Needless to say the debate rages on. The Zidisha discussion detracts from this central point, but on valid grounds. To some extent the shortfalls of microfinance may be due to the high interest rates charged, the restrictive loan conditions, the profit motivation, the aggressive intermediating MFIs etc., some of which Zidisha addresses. However, if you read the work of Milford Bateman, for example, he raises more fundamental questions about microfinance. In the P2P space profit is not really an option, and in Zidisha's case it is debatable whether one can even break-even. But look at the higher levels, the profits accumulating to Accion in the IPO of Compartamos; the millions made in the IPO of SKS, or at some of the microfinance investment funds - these are the big boys, and at that level yes, there are vast profits to be made. Motivation is a hard thing to prove, but you are right to suggest that profit has come to play an unhealthy role in the debate. The sector is extremely opaque, transparency is rare, there are rampant conflicts of interest and ill-aligned incentives. It is largely unregulated in practice, despite endless window-dressing to the contrary. And when some new start-up promotes microloans for poor people to connect their houses to the drinking water supply, as sensible and this might superficially seem, isn't this a public good that the government should provide?
Reading the standard microfinance mantra one could be forgiven for concluding that all the world's problems can be solved with loans, and that profit maximisation is the best structure to arrange this. We are familiar with the theory, but the evidence is sorely lacking.
Sure, Banco D-Miro in Ecuador. When I say I face no forex risk, this is because Ecuador is a dollarized country and I operate in dollars, obviously Europeans etc. would face forex risk. The interest rates are visible here:
My actual rate is marginally below 8% (alas I couldn't do a $20.000 deposit!), and foreigners basically have to deposit for 1 year, or it gets complicated. All deposits up to $30.000 approximately are guaranteed by the government. The bank is BBB+ rated I believe, and Ecuador is not a risk-free country, so when I say this is "relatively risk free", this is somewhat subjective. I live in Argentina, and that is a whole load riskier.
Completely accurate, which is why I state "I am basing this entire analysis on a number of large assumptions. Firstly, that my current outstanding loans will be repaid with the same reliability as my completed loans", and a number of other assumptions in addition. Nor do I know the extent to which previous loans, which may now be 100% repaid, may have dipped into default and subsequently recovered, which also incur a marginal cost in terms of opportunity cost of capital etc. My analysis is very much "cash based", which is limited, but this is the data I have available. A quick look through some of the 100% repaid loans suggests that at certain points in the cycle these were also overdue. Perhaps I could obtain the individual repayments versus due dates for each loan and work this out with greater accuracy, but I haven't bothered so far. To what extent are my current loans likely to default at the 18.65% rate you suggest? I have no idea. To what extent did my previous loans default to this extent? I have no idea. This is why I simply take a bird's eye perspective and look at net cash flows. However, as you accurately imply, for me to genuinely do this with precision I should wait until all current loans have been repaid, which either means I have idle funds on the platform, or I have to complicate the analysis by considering my average outstanding balance, which I have assumed for simplicity is $1000, which it has been to date roughly.
However, are you considering the gross interest income I will earn on the loans that don't default over the forthcoming period? This will partially offset some of the defaults, I can't say with precision to what extent. I count only principal outstanding, not interest income due. So, to some extent this will lessen the impact of possible defaults.
Finally, I have done a couple of new bids since my cut-off date of April 6th when I downloaded the data, so the numbers might not add up 100%. Oh, and Jessica is my wife, she started on Zidisha first, before I took over!
At the end of the day I am hesistant to say resolutely that Zidisha is a break-even venture. It has been so far, but subject to certain assumptions which I hope I have stated clearly. 18 months is a decent trial-period, but it's not a perfect analysis. And if I do subsequently lose a few percent on $1000 that is tolerable. Would I put $10.000 on the platform? No. And comments warning that Zidisha needs to tighten delinquency are completely accurate, and to an extent I am gambling on their ability to do just that - progress seems good so far (in 2014). What intrigues me is the innovation in the business model. It is disruptive. It is a first-mover in this space. Does that mean the model is perfectly refined and cannot be improved? I doubt it, and I look forward to seeing how they develop. But I think it is worth giving them a chance, which is what I have done (to a modest extent), and waiting to see what happens. I will update my blog periodically when more data comes in.
