You are on page 1of 16


Mohammad Kamil
Arshiya Sherwani

Evaluation of Micro finance - Perspectives

Discussion paper for International Seminar on Re-Engineering of Indian Banking
Industry in Global Perspective, Bareilly
February 21, 2008


Microfinance generally refers to the entire gamut of financial services, such as micro credit deposits,
offered in small quantities to those at the bottom of the pyramid. It could range from poverty
alleviation to improvement in women welfare or merely to provide assistance of financial services to
the poor.

In India Microfinance has been widely encouraged by boffin Banks essentially NABARD and other
financial alliances. Technological progress is also providing a window of opportunity for the MFIs,
by reducing transaction costs and enabling microfinance to be a commercially viable and profitable
activity. Hence, the advantage of microfinance is its simple disbursement procedure and
accountability of the borrower whereby livelihood finance will be a national strategy, with much
large levels of resource allocation, both from the public sources and the capital finance. Thus, macro
realization in terms of livelihood finance is eventually the goal of microfinance with social
responsibility of deprived and efficient poor management.

The foreseeing issues will be indulgement level of commercial banks, commercial viability,
representation of self help groups, delivery of non credit financial services by microfinance
institutions and Non Banking Corporations. The evaluation of Micro Finance will transact with
sustainability, efficiency, portfolio quality and control while answering the conflict reduction
process between commercial and social objectives. In addition to this the authors will address the
perspectives of Micro finance, assess the critical success factors with respect to micro credit, sense of
social entrepreneurship, innovation and all other anecdotal facets necessary for benefitting the less
privileged. The study will also guide as to how the rural poor if carefully nurtured can be an asset
rather than a liability.

 Prof. Mohammad Kamil is a Management Graduate from University of Bombay, D.Phil from
University of Allahabad, and has 16years of rich experience in Industry, with a major contribution to
Equity Research and Market Analysis while he was engaged with A.C Nielsen, India and then shifted
to Research & Academics. He has taught in S.P Jain Institute of Management, Mumbai besides JBIMS,
NMIMS and Welingkar and K J Somaiya, Mumbai, Dr. VSIMS, Kanpur and KCMT Bareilly. Currently
he is a Professor in Management in IEM, Lucknow and visiting fellow of other prestigious institute’s of
the nation. His area of D.Litt is Investment Behaviour – Risk & Return Tradeoff vis-à-vis Psychological

 Ms. Arshiya Sherwani, Research Scholar, IEM, Luckow

Prof. Mohammad Kamil
Arshiya Sherwani

Microfinance generally refers to the entire gamut of financial services, such as micro credit
deposits, offered in small quantities to those at the bottom of the pyramid. Effective
availability of microfinance could, therefore, be an important factor in improving the welfare
of the poor and the underprivileged.
The rural finance policy pursued in most developing countries beginning from 1950s was based
on providing subsidized credit through state controlled or directed institutions to rural
segments of population. The distortions and imperfections such as low saving and absence of
modern technology in rural credit markets were sought to be addressed through Government
interventions. On account of the above developments, the resultant shift in rural finance
discourse and operational paradigm is shown in Table.

Features Old Paradigm New Paradigm

Problem Definition Overcome market imperfections Lower risks and transaction
Role of Financial Markets Implement State plans Intermediate resources
Help the poor efficiently
View of users Beneficiary Clients
Subsidies Create subsidy dependence Create independent institutions

Sustainability Largely Ignored A major concern

Evaluations Credit impact on beneficiaries Performance of financial

Emergence of micro credit 7 in late 1970s and early 1980s in the backdrop of growing world attention
on deficiencies of earlier approach in rural finance explains much of its dominant theoretical
underpinnings. The universal appeal of microfinance stemmed from its ability to reach the poor
without social collateral and generation of near full recovery rates through what has been described
as a Win-Win proposition. The mainstreaming of microfinance worldwide and its global acceptance
in development community is based on this Win-Win proposition. This concept of provision of
sustainable financial services at market rates has been termed as ‘Financial System’ approach or
‘Commercial microfinance’.
It is claimed that this new paradigm of unsecured small scale financial service provision helps poor
people take advantage of economic opportunities, expand their income, smoothen their
consumption requirement, reduce vulnerability and also empowers them. Former World Bank
President James Wolfensohn said “Microfinance fits squarely into the Bank's overall strategy. As
you know, the Bank's mission is to reduce poverty and improve living standards by promoting
sustainable growth and investment in people through loans, technical assistance, and policy
guidance. Microfinance contributes directly to this objective. The emphasis on microfinance is
reflected in microfinance being a key feature in Poverty Reduction Strategy Papers (PRSPs)

