Financial Access Initiative, 5 July 2011
What does a microlending operation look like? Well, it may be a bank or an NGO (and many others in between), it probably has some branches, branch managers, loan officers. The funding of the MFI may come from deposits or from debt, whether from a local or foreign institution, including from online platforms such as Kiva. There may be variations on these themes, but that pretty much describes microlending as we know it.
What if you took all that away – the branches, the loan officers, the institutional funders? Could the lending still work? Well, one model is that of Zidisha Microfinance, an online lending platform that connects lenders in (mostly) developed countries with borrowers in developing ones. And, unlike Kiva, the connections are real – borrowers create their own online profiles, post their own loan applications, and make their own repayments. They also post their own comments, as do the lenders. There is no local MFI intermediary – it is literally the first true person-to-person (P2P) microfinance lending platform in the world.
Naturally, getting there has required a lot of innovation and experimentation. But before I describe the novel business model behind Zidisha, let me first put to rest your skepticism. Zidisha works. Since its first disbursements in September 2009, Zidisha has issued 107 loans totaling $62,000, of which it collected $25,000 in repayments. In Kenya, where most of its lending is done, delinquent loans (PAR 30) are 6.6% of portfolio, lower than the country median of 7.7% reported on MIX.
The inner workings
So how does Zidisha work? How does a borrower in Maasai Mara, Kenya, receive a loan from a group of lenders in Germany and US, and, perhaps more importantly, how does she pay it back? Here’s the story of an archetypal borrower I’ll call Sarah.
Sarah first hears of Zidisha from a visiting student volunteer. She does not borrow then, but some of her friends do. Over time, she comes to appreciate Zidisha’s transparent process that leaves no room for a corrupt loan officer to abscond with any of her friends’ money. She decides to try it. Being computer literate, Sarah uses the local school computer to create her profile, providing information about herself and her business. Because this isn’t her first loan, she also provides information about her earlier loans from MFIs and others. She also provides her Safaricom phone number. Then she waits.
Back in US, on the outskirts of Washington, DC, Julia (Zidisha’s volunteer founder and director), sees Sarah’s application and forwards it to a credit reference partner in Kenya – a subsidiary of the credit bureau CRBAfrica. They check the borrower’s credit history, verify references, and post the results on Zidisha. Seeing Sarah’s good history, Julia approves her for borrowing. Sarah then chooses her desired loan amount (within limits set by Zidisha) and 7% as her maximum interest rate, providing an explanation for how she will use the loan. Then she waits one more time.
A lender in Australia, waking up to find a new borrower online decides that he likes her profile and bids $50 (US, not Australian) at 5% interest. Within the next few days, lenders from around the world bid their desired amounts and interest rates. After her loan request is met, Sarah decides to end the bidding when her interest rate reaches 2.5%.
The process is nearly done. Zidisha transfers the funds to its bank account in Kenya, then sends them via M-PESA to Sarah. A message appears on her phone. Sarah smiles – her money is here! She only needs to pick it up from the local M-PESA agent. Next month her first payment is due, and she again goes to the M-PESA agent to send the installment back to Zidisha. Sometimes she’s late by a few weeks, but not often. From time to time she posts updates on her business and responds to lender comments. When she repays her loan, she can start again whenever she wishes.
This then is the mini-version of the Zidisha model. A few notes worth pointing out. Zidisha’s interest rates are remarkably low, ranging between 7-8% annually (quoted flat), of which 2-3% is charged by lenders and 5% is levied by Zidisha to fund its operations. An additional fee of some $10-20 is charged for initial registration (but not for subsequent loans). This is far below the local prevailing rates – MFTransparency places similarly-sized loans (50,000 KES) at about 35% APR (Zidisha’s loans are 15-18% APR). And it’s all the more noteworthy, considering that Zidisha’s borrowers are largely in rural areas, where credit tends to be more expensive. The key to the low rates rests on Zidisha’s avoidance of costly staff and operations on the ground, and its ability to leverage low-cost funds from socially-motivated lenders.
The one physical connection Zidisha has to borrowers is the use of college student interns who spend periods of several months on the ground holding Zidisha workshops and visiting borrowers. They never collect nor disburse any funds. And they’re not always present – Kenya has not had a student volunteer since late 2010. However, in addition to marketing Zidisha, these volunteers play another important role – because they visit the borrowers at their place of work, they can help prevent fraudulent applications.
The challenges of being first
Of course, like any organization charting a new path, Zidisha has had its share of obstacles. It has had an unusual case of fraud in one village in Kenya, where a neighbor of several Zidisha applicants had used their passwords (which they didn’t know shouldn’t be shared) to change the target M-PESA accounts on their profiles. Thus, the loans were incorrectly disbursed to him instead of the rightful borrowers, and with these proceeds he fled the village. Working with the Kenyan police authorities, Zidisha was able to find him and he’s currently in jail, while awaiting trial – the lost money may yet be returned. An interesting twist to this story is that the police were able to use M-PESA transaction records to track down the suspect – something that is usually not possible when transactions are done in cash. Also, since this incident, Zidisha has modified its security procedures, requiring double-verification of changes to critical account information.
Although Zidisha’s Kenya portfolio has been performing remarkably well, it has had more trouble in Senegal. The Senegal portfolio is more recent than Kenya and comprises some 40% of its combined portfolio. It also carries a 33% delinquency rate (PAR 30), which is naturally a serious concern. However, the high number can be also somewhat misleading, as many borrowers in Senegal frequently pay late. Mobile banking is as yet unavailable in Senegal, so the payment process for borrowers is via direct bank deposit, requiring them to make a trip to Zidisha’s bank and often deal with long waits. In addition, a number of other internal and external factors may also be responsible for its portfolio issues in Senegal, which Zidisha has been monitoring closely. It has already acted by lowering allowable loan amounts, and is likely to make further changes. And of course the troubles in Senegal don’t take away from the obvious successes in Kenya.
The brave new online world in Maasai Mara
The new path blazed by Zidisha raises some interesting questions. How critical are the armies of loan officers employed by MFIs? Are there additional ways to leverage credit bureau data? What other novel approaches may there be for mobile banking?
But besides these questions, Zidisha deserves attention for its own achievement, which is indeed remarkable. It has brought traditional microfinance clients into the same virtual space commonly thought to be the domain of the middle class, and demonstrated that this can be done sustainably. To do so, Zidisha is leveraging every recent connectivity innovation that has made its way to some of the poorest areas in the world: widespread telephony, internet connectivity, mobile money transfers.
All this, from a tiny start-up on the outskirts of Washington, DC, with no office or paid staff.
Daniel Rozas is a lender on Zidisha, and has been informally advising the organization for several years.