Where Angels Prey

Where Angels Prey is a novel by Ramesh S Arunachalam. Please refer to www.whereangelsprey.com for more information

Tuesday, December 7, 2010

Understanding Micro-Finance Interest Rates: Some Fundamental Issues For The RBI Sub-Committee to Consider...

Ramesh S Arunachalam
Rural Finance Practitioner

There has been a lot of debate about interest rates in micro-finance and rightfully so and in this post I look at what I see as some of the fundamental issues in this debate...I hope the RBI sub-committee looks into the key issues highlighted here...

Let us start with the different costs that are incurred in delivering financial services to low income people. I see four major costs and they are:

1.      Transaction/Financial Costs[i] Institutions        =          TFC Institution          
2.      Transaction/Financial Costs[ii] Intermediary       =          TFC Intermediary
3.      Transaction/Financial Costs Clients            =          TFC Clients
4.      Transaction/Financial Costs Total             =          TFC Total

TFC Institutions   + TFC Intermediaries  + TFC Clients = TFC Total         (This is the Basic Equation)

Premise # 1 - Total Costs (TFC Total ) and Apportioned Costs(TFC Institutions   + TFC Intermediaries  + TFC Clients) of Delivery Will Vary Across Models, Contexts and Related Parameters: First, the total and apportioned costs will vary by:

ð   the model used - group versus individual, bank/MFI branch based Vs centre meeting based Vs agent based,
ð   the number of years in operation and life cycle stage of the different channel partners – the experience/learning curve aspect should reflect here,
ð   the economies of scale and scope available including aspects of fixed/variable costs – enhancing major product outreach to larger number of clients as well as offering them a range of other products as well,
ð   the trade-off between risk, efficiency and controls in deliveryefficiency can be gained by reducing controls by that level of risk will have to be tolerated,
ð   the product strategy - in terms of savings alone, credit alone, savings + credit + others including risk management products,
ð   the number of channel partners (or intermediaries) – this decision is very critical to the cost and more intermediaries should just the additional costs
ð   the strategic context - including clients, geography, etc,
ð   the basis for the competitive strategy - in terms of differentiation vs quality vs cost leadership (this can also be thought of the overall strategy of managing total costs) and other such factors

Thus, we have the following:

Total and Apportioned Costs of Financial Services Delivery to Low Income People = Function of Model Chosen, Life Cycle Stage and Age of Channel Partners, Economies of Scale and Scope Available, Trade-Off Between Risk, Efficiency and Controls, Product Strategy, Length of Channel, Strategic Context and Competitive Strategy

And we need to recognise that all of the above are INDEED strategic choices exercised by organisations, depending on various factors...

Therefore, we must understand that it would not be appropriate to expect all institutions to be able to deliver financial services to low income people at the same interest rate/level of fees.

That said, I am not arguing for any (high) level of interest/fees to be charged from low income people – all I am saying is that, we need to be sensitive to the fact that, it may not be possible for all different models to come under a specific SINGLE rate.

The most appropriate strategy here would be to get answers to the following question for a typology of contexts, models and related parameters:

1. “Given a typology of contexts, models and related parameters, what constitutes the optimal range of interest/fees that need to be charged from low income clients to fully cover total costs?”

The above critical aspect is best understood through the following example: For a Grameen MFI 36% could be the Full Cost as it is operating in a hilly and difficult terrain and providing doorstep services; For an SHG MFI B, 22% could be the full cost as it lends directly to SHGs and thereby is almost a semi-wholesaler; For SHG federations or Cooperatives, 18% could be the full cost because they accept deposits which are the cheapest source of non-subsidized capital and so on. [There numbers given above are merely illustrative and they can vary from context to context and model to model – so, please do not join issue with me on this. Thanks] 

The above strategy would also be fair approach in my opinion and the RBI Sub-Committee must try and recognise this, study this aspect[iii] and make recommendations accordingly. Please note that my FIRST emphasis is on understanding what the full (total and apportioned) costs are and this could be very different from full (total and apportioned) cost recovery (taken up next).

Now, with the total and apportioned costs – for different contexts, models and related parameters - that need to be charged for a full cost recovery out of the way, let us get to next aspect...how to recover these full costs?

