Muhammad Yunus on the microfinance collapse in A.P.

Along with commercialization of credit, lack of savings accounts with the organizations may be one the reason for the failure; see Sacrificing Microcredit for Megaprofits :
"Grameen Bank, where I am managing director, has 2,500 branches in Bangladesh. It lends out more than $100 million a month, from loans of less than $10 for beggars in our “Struggling Members” program, to micro-enterprise loans of about $1,000. Most branches are financially self-reliant, dependent only on deposits from ordinary Bangladeshis. When borrowers join the bank, they open a savings account. All borrowers have savings accounts at the bank, many with balances larger than their loans. And every year, the bank’s profits are returned to the borrowers — 97 percent of them poor women — in the form of dividends.

More microcredit institutions should adopt this model. The community needs to reaffirm the original definition of microcredit, abandon commercialization and turn back to serving the poor. "
He also has suggestions about regulation.
P.S. There is some criticism of the above By David Roodman in Professor Yunus’s Opinion :
"So if all the microcreditors could raise all their capital from their clients then, I agree, that is the way to go.

But almost none has—not even the Grameen Bank.

Yunus’s achievements should not be slighted. In its pioneer days, the way forward for the Grameen Bank was far tougher and more uncertain than for those that followed. That said, thanks to his pioneering status and his abilities as a salesman, Yunus had help: a couple hundred million dollars from the Ford Foundation, the United Nations, Japan, and Western donors. As Vijay Mahajan, the father of commercial microfinance in India, has pointed out, microcreditors today cannot expect the same help, whether because of limited funds among private and public donors or the donors’ sense that microcredit has graduated from grants. If microcreditors today want such big chunks of capital from outsiders, they will have to buy it.

In particular, while it is true that Grameen members hold legal claim to 97% of the Grameen Bank’s net worth, they only contributed about $7.5 million in capital, at 100 taka ($1.40) per member. The Grameen Bank has not shown that microfinance can grow large purely through cooperative ownership.

In fact, as I wrote last summer, an irony in Yunus’s criticism of for-profit microlenders for going to the capital markets is that Grameen Bank is itself running low on capital, by which I mean risk-absorbing, profit sharing funds that banks are required to keep on hand in case of losses. And it is not clear (to me at least) how Grameen will get more. Maybe the government will step in…"

Mahahan said in Vijay Mahajan speaks about new economics and micro-finance
"Microcredit is a necessary but not a sufficient condition for microenterprise promotion. Other inputs are required, such as identification of livelihood opportunities, selection and motivation of the micro-entrepreneurs, business and technical training, establishing of market linkages for inputs and outputs, common infrastructure and some times regulatory approvals. In the absence of these, microcredit by itself, works only for a limited set of activities – small farming, livestock rearing and petty trading, and even those where market linkages are in place.
.....
Moreover, there is serious evidence that like all other “single” interventions, microcredit works less well for the poorer clients. As David Hulme and Paul Mosley have shown in their important work Finance Against Poverty (Routledge, London, 1996), the increase in income of micro-credit borrowers is directly proportional to their starting level of income - the poorer they were to start with, the less the impact of the loan. One could live with this finding in an imperfect world, but what is really troubling is that a vast majority of those whose starting income was below the poverty line actually ended up with less incremental income after getting a micro-loan, as compared to a control group which did not get the loan. This should stop recent converts from offering microcredit as the solution for poverty eradication, since it can do more harm than good to the poorest."

The other interventions that Mahajan spoke of seem to depend on the context. From Profile of Social Entrepreneurs:
"BASIX, the brainchild of Mr. Mahajan aimed at livelihood promotion, rather than microfinance (microfinance being only one tool he uses to reduce poverty and create economic opportunities for the poor). His organization, BASIX, assists a third of poor rural households directly using micro-loans and importantly, companion savings programs, insurance plans, and agricultural and business development services. The other beneficiaries are supported by over 100 citizen organizations (COs) or community based microfinance institutions to which BASIX provides assistance in funding, training, development, and operations."
The reasons of scale of operations seem to have made Mahajan convert from NGO to a company. From Vijay Mahajan on the Risk of a Microfinance Bubble
"I ran BASIX for almost ten years as an NGO. But as an NGO you have a lot of limitations in terms of how much money you can raise, what scale you can achieve - and banks don’t lend to NGOs easily. And so we decided to run BASIX as a company and be financially sustainable and indeed, then we also applied for a banking license and part of BASIX is also a bank now. So, it all evolved in pursuit of the goal. But the goal remains to help poor people to achieve a life of dignity and a higher quality of life."
In the same interview Mahajan said "We need to distinguish between investors and microfinance entrepreneurs. And in both categories, there are those who are doing it for a social motivation and those who are doing it because it’s a good business proposition. So, the sector is a mixture of these two. You will simultaneouly have MFIs which barely make a small profit and their investors make no return. And you’ll have other MFIs which are making a lot of profits and therefore are able to reward their investors. But the sector will settle down to a middle path because if you make too little profits, you cannot grow and will have a problem with capital and attracting loans. If you make too much profit, almost usurious, then eventually someone will raise their voice against you and you will have to stop that."
Mahajan's advice for avoiding the bubble have not been heeded "To avert a bubble, lenders and equity investors should be careful not to ask for growth rates which are unrealistically high. Demand is very large. But the supply should be limited to the institutional capability to do it well. You can’t increase that capability by 150 or 200 percent per annum. You can do that in the early years but you can’t do it later. So, investors have a role in saying “okay, cool it. Whatever the locally appropriate growth rates are, just stay with that”. Then regulators have a role. And then the managers in those MFIs have a role in ensuring that the incentive structures for the field staff are designed in such a way that they engage in prudential lending and are not just throwing the money away. They should be incentivized under recovery rate and not just on how much they disburse. Through all these means together, the chances of a bubble can be significantly avoided." And added to that are competetion between MFIs, multiple lendings piggybacking on the existing SHGs. It seems that only that many players did not have the original mission of helping the poor in mind but entered a period of disaster economics as in the case of subprime lending. Finally Mahajan says in the first interview quoted above:
"Recent studies by CGAP show that only about a 100 of the 10000 odd MFIs round the world are financially self-sufficient. Thus the dual promise that microcredit is able to serve the very poor, and in a financially sustainable manner, is not borne out in practice. Experience shows that either one of these two mutually contradictory goals can be achieved, but not both together."