}}_{{\varvec{L}}{\varvec{o}}{\varvec{a}}{\varvec{n}}{\varvec{D}}{\varvec{e}}{\varvec{f}}{\varvec{a}}{\varvec{u}}{\varvec{l}}{\varvec{t}} {\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\), we denote it as the default risk coefficient score. To calculate this score, we first selected five variables from the dataset that are highly likely to affect the possibility of default on each loan, which include the percentage of loans that defaulted among all the loans, car loan, education level, income level, mortgage of each borrower. We included these variables in a weighted average equation with given the same percentage of weights to each variable, and then we calculated the reciprocal value to generate a continuous variable as loan default risk of each loan.
Furthermore, secondly, to measure loan liquidity risk \({{\varvec{y}}}_{{\varvec{L}}{\varvec{o}}{\varvec{a}}{\varvec{n}}{\varvec{L}}{\varvec{i}}{\varvec{q}}{\varvec{u}}{\varvec{i}}{\varvec{d}}{\varvec{i}}{\varvec{t}}{\varvec{y}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\), we initially measured the funding time by subtracting the time of loan issuance from the time to apply for loans. We selected the amount of the loan as the funding quantity. Next, we created a weighted average equation (Athey & Imbens, 2017) to generate a continuous variable as the loan liquidity risk by giving fifty percent of the weight to each of these variables and adding them together and then calculating the reciprocal value. Third, to measure loan investment concentration risk \({{\varvec{y}}}_{{\varvec{I}}{\varvec{n}}{\varvec{v}}{\varvec{e}}{\varvec{s}}{\varvec{t}}{\varvec{m}}{\varvec{e}}{\varvec{n}}{\varvec{t}}{\varvec{C}}{\varvec{o}}{\varvec{n}}{\varvec{c}}{\varvec{e}}{\varvec{n}}{\varvec{t}}{\varvec{r}}{\varvec{a}}{\varvec{t}}{\varvec{i}}{\varvec{o}}{\varvec{n}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\), we examine how diverse the investors manage their loan purchases by calculating the number of loans lenders purchase, and the amount of investment in each loan divided by the total amount of money they invested, then we calculated its reciprocal value that refers to loan investment concentration risk score.
Apart from the dependent variables, we also measured other risks and policy enactments as independent and control variables. The independent variables include platform regulatory risks, information transparency risk, and platform market entry risk. First, we measure platform regulatory risk \({{\varvec{\chi}}}_{{\varvec{P}}{\varvec{l}}{\varvec{a}}{\varvec{t}}{\varvec{f}}{\varvec{o}}{\varvec{r}}{\varvec{m}}{\varvec{R}}{\varvec{e}}{\varvec{g}}{\varvec{u}}{\varvec{l}}{\varvec{a}}{\varvec{t}}{\varvec{o}}{\varvec{r}}{\varvec{y}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\) by creating a weighted average equation to generate a continuous variable. This equation includes the dummy variable indicating the time when regulations were enacted by the China Banking Regulatory Commission to clarify the feature of the P2P lending platform being intermedia. Second, since we have limited information about the transparency of the loan information of Renrendai, we measured the information transparency risk \({{\varvec{\chi}}}_{{\varvec{I}}{\varvec{n}}{\varvec{f}}{\varvec{o}}{\varvec{T}}{\varvec{r}}{\varvec{a}}{\varvec{n}}{\varvec{s}}{\varvec{p}}{\varvec{a}}{\varvec{r}}{\varvec{e}}{\varvec{n}}{\varvec{c}}{\varvec{y}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\) using a dummy variable. If the dummy variable equals ‘0’, meaning the regulation had been enacted by China Banking Regulatory Commission to require P2P lending platforms to keep loan information transparent, therefore the risk level is low. If the dummy variable equals ‘1’, meaning the regulation had been enacted and the risk level is high.
The information asymmetry risk \({{\varvec{\chi}}}_{{\varvec{I}}{\varvec{n}}{\varvec{f}}{\varvec{o}}{\varvec{A}}{\varvec{s}}{\varvec{y}}{\varvec{m}}{\varvec{m}}{\varvec{e}}{\varvec{t}}{\varvec{r}}{\varvec{y}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\) is measured by the calculating the difference between the credit score generated by Renrendai and the textual information of loan title left by borrowers indicating the reasons for applying for the loan. We defined that the more the words and vocabulary related to ‘investment’, and ‘repaying for a second loan’ in Chinese shown in the loan title, the risk level of the borrower is high. We generated a new score according to the text analysis results, and we calculate the difference between the credit score and this new score to indicate information asymmetry risk. In addition, there are three control variables that we consider might affect the dependent variables, which are all dummy variables. These control variables include whether the Renrendai’s mobile App has been launched \({{\varvec{\chi}}}_{\boldsymbol{ }{\varvec{L}}{\varvec{a}}{\varvec{u}}{\varvec{n}}{\varvec{c}}{\varvec{h}}{\varvec{o}}{\varvec{f}}{\varvec{M}}{\varvec{o}}{\varvec{b}}{\varvec{i}}{\varvec{l}}{\varvec{e}}{\varvec{A}}{\varvec{p}}{\varvec{p}}}\), whether the bidding service has been launched \({{\varvec{\chi}}}_{{\varvec{L}}{\varvec{a}}{\varvec{u}}{\varvec{n}}{\varvec{c}}{\varvec{h}}{\varvec{B}}{\varvec{i}}{\varvec{d}}{\varvec{d}}{\varvec{i}}{\varvec{n}}{\varvec{g}}{\varvec{T}}{\varvec{o}}{\varvec{o}}{\varvec{l}}}\), and whether the platform has been collaborating with the custodian bank \({{\varvec{\chi}}}_{{\varvec{C}}{\varvec{o}}{\varvec{l}}{\varvec{l}}{\varvec{a}}{\varvec{b}}{\varvec{w}}{\varvec{i}}{\varvec{t}}{\varvec{h}}{\varvec{C}}{\varvec{u}}{\varvec{s}}{\varvec{t}}{\varvec{o}}{\varvec{d}}{\varvec{i}}{\varvec{a}}{\varvec{n}}{\varvec{B}}{\varvec{a}}{\varvec{n}}{\varvec{k}}}\) (MinSheng Bank). If the dummy variable equals ‘0’, it indicates that the policy or service has not been launched. Conversely, it the dummy variable equals ‘1’, it indicates that the policy or event has been launched or occurred.