Premise # 2 Full costs are Always Recovered As A Combination of Interest/Fees and Different Kinds of Subsidies: This is again a strategic choice aspect and different models do it differently. There are two major ways in which this cost recovery can be handled: a) Apportion the same across the institution, intermediaries and clients – this is a creative strategy as it transfers the costs from one stakeholder to another; and b) Decide on the extent to which costs will be actually recovered and balance will be subsized.  We will look at each of these issues separately.

As you can see above, the full (total and apportioned) costs comprise of two portions – Recovered and Unrecovered costs. And the Unrecovered Cost contains different types of subsidies – Direct, Cross, Indirect and Hidden Subsidies.

Therefore, in some sense, there is always full cost recovery through interest plus fees and a range of subsidies provided. This is a very critical aspect to note and much of the arguments over interest rates can be better understood, if this crucial aspect is noted.

Therefore, the second strategic choice entails decision making within the organisation on: a) the extent to which full (total and apportioned) costs are to be actually recovered; and b) the different kinds of subsidies that (need to be and) are provided to cover the balance portion of total costs minus recovered costs (i.e., interest plus fees etc).

The most appropriate strategy here would be to get answers to the following question for a typology of contexts, models and related parameters:

2. “Given a typology of models, contexts and related parameters, what is the proportion of costs that are actually recovered (from clients etc) and how are the balance (unrecovered) costs met through various subsidies?”

The above critical aspect is best understood through the following example: MFIs perhaps charge 24-36% or more, recover all/most costs and sometimes even have a surplus. Banks charge 12% and cross-subsidize costs; SHG federations or Cooperatives charge 24% and recover full costs and Government programs perhaps lend at 3-6%, with the major costs being subsidized. [There numbers given above are merely illustrative and they can vary from one organisation to another – so, please do not join issue with me on this. Thanks] 

So, from the above discussion, it is clear that in some organisations, the whole cost could be recovered where as in others, there is only partial cost recovery and the rest is perhaps subsidized. MFIs perhaps charge what they charge because they have less of subsidies and practice door step banking; Cooperatives and Community models perhaps charge what they do because of using local and low cost staff and community for various aspects and also have access to savings (which is the cheapest source of non-subsidized capital); Banks charge what they do because of the norms set by regulators and supervisors and manage the actual (unrecovered) costs differently through cross subsidies, outsourcing etc; and Governments directly subsidize clients and charge as low as they do for various reasons

To summarise, the aspect of transactions and financial costs primarily centres around strategic decision making that organizations make on the following basic aspects.  What brings diversity in terms of the costs is the strategic choice that organizations exercise with regard to the following basic decisions:

ð   Whom to Serve? – Clientele, especially, with decision making on whether to serve the poor, not-so-poor, excluded, included, men, women etc
ð   How Many Clients to Cater To? Where to Operate? And How to Expand? – Outreach, Geographic Dispersion and/or Growth Strategy (Incremental, Quantum etc)
ð   What Specific Services to offer to the clients? – Products (financial intermediation encompasses a large number of products and combinations there of)
ð   What Methods of Service Delivery to Employ? – How to organize these channels like Groups, Individuals etc and their Tasks/Roles and outsourcing if any and the implications there of
ð   What Organizational Mechanisms to Use? – Legal/Institutional Forms
ð   How to Communicate the Availability of Various Services? - Promotion
ð   What the Medium/Long Term Objectives Are? – Single versus Double versus Triple bottom  lines?

And while, as noted earlier, full costs are always recovered from various sources, the strategic choices exercised to above questions result in some models choosing to recover costs fully from clients whereas other models may recover these partially from clients and cover the balance through different subsidies.

Therefore, it is humbly submitted to the RBI sub-committee that interest rates should NOT be viewed in rigid terms. They must be understood in terms of their broader context and its implications in terms of full cost recovery from clients versus partial cost recovery from clients plus subsidies. However, at the end of the day, there must be sufficient justification for pursuing either of the above strategies and that must be ascertained and understood...

