In a recent post, Daniel Rozas and Vinod Kothari have argued that microfinance securitisation does not, in reality, separate ‘pool risk’ from ‘originator risk’ and hence should not be rated very differently from the originator of the portfolio.
Some of the arguments are well reasoned out such as the lack of control over co-mingling of cashflows and lack of regulatory supervision on ‘bilateral assignments’. However, some of the other arguments presented beg greater scrutiny.
Microfinance securitisation has really taken off in the last one-and-a-half years. Prior to that, MFIs, like other financial institutions, sold assets on a bilateral basis to a banks, which were largely incentivized by regulatory benefits. Bilateral assignments still constitute a majority of off-balance sheet transactions. Such deals are opaque, largely unrated and mostly occur in the last quarter of the financial year.Such bilateral deals do not constitute ‘securitisation’. Securitisation is a market-driven transaction, priced against comparable rated benchmarks, tracked by independent agencies and transparent in nature. As such, to tar bilateral deals and market-linked securitisation transactions with the same brush is to do injustice to the tightly structured microfinance securitisation market that has recently developed.
Let us focus on the fundamental question that the authors have raised about microfinance securitisations in this context.
How can assets be rated better than the originator?
The premise of any securitisation is that the underlying assets can be ring-fenced from the originator – cash flows from such assets are pooled into a secure account controlled by a trustee. Further, it is possible to select high quality assets, originated in less risky geographies from borrowers who have historically had good credit history. Third, the assets are not exposed to servicing risk precipitated by a bankruptcy event.
The rationale for a securitisation of assets to have a better rating than that of the originator is that the assets being securitised are of higher credit quality than that of the originator’s own credit quality (which comprises its entire balance sheet).
Vinod Kothari describes securitisation aptly as “the process of de-construction of an entity. If one envisages an entity’s assets as being composed of claims to various cash flows, the process of securitisation would split apart these cash flows into different buckets, classify them, and sell these classified parts to different investors as per their needs. Thus, securitisation breaks the entity into various sub-sets.” Further, he describes receivables securitisation as a way to “originate an instrument which hinges on the quality of the underlying asset. As the issuer is essentially marketing claims on others, the quality of his own commitment becomes irrelevant if the claim on the debtors of the issuer is either market-acceptable or is duly secured. Hence, it allows the issuer to make his own credit-rating insignificant or less-significant, and the intrinsic quality of the asset more critical.”
The arguments in the afore-mentioned article against superior rating of the asset pool are twofold:
(a) Collections are manual, hence co-mingling risk is high and effectively removes ring-fencing (in fact the authors have gone so far as to say that MFIs may pay the investor the scheduled amount regardless of the actual collection) and
(b) if the servicer fails, there is no alternate mechanism to collect from the end-borrowers.
However, in the microfinance securitisation transactions, several structural mitigants have been built in to counter servicer risk.
First, it is mandatory for the MFI to pay the trustee of the SPV on a weekly basis. Hence, the co-mingling risk is reduced to a week’s cash flows. This risk is accounted for while taking into account the rating of the transaction (Please refer to the rating rationales available on IFMR Capital’s Deal Portal)
Second, the MFI maintains a first loss cash collateral from the onset of the transaction. It is difficult to understand why an MFI would make scheduled payments to the trustee despite receiving lower collections (and hence ‘mask’ the true pool performance), while at the same time maintaining a cash collateral – a double whammy! Would it not make better financial sense to let the cash collateral be debited at the end of the transaction?
Third, it is mandatory for MFIs to provide a detailed MIS on collections to the trustee with every payment. The Arranger and the trustee play a key role in in terms of diligence of the MIS. This reduces the risk of fraud.
Fourth, one of the key benefits of the structure is to mitigate the risk of systemic default. The structure provides diversification across branches, geographies, loan types such that the risk of an external shock is minimal. Such diversification can be achieved across multiple originators, where the underlying diversity in a single originator is not sufficient.
Fifth and most important, IFMR Capital’s monitoring and surveillance setup is a critical early warning signal system that has been set up to counter this risk. The monitoring team consists of microfinance professionals with deep experience on the field and strong networks across the sector. The team holds both close relationships with the top management and has indepth knowledge of field operations through these visits. It diagnoses weaknesses in the MFI’s operations and/or systems and provides inputs on corrective measures to the MFI. This continued involvement and diligence by the Arranger post transaction settlement differentiates microfinance securitisation as it is today from securitisation of other asset classes.
In rating microfinance asset backed securitisations, CRISIL has indeed been conservative in its assumptions, due to the limited track record of microfinance securitisations in capital markets. On a more careful analysis of the securitisation structure, the authors of the “Hidden Risks” article should note that the underlying pool of micro-loans securitised, despite their low default rates, have not been assigned a higher rating than that of the originator. After a detailed analysis of default rates of the underlying micro-loan pool, its volatility, the sustainability of the originator and commingling risk, the rating agency arrives at an assessment of base case default rates and the amount of additional credit protection required to achieve the ratings of the senior and junior tranches of securities. The authors’ contention that the rating agency in question “apparently does not understand” the underlying sector and the assets is spurious as CRISIL has assessed more than 140 MFIs over the past eight years and has in fact developed a separate methodology for this sector. CRISIL does not only rate microfinance securitisations, they also provide risk assessments, ratings on bank facilities and debentures. This is accompanied by the arranger, IFMR Capital’s due diligence, monitoring and supervision processes of each MFI that is a potential candidate for securitisation. This is the crux of what goes behind the “P1+ or AAA” rating!