where i indexed each individual loan applied by the borrower, \({{\varvec{y}}}_{\varvec{L}\varvec{o}\varvec{a}\varvec{n}{\varvec{D}}{\varvec{e}}{\varvec{f}}{\varvec{a}}{\varvec{u}}{\varvec{l}}{\varvec{t}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\) denotes the dependent variable of loan default risk. \({{\varvec{\chi}}}_{\varvec{P}\varvec{l}\varvec{a}\varvec{t}\varvec{f}\varvec{o}\varvec{r}\varvec{m}{\varvec{R}}{\varvec{e}}{\varvec{g}}{\varvec{u}}{\varvec{l}}{\varvec{a}}{\varvec{t}}{\varvec{o}}{\varvec{r}}{\varvec{y}}\varvec{R}\varvec{i}\varvec{s}\varvec{k}}\) represents the platform regulatory risk as independent variable. \({{\varvec{\chi}}}_{ {\varvec{i}}}\) represents the control variables which are all dummy variables that may affect the dependent variable. These include the launch of the mobile App in the platform, investment assistant services in the platform, and collaboration with custodian bank MinSheng Bank. If the dummy variable equals “1” indicating the policy was enacted or the service was launched, and the value is “0” indicating the policy was not enacted or the service was not launched. Our primary interest was β1 which captured the impact of platform regulatory risk on loan default risk on Renrendai.
where i indexed each individual loan applied by the borrowers, t indexes the loan requesting time, and \({{\varvec{y}}}_{{\varvec{I}}{\varvec{n}}{\varvec{v}}{\varvec{e}}{\varvec{s}}{\varvec{t}}{\varvec{m}}{\varvec{e}}{\varvec{n}}{\varvec{t}}{\varvec{C}}{\varvec{o}}{\varvec{n}}{\varvec{c}}{\varvec{e}}{\varvec{n}}{\varvec{t}}{\varvec{r}}{\varvec{a}}{\varvec{t}}{\varvec{i}}{\varvec{o}}{\varvec{n}}\varvec{R}\varvec{i}\varvec{s}\varvec{k}}\) demotes the loan investment concentration risk as the dependent variable. \({{\varvec{\chi}}}_{{\varvec{I}}{\varvec{n}}{\varvec{f}}{\varvec{o}}{\varvec{T}}{\varvec{r}}{\varvec{a}}{\varvec{n}}{\varvec{s}}{\varvec{p}}{\varvec{a}}{\varvec{r}}{\varvec{e}}{\varvec{n}}{\varvec{c}}{\varvec{y}}\varvec{R}\varvec{i}\varvec{s}\varvec{k}}\) represents the loan information transparency risk as the independent variable. \({{\varvec{\chi}}}_{ {\varvec{i}}{\varvec{t}}}\) represents the control variables that may affect the dependent variable, which includes whether mobile App in the platform has been launched, whether the investment assistant bidding tool has been launched in the platform, and whether the platform has been collaborating with a custodian bank. Our primary interest was β1, which captured the impact of loan information transparency risk on loan concentration risk on Renrendai.
where i denotes each individual loan applied by the borrowers, t indexes the loan requesting time, and \({{\varvec{y}}}_{{\varvec{I}}{\varvec{n}}{\varvec{v}}{\varvec{e}}{\varvec{s}}{\varvec{t}}{\varvec{m}}{\varvec{e}}{\varvec{n}}{\varvec{t}}{\varvec{C}}{\varvec{o}}{\varvec{n}}{\varvec{c}}{\varvec{e}}{\varvec{n}}{\varvec{t}}{\varvec{r}}{\varvec{a}}{\varvec{t}}{\varvec{i}}{\varvec{o}}{\varvec{n}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\) represents the investment concentration risk of each loan as the dependent variable. \({{\varvec{\chi}}}_{{\varvec{I}}{\varvec{n}}{\varvec{f}}{\varvec{o}}{\varvec{A}}{\varvec{s}}{\varvec{y}}{\varvec{m}}{\varvec{m}}{\varvec{e}}{\varvec{t}}{\varvec{r}}{\varvec{y}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\) represents the loan information asymmetry risk as the independent variable. \({X}_{it}\) is a vector of control variables that may affect the dependent variable, including the launch of the mobile App, the launch of the bidding tool in the platform, and the collaboration with the custodian bank. Our primary interest was β1, which captured the impact of loan information asymmetry risk on investment concentration risk on Renrendai.
where i denotes each loan applied by the borrowers, t indexes the loan requesting time, and \({{\varvec{y}}}_{{\varvec{l}}{\varvec{o}}{\varvec{a}}{\varvec{n}}{\varvec{L}}{\varvec{i}}{\varvec{q}}{\varvec{u}}{\varvec{i}}{\varvec{d}}{\varvec{i}}{\varvec{t}}{\varvec{y}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\) represents the loan liquidity risk of each loan i as the dependent variable. \({{\varvec{\chi}}}_{{\varvec{P}}{\varvec{l}}{\varvec{a}}{\varvec{t}}{\varvec{f}}{\varvec{o}}{\varvec{r}}{\varvec{m}}{\varvec{M}}{\varvec{a}}{\varvec{r}}{\varvec{k}}{\varvec{e}}{\varvec{t}}{\varvec{E}}{\varvec{n}}{\varvec{t}}{\varvec{r}}{\varvec{y}}{\varvec{R}}{\varvec{i}}{\varvec{s}}{\varvec{k}}}\) represents the platform market entry risk as independent variable. \({X}_{it}\) is a vector of control variables that may affect the dependent variable, including the launch of the mobile App, and the collaboration with custodian bank. Our primary interest was β1, which captured the impact of platform market entry risk on loan liquidity risk on Renrendai.
