Independent Research and Policy Advocacy

Why small is not beautiful when it comes to savings?

Save Post

A recent article in the Economist notes approvingly about the growing phenomenon of Village Savings and Loans Association (VSLA) as a means for low-income clients to save securely and earn high returns. In brief, the mechanism entails a group, typically at a village level, pooling their savings together and lending out of this corpus to each other. The underlying principle is similar to that of chit funds and ROSCAs.

This whole design of VSLAs is missing a very fundamental point about the way in which savings operates and the value of financial intermediation. When an individual saves with an institution, she takes risk on the solvency of the institution (unlike in credit where the institution takes risk on the individual/borrower). She should be able to withdraw her savings at any time which in turn entails that the savings has been channelled and managed in a way that enables the institution to honour this. A VSLA tries to replicate this at the level of a village/community. Now imagine there is a drought in the village and there is a widespread need for liquidity. The savers will want to withdraw their money, but equally the borrowers will find it a hard time to make their repayments. Both groups are impacted by the same event (a covariate shock). Effectively, there will be a “run” on the VSLA. The VSLA, unlike a bank faced with a similar situation, does not have the diversification or the deep pools of capital to simultaneously bear the default and honour the savings withdrawals.

There is no question that access to traditional savings accounts is hard for low-income clients in all countries and innovations like the VSLAs are well-intentioned attempts to address that. However, we need to ensure that the financial inclusion designs we choose to scale do not end up creating newer types of systemic risk. The Business Correspondent approach, despite several tactical challenges, scores very highly on this count – it leverages the access of village level institutions/agents but the savings is channelled into the formal financial system through a bank.

Authors :

Tags :

Share via :

11 Responses

  1. This is an excellent note.  I also feel that given the state-of-the-art of technology in Africa and Asia where the mobile network as become an effective communications network for financial transactions and the mobile phone a low cost data entry terminal, it is no longer, at least for techological reasons, to continue to create these discrete, very high risk, savings pools. 

    It is my understanding that the Self-Help Group in India is not legally approved as a savings deposit taking institution (because as a quasi-bank it would run afoul of the Banking Regulation Act if it were) but as a quasi-equity pooling vehicle (with fewer than 50 members in a self-help group it effectively constitutes a private placement of equity and not a public issue thus not running afoul of Securities Regulation) and reflects the true nature of the underlying risk. As the Blog Points out models in which either well capitalised commercial banks through their Business Correspondents or SEBI regulated Money Market Mutual Funds through their Distributors or PFRDA regulated Pension Funds through their Accredited Intermediaries collect savings, has the effect of offerring both deep penetration and high levels of safety.

  2. In addition, while credit has come under a great deal of fire as having a causal linkage to severe financial distress, it often constitutes a savings tool with an extremely high degree of pre-commitment (for example mortgages are often viewed in this manner even by high income households).  It also has the characteristic that the financial institution or the Self-Help Group or the Local Lending Association is taking the risk on the individual honouring her commitment to “save” rather than the other way around.  From a safety and spread of access perspective this “pure credit model” then represents a superior design to that of VLSAs.

    There is of’ course the very grave concern with credit (or any other type of commitment savings product) that households with highly uncertain incomes or with steady incomes but no surpluses may not be in a position to honour those commitments.   In some cases financial savings may even be undesirable, particularly when they take the place of consumption expenditures such as good nutrition and schooling. A high degree of pre-commitment as implied by fixed loan repayments or recurring deposits, can then be very harmful.

    An ideal solution therefore is one in which both strong-commitment savings products such as credit and recurring deposits are available alongside flexible savings products such as high quality savings bank accounts and money market mutual funds, combined with customised wealth management guidance for each household.  The KGFS model of IFMR Rural Finance is an effort to create exactly such a model.

  3. Agree with the views on the Village Savings and Loan associations (VSLA). However, the issue here is whether poor villagers should be in standstill mode while they are waiting for financial inclusion to come by their way. The VSLA  may not be ideal but possibly is a trade off solution while they wait for  the better models (as proposed) to become available to them. Can we help them improve their model and mitigate some of the risks until then?

    Incidentally – where can I get access to a copy of the KGFS model of IFMR Rural Finance. Thanks

    1. Vivek, thanks for your comment and I fully agree with your point on urgency. I am concerned that a few VSLAs failing will set us back on the overall financial inclusion agenda though, with people concluding that this is all too risky. The way to mitigate VSLA risk is to make sure they are transferring risk to an entity that has the capital and size to manage it better. There are several good models out there as Nachilet mentions in his response, I think a debate on which models meet the goals of inclusion and financial stability is important which is what I wanted to achieve through this post.