Prof. Mohammad Kamil
Arshiya Sherwani

The current literature on microfinance is also dominated by the positive linkages between
microfinance and achievement of Millennium Development Goals (MDGs). Microcredit Summit
Campaign’s 2005 report argues that the campaign offers much needed hope for achieving the
Millennium Development Goals, especially relating to poverty reduction
Indian public policy for rural finance from 1950s to till date mirrors the patterns observed
worldwide. Increasing access to credit for the poor has always remained at the core of Indian
planning in fight against poverty. Development Goals, especially relating to poverty reduction
started in late 1960s, where India was home to one of largest state intervention in rural credit market
and has been euphemistically referred to as ‘Social banking’ phase. It saw nationalisation of existing
private commercial banks, massive expansion of branch network in rural areas, mandatory directed
credit to priority sectors of the economy, subsidised rates of interest and creation of a new set of
rural banks at district level and an Apex bank for Agriculture and Rural Development (NABARD) at
national level. These measures resulted in impressive gains in rural outreach and volume of credit.
As a result, between 1961 and 2000 the average population per bank branch fell tenfold from about
140 thousand to 14000 and the share of institutional agencies in rural credit increased from 7.3%
1951 to 66% in 1991.
These impressive gains were not without a cost. Government interventions through directed credit,
state owned Rural Financial Institutions (RFI) and subsidised interest rates increased the tolerance
for loan defaults, loan waivers and lax appraisal appraisal and monitoring of loans. The problem at
the start of 1990s looked twofold, the institutional structure was neither profitable in rural lending
nor serving the needs of the poorest. Microcredit emergence in India has to be seen in this backdrop
for a better appreciation of current paradigm. Successful microfinance interventions across the
world especially in Asia and in parts of India by NGOs provided further impetus. In this backdrop,
NABARD’s search for alternative models of reaching the rural poor brought the existence of
informal groups of poor to the fore. Under the programme, popularly known as SHG-Bank Linkage
programme there are broadly three models of credit linkage of SHGs with banks.

The programme has received strong public policy support from both Government of India and
Reserve Bank of India. The importance attached to it by Government is exemplified by mention of
yearly targets by Finance Minister in his annual budget speech as well as introduction of similar
group based lending approach in government’s poverty alleviation programme. The success of the
programme in reaching financial services to the poor has won international admiration“

The growth of microfinance in India has also to be seen in the light of financial sector reforms in
India starting from 1991 and the global emphasis on commercialization of the sector. The financial
sector reforms in India have focused on fostering a market based financial system by increasing
competition and improving the quality of financial services. The new approach has been deeply
influenced by the reorientation among international rural financial policy makers centering on
concepts such as self-help, self sustained growth and institutional viability. Under the new
approach, institutional viability is of prime concern and instruments of directed credit and interest
rate directives have been totally diluted or been done away with. As a consequence, banks are
increasingly shying away from rural lending as well as rationalizing their branch network in rural
Prof. Mohammad Kamil
Arshiya Sherwani

areas flowing out of negative experiences of the earlier state intervention, institutional viability has
become the focal point for evaluation of success of credit interventions. The philosophy and design
of SHG-Bank linkage programme reflects this new concern vividly by emphasising on full cost
recovery in order to become an attractive proposition for banks. The design features of the
programme emphasise that it does not envisage any subsidy support from the government in the
matter of credit and charges market related interest rates based on the premise that sub-market
interest rates could spell doom; distort the use and direction of credit. High recovery rates under the
programme are used to justify the dictum that ‘poor need timely and adequate credit rather than
cheap credit’.
The field research was undertaken to understand the clients perspective and analyse the factors
behind repayment rates as well as impact of credit on socio economic well being of clients. The
research covered 93 client households from 5 Self Help Groups from three different locations in
Western and Central part of India As the size limitations of paper constrain a detailed enumeration
of field research findings, only the key findings of the field research having a bearing on the central
aspect of the paper are listed -
- All clients were saving regular amounts of money at monthly/bimonthly interval building
up the group savings
- Internal loaning of group funds was very high resulting in significant waiting time for
members interested in borrowing
- Social awareness index of group members as measured on Likert Scale showed a definite
positive trend after joining the group
- Reliance on moneylenders for credit eliminated or decreased in case of approx 2/3 rd of
- No specific benchmarks for group membership leading to inadequate poverty targeting
- Only 6% clients had taken up any economic activity post group formation
- Marginal increase in real term income level after joining the group
- Bank credit to group often a result of banker’s zeal to achieve targets rather than based on
group demand
- Bank credit as well as loans used overwhelmingly for consumption purpose
- Group members not willing to borrow to take up economic activity on account of credit risk
and absence of skills
- Prompt Repayment a factor of group pressure and sourced out of reduced consumption,
extra work and borrowing from other sources
- High rates of interest in internal lending among group members (2-3%) was seen by
members as prescribed by the group forming agency and accepted as being better than even higher
rates of informal sector.
Further summarizing the findings at the cost of over simplification, it can be said that while
the programme had definite impact on building of social capital, it had marginal impact on income