And it goes without saying that, without understanding the above issues, condemning seemingly normal interest rates (I am sure we can discern exhorbitant interest rates straight away)  charged by institutions would be rather unfair and perhaps even unjustified...and it is sincerely hoped that decisions on (regulating) interest rates would be made only after undertaking a rigorous national study...encompassing alternative models in various contexts...and bench marking a range of interest rates for different contexts and models...And that alone will bring the interest rate controversy to its logical conclusion…

Tomorrow: Examples of how organisation’s reduce/transfer their (apportioned) costs including transactions costs

[i] Includes financial costs plus operational costs, loan loss provisions plus inflation adjustment etc
[ii] Same as above
[iii] I will also try and outline the methodology, for under taking such a study, to determine full costs for different models, in a separate post.


  1. Dear Ramesh,

    Excellent post; the only issue is we do not have adequate information on the total costs under different models and how they are met. In early nineties there were several studies on TCs. There is a near vacuum of information on TC on different models,in different contexts and in different cost recovery principles and practices. There is a need for a nation wide rigorous study to feed into policy.

    Warm regards

  2. Excellent post Ramesh. In nineties a few comprehensive studies were conducted on the TCs. Now there is near vacuum of such studies and this gap leads to comparison of apples and oranges. A thorough study on total costs and how they are met across different methodologies and different contexts is the need of the hour. This can help formulate policy. I would even go one step further; can we also combine benefits and impacts on clients to the scope of a nationwide study?It may be ambitious but there is very little information on costs versus benefits across various models.

    Warm regards

  3. I think u have made an excellent overview of the present malaise of MFIs and also brought out very vividly some of the pertinent areas that have been overlooked so far but need to be examiined thoroughly by all those who would like to see the revival MFIs once again. But I am more and more convinced that the present business model adopted by MFIs is not an approriate and sustainable model as it overlooks risk factor and also income generation factor and for that reason the present coercive method leading ultimate collapse of these institutions have arisen. I have been arguing for PPP model but who would listen ass people and institution look for easy money and try to make hay till the sunshine.

  4. Dear Ramesh,

    Good initiative!

    Richard Rosenberg (CGAP), following the Compartamos (Mexico) IPO in 2007, also has produced a paper for discussion -- which has helped to gather ideas on this controvercial issue. ... I believe your paper will do the same, and help us reach on a kind of 'conensus'. ... Please also note that Dr Muhammed Yunus has his own ideas on how 'microcredit interest' should be set.

    Again thank you.

    Getaneh Gobezie

  5. hi Ramesh, interesting and thought provoking article on a critical issue. my view on interest rate is:

    1. The company must make a 5 year projection with certain growth assumption
    2. it must find out what is its cost of operation likely to be when it has grown to a reasonable level and its growth plateaus off to around 15-20% p.a level
    3. this cost of ooperation (which i call optimum cost) must be taken as the input for deciding pricing.
    4. to this, we can add cost of bank funds, 2% for delinquency and 2-3% for RoA and fix the lending rate
    this will ensure that as a company starts and grows, the start up cost and growth cost is borne not by the borrower but by the investor. it is unfair to load the full cost of the company when it is either new or growing at a blistering pace to the client, since under these circumstances, the cost is boudn to be very high. the benefit of this growth is ultimatley going to accru to the investor and hence he should be the one to bear the cost of such growth too and not the borrower.
    At Equitas, we have been practising this from the day we extended our first loan. in Dec 2007 when we started our lending rate was 26% all inclusive reducing balance cost at a time when there was no other MFI in the country lending at less than 32%! this was because we assumed our operating cost to be 7.5% which is what we expect it to be when we reach a steady state and only then we will make the full expected Return on Equity of around 20%. and as on date our Operating cost is around 9% and hence we are at an RoE of around 14%. thus the investors have always borne the cost of growth of the company and we have shielded the clients completely from this cost.
    if only we had access to subsidies it coudl have enabled us to lower the rate to our members further...
    and the other element of lending rate is transparency. are we transparently printing and giving in writing to the members the real all-inclusive reducing balance cost? if we are feeling ashamed to show this to the client transparently then our lending rate is unfair ... and we wouldn't need anyone else to tell us this!

  6. Great post. Very informative data on MF interest rates.

    I also like this breakdown of the microfinance interest rates for a better understanding of the costs that go into rates.