      You can read more about the KGFS model here:

  4. Bindu:

    I don’t think anyone would argue that poor people don’t need a safe place to safe and that covariant risk is among the risks savings groups present. However, I think those of us seeking to support the poor need to recognize savings groups–SHGs, VS&Ls and their variants–for what they are, a hugely popular ‘good enough’ solution for hundreds of millions of poor people in rural areas that banks and microfinance institutions are unwilling or unable to serve. Until there is a meaningful alternative that the rural poor recognize as valuable, arguing that savings groups aren’t what financial services *should be* seems a rather academic point.

    Sean Kline

    1. Thanks for your comment Sean. I am disagreeing that this is “good enough” a solution for such an important problem. The bare minimum definition of a good savings product for me is that your principal is safe and available “on demand” (Inflation protection and similar features are highly desirable but arguably not bare minimum). The VLSA design does not guarantee this bare minimum because the savings is pooled and intermediated locally and therefore puts savings of poor people at risk. If we insittutionalise this bad design, where will the pressure for meaningful alternatives come from? 

    2. Sean, I know there is a tendency globally to count SHGs into the savings groups, perhaps to bump up the numbers. But nothing could be farther from the truth. Joanna and I visited some of the most acclaimed SHGs and their federations to find out for ourselves. While there is a growing literature in this regard now, my assessment is as follows:

      (1) SHGs do not represent a model. It means different things to different people, and it has evolved a lot since it was first initiated. I asked Al Fernandez of MYRADA if he was proud to see the scale of SHGs today. He was very sad to see the way it has evolved. He went on to say that the savings function of the groups has largely been compromised.

      (2) SHGs are today “credit-led” models, where the focus is on lending, while the VSLAs/ savings groups are “savings-led” models, with an emphasis on individual savings. C S Reddy of APMAS believes that the savings groups are “setting the clock back”, by not allowing the credit functions (presumably a higher order of financial service) to emerge strongly. However, listening to Prof Yunus in Delhi several years ago, it is clear that Grameen Bank could not have been what it is today without accessing  savings of the poor. He went on to say that the paranoia of banking regulators in India towards this was stifling economic development in India, particularly where it matters most. People need saving instruments in India, and these are woefully lacking. Post-office and LIC has done a bit, chit funds often flourished under such conditions (such as the recently closed Sahara) and then there is the “friendly neighbourhood” (whichever way it is used).

      (3) I dare say today that SHGs are little more than vehicles for expanding political patronage, particularly evident in Andhra Pradesh where state patronage of SHGs and their federations happened at the cost of constitutional bodies. Linkages with banks, particularly those managed by the State, focus on distribution of increasing volumes of loan. The focus is on repayment, rather than livelihood generation. Often this repayment is from another, larger, loan, rather than from any profitable enterprise. When the bubble gets too big, it is written off.
      While this may provide political dividends but severely dents the economy at that level, making poor rural communities ever more dependent on the state for subsidies and hand-outs. VSLAs/ savings groups are a threat to the perpetuation of this vicious cycle of poverty-handouts-votebanks by freeing the poor from external patronages of any kind by truly promoting “self-help”, an expression that is not available today for proper use. I think that is the reason why it will also not find a true champion.Let’s meet up if you happen to pass through Delhi. Regards, Somnath

  5. This is a classic conversation among individuals who are desperate, and even sensing a threat! I am neither a banker nor a finance wizard, and hence have no hang-ups on either the idiom or the institution. However, I did experiment with VSLAs, perhaps the only example in India till date. I had to convince a very large number of similarly opinionated colleagues, motivate a string of field staff who only understood SHGs and even convinced the Indian High Commissioner in Bangladesh to provide a work permit to a Bangladesh national to come over and help us launch the program in Bihar. Today, there are over 1,500 groups spanning across three districts, including peri-urban, large villages and remote rural areas of flood-prone north Bihar. More than 1,000 groups have completed their “share-outs” and are on their own and the program is now getting more ambitious by trying out a Village Agent model in which the beneficiaries pay for the training! Yes, at my personal request, Vijay Mahajan has also visited some of these groups and he was surprised at the individual corpus created at the end of the annual cycle and felt 5% of this could actually be paid back in the first year to the facilitator/ trainer.