The rating agencies’ approach has been consistent across other retail asset classes such as personal loans, commercial vehicle loans, car loans and small business loans. For instance, Tata Motors executed more than Rs. 2000 crores of rated securitisations in the year 2009. The senior tranches were assigned a AAA rating, while Tata Motors’ senior secured loan rating is at A+. While there is capital markets supervision of retail assets by investors and rating agencies, there is no back up servicer in place for any of these retail loans. What is important in the securitisation of any retail asset is the risk assessment of sustainability of the originator over the life of the underlying asset, a detailed analysis of the underlying pool and a robust monitoring and surveillance system that can monitor triggers that may indicate increasing servicer risk.
So how have the ‘AAA’s’ performed till date?
IFMR Capital has structured, arranged and co- invested in six rated microfinance securitisations comprising a total financing amount of Rs 195 crore
|Type of Deal
|Size (Rs crore)
|Sahayata, Satin, Asirvad, Sonata
|Grameen Financial Services
|Sahayata, Satin, Asirvad
|Grameen Financial Services
While this is a drop in the ocean in the context of the entire banking and finance sector, this is important as it signifies a way for high quality microfinance institutions to access debt capital markets, in a framework that builds in tight supervision, and rewards better access at better pricing to those microfinance institutions that are underwriting loans well.
Micro-loan securitisations have shown remarkably strong performance with 0.6% defaults in the underlying pool till date. Four of twelve tranches of securities have been upgraded. However, to maintain this performance, IFMR Capital recognizes that “event risk” and systemic risk such as the originator’s or the arranger’s incentives not being aligned must be carefully considered and incorporated in the design today itself.
The author quotes the example of a political risk event in Krishna district, in which MFIs suffer significant losses in their micro-loan portfolios. However, the largest lenders to MFIs operating in Krishna, have four years later, entirely recovered their dues from this district. This is a promising indicator that recoveries are possible, and that setting up an independent backup servicer that monitors & supervises performance and “steps in” in the worst case, is a probable future reality.
To mitigate systemic risk, the off balance sheet micro-loan securitisations are structured in a manner that the incentives of all the participants are aligned to those of the investors. Securitisation structures “pre- global financial crisis”, by and large did not have sufficient incentives for the originator or the arranger to underwrite and monitor well post sale, this is one of the important lessons we have learnt from this crisis.
Incentivizing the Originator: The MFI or the Originator has to ‘originate wisely’ and ‘collect well’. The investor carries the risk of being sold the riskier assets in the MFI’s portfolio as well as having to depend on the MFI to collect from assets that no longer belong to the MFI.
This risk is mitigated by the MFI retaining the first loss portion of the risk under the pool in the form of cash collateral or an irrevocable bank guarantee. The size of the cash collateral depends primarily on the following factors – historical default rates and their volatility, the extent of co-mingling of cash, an assessment of the originator’s sustainability,and prepayment rates.
The first loss default guarantees in the securitisation transactions till date have been in the range of 10-14% of the transaction size, while the observed default rates on the pools have been less than 0.6%.
The Reserve Bank of India has mandated that the first loss default guarantee is deducted from the Tier I and Tier II capital of the originator. Thus, this provides a powerful incentive for the MFI to ensure that there is no slip-up in its efforts to collect from the underlying pool.
Incentivizing the Arranger: An important risk that is not mentioned in the quoted article is mis-selling. This is even more pertinent for microfinance where there are very few listed organizations, very little public information and very little understanding of how microfinance actually works.
The Arranger in a capital market transaction is best placed to understand the originator and has the responsibility of effectively communicating its strengths and weaknesses to target investors. However, arrangers in debt capital market transactions usually have little involvement with the transaction once the placement is concluded.
However, in each of the microfinance securitisation transactions highlighted above, the Arranger (IFMR Capital) is also an investor in the subordinated tranche of the issuance. The Arranger thus takes the risk of having a junior claim on the asset cash flows. This is a strong incentive for the Arranger to conduct due diligence, monitoring and supervision properly.
IFMR Capital also takes on the responsibility of detailed credit surveillance of the Originator. This is in the form of quarterly visits that cover not only field operations in multiple geographies, but detailed financial analysis and operations review as well. These monitoring and surveillance visits highlight early warning signals well ahead of any potential stress event. Documentation signed with the MFI gives substantial powers to the Arranger to step in and conduct such discretionary audits.
Microfinance securitisation is an important alternative for financing that has opened up to MFIs in India. This permits smaller MFIs to reduce their dependence on a single source of financing, access the capital markets and avail of transparent, market-linked financing, as opposed to opaque, bilateral transactions. These transactions compel the MFI to adhere to the rigor of financial markets, share data, increase external oversight and improve internal systems and processes to meet capital market standards.
Risks in microfinance securitisation are recognised, have been sufficiently highlighted by both the rating agency and the Arranger, are similar to any other asset class securitisation in India and investors are compensated for the risks in the returns on such investments. Perhaps we should devote our energies on strengthening the mitigants instead.