Another study commissioned by NABARD in 2002 with financial assistance from SDC GTZ and
IFAD covered 60 SHGs in Eastern India. The findings of this study also corroborate the findings of
Prof. Mohammad Kamil
Arshiya Sherwani

earlier evaluation with 23% rise in annual income and 30% increase in asset ownership among 52%
of sample households. World Bank policy research paper (ibid) 2005 details the findings of Rural
Finance Access Survey (RFAS) done by World Bank in association with NCAER . The RFAS covered
736 SHGs in the states of Andhra Pradesh and Uttar Pradesh and also points to positive economic
impact. The findings indicate 72% average increase in real terms in household assets, shift in
borrowing pattern from consumption loans to productive activities and 33% increase in income
In such a scenario, while loan volume, outreach and repayment may outwardly justify the
intervention and make it attractive for bankers, its impact on economic gains for clients gets missed
out. The common underlying assumption behind reliance on such parameters is belief in the linear
cycle of credit, starting from credit offtake followed by economic activities, rise in income/assets and
repayment out of additional income.

The figure below illustrates this :

The focus on financial sustainability has meant that much of the evaluation criteria for microfinance
interventions are based on institutional performance. Weber says that while the virtuous impact of
microfinance is used to justify its expansion, much of this assessment is based on institutional
success and avoidance of engaging with impacts. While microfinance may be a winning proposition
for banks, the winning evidence on client’s side seems doubtful. In this scenario, it can be said with
certainty that potential of microfinance to contribute to achievement of MDGs in India, especially
reduction of poverty remains suspect. Greeley (2005) rightly notes that in absence of specific poverty
targeting and mainstreaming of impact assessment, the claims about the impact of microfinance on
the achievement of MDG lacks credibility.

Indian rural finance sector is at crossroads today. Following the financial sector reforms with its
emphasis on profitability as the key performance benchmark, banks are increasingly shying away.
The SHG-Bank linkage programme has witnessed phenomenal growth and the current strategy is to
focus on 13 underdeveloped states as also graduate the existing SHGs to the next stage of micro

Prof. Mohammad Kamil
Arshiya Sherwani

At this stage, the paper argues that if SHG-Bank linkage programme has to contribute to poverty
reduction, there is an imperative need for integrating impact assessment as a necessary design
feature of the programme. The significance of bringing the focus back to ‘people’ from ‘institutions’
and adoption of localized people centric approach can hardly be overemphasized.
Adequate emphasis on impact assessment is an integral part of the triangle of factors necessary for
judging microfinance intervention.

I. Financial sustainability

II. Outreach to the poor

III. Institutional innovations

Irrespective of socio economic status, credit can be put to little productive use in resource deficient
and isolated areas. In such areas, credit flow has to follow public investments in infrastructure and
provision of forward and backward linkages for economic activities. Homogenization of service
delivery without fully taking into account situational context and client needs will continue to have
limited impact.

The Indian economy at present is at a crucial juncture, on one hand, the optimists are talking of
India being among the top 5 economies of the world by 2050 and on the other is the presence of
260 million poor forming 26 % of the total population. India’s achievement of the MDG of
halving the population of poor by 2015 as well as achieving a broad based economic growth also
hinges on a successful poverty alleviation strategy. In this backdrop, the impressive gains made
by SHG-Bank linkage programme in coverage of rural population with financial services offers a
ray of hope.

