MOSEC in Microfinance SSF 2012 13
MOSEC in Microfinance SSF 2012 13
During the fiscal year 2011-12, Indian MFIs raised more The MOSEC TM has revolutionised the way MFIs access
than INR22bn through securitisation transactions . 1
capital markets in India.
However, small MFIs (having assets worth less than
Fernandes provides a detailed discussion on the
INR1bn) accounted for less than 15% of the funds raised 2. microfinance model and how the structured finance
A securitised pool must have a minimum critical size in approach was applied in microfinance in India to provide
order to make these transactions financially viable. Given MFIs access to reliable debt capital (Fernandes, 2011).
this constraint only large MFIs can access capital via the
In this chapter, we discuss the MOSECTM transactions in the
securitisation route. The smaller MFIs find it difficult to
Indian microfinance sector and how multi-originator
provide a critical size portfolio for single-originator
structures have helped MFIs to access debt capital and
securitisation. However via the IFMR Capital multi-
enabled investors to achieve portfolio diversification. In
originator securitisation (MOSEC TM) structure, small and the first section, we briefly discuss securitisation
medium originators can combine their loan pools to structures in microfinance and the key differences between
attain a critical size portfolio that can then be taken to single-originator and MOSECTM transactions. In the next
the market. Pooling loans across originators and section, we explain how MOSECTM structures help to attain
geographies results in a well-diversified portfolio; this critical portfolio size, enhance diversification and reduce
provides an attractive risk-return trade-off to the investor. overall risk of the transaction. Next, we discuss the
600
䊏 Total transaction size 䊏 Average pool size per originator
500
400
INR m
300
200
100
0
ICM-XI (8)
ICM-VII (7)
ICM-X (6)
ICM-XII (9)
ICM-XIII (6)
ICM-III (3)
ICM-II (3)
ICM-I (4)
ICM-IV (3)
ICM-V (4)
ICM-VI (3)
ICM-IX (5)
ICM-VIII (4)
Source: IFMR Capital
31
TM
MOSEC structures allow smaller MFIs to participate
actively in securitisation transactions. Exhibit 2 shows the
distribution of single-originator and MOSEC TM transactions
by total assets of participating MFIs in all microfinance
securitisations arranged by IFMR Capital. Single-
originator transactions are skewed towards bigger MFIs
(assets more than INR1bn) whereas smaller MFIs (assets
less than INR1bn) show active participation in MOSEC TM Vaibhav Anand Dr. Kshama
Fernandes
transactions. Smaller MFIs accounted for only eight out of
Vaibhav Anand, Senior Risk Analyst
42 pools securitised in single-originator transactions
IFMR Capital
arranged by IFMR Capital. However, similar sized MFIs
accounted for 47 out of 64 pools securitised in MOSEC TM tel: +91 (0) 22 6668 7375
5
transactions . fax: +91 (0) 22 6668 7010
email: [email protected]
12
䊉 Single originators
Number of transactions closed
10 䉬 MOSEC originators
0
0 2,000 4,000 6,000 8,000 10,000
*multi-originator securitisation Asset size of Microfinance institution in INR m
**Microfinance institution
32
originator is a small portion of total transaction size, For a given set of microloan pools, this diversification results
diversification across servicer reduces the overall servicer in superior performance for MOSECTM transactions when
risk of the transaction. compared to the performance of single-originator
transactions with the same microloan pools. In this section,
MOSECTM transactions also offer a high degree of
to illustrate the effects of MOSECTM diversification, we use
geographic diversification in the portfolio. Typically small
the historical default-instance distributions for three MFIs.
MFIs have regional focus limited to certain geographies or
6 TM The default-instance distribution is a probability distribution
states . MOSEC allows investors to reduce their
geographical concentration risk while allowing regional of observed default-instances7 for each microloan portfolio
originators to focus on operational efficiencies in their originated by the respective MFI. We compare the
respective geographies. Also, diversifying across states performance of three hypothetical single-originator
and districts helps to reduce political and event risks in a transactions to the performance of a hypothetical MOSECTM
transaction. transaction with microloans originated by three MFIs
underlying it. The MFI default-instance distributions used
Exhibit 3 compares diversification provided by select
here have been estimated using the performance of
MOSECTM and single-originator transactions. MOSECTM
securitised microloan pools of the three MFIs (Anand &
transactions provide higher diversification across
Fernandes, 2012). We refer to these MFIs as MFI-A, MFI-B
originators and geographies.
and MFI-C. The first panel of Exhibit 4 shows the estimated
default-instance distribution for three MFIs8.