Correct, I failed to mention this in my original post. My fee was $35 (3.5% of $1000), and my total bids have been $1818 over the period, so this equates to about 2% overall. Unsure if I have to pay this 3.5% to get the funds out again. Have you thought of speaking to PayPal to see if they would waive this fee as they did with Kiva? They seem fairly sympathetic to microfinance, and this is an innovate model. However, they may perceive you as a competitor to Kiva, particularly as they now have Kiva Zip, and I imagine Kiva themselves would prefer to retain exclusivity in this waiver of trasnaction fees. No harm in asking. Post their response here!
I saw the HN article questioning Zidisha - Modern Microcredit I think. My experience has differed from that author's impression, and I have tried to be as transparent as possible to explain with hard data why. But I cannot explain HOW this difference in experiences with Zidisha occurs. My concerns with Kiva are more far-reaching, but no point going over that now. I receive many emails asking precisely this question, which prompted me to embark on an overview of the sector, explain the options available (there are more than you might think!), and to examine the individual players in each option. Ultimately you need to decide what you actually seek to achieve, and which model you prefer. Do you need to use a P2P at all, for example? What about simply opening a savings account in a regulated microfinance bank and leaving your funds there, potentially earning a small interest rate? Your funds will be used for the same purpose, lending to microfinance clients, but you won't know which? And if you prefer the P2P mechanism, as far as I can see the key questions are: lend or donate? Lend at interest or interest-free? Lend via an MFI or directly? Once you answer these questions you will have narrowed your search down significantly, and can then do a more detailed analysis of the options available. I think you should think about these questions, and be aware of the options, before you try to select who you should invest with.
You are right that MyC4 is euro denominated. However, I had a quick look at their FAQ:
I would only need to raise my average APR by about 0.2% to cover the PayPal fee. Interestingly Kiva gets around this: "Kiva is the first organization that PayPal has supported by providing free payment processing. This support saves Kiva up to one million dollars each year, which helps us to keep our operational expenses low and send 100% of your loan contributions to the field." The CEO of Kiva is formerly from PayPal. This is a serious advantage to Kiva, and disadvantage to Zidisha. I have no idea if there is an alternative way of uploading funds to Zidisha without incurring these fees.
Regarding my profit motive, I don't see these platforms as an effective means to profit. I seek to break-even. By seeking to break-even this permits side-by-side comparisen. It creates a level playing field to compare, and from this, one can modify the break-even criteria to measure the impact. For example, on MyC4, if I see a loan is fully-funded at 12.2%, but the interest ceiling is 12%, I have a simple Excel tool that tells me how much I need to bid at 5%, or 0%, in order to displace enough of the highest-cost bidders in order to faciliate the loan. So, I might do a modest loan of only $25, at a very low or zero interest rate, but as a result of this a loan for $500 which may not have otherwise been disbursed is able to be disbursed. I consider this a good use of my funds. Also, by constantly displacing the highest bidders, at the margin this may frustrate them by constantly having their bids removed at the last minute, and persuade them to lower the rates they charge - who knows if this is effective or not, but this is one strategy I use. But, overall I seek to break-even on all these platforms.
This is not entirely altruistic. I also have a one-year fixed-term deposit at a regulated microfinance bank that I know very well, which yields 8% APR and suffers no forex risk. My funds are also used for lending to microfinance clients, only I don't know precisely which ones. My logic is that this generates a relatively risk-free return (the credit risk is with the bank itself, not the individual clients), and means that if I break-even, or take a few hits on my P2P lending, then in the long-run this all cancels out, and I might cover inflation hopefully. Such fixed-term deposits might be considered an alternative mechanism for microfinance lending, but it is relatively undiscovered in the mainstream retail sector, also invovles transaction costs, and requires more paperwork to set up. And of course, it doesn't harness the feeling (imaginary or otherwise) of P2P lending.