It is the process of structural change that involves financial inclusion, bankarization, or the
regulation of informal financial practices, and the utilization of voluntary sector and third sector
capabilities in the provision of financial services to people who are excluded from financial and
banking institutions – i.e., from 60 to 90 % of the entire population. It is one of the most fascinating
features of financial economics today.

India is experiencing a huge expansion in terms of households linked to microfinance, more

specifically linked to Self-Help Groups (SHGs). An average annual growth rate of 82 % was
observed in the period from March 1993 to March 2006, in relation to a 110 % growth rate in terms
of credit amounts.

Relative Strength of the SHGs among States

Prof. Mohammad Kamil
Arshiya Sherwani

The ratio of the number of SHG members to the total households of the states reveals a different,
although continuing, pattern in regional variations as compared to the relative strength of the SHGs.
In March 2001, there were less than ninety households participating in SHGs for every 1,000 of
Andhra Pradesh’s households. In the states Sikkim, Assam and Punjab, however, there were six,
five, and three households participating in SHGs for every 10,000 of the total households
respectively. The irregular pattern continued in 2003 and 2005. Nevertheless, the relative strength of
the SHGs slightly converged among states, as evidenced in the decline of the coefficient of variation
from 1.99 in 2001 to 1.15 in 2006 - a spatial pattern that is confirmed by inspecting the related maps
in Figure 1.
Figure 1: State variations in relative share of SHGs (measured by standard deviation)

In Figure 1, the states are grouped using the standard deviation of relative share of SHGs. In March ,
Andhra Pradesh was ranked first, with more than three units of standard deviation above the mean.
Tamil Nadu, Himachal Pradesh, Puducherry and Karnataka formed the second leading group,
which had a standard deviation above the mean, i.e., with nineteen, sixteen, twelve and nine
households participating in SHGs for every 1,000 households respectively. The rest of the states had
one unit of standard deviation below the mean and formed the weakest states in the process of

Relative Strength of the SHGs among Districts

The development of the microfinance sector through the SHG model reflects the relative importance
of this movement for the population, as depicted in Figure 2. It shows a map indicating the relative
strength of the households involved in a SHG provided with a bank loan during the financial year
2005-2006. The inequality pattern in the microfinance sector in India is also verifiable at the district

During this financial year, the relative share of SHGs was more than 35 % of the total households in
Nuaparha district (Orissa). In other words, more than one third of the total households in this
district counted a person involved in a SHG provided with a bank loan during the last year.
Prof. Mohammad Kamil
Arshiya Sherwani

Figure 2: District distribution in relative share of SHGs (measured by natural break – percentage)

Between 2000 and 2002, the number of SHGs in Rajasthan increased by more than 12 times. of U.P.
The number was only multiplied by 1.5. Between 2002 and 2004, there was a slowdown in the
growth of microfinanciarization except in some areas, where the number of SHGs increased by more
than 8 times from 2002 to 2004.

In a context of growing financiarisation, the poor more than anybody else need microfinance
services. In the same vein, in a context where democracy remains mainly formal and inaccessible to
the poorest, the collective approach (which is at the core of Indian microfinance through the Self-
help-group concept) undeniably represents a tool for democratic practices and therefore for grass
roots development, especially for women.

After a first cycle of growth where the number of clients went from a few thousand to several
millions, microfinance is nowadays at the core of many agendas, be they public or private. Indian
microfinance, both in terms of the number of clients and the volume of credit disbursed, is not
anecdotal any more. The SHG Banking Linkage Programme since its beginning has been
predominant in certain states, showing spatial preferences especially for the southern region –
Andhra Pradesh, Tamil Nadu, Kerala and Karnataka. These states accounted for 57 % of the SHG
credits linked during the financial year 2005-2006. Preliminary results from this communication

Prof. Mohammad Kamil
Arshiya Sherwani

show that the spread of number of SHGs did not evolve evenly over time within districts and states
of India.

Pricing Microfinance Loans and Loan Guarantees Using Biased loan write-off data

We present a simple, easy to implement methodology for pricing microfinance loans and loan
guarantees using publicly available data on loan write-offs by Micro Finance Institutions (MFIs).
Our methodology takes into account the selection bias inherent in available data in those MFIs that
do not report loan write-off data are less likely to be better performers. Our quantitative analysis is
consistent with pricing seen in a recent securitization deal. Our analysis suggests how securitization
and loan guarantees can greatly expand the supply of funds for microfinance loans.