MOSEC : understanding underlying
TM
Assume that each MFI participates in a single-originator
default risk transaction and each transaction has a microloan pool of
MOSECTM structures reduce overall transaction risk through size INR300m underlying it. Also, assume that all
increased diversification across originators and geographies. underlying microloans in a pool are of equal size. Based on
Originator-wise
diversification
Geography (state)-
wise diversification
33
*multi-originator securitisation
80% 80%
—— MFI-A** —— MOSEC (0)
—— MFI-B** —— MOSEC (0.25)
Probability
Probability
60% 60%
—— MFI-C**
40% 40%
20% 20%
0% 0%
0% 1% 2% 3% 0% 1% 2% 3%
Default instances (%) Default instances (%)
*multi-originator securitisation **Microfinance institution
the historical default-instance distribution of each performance of MOSECTM transaction is simulated using the
originator, the default-instance distribution of underlying default-instance distribution of each originator. The
pools can be estimated as shown in first panel of Exhibit 4. simulations are performed for two scenarios; first with the
assumption that the three microloan pools have zero
Using this distribution, we calculate the maximum number
correlation among them and the second with the
34 of default-instance to be observed at a certain confidence
level for three single-originator transactions. The
assumption that pools have a correlation of 0.25 10. The
default-instance distribution for the MOSECTM transaction is
calculated values are tabulated in columns two-four of
estimated for different correlation values: correlation=0
Exhibit 5. For the single-originator transaction with MFI-A,
there are 99.99% chances that the observed defaults will and correlation=0.25. The estimated distributions for the
not exceed 2.24% of the underlying pool. At this MOSECTM and single-originator transactions are compared
18
16 䊏 MOSEC (Corr = 0)
Credit-enhancement to provide
2.5 times credit cover (INR m)
35
Centralised structuring, reporting originators’ MIS and IT capabilities to ensure their
post-transaction compliance with reporting requirements.
and monitoring
Post-transaction monitoring of originator-cum-servicer
In a MOSECTM transaction the microloans from multiple
operations is essential to ensure consistently high
originators are pooled into a single Trust. The securities
performance. It enables the detection of early warnings
issued to investors are backed by cash flows from the
and stress signals which can then be addressed in time. In
pooled microloans. These cash flows are tranched
MOSECTM transactions, the required monitoring effort
depending upon the investors’ risk-return requirements.
increases many fold due to the presence of multiple
However, the timing and frequency of the cash flows may
originators. However, MOSECTM transactions help build
vary significantly from one originator to another. Thus an
efficiencies through centralised processing of the
effective MOSECTM structure requires adequate structuring
transaction. The structuring and performance management
and underlying sector expertise.
functions are typically centralised and offer economies of
Reporting and monitoring are two of the key components scales when pools of multiple originators are combined.
of risk and performance management in a securitisation Since the closure of the first MOSEC TM transaction in
transaction. Servicers report the collections, pre-closures, January 2010, IFMR Capital has consistently demonstrated
prepayments and overdue cases for every collection the usefulness, performance and sustainability of the
period. These reports form the basis of all risk and structure for enabling access to capital for small and
performance analysis for the transaction. In MOSEC TM medium sized MFIs. The same model with similar outcomes
transactions, the challenge is to ensure uniform reporting could be used for any asset class provided the same
standards across the servicers to maintain data integrity. It standards of high quality origination, structuring and risk
is essential to perform a pre-transaction due diligence of management are applied.
Transaction Collection
Date Transaction name Number of MFIs* Rating agency size (INR m) efficiency
15-Jan-10 IFMR Capital MOSEC-I 4 CRISIL 308.6 99.03%
14-May-10 IFMR Capital MOSEC-II 3 CRISIL 353.7 97.54%
16-Sep-10 IFMR Capital MOSEC-III 3 CRISIL 368.8 98.95%
17-Jun-11 IFMR Capital MOSEC-IV 3 ICRA 272.3 98.54%
30-Jun-11 IFMR Capital MOSEC-V 4 ICRA 269.0 99.67%
02-Aug-11 IFMR Capital MOSEC-VI 3 ICRA 276.3 98.37%
21-Aug-11 IFMR Capital MOSEC-VII 7 ICRA 511.9 99.77%
07-Oct-11 IFMR Capital MOSEC-VIII 4 ICRA 200.9 98.73%
09-Nov-11 IFMR Capital MOSEC-IX 5 ICRA 248.1 99.98%
12-Dec-11 IFMR Capital MOSEC-X 6 ICRA 492.8 100.00%
27-Dec-11 IFMR Capital MOSEC-XI 8 ICRA 574.6 99.91%
23-Feb-12 IFMR Capital MOSEC-XII 9 ICRA 472.1 99.87%
30-Mar-12 IFMR Capital MOSEC-XIII 6 ICRA 351.2 100.00%
4. Sane, R. & Thomas, S. (2011). A policy Response to the features of structured finance: pooling of assets, tranching of
liabilities backed by pool cash flows and delinking of credit risk
Indian microfinance crisis (Working paper-2007-11).
through an SPV.
Retrieved from Indira Gandhi Institute of Development
4 Average portfolio size for all securitisation transactions arranged by
Research website: http://www .igidr.ac.in/pdf/publication/ IFMR Capital is INR250m.
WP-2011-007.pdf 5 Source: IFMR Capital.
37