    Having nurtured this model in India, in the face of intense pressure from the World Bank sponsored BRLP program trying to replicate the AP model, I am today fully convinced of its role and utility, certainly till such a point when each individual in India becomes “bankable”. Its greatest strength is its simplicity, something that allows the poor to understand it and take charge. (For instance, I did not understand half of what Nachiket says above, and I think the onus is equally on him.) The second feature of savings groups – this is the official name for VSLAs now – is its flexibility for individuals within rigid, minimalist rules of the institution that is understood by all. Individuals can plan their savings (particularly in an environment of intermittent remittances). The corpus available to them after a year is planned, and used, for agriculture, housing improvement or life-cycle events.

    The author talks about catastrophe as an imaginary phenomenon. We worked in north Bihar where floods occur with devastating regularity. The groups not only continue but find it a great support under conditions of extreme stress. A study in Nepal on the WORTH program (VSLAs) suspended for eight years during the Maoist insurgency revealed that the number of VSLAs remained almost the same. True, some had disappeared but there were others that had spontaneously emerged. Digging a little deeper, I realised that such informal groups existed in the past among extended families (when banking institutions were even less pronounced) and there was a natural tendency towards adoption. It not only provided an avenue to smooth incomes but also provided an opportunity to bond with neighbours and members of the extended family. This is a form of risk-sharing among the poor that is not understood by the people who are used to “relationship managers”.

    Finally, I have often pondered the issue of using institutional solutions for the more economically advanced on those who are typically below the poverty line. If we visualise a graph with increasing wealth on the x-axis and debt as a proportion of income on the y-axis, I believe we should get a U-curve, the poorest and the richest being most indebted as a proportion of their incomes, while this should be typically lowest for the middle-income groups. However, while we would consider the ultra-rich to be “highly leveraged”, we would not use the same terms for the poorest for obvious reasons, and hence the institutional management that is available to the richest does not hold true for the poorest. Or, maybe the institution is different – not a bank but the extended family (does it explain why castes continue to be important for the poor?).

    Yes, I am with Vivek. We need to modernise and change society, and one of the indicators will be a bank account for every individual that is active (and not like the BASIX experiment in Rajasthan). But until that happens there has to be a way to provide support to the poor. It may not be paying for any institution but there are huge payoffs for society at large.

    1. Dear Mr. Bandyopadhyay, thanks for your detailed comment and your work in North Bihar sounds very impressive. My intention is not to take away from the efforts or achievements of people involved in the SHG/chit funds/savings groups. Surely, this has proven that there is a vast unmet need for good flexible savings services and puts more pressure on Regulators to address this need.

      I am questioning the robustness of the design of the VSLA/savings group programme. Informal savings and risk-sharing arrangements, as you point out, have existed for centuries. However, as a policy objective, why do we think that this is not good enough and we need “formal” financial services? Informal arrangements often have the benefit of flexibility and simplicity but are unable to provide protection against covariate shocks.  A community-based savings or insurance program suffers from the inherent lack of diversification and risk pooling that gives insurance and savings its true power. So, it might be easier for me as a poor person to buy livestock insurance from my friendly neighbourhood NGO than a formal insurance provider. However, when there is an epidemic in the village that impacts all livestock, the community based program cannot honour claims because it effectively goes bankrupt. What is the point of the insurance cover if it does not pay out when I need it the most? Designs that combine the approachability and simplicity of the village group with the diversification of formal institutions seem like a great way forward. So, in this example, the community based insurer could reinsure its portfolio with an insurance company that allow it to pay out when there are village level shocks. There might be something similar that a savings group can do to diversify.

      Until we have good answers to these fundamental issues, it would be dangerous to scale up these designs. Small pilots are fine but you will agree that is not going to make a dent on the problem at hand.

      1. Dear Ms Ananth: I am indeed glad that the concept of savings groups is being discussed by leaders among mainstream financial institutions. May be this stems from a desire to unlock the proverbial “fortune” at the bottom of the pyramid. Personally, however, I believe expansion to emerging markets is a win-win approach. My experiments with savings groups was for the purpose of development at scale, not a small philanthropy. We worked on improved cost-effectiveness, full-cost recovery models and, as I alluded earlier, a completely self-sustaining model through “village agents” that are fully paid by the beneficiaries. The idea was also to introduce other market-based approaches to livelihood enhancement on this initial platform, which, we believed, provided an improved social and financial capital.

