Peer-To-Peer Lending, P2p, Crowdfunding, Default Risk, Loan Performance
Essay Preview: Peer-To-Peer Lending, P2p, Crowdfunding, Default Risk, Loan Performance
Report this essay
Abstract: We exploit a large sample of loans observation from the Lending Club to identify the main determinant of borrowers’ default in P2P lending within each loan term and quantify the default risk of P2P loans based on two different maturity lengths. This allows for a more thorough evaluation of credit risk and a more specific forecast of the probability of default in comparison to previous generalized models. We apply binary logistic regressions to investigate the relationship between loan default and important risk factors. Our finding suggests that default determining variables and their significance are not similar in all loan duration. Credit Grades and Annual installment-to-Income ratio are the most significant factors in forecasting borrowers default for all loan length. (Add to results)Keywords: Peer-to-Peer lending, P2P, Crowdfunding, Default Risk, Loan Performance1- IntroductionPeer-to-peer(P2P) lending is a new platform of financial transactions that bypass conventional intermediaries by directly connecting borrowers and lenders (Magee, 2011). P2P lending companies offer unsecured consumer loans mainly to individuals. Through disintermediation, investors wanting diversified exposure to a fixed income asset class of consumer loans need not to go through asset-backed security (ABS) markets, removing layers of intermediation and opening the asset class to smaller investors (Morse, 2015). Since 2005, P2P lending sites have cropped up all over the world_ PPDai, Zopa, Lending Club, and Prosper are a few examples. The global P2P lending market is estimated to exhibit a CAGR of 48.2% within the forecast period of 2016 to 2024 in order to reach a valuation of USD 897.85 billion by 2024 (TMR research).  The most beneficial advantages of P2P lending for lenders are the possibility of higher returns than those achievable with investments available in traditional markets, and further diversification of each portfolio by offering new asset class (Magee, 2011).  Furthermore, loans to low-income borrowers will often be at lower interest rates in comparison to those available through traditional sources, as P2P lending largely cuts out the financial intermediary (Slavin, 2007).The main drawback of P2P lending is information asymmetry that can result in a moral hazard (Stiglitz and Weiss, 1981) and adverse selection (Akerlof 1970) which ultimately impacts the viability and success of individual P2P lending platforms (Cummins et al, 2018). Despite mentioned benefits, P2P platforms have some fundamental problems. P2P lending platforms seek to minimize the impact of information asymmetries through designing mechanism into their platforms, such as provision point mechanisms, general platform rules, feedback systems, crowd due diligence, and safeguard funds (Cummins, Lynn, Mac an Bhaird C, & Rosati, 2019). For instance, Lending club reduces lending risks associated with information asymmetry through categorization of a loan based on some platform assessment of the creditworthiness of the borrower represented by credit grades. However, recent research suggests that credit grades may not represent accurate estimates of borrowers’ creditworthiness (Serrano-Cinca, Gutiérrez-Nieto, & López-, 2015) and P2P lending is inherently associated with a greater amount of risk compared to traditional lending (Emekter, Tu, Jirasakuldech, & Lu, 2015). Moreover, credit risk of P2P lending can be high, which leads to lenders’ investments being in high risk of default (Mezei, Byanjankar, & Heikkilä, 2018). The hazard rate or the likelihood of the loan being defaulted increases with the credit risk of the borrowers (Emekter, Tu, Jirasakuldech, & Lu, 2015).
Previous studies with the purpose of defining credit risk of P2P platform (Emekter, Tu, Jirasakuldech, & Lu, 2015); (Polena & Regner, 2018); (Serrano-Cinca, Gutiérrez-Nieto, & López-, 2015) have found a positive significant correlation between a loan’s default and Credit Grade, and mentioned to some determinants of loan default such as Debt-to-Income ratio, Revolving Credit Utilization, Income and so on. However, there is some discrepancy between a number of findings due to applying different approaches and time span. Moreover, main previous literature is based on 36-month loans, due to the fact that 60-month loans which issued in May 2010 for the first time had not reached to the maturity at the time of previous studies. Hertzberg et al (2018) showed that borrowers who are less creditworthy choose the longer maturity loan, indicating different credit risk for each loan term. We conjecture that the significance of the default factors depends on the loan’s repayment length. Thus, we study borrowers’ default determinants of 36-month and 60-month loans separately, which allows for a more thorough evaluation of credit risk in comparison to previous generalized models. Moreover, we need to quantify the default risk of P2P loans based on two different maturity lengths, so that investors will be able to estimate the probability of default of each loan and make more informed investing decisions. In addition, separating forecasting models of loan default based on its length leads to a more accurate and specific forecast of default rate in comparison to other studies. Finally, we study all loans together without distinguishing them based on the repayment term to realize whether the default determining variables and its corresponding default prediction model is significantly different from the accomplished results for each loan term. Based on the mentioned points, we defined research questions as follow.1- What are the significant variables having an impact on borrowers’ default for 36-month and 60-month loans in the Lending Club; and how these variables differ within each loan term?  2- What is the specific credit risk forecasting model for estimating the probability of loans default for each loan duration?In order to investigate these questions, we collected data from Lending Club website, in December 2018. The sample includes loans that were issued during the period of May 2010 to September 2013 for 60-month loans and September 2015 for 30-month loans. The data consist of all available published information, including Credit Grade, debt-to-income ratios, loan status and loan purpose. Lending Club is America’s largest online lending marketplace connecting borrowers and investors, based on approximately USD 9.0 billion in loan originations during the year ended December 31, 2017 (www.lending club.com). Lending Club enables borrowers to create loan listings on its website by supplying details about themselves and the loans that they would like to request. All loans are unsecured personal loans, classified to 36-month and 60-months and can be between USD 1,000 – USD 40,000.