Social Underwriting 2.0
When we launched Kiva Zip in late 2011, one of the hypotheses we wanted to test was to what extent social underwriting could complement, and even supersede, conventional financial underwriting in assessing businesses’ credit risk. Whether the strength of a borrower’s social network, relationships and character, could represent a better determinant of her repayment rate, than the strength of her FICO score, balance sheet and collateral.
To-date, this social underwriting has primarily been effected through the use of Kiva Zip trustees – organizations and individuals who vouch for the character of the borrowers they endorse on Kiva Zip, based on their relationships with them.
And while there have been many bumps along the road, and Kiva Zip’s repayment rate of 88% is certainly much lower than Kiva.org’s repayment rate of 99%, we have been very encouraged by the progress we have made over the last two years, and the results we have realized:
• We have now onboarded over 400 trustees across both Kenya and the United States, the vast majority of which have a repayment rate of 100%.
• Our U.S. data shows that when trustees have known borrowers for over a year, the repayment on these loans is 91%, compared to only 83% for borrowers that have known their trustee for less than a year.
• Thought leaders are starting to demonstrate excitement about, and commitment to, our innovations in the field of social underwriting. Speaking at our launch in Little Rock, Arkansas last year, Bill Clinton celebrated Kiva Zip's trustee model, saying “when someone vouches for you, your sense of obligation to repay the loan goes through the roof”.
As well as the concept of trustees, we’ve also seen social underwriting happening through the crowd. While we acknowledge that we cannot solely rely on the due diligence of our community of lenders, we strongly believe that there is wisdom in this crowd, and that having thousands of additional pairs of discerning eyes on a borrower’s loan request will lead to much better underwriting decisions than not having that additional due diligence check. And again, this belief is underpinned by the numbers. Looking at U.S. Kiva Zip loans, where there have been at least 50 lenders to a loan, the repayment rate has been 94%. This is compared to a repayment rate of 83% for loans with less than 50 lenders.
It could be that this difference in repayment rates is also partly caused by the increased social pressure on borrowers with a larger number of lenders to repay their loans – both in a negative sense, because borrowers do not want to let down such a large number of people who have put their trust in them; and in a positive sense, because borrowers want to keep this significant community of lenders engaged and motivated to support their business. Either way, it’s another encouraging affirmation of the concept of social underwriting.
But what is most exciting to us is that these elements of social underwriting, which we have been developing over the last two years, are only scratching the surface of this concept. The trustee model and the wisdom of the crowd idea have been social underwriting in black and white. The Private Loan Period that we’re rolling out in a concerted way over the coming months is social underwriting in high definition Technicolor.
The idea is that before a Kiva Zip borrower is listed on the public lend tab, they must first invite 15 lenders from among their own social network to lend to them. As you can read in the blog post linked to above, not only will this help us grow the number of lenders participating in the Kiva Zip community, but it also represents a significant increase in our emphasis on social underwriting as a tool for managing our risk. Conventional lenders take assets (houses, cars, machinery, etc.) as collateral on the loans they make. Kiva Zip doesn’t take collateral of this kind. But for borrowers that go through this Private Loan Period, we take social collateral, in the form of the money that the borrower’s own friends and family lend to them, through the Kiva Zip platform. Given that every small business owner needs to be able to “hustle” to ensure that their nascent business recruits customers and grows, we also believe that requiring borrowers to demonstrate this entrepreneurial spirit, by persuading a small proportion of their social network to fund their Kiva Zip loan, is an excellent criterion for the assessment of a borrower’s creditworthiness.
As we only disbursed our first “Private” loans in December, it’s too early to evaluate the effectiveness of this feature yet, but the data we do have is very encouraging:
• Where U.S. Kiva Zip borrowers have invited at least 10 people to fund their loans, our repayment rate is 99%, compared to 87% for borrowers that have invited 9 or less.
• Borrowers that have invited 5 or more lenders have a repayment rate of 96%, compared to 86% for borrowers that have invited 4 or less.
• And even 1 invited lender makes a difference. The repayment rate for borrowers that have invited at least 1 lender is 90%, compared to 85% for borrowers that have not invited anyone.
Over the next few months, we’ll be starting to put every U.S. Kiva Zip loan through a Private Loan Period. There will be some friction, as we strive to clearly communicate this policy change to trustees, borrowers and lenders, but we’re very confident in the benefits of this approach for the long-term sustainability and growth of the Kiva Zip program, and so we’re excited to commit to it in a wholehearted way.
At first we’re rolling this feature out in the U.S. only, but in the future we imagine its adoption in Kenya too – perhaps we will require Kenyan borrowers to get 15 people in their network to send us an SMS vouching for their character; or perhaps we will start enabling Kenyans to lend on Kiva Zip using M-PESA, as well as borrow money in this manner.
At Kiva, we aspire to ascribe more value to human relationships than financial transactions. Where banks are reluctant to risk lending $5,000 to a small business owner who lacks traditional collateral, and doesn’t have an established FICO score; we see a crowd of generous lenders being willing to risk $5 to give that same small business owner a chance, based on the social collateral that they demonstrated in persuading 15 people in their network to fund their loan, and the long-standing relationship that they have with their trustee. Where borrowers feel emotionally uncompelled to repay an impersonal financial institution that only communicates with them in stern, critical words; we see those same borrowers committing to repay the community of hundreds of lenders that put their $5 of trust in them, and became their customers, business advisers and supporters. A hundred years ago, banking was built on the foundations of relationships between banks and their customers. Now imagine a financial system that marries those human connections and relationships, with the power of twenty-first-century technology to expand access and reduce costs and interest rates. We’re on an extremely long, uphill, winding road with this Kiva Zip program. But that’s where it’s headed.