10.3 Results
The results are reported in Table 2. In Model (1), the platform regulatory risk is positively related to loan default risk (\({\beta }_{1}\)=0.002, p < 0.01), which reveals that the lack of precise regulation enacted by the financial regulatory institution to clarify the feature of the P2P platform in the market may attract more borrowers with lower creditworthiness to join the platform which causes defaults in the platform. This result highlights the importance of the element of Structure in the social-technical model.
Table 2
Results of main analyses
Model (1)
Model (2)
Model (3)
Model (4)
IVs:
y = loan default risk
y = Investment concentration risk
y = Investment concentration risk
y = Loan liquidity risk
\({\chi }_{PlatformRegulationRisk}\)
0.002** (0.080)
\({\chi }_{InfoTransparencyRisk}\)
0.038** (0.000)
\({\chi }_{PlatformMarketEntryRisk}\)
0.002** (0.000)
\({\chi }_{InfoAsymmetryRisk}\)
0.002** (0.000)
Controls:
\({\chi }_{Launch MobileApp}\)
-0.001** (0.048)
0.016** (0.048)
0.017** (0.000)
-0.002** (0.000)
\({\chi }_{LaunchBiddingTool}\)
-0.0248** (0.000)
-0.019** (0.100)
-0.019** (0.000)
\({\chi }_{CollabWithCustodianBank}\)
-0.002** (0.000)
-0.001** (0.000)
-0.004** (0.000)
-0.001** (0.000)
Cons
0.088** (0.000)
0.042** (0.000)
0.042** (0.000)
0.013** (0.000)
Obs
102,736
7,519,488
7,519,488
102,736
R-squared
0.0758
0.013
0.013
0.0758
Standard errors are reported in the parentheses
**p < 0.01; * p < 0.05; † p < 0.10
In Model (2), information transparency risk has a positive relationship with investment concentration risk (\({\beta }_{1}\)=0.038, p < 0.01). This result reveals if the loan information is disclosed on the platform transparently will make investors invest rationally and be more likely to diversify their investment portfolio, which highlights the importance of the element of Technology in the social-technical model.
In Model (3), we found that loan information asymmetry risk significantly increases the loan investment concentration risk (\({\beta }_{1}\)=0.002, p < 0.01). This result illustrates whether the credit score of each borrower generated by platforms to indicate their creditworthiness authentically can affect lenders’ investment decisions. This result also illustrates the relationship between Actors and Technology, in particular, how platforms use technologies to display the authentic characters of stakeholders in the platform.
Interestingly, both in Model (2) and Model (3), we found that the launch of the mobile app, considered as a control variable, is positively associated with loan investment concentration risk. This indicates that using a mobile app to purchase loans decrease the investment diversification of lenders. Consequently, this form of investment in the mobile App can even make platforms riskier. This result may help clarify the cause of failure in the Chinese P2P lending market, as the majority of platforms launched their mobile app to encourage users to invest in their mobile phones. This finding further highlights the interdependencies between Technology and Task.
In Model (4), platform market entry risk is positively associated with loan liquidity risk (\({\beta }_{1}\)=0.002, p < 0.01), which indicates that whether the financial regulatory institution establishes an entry standard for platform owners to enter the P2P lending market can be helpful to maintain the loans’ liquidity in the platform. This result highlights the relationship between the Structure and Task in the social-technical model.
To support the results obtained from empirical analysis, we also collected all the news and policies which can affect the state of operation in Renrendai. The policies enacted by the financial authority in China in 2014 specifically include market entry standards for platform owners, the clarification of the nature of P2P lending platforms, and the maintenance of the transparency of the loan information. We found that regulations enacted by financial authorities establish the foundation of P2P lending platforms’ sustainable operations in China. However, Chinese financial authorities required P2P lending platforms to exit the industry from 2019 to 2021 due to the appearance of fraudulent platforms, and public anger towards the financial loss in P2P lending platforms. This devastating situation occurs because the financial authorities in China cannot help establish a sustainable market that not only restricts platform owners' activities from conducting illegal financial activities but also maintain a reasonable operational approach in the long run.
To further evaluate the effectiveness of the regulations, we found that the establishment of a market entry standard for platform owners helps improve the loan liquidity in the platform, indicating that the lenders have more confidence in the platform that brings them good returns. This result clarifies the cause of the failure of the Chinese P2P lending market, in the sense that setting market entry standards guarantees the quality of operational services in the platforms and builds a better reputation. These will encourage investors in the market to invest more amount of money, which improves loans’ liquidity. The findings further support our theory that the interdependencies between Structure and Actors determine whether the platforms can succeed or not in the long run.
However, it still cannot reduce the occurrence of defaults and illegal financial pool funds in the platforms. According to our risk framework, regulation enactments are not sufficient for developing a P2P lending market, and a more comprehensive supervisory system that helps reinforce the relationship between Actor, Structure, Technologies, and Task is needed.
11 Risk Mitigation Strategies
Platforms and regulators have the means to apply strategies to mitigate risks in P2P lending. Based on the above three models, we identified risk, and propose potential risk mitigation strategies in the following, for each type of interdependencies between the social-technical factors, (see Table 3). Such strategies can be clustered into three subgroups/levels as follows, (1) Structure-Actor interdependencies: platform internal design, (2) Structure-Task interdependencies: internal structures of the platform and external regulatory structures (3) Technology-Actor interdependencies: the nature of transactions.
