Stress-Testing a Microfinance Securitisation and Choosing Between Structures
Pragati Microfinance (AUM ₹3,100 crore) is securitising ₹450 crore of 24-month microfinance receivables. The pool yields 19.5% to the originator on a reducing balance; expected credit loss is 3.0% over the life; servicing is 1.1% a year. Senior pass-through certificates of 82% are priced at 11.25%. The originator retains a 13% subordinated tranche and funds a 5% cash collateral, both from its own balance sheet at an incremental funding cost of 9.0%. The deal has a 12-month revolving period, an excess-spread trap triggered when 90+ dpd exceeds 4%, and a clean-up call at 10% of the original pool balance. Two quarters in, 90+ dpd has risen from 2.1% to 4.6%, concentrated in two Maharashtra districts hit by unseasonal hailstorm; the trustee has asked for a ₹6 crore cash-collateral top-up. The originator's bank lines cost 13.8%, and a bank has separately offered a ₹150 crore direct assignment at par with 10% retention.
1. Stress the waterfall at a 9% lifetime loss and determine whether the senior PTCs are protected and how much of the originator's own money is consumed.
Read the worked answer
Sources of protection.
| Layer | Amount (₹ crore) | % of pool |
|---|---|---|
| Cash collateral | 22.5 | 5% |
| Subordinated PTC | 58.5 | 13% |
| Total credit enhancement | 81.0 | 18% |
| Excess spread (annual, pre-loss) | ~41 | ~9.2% p.a. |
| Lifetime excess spread over ~1.2 yr average life | ~50 | ~11% |
Stress case: 9% lifetime loss = ₹40.5 crore.
The waterfall applies excess spread first, then cash collateral, then the subordinated tranche. A 9% loss is three times the 3.0% base case, so the base-case ₹13.5 crore loss becomes ₹40.5 crore - an incremental ₹27 crore absorbed entirely by the originator's layers. Rough lifetime cash flow: interest received ~₹105 crore plus principal ₹450 crore = ₹555 crore; senior principal and coupon ₹369 + ~₹50 = ₹419 crore; servicing ~₹6 crore; losses ₹40.5 crore; residual ~₹90 crore against the ₹81 crore the originator has at risk.
Conclusion. The senior is protected - total enhancement of 18% plus excess spread comfortably exceeds a 9% loss. But the originator recovers only about its stake, i.e. the subordinated tranche is effectively written down to near zero and the cash collateral is largely drawn. At the 4.6% dpd level the 4% trap has already triggered: excess spread is now being diverted to the cash collateral rather than released, which removes the cushion that would have absorbed the stress.
Crucially, the 12-month revolving period is the real risk. If the trigger suspends reinvestment, the originator loses a ₹369 crore funding line at short notice while its own loan book keeps growing - a liquidity event, not merely a credit event.
2. Compare the all-in economics of securitisation, the 13.8% bank line and the ₹150 crore direct assignment offered by the bank.
Read the worked answer
Securitisation, per year on ₹450 crore pool.
- Senior coupon: 11.25% × 369 = ₹41.5 crore
- Expected loss: ₹13.5 crore over life ≈ ₹6.8 crore a year
- Servicing: 1.1% × 450 = ₹4.95 crore
- Total ≈ ₹53.2 crore on ₹369 crore of cash raised = ~14.4%
That is slightly more expensive than the 13.8% bank line before any capital relief. The case for the structure therefore rests on three things bank borrowing does not give: (i) capital release - the pool is de-recognised, lowering risk-weighted assets and improving CRAR; (ii) diversification away from a single funding source that is already close to its single-bank exposure limit; and (iii) the originator retains the excess spread above the senior coupon, so the marginal cost to equity is lower than the headline 14.4% suggests.
Direct assignment. ₹150 crore assigned at par with 10% retention: the originator raises ₹135 crore of cash, keeps ₹15 crore of the first-loss position, and - importantly - avoids structuring costs and the trustee machinery of a full securitisation. Retention of 10% versus 18% here means less of the originator's capital is locked up.
Judgement. The two instruments solve different problems. Direct assignment is cheaper, faster and lower-risk for a ₹150 crore need; securitisation is the right tool for scale and capital relief. Given the current dpd trigger, this is the wrong moment to launch a fresh securitisation - the pool data will be re-underwritten by investors at a wider senior spread. The pragmatic sequence is to use direct assignment now to demonstrate clean performance, restore the dpd trend, and re-enter the securitisation market next year.
3. Should the originator top up the ₹6 crore cash collateral? What are the accounting, servicing and governance consequences?
Read the worked answer
Yes - top up, but with conditions. The arithmetic is decisive. Refusing a trustee call of ₹6 crore risks an event of default on ₹369 crore of senior paper. The consequences are not confined to this deal: a default would be disclosed, would impair the originator's ability to securitise for at least two to three years, would trigger cross-default clauses in bank lines, and would almost certainly cost a rating downgrade across the whole book. ₹6 crore against that risk is trivially cheap insurance.
Accounting. The top-up must be recognised as a financial asset at fair value, not as an expense, provided it is contractually recoverable from future excess spread. If the revised loss estimate shows the cash collateral is unlikely to be returned, an impairment provision is required immediately rather than being smoothed. Under Indian accounting standards this is also a credit-enhancement asset that must be disclosed as an off-balance-sheet exposure in the capital adequacy computation - it is not free.
Servicing. The originator is both servicer and residual holder, a structural conflict. The trustee should require district-level collection data for the two hailstorm-affected districts, a moratorium policy that is applied consistently to the on-book and securitised books (regulatory rules typically require non-discriminatory treatment), and independent verification of collections. Diverting good collections to the on-book portfolio at the expense of the pool would be a breach.
Governance. The board should treat this as a risk-appetite event, not a treasury entry. The relevant question is whether the underwriting standard in the securitised pool differed from the on-book standard, and whether the 4% dpd trigger was calibrated for a portfolio with geographic concentration. Two districts out of a dispersed pool should not have produced a 250 bps jump; that points to concentration limits that were never specified.