By: David A. Smith
I don’t want to cry panic, but if even half of what Daniel Rozas writes in this Micro-Finance Focus article is true – and it’s his beat, so it ought to be – then for the MFI industry, the question is not whether the cheap-money bubble will pop, but just how severe the popping will be.
Microfinance Focus, November 08, 2010: The crisis in Andhra Pradesh has highlighted how exposed MFIs are to mass non-payments. Industry insiders have suggested that even some of the largest MFIs simply might not survive if the crisis is not resolved soon.
Any lending institution takes on risk as measured by a common-sense equation, R = P x L
Risk = Probability of default x Loss if a default occurs
Like many great equations, a simple one
(All slides from my talk at the South Asia Housing Finance Forum in Delhi)
In a microfinance institution (MFI), the customer is a person and the asset lent has little collateral value, so L = 100%: if there is a default, the net loss is likely to be a complete writeoff of the outstanding loan amount. (Net loss, remember, so that recoveries are offset by the cost of collection.) With L so high, the MFI must minimize P, the probability of default. Microfinance institutions pursue this through:
· Small loan amounts
· Short tenor loans
· Know Your Customer protocols
· Credit denial as the enforceable threat
As Mr. Rozas puts it:
One of the greatest incentives for borrower repayment in microfinance is the expectation of future loans.
Like other loan collection systems, microfinance’s is predicated on infrequency, defaults being rare. But protocols like Know Your Customer can collapse when the business is rapidly expanding, and the threat of credit denial weakens if the industry is undergoing systemic credit freeze. Does this mean that the microfinance business faces a high correlation risk?
Danger young microlender!
And if [a large MFI were to go under], is the industry prepared to deal with the process of unwinding one of these giants?
The absolute most terrifying thing for a lender, even worse than an epidemic of foreclosures, is payment boycott. Collateral-based enforcement systems cannot cope with it. That’s why strategic mortgage default, whether real or unreal, looms over the industry as such a specter.
We’re patient
The top MFIs in India are large by any standard, with assets in the multiple $100s of millions, most of which are held in the form of outstanding microcredits.
Yet, as we saw with Fannie and Freddie, a lender is good only if its paper is good.
Once an MFI is hobbled to the point that it cannot survive as a going concern, what happens to these assets? Experience from other MFIs suggests that prospects for recouping them are not good.
When a bank goes under, nearly all of its balance sheet is in paper – loans receivable. They have to be transferred to a going-concern bank, a transfer induced and financed by the banking regulator – in the US, FDIC. The loan book can be priced within a range because the collateral is collectable. Real estate does not move, so you can track it down, and as long as the structure is physically sound and there is a valid title or similar evidence of possession, you can recoup some or all of the delinquency. Not so for microfinance loans:
Microcredits are unusual assets in that their value is tied to the MFI that created them.
Unlike mortgage finance – which is predicated on tangible collateral that informs the credit decision — microfinance is consumer finance, without collateral, and therefore L, Loss given default, can run away to 100%.
The empty space between microfinance and mortgage finance
When an MFI is closed, its borrowers will simply not pay anyone else unless provided with clear incentive to do so.
If Mr. Rozas’s view is true – and I have no reason to doubt it – then this is a huge systemic risk in MFIs absent from other forms of lending. Loans cannot be executory contracts, lest they be worthless as collateral, yet evidently many borrowers think of them so.
Examples where liquidations of MFIs through sale of their assets were successful are rare, and usually involve elements in the MFI’s business model that limit the importance of the borrower-MFI relationship, such as the use of collateral.
A collateralized microfinance loan is, if not an oxymoron, at least an extreme rarity.
None of these elements are used in any significant way by Indian MFIs.
Quality collateral drives down L to levels much lower than 100%, and that revamps the business model.
The home is an asset to figure in a credit decision
Thus, the commonly-used approach in liquidations – closing down the institution and selling off its assets – is guaranteed to result in near-complete loss of those very assets.
L = 100%, not just for one loan in the book but for all loans in the book. Shutting down the MFI thus destroys massive value – again, presuming Mr. Rozas knows his stuff.
A successful windup must thus provide for retaining the MFI’s operations for some period.
Much like using a Chapter 11 bankruptcy reorganization that then converts into a Chapter 7 bankruptcy liquidation.
However, that alone is not enough. One of the greatest incentives for borrower repayment in microfinance is the expectation of future loans. Once borrowers perceive that those loans are not forthcoming, they will stop repayments in droves.