One of the key considerations in pricing a loan is the probability that the borrower might default.
Popular and anecdotal accounts of the microfinance revolution boast of repayment rates that are in
almost all cases above 95 percent. Yet, amidst all the euphoria, there are those who urge caution by
carefully documenting that reported default rates, even for the most illustrious and visible cases
such as the Grameen Bank are underestimated.
In this part of the paper, we provide a simple methodology for estimating the true distribution of
loan returns using only observations that are self-reported by MFIs by assuming that those MFIs
who choose not to report the information are likely to have on average a higher loan default rate
than those MFIs who report the information.

Suppose that the natural log of the returns generated on loan portfolios for MFIs, denoted, are
drawn from a normal distribution with mean ¹ and variance ¾2. Formally, ln(R) » N(¹; ¾2): Since
ln(R) is distributed normally, we can define a standardized variable x ´ln(R) ¡ ¹ ¾ » N(0; 1); i.e., x will
have a standard normal distribution with mean 0 and variance equal to 1. Let Áx denote the
standard normal probability density function and fx denote the standard normal cumulative density

I. Let L and H denote two values of log returns such that L < H. We divide the loan
write-off data into three buckets:
1. Bucket 1 (R ¸ 100): MFIs that report zero percent loan write-off (equivalent to a return of one
hundred percent or more). We choose H = ln(100).
2. Bucket 2 (100 > R ¸ 100 ¡ ®): MFIs with loan write-off rates between zero percent and another pre-
specified cut-off rate (equivalent to some return that is less than one hundred percent, say 100 ¡). We
choose L = ln(100).
3. Bucket 3 (100 > R): MFIs with loan write-off rates greater than (equivalent to returns smaller than

Prof. Mohammad Kamil
Arshiya Sherwani

FIGURE 1- Normal Distribution for ln R N_2+n observations N_1 N_3+2n observations

observations L= ln (100-α) H = ln 100 . There are a number of MFIs that do not report any loan
write-off numbers. The main insight that we will use in our pricing analysis is that these MFIs
almost surely do not belong to Bucket 1 and are more likely to belong to Bucket 3 than Bucket 2. In
particular, we will assume that MFIs with missing data are twice as likely to belong to Bucket 3 as to
Bucket 2. Let N1, N2 and N3 denote the number of MFIs that belong to Buckets 1, 2 and 3
respectively based on their self-reported loan write-off numbers. Let 3n denote the number of MFIs
with missing loan write-off data. We will assume that n of these missing observations belong to
Bucket 2 and the remaining 2n to Bucket 3. (See Figure 1)

The parameters of the distribution of log returns, ¹ and ¾ are determined, by setting:
Pr [ln(R) > H] = Pr[x > h] = 1 ¡ h = N1 N
(1) Pr [ln(R) < L] = Pr[x < l] = l = N3 + 2n N
(2) Where N = N1 + (N2 + n) + (N3 + 2n); h ´H ¡ ¾; l ´ L ¡ ¾:

It is easy to see that the probability of default is given by Pr[R < 100] = Pr[x < h] = h: We now show
how loans and loan guarantees can be priced. Once the parameters and ¾ are determined, the
lognormal distribution of R is completely specified; let g denote the probability density function of
R. The mean and variance of the lognormal distribution are:

E[R] = exp(¹ +12 ¾2); (3)

Var[R] = exp[2(¹ + ¾2)] ¡ exp[2¹ + ¾2]: (4)

The expected payoff of the lender with face value of the repayment of 100 is given by:
1. Z 10 Minf100; zgg(z)dz: (5)
2. It is easy to consider other partitions of MFIs with missing data into the two buckets as well.
3 The above expression re°ects the fact that the lender never receives more than the full contractual
value of 100 but will sometimes receive less if the borrower is unable to pay the entire amount. If we
were to price a loan guarantee3 in which the lender is insured to receive its full payment of 100, the
price of such a guarantee can be calculated as:
Z 10 Maxf0; 100 ¡ zgg(z)dz: (6)

The guarantor is obliged to pay the difference between 100 and what is collected from the borrower
if that amount is less than 100.