[I posted this on the MMC website, should have posted here! Mildly edited to relfect context here]
I get bogged down in such interest rate debates, so decided to put my money where my mouth is and lend on Zidisha - MY loans, MY transactions, MY calculations. I uploaded $1000. 18 months later (in fact slightly less as I did this in two batches of $500, minor detail) my cash + pending loans + outstanding capital is $1006. For all practical purposes let's say I am at break-even. I made some money in interest, however calculated, and I lost some money in defaults, late payments and foreign exchange losses. Overall these cancelled out (in fact I am $6 up). My average interest rate that I charged, weighted by the amount I bid, was 4.4% (flat per year I believe). I had a few late payments, one outright default, and I have no idea how much forex losses cost me. To repeat, this all largely cancelled out.
I then looked at the average interest rates as stated by Zidisha only for the clients I had lent to. These were 9.31%, but included the 5% fee that Zidisha charged, so it appeared that the average investor was charging 4.31%, marginally lower than me. Indeed, I lent to a few people who were unwilling to pay any interest, and sure enough their stated rate was 5% - the Zidisha fee alone. So, in terms of Zidisha claiming the average LENDER interest rate is 5.3% seems reasonable from my personal lending experience (40 loans to date).
I agree that flat rates are inferior. I wish Zidisha would stop this practice. And it is a fair rule of thumb to double them to get a real APR. There is a fragment of truth in the claim that borrowers understand flat interest rates better than APRs, as these are still common in some countries (where they have not yet been outlawed). Claiming the world was flat a few centuries ago was acceptable and commonly accepted, although wrong! I agree with the author that converting to APRs would be better. But, I also agree that the one-off fee for a credit check, in this lending model, does seem reasonable. But I concede that this is a debatable point.
So, excluding this one-off fee, it does appear that the loans I have personally done have an APR of about 20% (9.31% x 2). What's more, by me charging 4.4% (flat, equivelent to 9% APR), this has covered forex losses and defaults over an 18 month period, almost perfectly (by coincidence). I don't lend on Zidisha to make money, but if I can protect my capital, that is fine by me. This is what most other P2Ps will try to offer. MyC4 offers a net return, Kiva is generally break-even. What fascinates me about Zidisha is that there is no intermediary, and the rates do genuinely seem lower. I accept this might not be the case for a first time client on a $50 loan having to pay $12 for the credit check. The one-off fee is the source of the problem. But where do we draw the line - what about the bus fare to get to the office? The cost of completing the forms? The opportunity cost of time in completing the Zidisha process? Yes, there are entry costs to join Zidisha, as there are in many services. Indeed, one could argue these present a barrier to entry to dissuade non-serious potential borrowers.
Do not mis-understand me, I am a fanatic for transparent pricing in microfinance.
In fact, I should also add that there is an additional fee which I (i.e. from a lender perspective) have to incur that wipes out my measly $6 profit - the PayPal fee, which was $34 in my case. So, in fact, I lost $28. But, a rate of 20%, or 25%, or 30%, is alas pretty reasonable, particularly in Africa. I agree that Zidisha should adopt APRs as soon as possible, but I would be hesitant to describe this as deceptive. There is no pre-funding, at least they make an effort to state the interest rate, which some P2Ps don't even attempt. I do hover the mouse over the blue buttoms and was aware that this is flat, and I know how to interpret this, but I may not be typical. But compare this to Kiva, whose greatest effort to explain an interest rate is to state the self-reported, unverified portfolio yield of the bank as copied from the MixMarket often years out of date, and this is not even a good proxy of the APR in my opinion. I did a blog post a year or so ago comparing the stated portfolio yields reported by Kiva compared to the actual APRs calculated by Chuck Waterfield, and the divergence is staggering. Is Zidisha perfect, no? Is it an interesting development, challenging the status quo of the current P2P market? In my opinion, yes. There is scope for improvement, and I hope they constantly remain aware of this, but so far I find this a promising venture. It will be interesting to see how it scales up.