Table 3
Social-technical factors and risk mitigation strategies
Level of Risk
Strategies
Risk Types
Structure-Actor: Stakeholders and regulation authorities
Optimize platform trading page design by exposing credit information about borrowers (eg. social media and third-party payment systems)
Loan information asymmetry risk
Apply deep learning techniques to optimize credit scoring systems
Loan information asymmetry risk
Enact regulations that requires platform intermediaries to expose up-to-date trading information of borrowers and lenders on the web page
Loan information transparency risk
Establish supervision systems by regulatory institutions to ensure lenders and borrowers receive funds as agreed
Platform users’ right risk
Building solid collaboration with custodian banks and SaaS systems
Loan liquidity risk
Establish multiple scenarios to meet users’ funding requirements such as (car loans; education loans) to remain users
Loan liquidity risk
Structure-Task interdependencies: Internal structures and external regulatory structures
Improving the effectiveness of policies by establishing supervision systems that observing each platforms’ implementation on regulations
Platform regulatory risk
Standardize market entry thresholds by conducting due diligence on new P2P lending platforms to keep the entry thresholds optimized for new platform owners
Platform market entry risk
Set interest rates based on borrowers’ credit level and prepare sufficient provision funds for lenders
Loan default risk
Regulatory institutions set the range of interest rates for platforms to control the maximum of interest rates
Loan interest rate risk
Technology-Actor interdependencies: Platform Service Design
Optimizing loan recommendation systems by using deep learning techniques to provide high quality loans to lenders who have sufficient funds
Loan default risk
Create automatic loan diversification services to lenders and provide financial investment knowledge to lenders on the platforms’ web page
In this study, Structure-Actor interdependencies illustrate how platforms should be designed to maximumly satisfy stakeholders’ profit and supervise user behavior and transactions. As platform internal design plays a key role in mitigating risk throughout the P2P lending platforms when Structure-Actor interdependencies are supposed to be robust. An effective partnership with a custodian bank, which would manage all the funds, is an efficient way to prevent platform providers from engaging in behaviors that would see them taking capital and escaping to reduce loan liquidity risk. Next, it is necessary for platform providers to maintain appropriate interest rates, enabling them to reduce the loan interest rate risk. If the interest rates are disproportionately high to attract lenders, the borrower may not be able to afford the interest, generating a substantial loss for lenders. Platform owners should prepare sufficient provision funds as compensation to lenders if default action occurs, mitigating liquidity risk. Furthermore, regulatory institutions may enact policies that fundamentally require platform owners to design internal structures to expose sufficient information for both lenders and borrowers to reduce loan information asymmetry risk.
11.2 Structure-Task Interdependencies: Internal Structure and External Regulatory Structures
Structure-Task interdependencies demonstrate how the internal structure within the platform and external regulatory system interact to maintain a safe, profitable, sustainable online lending market in different countries. Firstly, they may aim to maintain high levels of information transparency regarding the information pertaining to loan products and user data. In different countries, regulatory institutions may seek to standardize entry thresholds and regulate the registration process for each new platform provider in their start-up phase to reduce platform market entry risk. This serves to prevent platform providers from generating a negative impact on the online lending marketplace both ethically and financially. Secondly, to enable the successful implementation of regulations, the regulatory environment should be protected by enacting sustainable and suitable policies for platforms in different countries at both a national level and a local, and provincial level, with the policies enacted within both of these designed to support and facilitate each other. Furthermore, regulation institutions should maintain information transparency between themselves with platforms to play the role of supervising platforms to reduce loan information transparency risk. Lastly, regulatory institutions should encourage the participation of Institutional Investors in the online lending market, such as banks and investment banks, investing in P2P lending platforms garner multiple advantages. This will also mitigate loan liquidity risk, allowing the platform to gain an increased amount of cash flow, improving the liquidity of the platform.
11.3 Technology-Actor Interdependencies: Platform Services Design and Nature of Transaction
Technology-Actor interdependencies demonstrate how the technology in the platform should be applied to protect and benefit stakeholders. Firstly, platform owners may optimize the societal credit scoring systems for P2P lending platforms to enable the effective and accurate assessment of user information, and possess more data regarding their borrowers, therefore establishing an accurate credit assessment service. Platform owners should look to adopt advanced machine learning techniques and optimize internal software to accurately assess borrowers and calculate correct interest rates which allow the platform to offer lenders the most suitable loan products (Lee et al., 2020). They might collaborate with an external credit-scoring firm that forms the basis of mitigating credit assessment risk and information asymmetry risk. Second, installing a secondary trading platform provides lenders with the opportunity to transfer loans they have purchased to other lenders, thus mitigating concentration risk. This means that lenders transfer the risk to other lenders and receive repayment capital back before loan maturity, potentially helping to avoid financial loss. Furthermore, each platform provider should establish loan diversification mechanisms to separate each loan into several packages and sell them to multiple lenders. As a result, lenders only absorb a small percentage of the default risk generated by each borrower.
12 Risks and Development of P2P Platforms
In the following section, we focus on three specific platforms (1) LendingClub (US), (2) Upstart (US), (3) Renrendai (China), (4) Zopa (UK), and generalize the insights extracted from our theoretical frameworks to understand how risks or risk mitigation approaches affect the development of P2P platforms across different countries, and all the details can be seen in Table 4.
