729 karma · joined February 28, 2013
We did experiment with offering ACH payments via a different service (Check Gateway) a couple years ago. We discontinued that, because the automated address and other verification checks required for online ACH authorizations blocked many legitimate payers, such that it cost more in declined payment fees and hassle for our lenders than it saved in transaction cost. Also, almost half of our lenders are outside the US.
We did request free payment processing for Zidisha, and have also asked Kiva to request the same deal for Zidisha. Two years ago we even had an online petition for PayPal to waive fees for Zidisha loans - https://www.change.org/en-AU/petitions/paypal-stop-charging-....
Thus far, our applications for free or heavily discounted payment processing have been turned down, and we continue to seek more cost-effective options.
We had a discussion of this idea in the Zidisha member forum some time ago, and most participants indicated they would lend less if they did not have the option of receiving their funds back.
A greater concern is that turning the loans into donations, even though the individual borrowers are expected to repay them, could change the nature of the relationship between lenders and borrowers from one of mutually beneficial partnership, to that of patron and dependent. I don't think this development would be good for borrowers or lenders.
Our current position is, as you say, somewhere between the extremes of for-profit lending and outright donations. Some would call this subsidized provision of capital or "impact investing." As a pioneer in this space, we are still adjusting our positioning as we learn with experience what works best.
Zidisha is registered as a 501(c)(3) nonprofit only, and does not have a securities license. For this reason, our terms of use do not offer lenders any guarantee they will be able to withdraw repaid loan funds. Legally, Zidisha lending accounts are not financial assets, and any payouts to lenders are treated as voluntary promotional payments offered at Zidisha's sole discretion. That is one more reason we try to keep the emphasis on philanthropic impact, rather than the financial aspect of the loans.
It is true that the conclusion that higher interest has not led to more lending is not based on rigorous data. It is rather based on the general observation that during the past year or so, loan funding volume has not noticeably increased at times when the highest interest rates offered at Zidisha increased.
The generalization that Zidisha lenders are not primarily attracted by financial returns is likewise not based on data, but rather a subjective assessment resulting from many years of survey responses, comments posted on our website and conversations with lenders.
(An important caveat is that many of our lenders do care about preserving the value of their lending funds over time, and often try to lend at interest rates that they believe will help them achieve this goal. But this is a very different motivation than profit maximization.)
I agree with you that offering dramatically higher interest may well attract more capital, from lenders who are motivated by profit rather than philanthropy. This would not necessarily be a bad thing, but it is not our mission or core area of expertise.
It will be interesting to see if a profit-based variation of Zidisha emerges, and whether it ultimately has a greater impact on the opportunities available to entrepreneurs in developing countries than our philanthropic platform.
Thus far, we've opted to display only one kind of interest rate (the simple flat rate upon which our website calculations are based), along with a modal box note that the equivalent APR is roughly twice this, and a prominent display of the actual dollar costs to the borrower.
Thanks in part to this debate, we have opted to reduce the interest Zidisha borrowers may offer to lenders. You may view the full announcement and discussion here:
As of today, we've funded 5812 loans and have 4757 registered borrowers, so the majority of loans we've funded have been first loans. That said, we have almost no registered borrowers without loans. We don't have precise data but I would estimate over 95% of borrowers go on to raise additional loans after their first is repaid.
The reason for the high proportion of first loans is that our growth has accelerated in the past several months, and most of this growth has been a result of admitting new borrowers. We expect the large majority of these new borrowers will go on to raise multiple loans, as has historically been the case.
Zidisha's total revenue and operating costs are less than $10K per month. Only two people, myself and a web developer, receive a salary for working with Zidisha. (That salary is barely enough to cover the cost of a shared or low-income apartment.) Everyone else who works with Zidisha is an unpaid volunteer.
We do our best to ensure loans do not cause over-indebtedness by starting with very small amounts initially, and increasing credit limits only if high on-time repayment rates are maintained over time. In order to minimize the risk of borrowers taking out local loans in order to repay Zidisha loans, we do not allow credit limit increases if a large percentage of Zidisha loans are repaid in a short period of time.
In my experience, repeat loans have been a vehicle for increasing assets and revenue over time, not for increasing indebtedness. Here are a few among many examples:
http://p2p-microlending-blog.zidisha.org/2014/03/15/spotligh...
http://p2p-microlending-blog.zidisha.org/2014/03/29/spotligh...
http://p2p-microlending-blog.zidisha.org/2014/04/06/an-agele...
These are two separate costs. The fee for lenders offsets the cost of receiving their funds via PayPal or credit card. It is set to match as closely as possible the average transfer fees that our organization pays to PayPal and Stripe.
The 5% charged to borrowers covers international money transfer costs, and the cost we pay to local payment services in each borrower country. This fee is also set to match the average money transfer cost we pay across all countries.
We use two separate fees because it allows us to ensure that we receive income at roughly the same time costs are incurred. This helps ensure we always have enough cash flow to cover our costs, no matter how quickly the organization grows.
I see what you mean. This is confusing because our website is programmed to display the registration fee only for applicants who have not yet had a loan disbursed. Once a loan is disbursed, the fee is paid and is no longer displayed in the loan profile page as part of the amount due to be paid by the borrower.
The second and subsequent loans contain a link "View previous loans" under the "On-Time Repayments" record in the right panel. All the loans that do not have that link are first loans.
Thanks for commenting here. I wish you had asked for our side of the story before publishing these allegations.
I agree that flat rates should be compared to flat rates. It would indeed be misleading to compare Zidisha's flat interest rates with another microfinance provider's APR rates.