Now, consider pricing the loan portfolio to ICICI in the type of securitization deal between SHARE,
ICICI and Grameen described in the introduction. As long as the return is greater than 92, ICICI will
get paid 100 because Grameen has guaranteed a loss of up to 8. If the return is less than 92, ICICI
will receive that plus 8 from Grameen.

Thus, ICICI's expectedpayoff is given by:

Z 1 0 Minf100; z + 8gg(z)dz: (7)
Prof. Mohammad Kamil
Arshiya Sherwani

Similarly, Grameen's expected payment in such a deal can be calculated as:

Z 1 0 Minf8; 100 ¡ zgg(z)dz: (8)

This represents Grameen's payment of 8 if the return is less than 92; if the return is between 92 and
100, Grameen only needs to make up the di®erence. Obviously if the return is greater than 100,
Grameen doesn't need to pay anything.

II. Application to MIX Market Data

We downloaded the data (on March 26, 2005) - see the Appendix- on 420 MFIs from the website of
Mix Market ( 4 of these MFIs, data on 63 were not reported on MIX Market
site because microfinance represents less than 91% of their operations.
For expositional simplicity, we are only considering risk-neutral pricing and have normalized the
risk-free rate to equal zero.
The data and the supporting analysis in this section are available in an excel spreadsheet from the
authors upon request.
Financial analysis data were reported for 357 of these MFIs (N = 357). Of these, 78 MFIs reported a
loan write-off ratio of 0%. These firms belong to our Bucket 1. N1 = 78:

We then ordered the remaining MFIs by their loan write-off ratios and chose sd = 9:04% as the lower
cutoff for deciding the range for Bucket 2 because that provided a natural break-point.5 That
resulted in 160 MFIs with loan write-off ratios of between 0% and 9.04%.

Thus,N2 = 160:Seven MFIs reported write-off ratios of less than 9% and 112 MFIs had missing data.
Thus, N3 = 7; 3n = 112:
We assume that 37 of the MFIs with missing data belong to Bucket 2 and the remaining 75
MFIs belong to Bucket 3. Thus,
1 ¡ ©h = 78
357 ; ©l = 7 + 75
357 ; h = ln(100) = 4:60;
l = ln(100 ¡ 9:04) = 4:51:
Solving (using the \solver" function in Excel), we obtain:
¹ = 4:56;
¾ = 6:25%:
Using these parameter values, we perform a reality check as follows. If we only consider ob-
servations that are between l and h (observations in Bucket 2), then these observations could be
described as drawn from a truncated normal distribution with l as the lower truncation.

The choice of the lower cutoff point is somewhat arbitrary. It is sensible not to choose this cutoff too
close to 100 because we would like the sizes of all three buckets to be reasonable so that we can
approximate the probabilities by the relative frequency of observations in each of the three buckets.
5level and h as the higher truncation level. The mean of the truncated distribution, ¹T , and the
variance of the truncated distribution, ¾2 T , are given by (Johnson and Kotz, 1970): ¹T = E[x j l < x <
h] = ¹ + ¾Ál ¡ Áh fh ¡ fl ;¾2 T = Var[x j l < x < h] = ¾22 41 ¡hÁh ¡ lÁl©h ¡ ©l ¡ (Áh ¡ Ál ©h ¡ ©l)235 :
Prof. Mohammad Kamil
Arshiya Sherwani

Using the parameter values, the truncated mean is calculated to be ¹T = 4:558 which is less than
4:588, the arithmetic average of the N2 = 160 observations between l and h. This is reassuring
because we have argued that MFIs that did not report any write-off data are likely to have smaller
loan repayment rates. Thus, true truncated mean must be lower than the one calculated from biased
observations. Similarly, truncated standard deviation is calculated to be ¾T = 2:63% which is more
than 1:98%, the estimated standard deviation of N2 = 160 observations between l and h. Thus the
estimated standard deviation is biased downwards compared to the true truncated standard

The mean and the standard deviation of lognormal distribution of return R, using (3) and (4) are
calculated to be: E[R] = 95:45; stdev[R] = 5:97:

The pricing of loans and loan guarantees in expressions in (5) to (8) are calculated using numerical
simulations using Crystal Ball" in Excel.