Table 4
Case analysis and risk mitigation strategies
Status
Risk Identifications
Organizational Perspective
Risk Mitigation Strategies
LendingClub
Successful aspects
Investment concentration risk
Technology-Actor interdependencies
Operating second trading platform to allow lenders sell loans between each other
Loan liquidity risk
Structure-Actor interdependencies
Collaborating with custodian banks (WebBank) to manage funds
Platform market entry risk
Structure-Task interdependencies
SEC standardized the entry requirements for P2P lending platforms in the US
Problematic aspects
Platform regulatory risk
Structure-Task interdependencies
The “Quiet Period” is suggested to be shortened and the loan application procedure can be simplified
Loan information transparency risk
Structure-Task interdependencies
Investor advisory team (LCA) should maintain the authenticity of the loan products to lenders
Upstart
Successful aspects
Loan credit assessment risk
Technology-Actor interdependencies
Undertaking multi-types of businesses
Platform operation risk
Loan investment concentration risk
Structure-Technology interdependencies
Implementing AI Risk Management Systems; Combining multiple machine learning algorithms into credit scoring model
Collaborating with custodian banks (WebBank) to manage funds; Creating secondary platform for selling loans;
Loan investment concentration risk
Technology-Actor interdependencies
Building robust credit scoring/recording system
Loan information transparency risk
Structure-Task interdependencies
Establishing a mechanism that maintains the transparency of loans
Zopa
Successful aspects
Platform market entry risk
Structure-Task interdependencies
Regulatory authorities guarantee that platform owners need to attain a good level of financial management capabilities
Loan information transparency risk
Structure-Task interdependencies
Platform users’ rights risk
Structure-Actor interdependencies
Users must receive specific, predetermined levels of compensation if bankruptcy were to occur
Loan credit assessment risk
Technology-Actor interdependencies
Collaborating with the credit reference agency Equifax to accurately calculate borrowers’ credit scores on a scale of 1–10
12.1 LendingClub P2P Lending Platform
In 2007, the LendingClub (LC) Corporation was launched in the US. The stakeholders engaged with the platform included the lender, the borrower, and the platform provider. Also heavily involved was an industrial bank (WebBank), a subsidiary advisory firm (LCA), a regulatory institution (SEC), and a credit assessment firm (FICO). On LC, the industrial bank (WebBank) maintains operational control with regards to the offering of loans to borrowers and transferring loan obligations to the platform, providing evidence that the borrower has received the loan (Jin & Zhu, 2015). Institutional lenders constitute a grouping of investment banks, hedge funds, and insurance companies. These institutions securitized their loans and delegated platforms to sell loans to investors as institutions. This securitization strategy generates substantial profit for themselves and reduces the default risks, as the institutions have an excellent reputation in relation to the marketplace and have invested significant funds to the platform. The key differential reason for LendingClub’s success is the collaboration with social networks, such as Facebook (Jin et al., 2020), which established a reliable societal credit scoring system and established a healthy lending environment.
By using our social-technical risk relationship model, we find that LC possesses a comprehensive Structure-Actor and Technology-Actor interdependencies. Specifically, we found that the feature of the actor strongly matches with the structure and technology. For instance, for Technology-Actor interdependencies, LC operates a second trading platform to sell loans between lenders, which mitigates investment concentration risk. As secondary market strongly matches the feature of lenders’ requirements to diversify their risk. LC also diversifies its loans into multiple ‘Notes’ that further mitigate the concentration risk (Brummer & Gorfine, 2014). There is also an advanced internal risk management system that mitigates platform operational risk.
Secondly, LC has strong Actor-Structure interdependencies. It formed partnerships with custodian banks to manage funds to mitigate loan liquidity risk. Strong regulation control is given by the government and the typical regulated internal structure matches the features of the actor. Borrowers apply for loans via WebBank and submit their asset certification for credit assessment. WebBank then finalizes the assessment and issues the loan whilst transferring the ownership of the loan to LC. LC then uploads borrower data and loan purposes to the platform. Next, lenders can begin the process of choosing to buy payment-dependent ‘Notes’. Furthermore, the SEC standardized the entry requirements for P2P lending platforms in the US. LendingClub has had to securitize its loan arrangements and pay substantial registration fees, alongside waiting for a full year to finish the final application. According to our framework, this policy substantially mitigates platform market entry risk and prevents other platform owners from initiating their business without sufficient funds and management skills.
12.1.2 Failure of LendingClub
Concerning potential reasons for failures, we use a developed social-technical risk relationship model to identify the failures. Based on scams that occurred at LC in the US, the significant decline in stock price represents a temporal failure. First, the weakness of Structure-Task interdependencies leads to the failure of LC since the policies enacted by SEC failed to maintain the profit of platforms. The regulatory policy enacted by the US regulatory authority SEC named “Quiet Period” decreed that platforms must not issue new loans nor gather new lenders during this period (Verstein, 2012), which caused LC to lose valuable time during their application process. This policy hinders the development of LC and failed to match the main task of maintaining revenue and security of platforms, which generates severe platform regulatory risk.
In addition, the US regulatory authorities ensured that all P2P lending platforms pay registration fees in the individual states in which they issue securities, which is costly for platforms. Moreover, the regulatory authority in the US stipulates that only lenders who generate over $200,000 per year can invest, which excludes many individual lenders (Aguilar, 2014).
In 2016, the investor advisory team LendingClub Advisory (LCA) failed to uphold their responsibility to maintain the authenticity of the loan products to lenders, putting their interests ahead of user benefits and breaking the antifraud provisions of the Investment Advisers Act of 1940. It has also been reported that LC’s former CEO intentionally tampered with loan information, making the products with a value of more than $20,000,000 appear as more profitable than they were in actuality to attract lenders (Armstrong, 2018). This incident is a good example of loan information transparency risk seen in our model. The CEO was able to change product information privately as the regulatory authorities had not established a mechanism to maintain the transparency of each loan product to authorities.
12.1.3 Risk Mitigation Strategies for LendingClub
According to our developed social-technical risk relationship model, we found that Structure-Task interdependencies were not robust within LendingClub, as those policies failed to achieve the main task. Policymakers should dedicate their efforts to the building and maintenance of a robust regulatory environment in which the P2P lending platforms operate. The regulatory governance system in the US should mitigate platform regulatory risk through the balancing of regulation restriction and support of the P2P lending platform, avoiding overbearing control of these businesses to achieve the main task of the online lending market. This can be achieved by simplifying the registration process, shortening the application process, and reconsidering the necessity of loan securitization. Policymakers are hereby recommended to establish decentralized mechanisms between SEC and platforms to maintain the information transparency (Davis & Murphy, 2016), and to mitigate loan information transparency risk. The last recommendation is to enact policies that protect the rights of both lenders and borrowers, mitigating users’ rights risk and information asymmetry risk, ensuring that service users possess all the relevant details pertaining to the loan products and receive payment on time.