This is terrifying because it is so plausible. Any cluster of borrowers can be induced to boycott repayment if they collectively concluded that either (a) there is no penalty for defaulting, or (b) further ruin of their credit is irrelevant.
The observant herd will do together what the members will not do singly
Borrowers are especially likely to feel this way if they believe that their current delinquency is ‘not their fault’ because it was causes by arcane and hence malevolent outside factors. An external shock – say, a credit squeeze at the regulatory level – could seize up all the MFIs at once, preventing them from lending to each other’s customers. Conversely, an unopposed payment boycott could scare liquidity completely out of the sector, creating its own credit vacuum.
That poses an inherent dilemma – if an MFI is to be closed, it can’t be making new loans, which would in turn have to be followed by additional loans, and so on.
The result would be an MFI shutdown epidemic, followed by a regional or even national customer payment boycott. Such a trend would be so powerful few governments could stand against it – the more so as borrowers would see government as their benefactor and expect that government would ‘of course’ forgive the loans.
If you think this scenario farfetched, study the history of South Africa’s Gateway program, of new-build secured home lending in formerly black townships shortly after full democracy. Nationwide payment boycotts, enforcement strikes by municipal officials, and an entire program collapsed, wrecking lender confidence in inclusionary lending for a decade – a precious decade. As we wrote in 2003:
As a result of a range of factors including rising interest rates, and increasing unemployment, not to mention inappropriate loan approvals, many financial institutions faced a rising crisis of properties in possession in the early 1990’s.
Gateway prefigured our subprime crisis. The program was announced with great fanfare as a government-endorsed initiative to expand homeownership rapidly in the first flush of enthusiasm after full democracy. Loans were variable rate, which proved catastrophic when inflation and interest rates spiked – wiping out the borrowers’ equity and putting them into virtually guaranteed payment default, so they became economically worse off than renters. Those same macroeconomic factors threw out of work a lot of the very people who had been stretched into buying properties with Gateway loans – including a combination of rose-colored underwriting and strong government moral suasion to take an optimistic view of the credit decision.
Expanding the housing supply and homeownership … with interest rates that shot up above 20%
In June 1995, government and the banking sector established a private company known as Servcon Housing Solutions. Servcon’s initial, three year mandate was to deal with an estimated 14 000 repossessed properties and non-performing loans as a result of bond boycotts and consumer affordability problems.
When the borrowers, who for the most part were first-time home buyers, discovered they could not repay their loans, they defaulted en masse. As the defaults were concentrated in a few places, notably formerly black townships, both borrowers and local authorities jumped to the erroneous conclusion that this was simply racism disguised as economics. Hence the payment boycott was accompanied by an enforcement boycott – local authorities simply refusing to evict homeowners. The banks had the worst of both worlds – losing money on an epidemic of non-performing loans and being publicly vilified to boot.
Servcon was a hastily created South African equivalent of our Resolution Trust Corporation, with the banks selling their bad paper into Servcon at a severe discount (I recall it being 50%) that was nevertheless a good price relative to the loan’s practical fair market value.
In March 1997 the mandate was extended to include an additional 10 500 properties. Servcon’s current [2003 – Ed.] portfolio comprises some 33 089 properties which went into default before 31 August 1997.
Unlike other countries where the housing market is stable, the problem of properties in possession in South Africa is particularly difficult because households often refuse to vacate their property. Servcon’s job is to negotiate a solution that might include being “right-sized” into more affordable accommodation.
In both cases, India having no RTC or Servcon mechanism, there is only one solution:
The most obvious option is through a merger with a stronger institution, which can retain the MFI’s operational capacity and provide the needed capital to continue issuance of new loans.
Almost by definition, a consolidation-by-stronger scenario presumes that secondary-market liquidity credit denial afflicts not the sector as a whole, just its weaker members. In the US, we thought we could contain our problem to the subprime lenders. We couldn’t. Although microfinance’s mileage may vary, the sector has exactly the same problem, in miniature.
Prepare to defend your financial positions
Unfortunately, the breadth of the current crisis reduces the chances that other large MFIs would be in a position to effect such a buyout. It is possible that an outside party might seek to take over, but it is not a given – microfinance in India is not a sector whose future prospects are in any way clear at this stage.
How will India deal with the challenge?
[Continued tomorrow in Part 2.]