We obtain the following results:

The expected payoff of a lender with face value of the repayment of 100, using (5), is 94:59. In other
words, a lender without any guarantees expects to lose over 5%. Lenders anticipating this will gross
up the interest rate charged. However, if the lender's payoff is guaranteed, using for example the
guarantee provided by Grameen and SHARE to ICICI, the expected payo® to ICICI, using (7), is
calculated to be 98:97. In other words, with the guarantee ICICI expects to lose only about 1%, and
therefore it doesn't need to gross up the 6 interest rate charged as much. This is roughly consistent
with ICICI charging only 8.75% to SHARE, whereas SHARE usually paid nearly 13% on loans
without the guarantee.Let us now calculate the expected cost to Grameen for providing the
This, using (8), is calculated to be 3:58. This makes sense as Grameen loses all of its 8% collateral
only in some cases.

Finally, we calculate the expected cost of providing a complete guarantee to a lender or insuring
the lender completely against default. Using (6), this is calculated to be 5:41. In other words, an
insurer with deep pockets will only charge 5:41 per 100 to make the loans risk-less to lenders.

III. Concluding Remarks

We presented a simple, easy to implement methodology for pricing micro¯nance loans and loan
guarantees using publicly available data on loan write-o®s by Micro Finance Institutions (MFIs).
Our methodology takes into account the selection bias inherent in available data in that MFIs that
do not report loan write-o® data are less likely to be better performers.

Prof. Mohammad Kamil
Arshiya Sherwani

Innovations in Microfinance

“As with many ideas, most common spread of microfinance has been through imitation”
-Professor Sendhil Mullainatham
Some innovations include-
 Franchising microfinance
 A borrowing from MFI’s to become local capitalists.
 Introduction of equity investment and Venture capital for micro enterprises
 Securitization

Regulations of Micro-Finance Institutions

 Encouragement of micro credit through commercial banks
 Enhancement of Financial inclusion
 Increasing size of MFIs
 Organizational form of microfinance institutions appear varied, making them subject to
different legal frameworks
 Different state government take different approaches to the microfinance institutions-
including subsidizing interest rates
 Developments in technology seem to provide a window of opportunity for the MFIs,
reducing transaction costs and enabling microfinance to be a commercially viable and
profitable activity.

Justification for imposing regulation on MFIs on the present stage

 All financial institutions have to report to RBI, NABARD and other Agency.
 More and different forms of microfinance are likely to emerge.

Limitations of micro credit- five fatal assumptions

 Accredit is the main financial service needed by the poor

 Credit can automatically translate into successful micro enterprises
 Poor people want to be self employed and can be helped by the micro credit
 Those slightly above the poverty line do not need micro credit and giving them
amounts to mis-targetting
 Micro credit institutions can become financially self sustaining.

From micro credit to livelihood finance

Prof. Mohammad Kamil
Arshiya Sherwani

Livelihood finance should be a national strategy, with much large levels of resource allocation, both
from the public sources and the capital finance

Franchising Microfinance

Access to credit remains a matter of great importance in the rural sectors of developing economies
Significant efforts continue to be made to ameliorate the situation. Governments are undertaking
various schemes to facilitate the process, either directly or indirectly. In India, for instance, various
arms of the government offer both direct and indirect assistance in the form of subsidized interest
rates on loans, requiring lending to \priority sectors" such as agriculture, promoting self-help group
(SHGs), and so on. Despite years of such efforts, significant portions of the rural poor are unable to
access funds. Instead, the role of moneylenders continues to be critical in the provision of credi to
the rural poor.

The Model

We assume a local area, e.g., a small village economy, with three sets of players - borrowers,
moneylenders with enforcement technologies, and local capitalists with (possibly illiquid) collateral
against with they can borrow funds from ¯nancial intermediaries to potentially lend to borrowers
with no collateral. All players are risk neutral.

In every period, a borrower without any collateral and without any funds of he own, needs $1 to
fund a project. In the next period, the project generates a gross payo®, R, with probability p, and 0

The project is a positive NPV project, pR 1+rf > 1 where all risk-neutral agents discount future cash
flows at the risk-free rate rf . Let rd denote the discount rate adjusted for the condition that the
positive payoff R is obtained only with probability p < 1. Thus 1 + rd =1 + rfp: (1) the only contract
that moneylenders can o®er is a standard loan contract. For each borrower, there is a single
moneylender in the village with two distinctive characteristics. One, he has the ability to enforce the
loan contract.2 We de¯ne enforcement as the ability of the moneylender to impose a cost on the
borrower, explicit or implicit, either through social sanctions or physical sanctions, equal or greater
than what is owed to him by the borrower. This ensures that the borrower obtains no benefit from
voluntarily defaulting against the moneylender.