12.2 Upstart P2P lending Platform
Upstart P2P lending platform was founded in 2012, which is in California, the United States. This company builds an online lending platform to match individual unsecured loans, and upgrade the risk control model by using Artificial Intelligence (AI) technology. On the basis of reducing the credit risk of cooperative banks, Upstart is dedicated to providing loan channels at acceptable prices for customers with less credit information but stronger repayment ability.
In this study, we summarised three main reasons for the success of this platform and analyzed them by using our social-technical risk relationship model. First, the company owns a series of unique AI risk management models, which is the asset-light SaaS model that differentiates itself from applying traditional consumer credit companies. Second, its foray into the auto loan market through the acquisition of auto sales software provider Prodigy. Third, the collaboration with different types of banks helps them obtain a tremendous number of users which provides them with opportunities to optimize their SaaS model and quality of services. The above strategies can be used for reference by domestic consumer credit companies.
12.2.1 Success Aspects of Upstart
Multi-types of Businesses
One of the key reasons for making Upstart successful is that it launched multiple types of loans for different groups of borrowers to satisfy their needs in the market. For instance, it entered the auto loan market successfully through the acquisition of the company Prodigy, which is a famous supplier of auto sales software. After the acquisition of Prodigy, Upstart integrated its AI lending platform with Prodigy's software to quickly acquire a large number of customers from major auto dealers. The business strategy helps this firm gain much more amount of user data which is more than one billion users, resulting in a rapid increase in the volume of auto loans. Furthermore, this large amount of user data was utilized to help build risk management systems API in the platform. According to our social-technical risk relationship model, Upstart as an example identifies the strong relationship between the Actor and Technology that can strengthen the outcome of the Task.
AI Risk Management Systems
Most P2P lending platform in the US which are all depends on FICO (credit reporting system) to offer borrowers credit scores. Whereas, instead of relying on FICO scores, Upstart builds AI risk control models by applying massive data collection, more variables, and deep learning algorithms.
In particular, Upstart's data sources include obtaining borrower data from partner banks, obtaining data from national credit bureaus, monitoring borrowers' repayment performance, and other third-party data. In addition, Upstart will obtain its credit report from credit investigation agencies such as Equifax, Experian, and TransUnion. The platform also learns details about the borrower's education and graduation date, occupation, company and income, deposits, and recent loans. Upstart works with third-party verification agencies to verify the authenticity of this information.
In terms of AI techniques, the core algorithm of big data risk control systems has experienced a well-developed path. It was updated from the expert scorecard to logistic regression, and toward using machine learning, and ends with using deep learning techniques. Most traditional credit institutions still use a FICO scorecard model, while Upstart combines multiple machine learning algorithms into its model. Specifically, the main difference between the scorecard model and the machine learning algorithm is that the parameters of the former are not adaptive, however, the parameters of the latter are adaptive, hence, the optimization model is automatically adjusted after data processing. As a result, the credit level assessed by Upstart can better identify borrowers’ credit levels to help lenders diversify their loans and reduce the loan investment concentration risk. By using our social-technical risk relationship model to analyze the advantages of Upstart, we found that the occurrence of its success due to the interdependencies between Structure and Technology.
12.2.2 Risk mitigation strategies for Upstart
Establising SaaS Systems through Bank Partnerships
Upstart has forged partnerships with various kinds of banks, including community, commercial, regional, and credit unions to create an efficient SaaS system. This system bridges the gap between borrowers (the demand side) and the banks (the supply side), utilizing bank-owned user data to continuously enhance the services offered to banks. Based on our social-technical risk relationship model, we infer that the quality of the Technology is directly influenced by the relationships between each of the Actors within Upstart. Furthermore, collaborating with banks ensures the sustained development of Technology of Upstart, as the ample customer base derived from these banks allows for the refinement of the deep learning model used in their AI risk control systems.
12.3 Renrendai
In the Chinese online lending market, two unique formats of P2P lending platforms exist, which are the online version and the online-to-offline version. The online P2P lending service takes on the responsibility of connecting lenders and borrowers without other credit obligation transformations. However, policymakers in China failed to effectively ensure that those platform providers acted solely as ‘information intermediaries’ rather than as active fund managers (Deng & Yu, 2020). Since the launch of the first P2P lending platform in China, lenders pursued high returns which are along with high risk, and these loans were not diversified in the platform. This practice resulted in substantial levels of financial loss (Liu et al., 2019a, b).
12.3.1 Failure of Renrendai
Our social-technical risk relationship model can be employed to analyze the most harmful risks in the Chinese market, including default risk, loan liquidity risk, and platform regulatory risk, as well as the Chinese regulatory environment. We found that the weakness of Structure-Task, Technology-Actor, and Structure-Actor interdependencies caused the failure of Renrendai.
First, the weakness of Structure -Task interdependencies demonstrates the failure of Chinese online lending regulatory settings. The regulatory environment was initially plagued by loopholes since the launch of the first Chinese P2P lending platform. As a result, platform information transparency, the security of transaction information, and users’ rights were not protected. There was no standardized entry requirement and process for platform intermediaries, resulting in a P2P system that was primarily comprised of platform intermediaries which did not prioritize strong risk mitigation practices. It is common practice among P2P lending platforms in China to first identify borrowers before identifying potential lenders and funding, which generated platform regulatory risk. In China, several intermediaries reversed this process, and gathered funds from lenders before identifying borrowers, often asking them to pay higher interest rates with the intention to increase returns, but often resulting in an inability for borrowers to repay. In some instances, platform intermediaries received funds from investors, created pool funds, and used these funds to invest in other projects (Wildau & Jia, 2018), which generates a financial loss. In several scenarios, these platforms failed to return these funds to investors, with the providers escaping cost because they are unable to repay investors.