In my experience, flat interest and fee rates of over 30% are common in microfinance, and that was the basis for the comparisons I have made with the cost of Zidisha loans. It is hard to find reliable data on the average APR of microloans worldwide, because many microfinance organizations underreport the full cost of their loans, and those that do opt for full disclosure on sites such as Microfinance Transparency and the MIX Market tend to be those with the lowest rates.
This article provides a good overview of some of the complications involved in measuring the cost of traditional microfinance loans: http://blog.givewell.org/2010/04/02/microfinance-interest-ra...
Thanks for sharing this. What if we were to continue to calculate interest and fees using our current flat method for the time being, but display the equivalent APR to lenders and borrowers? As a prospective lender, would this have addressed your concerns about transparency?
You can propose any interest rate between 5% and 30%. The first 5% goes to Zidisha to cover money transfer fees, and anything above that goes to lenders. Lenders who find the proposed interest rate reasonable can choose to fund the loan at or below your proposed interest rate.
In addition, new applicants who succeed in raising a loan pay a one-time registration fee, which is deducted at the time the first loan is disbursed. View current registration fees
"View current registration fees" displays a modal box with the registration fees applied in local currency in each borrower's country. (These are currently about $12 in Kenya and Burkina Faso, and zero in the other countries in which Zidisha operates.)
Our loan application form does provide a cost calculator to borrowers, though it does not use the term APR. The loan application form works as follows:
1. The applicant specifies the amount he/she would like to borrow, and the combined interest and fee rate he/she would like to offer.
2. The applicant specifies the local currency amount he/she prefer to pay each week or month. (Our website advises selecting the largest installment amount one can comfortably pay in order to minimize interest cost.)
3. The applicant chooses the day of the week or month he/she would like to make payments.
4. Based on this, our site generates a display of the local currency principal and fee/interest amounts that will be due, along with a projected repayment schedule.
5. The applicant can go back and modify the proposed loan parameters until he/she is satisfied with the cost and repayment schedule displayed, and can then publish the loan application on our website.
6. Lenders can fund the loan at rates between zero and the applicant's offered rates. In practice, many lenders opt to fund loans at low or zero interest, such that most applicants receive the loans at lower interest than what they had proposed.
This makes sense, and we are already doing some of that.
In the loan application form that borrowers see, the 5% fee to Zidisha and the interest offered to lenders is in fact combined. Borrowers choose any combined rate from 5% up to a maximum of 30%, with a text description indicating that the first 5% of this is paid to Zidisha and the rest is paid to lenders.
The loan application page also indicates the registration fee, as a cost separate from the combined interest / fee rate that the borrower selects.
We separate the service fee from the lender interest in the loan applications that are displayed to lenders, because the intent is to indicate the maximum interest that they can expect to receive from each loan proposal.
Although I do not agree with Modern Microcredit's allegations that we are intentionally misleading, the discussion that has resulted is a useful one.
As I see it, we have a several options:
1. The status quo: Continue to quote flat interest rates in our website along with an explanation of the calculation in a modal box, and display the dollar amounts paid for each loan for extra clarity.
2. Transition from a flat rate methodology to APR. This would be more in line with the standard used in most lenders' countries, but would be more difficult to understand for borrowers. (It would also require a rewriting of much of our website. Zidihsa has only one web developer, so implementing this option would probably have to wait until have more programming resources.)
3. Continue to use flat rates as the basis for calculating the cost of Zidisha loans, but display the equivalent APR rates along with them. Perhaps Microfinance Transparency could help us develop a tool to facilitate this.
4. Continue to use flat rates, but use the term "fee" instead of "interest." This would be easy to implement and could help lenders avoid mistaking our flat rates for APR, but makes cost comparison with loans in borrowers' countries less straightforward.
Zidisha has little value without transparency, and those who have commented here obviously care deeply about it. I'd welcome everyone's advice on which option Zidisha should adopt, or alternative proposals.
Am I understanding correctly that you are suggesting we not use the term "interest", but instead call the amounts borrowers pay to lenders a fee?
That is worth considering. It would avoid much of the confusion over flat rates versus APR, and would still be clear to borrowers that the amount represents the cost of the loans.
I'd be interested to hear others' perspective on whether Zidisha ought not to use the term "interest" at all, but instead replace it with "fee" or some other term.
Thanks for this comment. I'd welcome the help of MFT to develop an APR calculation tool for Zidisha loans. I've just sent you an email, and look forward to discussing further.
Thanks for noticing that. I am not finding the quote "And that's just for repeat borrowers (by sheer chance, none of the 20 random loans was from a first-time borrower)." in the blog post now.
I did take a look at the borrower profiles that were linked below the photo collage, and the majority of them were indeed first-time loans. These can be identified in that only second and subsequent loans include the link "View previous loans" below the on-time repayment rate displayed in the profile pages.
Yes, that is true: loans held for less time pay less, and a loan at 5% held for 2 years would pay 10% of the principal.
Using flat rates instead of APR does result in loans that cost disproportionally little when held for short periods, and disproportionally much when held for long periods. The greatest concern are cases in which a loan for an initially large amount is held for a long time after most of the initial amount has been repaid, resulting in interest calculated as a flat percentage of the original amount being applied even though most of it is no longer outstanding with the borrower. Avoiding such situations is probably the strongest argument for changing from a flat rate to an APR methodology.
We have not done so because such cases are not very frequent, and because we judged that the value of having a simple and intuitive cost calculation method outweighed the value of having a more precise model that the majority of our members would not understand in practice.