Second, the moneylender incurs zero fixed cost of screening and monitoring the borrower. There are
a number of local capitalists in the market who own assets that they can pledge as collateral to
borrow from banks at the risk-free rate. However, they do not have an enforcement technology of
the type that is available to the moneylender. In addition, each such local capitalist incurs a unique
cost of screening and monitoring the loan (let us call it transaction cost of lending) to a given
borrower every time he enters into a loan contract with the borrower.

Prof. Mohammad Kamil
Arshiya Sherwani

Local capitalists are differentially familiar with a borrower. The more familiar is the local capitalist;
the lower is his transaction cost. Note that these costs are specific to the borrower. In other words, a
local capitalist will have di®erent transaction costs for di®erent borrowers. Without any loss of
generality let us denote these costs as 0 < c1 < c2 < c3 <::::::: (For simplicity, we will not carry any
notation denoting the borrower.) One way to think about this is that the transaction cost is based on
the distance between the lender and the borrower. Lenders incur a cost of monitoring to ensure that
the funds are being used for the declared purpose. So, the closer the borrower, the lower the
transaction cost. The moneylender is the extreme form where he incurs no transaction cost and is
also able to prevent a willful defaulter from enjoying the spoils of default.

We assume that project return, success probability, discount rates and lenders' transaction costs are
all common knowledge. We further assume that borrowers have limited liability and funds that the
borrower does not commit to a project cannot be laid claim upon in case of loan default. In other
words, enforcement costs are implicitly assumed to be prohibitive for the local capitalists.

The borrower can consume the funds after defaulting voluntarily but cannot default and invest
these funds in future projects (in essence become her own banker for future projects).

3 In case the project fails, the borrower would be forced to default. However, a borrower can also
default voluntarily when the project is successful. We assume that lenders cannot identify the reason
for defaulting. Each lender chooses never to re-lend to a borrower who has defaulted on his loan to

4 In every period, there are two stages. In stage one, a lender decides whether to lend to a specific
borrower or not, and what interest rate (r) to charge. Then the borrower decides whether to
undertake the project. If she decides to undertake the project, she chooses the lender to borrow
from. In stage two, payoffs are realized and a borrower decides whether or not to default on the
loan. The same game can be repeated over multiple periods for every borrower. Borrower's
participation constraint is simply that the total amount owed the principal payment of 1 plus the
interest payment of r amount owed cannot exceed the gross payoff from the project. Formally,1 + r ·
R: Lenders' participation constraints are that they must expect to recover the amount owed plus the
transaction cost of making the loan, c. For the moneylender c = 0. For local capitalists c is positive.
Let r0(c) denote the minimum interest that a lender must charge to break-even (make zero expected
profits). Thus 1+r0(c) represents the total cost making a loan of 1 to the borrower.
For the money lender, the total cost of lending is given by
1 + r0(0) = 1 + rd: (2)
For a local capitalist with transaction cost c,
1 + r0(c) = (1 + rd)(1 + c): (3)
It is easy to see that
r0(0) < r0(c1) < r0(c2) < r0(c3) < ::::::: Lenders' participation constraints, then are that the interest
charged must be such that it exceeds the cost of lending. Formally, 1 + r ¸ 1 + r0(c): (4)

Prof. Mohammad Kamil
Arshiya Sherwani

Franchising Solution
Lending by local capitalists fails to occur in equilibrium because of a coordination failure on the part
of these potential lenders. Each local capitalist has an incentive to make a loan to the borrower who
may have defaulted against another local capitalist lender. If it were possible for local capitalists to
credibly commit to not extend credit to a borrower who has defaulted on another lender's loan, the
borrower would effectively face a single typical local capitalist lender and will have an incentive not
to default voluntarily.

Concluding Remarks

We have shown that a franchising contract in which each franchisee is required to refuse credit to a
borrower who may have defaulted against another franchisee can attract local capitalists as lenders
belonging to the franchise. This has the advantage of providing a competitive alternative to
moneylenders who are notorious for charging usurious interest rates when they are able to act as
monopolists. Financial intermediaries such as banks do not have local information to be able to
profitably compete with moneylenders whereas local capitalists have local information but may not
have liquid funds.