In addition, the weakness of Technology-Actor interdependencies generates risks in the platforms. The credit assessment systems across the sector failed to accurately assess borrowers due to the lack of robustness of machine learning techniques used to evaluate borrowers’ creditworthiness. This makes lenders regularly overestimate borrower credit levels and make irrational decisions. Consequently, online credit assessment risk and loan default risk associated with P2P lending were extremely high. These structural risks and disadvantages combine and directly cause the failure of P2P lending platforms in China. Meanwhile, the Central Bank of China issued policies to limit the leverage of all financial activities, which decreased the lending capabilities of SMEs in China. Therefore, many institutional borrowers found themselves in a position where they were unable to repay loans to the platforms, resulting in loan liquidity risk and bankruptcy.
In terms of the weakness of Structure-Actor interdependencies, many platform intermediaries such as Renrendai utilized high-interest rates to attract lenders to establish private pool funds such as ‘shadow banks’, allowing them to conduct private investments which generates loan interest rates and loan default risk (Michels, 2012), presenting severe dangers to the platforms. This also resulted in platform users’ rights risk and loan information transparency risk. This happens because there is a lack of regulations (Structure) to govern the platform owners (Actor)and manage the fund of stakeholders.
Renrendai is well known as one of the top P2P lending platforms in China. Several key factors combined attribute to the platform’s success. Drawing on insights from our model, one could argue that, in recent years, the regulatory environment in China has been undergoing a process of optimization. The regulation authority has issued laws that request P2P lending platforms to implement reasonable interest rates set within a specific range (below 36%), thereby, reducing the loan liquidity risk as borrowers are able to repay funds on time. This in turn reduces the default risk that originates from borrowers. The platform intermediaries do not receive collateral or deposits from the borrower. Instead, their role is largely limited to the facilitation of direct exchanges between lenders and borrowers. Custodian banks are responsible for issuing all loans, which significantly help reduce loan liquidity risk. Finally, Renrendai has developed a comprehensive risk management system through the generation of a substantial amount of user data.
12.3.2 Risk Mitigation Strategies for Renrendai
Next, the platform improves Structure-Actor interdependencies by optimizing online credit assessment systems that are upgraded to assess both hard and soft information of borrowers, mitigating the risk of loan information asymmetry. In addition, the platform should consider the creation of a secondary platform for lenders to sell their loans, as this will mitigate the loan default risk, allowing lenders to withdraw from the platform if there exists a need to withdraw cash. Finally, the platforms should consider establishing partnership relationships with custodian banks to manage funds as opposed to engaging in this activity themselves to prevent the occurrence of a ‘shadow bank’. This will reduce liquidity risk and default risk.
To strengthen the Technology-Actor interdependencies, policymakers in China should initially place their focus on the creation of regulations to be categorized and detailed instead of disbanding all the P2P lending platforms across the market. We recommend they consider building a robust credit scoring/recording system (Lin et al., 2017) which in turn helps online lending firms access the historical performance of an increased number of borrowers. This allows for the successful assessment of their credit levels. They can, for instance, begin a process of collaboration with other Fintech applications to gain more users' credit information, also potentially mitigating loan information asymmetry risk. In addition, platform providers should maintain low-interest rates to avoid scandals and errors and establish a diversification system to diversify the loans and mitigate the investment concentration risks, and loan default risk, and avoid financial loss.
Further, to improve the interdependencies between Structure-Task interdependencies, it is important for platforms to establish a mechanism that maintains the transparency of loans to regulatory authorities to assist in reducing loan information transparency risk. Regulations and policies should be enacted to clarify the position of P2P lending services in that they are to operate as an information intermediary instead of fund managers to reduce the potential financial loss generated by the platforms.
According to the developed social-technical risk relationship model, P2P lending platform owners need to launch a loan diversification system to help lenders diversify loans into multiple packages to reduce investment concentration risk and financial loss. Finally, platforms should keep optimizing risk mitigation systems and evaluate both soft and hard data (Lin et al., 2017) relating to borrowers to improve credit assessment efficiency and accuracy. This enables the comprehensive knowledge acquisition regarding borrowers from a multitude of aspects and matches the appropriate loan products to appropriate lenders, serving to gradually mitigate default risk.
12.4 Zopa
The Zopa Corporation, which was launched in 2005, was the first platform to enter the local P2P lending market. Its successes have been the result of multiple aspects over the last decade. Zopa has been subject to effective governance by both the Financial Conduct Authority (FCA) and the Prudential Regulation Authority (Pooley, 2019), allowing it to operate within a robust regulatory environment.
12.4.1 Success Aspects of Zopa
According to our framework, the Structure-Task interdependencies are the key to making the platform sustainable and successful in the long run. The policies introduced by the regulatory authorities guarantee that platform owners need to attain a good level of financial management capabilities, and users must receive specific, predetermined levels of compensation if bankruptcy were to occur (Zopa, 2018). It also allows for the rights of users to possess knowledge of all relevant details regarding the company’s investment activities, a key strategy in tackling platform users’ right risk. The regulatory policies also allow users to easily come to their own investment decisions. Regulations in the UK also stipulate that a lending platform must publish the expected default rate, alongside the actual default rate, and share up-to-date lending information with the P2P financial authorities quarterly, helping mitigate loan information transparency risk. The FCA determines the fundamental requirements for IT systems, as well as requirements regarding management teams, initial capital, and minimum operational funding of P2P lending platforms in the UK, therefore reducing both entry risk and information security risk. The FCA also announced rules and regulations that place a particular focus on credit assessment, risk assessment, and lender protection.
