Peer to peer lending first started appearing in 2006-07 initially working off as applications running on top of popular social networks like Facebook. These platforms sought to utilize the information contained in a person’s networks and connections to estimate the credit risk with more accuracy and offer fine tuned terms to borrowers and investors. The foremost of these platforms is Lending Club with a market cap of $3.21 billion. Lending Club operates an online platform which enables borrowers to obtain a loan and investors to purchase notes based on the payments made to those loans. In this blog we explore if network effects exist in the peer to peer lending business model and its implications for Lending Club.
Business model: The business model for Lending Club is summarized in the following graphic.
The loan process can be summarized in the following sequence of steps: (a) Borrower opens account with Lending Club providing required information (b) A bank provides borrower the loan (loan terms are based on borrower’s FICO score, debt to income ratio, term, purpose of use etc.) (c) The bank sells the loan to Lending Club (d) Lending Club keeps the loan on its balance sheet (e) Lending Club creates an unsecured structured note linked to that loan (notes are obligations of Lending Club not of the borrower, thus investors also stand to loose if Lending Club becomes insolvent even if borrowers continue to service the loans) (f) Lending Club sells the note to an investor (g) Borrower makes payments on the loan which are transferred to note holder after Lending Club keeps a 1% fee for providing loan to borrower and service fee from investor.
The key ways this is different from a traditional bank is that 1. Lending Club’s assets (loans) and liabilities (notes) are perfectly matched in duration (i.e. notes become due only when the loans mature) and 2. Loss is born by the investors and not by Lending Club i.e. every dollar that a borrower doesn’t pay to Lending Club is a dollar that Lending Club doesn’t pay to the corresponding note holder. Thus Lending Club closely resembles a two sided market place comprising of borrowers and investors. We now explore (a) Same side network effects for borrowers (b) Same side network effects for investors (c) Cross side network effects for borrowers and investors.
- Same side network effects for borrowers: In some ways borrowers on Lending Club are competing against each other to get investors to finance their loans. This leads to a negative network effect for borrowers. However, since the investors are transparently exposed to borrowers they need to diversify their investments on their own i.e. they will seek to buy notes linked to many different loans to diversify the risk of default by an individual borrower. This is not possible with only a few borrowers. Since diversification decreases the portfolio risk for the investors it makes it possible for them to offer credit at a lower interest rate. This is a positive network effect for the borrowers.
- Same side network effects for investors: The network effects for investors are similar to those of borrowers, they gain from opportunity to co-invest with other investors and diversify (positive network effect) while also competing for the most credit worthy borrowers at a given interest rate (negative network effect)
- Cross side network effects for borrowers and investors: In the two sided marketplace that Lending Club represents there are positive cross side network effects with more borrowers increasing investment opportunity for investors and more investors driving the interest rate lower for the borrowers. In its early days Lending Club used its algorithm LendingMatch to match borrowers and lenders using common factors between them like geographic location, educational background, similarity of social networks etc. with the logic that borrowers taking loans from lenders with similar background as theirs will have lower default rates. This has been discontinued since Lending Club shifted to note based model. Since there is a loss of this information as the rate at which borrower borrows is independent of lenders lending her we can argue that there is negative network effect mechanism in a bigger diverse borrower set. However, a diverse borrower set also leads to better diversification which is a positive network effect.
Lending Club has provided great risk adjusted returns over the past years e.g. average return net of defaults and fees were 8.01% for the 2010 vintage.
Along with net positive network effects Lending Club benefits over commercial banks in (a) Having no reserve requirements (banks essentially provide risk free like claim to investors on risky investments like loans to people and business, hence they are susceptible to runs and need a reserve to absorb some shocks) (b) low overhead costs due to online only presence. This has helped them to outperform traditional lending institutions like banks.
However, though the network effects contribute to making the two sided marketplace model succeed they still do not guarantee a winner takes all market. Since the loans are for a 3-5-year period the borrower and seller interaction with the platform are low; additionally, multi homing costs are very low leading to a plethora of peer to peer lending companies. Many firms have also started in niche segments such as offering loans to students of culinary classes etc. to mobilize borrowers and lenders to their platform. The participation of well diversified professional funds also reduces the need for diverse consumer base in peer to peer lending making platforms with less borrowers also viable for investing. Though there are strong network effects and also some benefits from economies of scale we project that peer to peer lending will remain fragmented in absence of major regulatory changes. In such a scenario the key growth driver for Lending Club is likely to new loan issuance to new borrowers making the competitive dynamics of peer to peer lending not very different from commercial banking.