Reserve Bank · Essay № 18

FLDG, after the FLDG died — and the cap of five per cent that brought it back.

A reading of the June 2023 circular that restored a banned credit-enhancement mechanism, with a few new sentences attached.

In September 2022, the Reserve Bank of India effectively ended the practice of first loss default guarantee — the arrangement in which a fintech platform agreed to absorb, up to a contractual percentage, the losses on loans that an NBFC made through the platform. The regulator's objection was structural: credit risk that ought to sit on a regulated balance sheet was sitting, instead, on an unregulated platform's balance sheet, where the regulator could not see it. The ban was applied without ceremony. The industry restructured.

In June 2023, the same regulator restored the practice. The restoration was careful, narrow, and accompanied by a new vocabulary. The instrument was renamed Default Loss Guarantee — DLG, not FLDG — to mark the distinction. The permissible forms were limited: cash deposit, bank guarantee, or lien on a fixed deposit, each held with a scheduled commercial bank. The quantum was capped at five per cent of the underlying loan portfolio. The arrangement could be entered into only between a regulated lender and an entity falling within a specified definition. The reporting obligations expanded.

The cap that did the work

The five per cent cap is the heart of the new regime. Before September 2022, FLDG percentages of twenty, twenty-five, even thirty per cent were not unusual; in some cases the platform was bearing more credit risk than the lender. The five per cent cap reverses the polarity: the lender carries ninety-five per cent of the loss, the platform absorbs the first five. The platform has skin in the game; the lender still owns the underwriting decision. The regulator can see all of it.

The DLG regime is what happens when a regulator decides that an arrangement is acceptable in moderation but unacceptable in scale.

What changed, between September and June, was not the regulator's mind about the principle. The principle held: credit enhancement by unregulated parties is dangerous when it crowds out the lender's own underwriting. The change was a recognition that, properly capped, the arrangement can sit alongside a careful lender's underwriting rather than displace it. Five per cent is enough to align incentives. It is not enough to substitute for them.

The industry's settling

The industry's response has been measured. The largest platforms, which had restructured in the intervening months toward pure sourcing and co-lending models, were not in a hurry to undo the restructuring. Smaller platforms returned to DLG arrangements quickly. The aggregate credit risk on Indian fintech-sourced lending today sits, in the regulator's estimation, considerably closer to where it ought to than it did three years ago. This is, by the standards of financial regulation, a complete sentence with no further pages.

Further reading

Three essays that follow on.

All essays