Regarding the Technology-Actor interdependencies, the internal structure of the platform proved to be successful and more applicable compared with the Chinese P2P lending platforms. Zopa collaborated with the Equifax, a credit reference agency, to accurately assess borrowers’ credit scores ranging from 1 to 10. This assessment reflects borrower’s past credit performance based on their borrowing history, thereby reducing loan credit assessment risk. Zopa has also been instrumental in assisting lenders in diversifying their funds, ensuring that no single borrower possesses more than 1% of a lender’s total investment, which decreases the investment concentration risk. Furthermore, since its inception in 2005, Zopa has been managing funds through a trust account with the Royal Bank of Scotland (RBS). The concurrent establishment of a custodian bank prevents platform owners from unauthorized usage of the capital, which consists of lenders’ investments and loan repayments, separate from Zopa’s internal revenue. This tactic further minimizes the loan liquidity risk within the platform.
12.4.2 Risk Mitigation Strategies for Zopa
According to Technology-Actor interdependencies, the platform owners in the UK should maintain their commitment to work with the regulatory authorities and optimize their online credit assessment systems by collecting borrowers’ soft and hard information to mitigate the platform information asymmetry risk and loan default risk. According to Task-Structure interdependencies, the policymaker in the UK should maintain the robust P2P lending business model, keep optimizing the entry requirements for start-up companies to mitigate platform market entry risk, and encourage both individual and institutional lenders to invest in the lending platforms in the UK by lowering the asset requirements of lenders.
13 Future Research
There are the following aspects that we plan to focus on exploring in the future. First of all, we aim to enhance the theoretical contribution by including and categorizing risks in more scenarios to understand how risks are transformed and mitigated at the organizational level, such as cryptocurrency trading mechanisms and blockchain-based lending systems. We have considered that the collaboration between giant fintech companies is becoming more common, and risks in the platforms are no longer easy to examine as before in the P2P lending platforms. Therefore, we plan to use developed social-technical models to enhance our current model by carefully categorizing the key elements and including other types of interdependences between the elements. According to our model, it would be more meaningful for qualitative researchers to conduct in-depth interviews and case studies to understand the risks in different fintech firms. It would also be valuable for quantitative scholars to measure the risks in fintech companies with different internal structures. Secondly, we further plan to optimize the risk categorization model and risk mitigation framework to demonstrate more potential relationships between risks that have not been highlighted in this study.
Additionally, to optimize our current study, we also plan to examine whether the credit scores authentically reveal the creditworthiness of borrowers and their intention of repaying the loans by using text mining techniques to explore borrowers’ textual comments and investment behavior in the platforms. Besides, we consider collecting a dataset of another Chinese P2P lending platform and utilizing a difference-in-difference approach to examine whether the regulations that confine the market entry standard of platform owners by financial authorities can reduce the loan default risk in the platform. Moreover, we plan to explore the users’ historical activities in other financial payment applications and social media App, to explore whether the usage of multiple applications can build trust between lenders and borrowers and enhance credit assessment accuracy in the P2P lending platforms.
14 Conclusion and Contributions
P2P lending platforms, as a mature market, need to be understood from not only the platform level but also from an organizational perspective instead. While they offer an innovative lending format that reduces transaction costs and generates growing returns, they also create new forms of risk. Surprisingly, previous literature has not sufficiently understood and addressed risks in the context of P2P platforms IS perspective. Theoretically, our research contribution is to develop the social-technical risk relationship model and a systematic framework that classify all the relevant risks on P2P lending platforms from both macro and micro levels and categorized them from an innovative hybrid financial and organizational perspective, proposed applicable strategies for regulators and platform intermediaries to mitigate the risks in such platforms.
Our model provides a novel approach for assessing risks in P2P lending platforms’ risks by utilizing the social-technical and organizational perspective. It underscores those risks are not solely attributable to stakeholders but rather, result from weakened interdependencies between the market’s Structure, the Actor and Technology, and the target as the Task in P2P lending platforms. A deficiency in any one of these factors cannot fully account for the generation of risks in the P2P lending market. It specifically emphasizes that the enactment of regulations determines the interdependencies between the market’s Structure and the other three factors. This indicates that the internal and external structure of the P2P lending market requires a better alignment with all the stakeholders, platform operational systems, and the goals of the platform owner. This serves as a valuable starting point for researchers to explore how risks are generated due to the ever-changing regulatory environments and online lending markets, and to suggest necessary adjustments in the business models of P2P lending platforms to ensure market success.
In practical terms, this paper opens possibilities for helping platform owners, managers, and policymakers with insights into the interplay between risks, regulatory environment, and organizational components. These insights help them to apply strategies to mitigate risks and build sustainable regulatory environments for online lending markets across the world. Senior managers can leverage this model to assess risks not only from a financial perspective but also from an organizational standpoint, emphasizing the interdependencies between actor, structure, task, and technology in P2P lending platforms. They can utlise our findings to evaluate the successes and failures of the global online lending market. Moreover, the model can assist them in adapting their internal business models to various regulatory climates. Regarding future research, we encourage researchers on quantifying the degree of interdependencies between the organizational components and how policies align with interdependencies to reduce risks. Additionally, attention should be given to the creation pf innovative financial tools on P2P lending platforms to mitigate risks.
Policymakers across various nations should aim to refine the roles and responsibilities within their respective P2P lending services, while also assessing these platforms from an organizational standpoint. Their focus should be on devising regulations that not only restrict the occurrence of illegal financial activities while also taking into account the specific characteristics of the platforms to mitigate risk in the long run. We advocate for their continued efforts to discover the ideal equilibrium between regulating and bolstering P2P lending platform markets.
Acknowledgements
We wish to thank the Gillmore Centre for Financial Technology at the Warwick Business School for supporting this research.
Declarations
Competing Interests
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript, and they have no competing interests to declare that they are relevant to the content of this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