        Importantly, we focused only on savings and, although the VSLA model provides for a small risk fund, did not introduce the insurance element in order to keep it simple. In any case, people get back a substantial sum within a year. Also, in case of emergencies, we’ve seen that the social capital generated through the process enables individuals to borrow outside the group activity which may or may not be linked to the individual’s fiscal prudence. Large shocks, of the kind you are alluding to, is certainly beyond the control of poor communities and need public intervention. But the more regular, day to day shocks to individuals (such as the loss of wage) or need for small expense (tuition fees for a child, for example) can be managed from this service. Very often I’ve heard the argument that poor need a lot more money in order to meaningfully derive a livelihood from an asset. Talking to beneficiaries, on the other hand, provides a very different picture. They often mention that loans from micro-finance agencies might help them procure a cow, but the cost of fodder, vet and other services often come from the savings groups. The point I’m trying to drive at is that the savings group mechanism can, and will, co-exist with other forms of financial services. Another example is its seamless existence with the SHGs in Bihar. People are savvy enough to form SHGs to claim their share of government subsidies while continuing with their own savings groups for the small, daily needs that are met through it. Adoption, and retention, particularly beyond the period of external intervention, is the biggest proof of its effectiveness.

        Another important gain from the savings groups, we realised, was “basic financial literacy” obtained through a very hands-on process, usually over a year (52 weekly meetings). They understood the principles of savings and lending, the concept of interest and the need for timely repayment, and developed a basic discipline for savings. Admittedly, these are largely women members, but as and when they begin to deal with mainstream financial institutions, I’m sure these lessons would come in handy not only for the people but also for the institutional service providers. We even tried to add on a separate “financial literacy” package to the basic VSLA training module since the market for financial services at that level in Bihar was simply ruthless and skewed entirely against the interests of the poor and uninformed. Every village, every household had a horror story of duping, and yet they were so keen to save.

        You are questioning the robustness of the design, which is what I see as its greatest strength. Its strength is its simplicity and adaptability. Its strength is its ability to co-exist with other programs, its ability to reinvent itself as per the demands of the context and its ability to work “under the radar”. On the other hand, the large financial institutions are often like large animals (think tigers, dynosaurs) – powerful but extremely vulnerable, as recent economic events in the world testify. I am a biologist/ ecologist, and hence I see things differently when it comes to resilience and adaptation. If you study the mass extinctions in natural history, it is the smaller animals (typically rodents living inside caves) that survive extreme climatic change and develop their own lineage, including the larger ones.

        Finally, I am not dealing with savings groups any more and have moved on to providing safe drinking water through market-based approaches. Still, I think there is something in this intervention that is of great value to India as it transits from a traditional mould to a new economy. If we want to bring financial inclusion for 70% of our population within the next two decades, there is need for all to join hands and look at every opportunity. There may not be enough returns each quarter for every investment, but we may be avoiding economic (or even social and political – think of the Maoist insurgency) disaster and laying down the foundations for a more robust economic growth and prosperity for all.

        I suggest you send a team from your institution to Bihar to take a look at the program. Maybe you will find your own answers from a more thorough study. In case you approach this as a researcher – that is, with an open mind – you’ll find my whole-hearted support. Best regards, Somnath

  6. SHGs are now 20 years old institutional arrangement in the country. During the last decade, it was ignored at the cost of promoting MFIs. Post AP crisis, many of us have started finding  a great virtue in SHG linkage banking again. See the way State of the Sector Report 2010 and 2011 have reported about SHGs.  I would also like to make a small correction to what  Mr Nachiket Mor has observed about the savings of SHGs and the number of members etc. RBI as early as 1991 /92 has recognised the way savings (rather thrift which for the poor is defined as ‘expenditure foregone’ and not any surplus from out of labour / wages or enterprise operations) are made and pooled by the SHGs and parked with the banks. So, there is very little risk about the loss of money saved by the poor so long as they  get linked with banks. NABARD has also made it through their policy guidelines that a group not exceeding 20 members only should be formed as otherwise it would attract some of the provisions of the Company’s Act. As such, all SHGs are supposed to have a maximum of 20 members only.  But, one should also appreciate the innovative way KGFS has been spearheading new generation financial institutional model. Unfortunately, in India, no one wants to learn from others. Even though hundreds of MFIs are there in the country, no MFI will follow the good features of another MFI and recognise such contribution. 

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Posts :