I am not sure what to make of this comment - are you suggesting that Kiva is a "poison on humanity" for doing this, or that I am for saying it?! Please clarify! For further reference on this specific point you might find Dr. Phil Mader's post, which I reference, useful:
This article does not proove Kiva's interest rates are higher than those stated, as we don't have Kiva's interest rates (for some reason), but they proove that actual APR interest rates are higher than the portfolio yield. This is ALWAYS the case, because portfolio yield does not incorporate some aspects of interest cost. It is not merely that these 10 banks demonstate this phenomenon, but EVERY bank has this, and it is generally worse in Africa where practices such as forced savings are more rampant. Also, you suggest that this might be okay because Kiva loans are lent at lower rates that the "regular" loans. How can we assume that? The banks are under no obligation to do that. The loans were made some months ago, from their regular loan portfolio, and then the Kivans come in and retrospectively buy this loan from the bank, in effect. What you are suggesting is that the bank gives out certain loans at a cheaper rate than its regular loans, and then hopes that the Kivans buy those specific loans from the bank. I find this extremely unlikely and have seen no evidence of this, either in the field or on the Kiva website.
But, perhaps we might not agree on this, but I am deeply worried about the use of the portfolio yield rather than stating an actual interest rates. In part this is because I think the poor are paying more than we are being told. Also I find this poor transparency, and Kivans are offering their money interest free, the least Kiva could do is be open with them. Other lending platforms are perfectly willing and able to do this. I also find this a worrying sign regarding Kiva's ability to control and monitor what is going on - if they can't even report an interest rate how good are the internal controls. They have had a number of "problems" with rogue banks - this doesn't surprise me given what I see of their controls. But most interestingly perhaps, this is perhaps a function of a flawed business model. It is not that I think Kiva CHOOSE not to publish these rates. I believe they simply CANNOT, and that is more worrying.
Kiva is, in some regards, similar to a retail bank (this is an analogy, not literal). It takes money from some people and lends it to others. Obviously there are major differences, but in essence these are the two key transactions - get money from A (you), lend to B (poor person) via agents (banks). It intermediates. And yet it doesn't know the interest rate in what is meant to be a P2P model? It can get you the photo and the story and some of the loan details like amount and number of months. Isn't this a little strange?
However, broadly I think we agree here - this is ultimately about transparency and reporting obligations, we disagree on the severity of the problem. But I would urge you to look into this further.
Another interesting comment is that you suggest change you personally would make - improved reporting, greater transparency, improved control, more accurate portfolio yield (by which you are referring to some proxy for interest rate) etc., good points. What is interesting is that there are other institutions already out there that have managed to overcome these problems, albeit to varying extents, and thus demonstrated that these are not deal-breakers, but relatively simple hurdles. Kiva has substantial funding and is heavily subsidized by volunteers, and yet can achieve none of this. And don't forget, it is not even as complicated as a bank. The most expensive stage of the chain is not collecting money from US citizens via PayPal and a website, nor sending money over to the banks in other countries, but doing the actual loans to the poor people in the field, and Kiva doesn't do this final stage, so has no actual costs of due diligence of clients, visiting their businesses, completing the forms, paying the loan officers, collecting repayment and dealing with non-repaying clients. It is only dealing with the relatively simpler and less onerous parts of the chain of money from your pocket to the pocket of the end client, and yet cannot even do this well.
It is interesting to see how readers come away with different gists of this article. I do think that Kiva could be a lot better than it currently is, and other similar companies have addressed some of the issues, only they are not as well-known, perhaps because their PR/marketing is not as effective. See some of the later comments in the Next Billion article when Zidisha is mentioned, for example.
But I think you touch on a CRUCIAL point - it helps our conscience. Phil Mader's original post, which I reference, is well worth reading:
It is a short piece and eye-opening. I cite a phrase from this article in my piece:
Kiva enables us "to consume the feeling of charity without financial loss"
To say that Kiva simply "just doesn't work" is tricky, because we have to define "work". It depends what the objective is. If the idea is to make people feel that they have done something useful with their $25, which they get back a few months later, it works wonderfully. If the idea is to efficiently transfer spare capital in "rich" countries to "poor" countries (forgive the simplification of terms), it doesn't work very well, as there are far more efficient mechanisms. If it is to provide low-cost funding to entrepreneurs, then it has pretty mixed results which are hard to verify as Kiva (mysteriously) don't publish the actual interest rates. In some cases you can find them out though.
In summary, you are very wise to suggest that an endorsement requires you to be convinved that this is a good business. It appears Kiva did not meet your standards, and I am inclined to agree with you.
These are valid points. If I could focus on just one issue of legality, what about insisting that child labor laws are obeyed? I know you could question this on the basis of one person is pro-hunting, another is anti-hunting, so it is subjective. But I think most people are not in favour of child labor. If we could just insist on one law, would this be one that is worth Kiva incorporating? It wouldn't be perfect, monitoring is a problem, definitions of child labor vary from country to country, but perhaps a mild attempt to at least only partner with banks that have a stated policy of obeying local child labor laws with their loans would be a good start? Some microfinance players do have such policies, one of the best is Oikocredit, but in fact many have these policies. Would that be one law that is worth insisting on?
Kiva does not state interest rates. You are viewing portfolio yields, which are not interest rates and consistently underestimate the actual rates. For a post specifically discussing this topic with 10 examples using actual data comparing actual interest rates versus the portfolio yield stated by Kiva, see:
If you want to see an MFI with rates approaching 100%, look at Mexico in particular. Before the MFI quit Kiva, Brac in South Sudan had a portfolio yield of 88%:
This MFI quit Kiva, but you can see it here: http://www.kiva.org/partners/107 note that Kiva now states the portfolio yield is "only" 69.03%, but you can't lend through them anymore as they are one of the many MFIs that have quit Kiva.
Portfolio yields do not include the impact of forced savings, some fees etc. and can be manipulated, and for a discussiong on why these are absolutely critical in understanding how much a loan costs, see the excellent website www.mftransparency.org, an NGO dedicated entirely to revealing the actual interest rates charged by microfinance banks, many of which exceeed 100% (but are not all Kiva partners). It is not possible to generalise about the ratio between the portfolio yield and the actual interest rates, but as a rule of thumb, a yield over 50% is likely a little too close to an actual interest rate of 100% for me, but this is a very broad rule of thumb that I use if I can't get more accurate data.
But it is good that you look at this, albeit flawed statistic, and try to only make reasonably priced loans. If everyone does this then the MFIs offering lower rates will receive more funding, encouraging those charging high rates to reduce them. Competition basically.
Please do not think this is a trivial detail. As the post above demonstrates, this can make a big difference in the overall cost of the loan that the poor person actually pays, so if you are trying to avoid exploitative interest rates you really must consider this.
But, the more interesting question that is harder to answer, and has far reaching implications, is why Kiva does not publish these rates. See the comments on the Next Billion article. Is Kiva unable, or unwilling to do so? They manage to get all sorts of other information - why not the interest rate? How hard can it be? The bank presumably knows this crucial piece of data, and yet this is replaced with a knowingly flawed portfolio yield statistic. Why?
"On current evidence, the best estimate of the average impact of microcredit on the poverty of clients is zero"
http://content.time.com/time/world/article/0,8599,2103831,00...
Needless to say the debate rages on. The Zidisha discussion detracts from this central point, but on valid grounds. To some extent the shortfalls of microfinance may be due to the high interest rates charged, the restrictive loan conditions, the profit motivation, the aggressive intermediating MFIs etc., some of which Zidisha addresses. However, if you read the work of Milford Bateman, for example, he raises more fundamental questions about microfinance. In the P2P space profit is not really an option, and in Zidisha's case it is debatable whether one can even break-even. But look at the higher levels, the profits accumulating to Accion in the IPO of Compartamos; the millions made in the IPO of SKS, or at some of the microfinance investment funds - these are the big boys, and at that level yes, there are vast profits to be made. Motivation is a hard thing to prove, but you are right to suggest that profit has come to play an unhealthy role in the debate. The sector is extremely opaque, transparency is rare, there are rampant conflicts of interest and ill-aligned incentives. It is largely unregulated in practice, despite endless window-dressing to the contrary. And when some new start-up promotes microloans for poor people to connect their houses to the drinking water supply, as sensible and this might superficially seem, isn't this a public good that the government should provide?
Reading the standard microfinance mantra one could be forgiven for concluding that all the world's problems can be solved with loans, and that profit maximisation is the best structure to arrange this. We are familiar with the theory, but the evidence is sorely lacking.