Lending without risk and risk without lending:

The new paradigm of lending partnerships in India

– Vinod Kothari | finserv@vinodkothari.com

In the world of lending, there is a new buzzword – sourcing partnership. This partnership entails the coming together of two entities, both which, let us presume, are financial entities. The one which has strong origination abilities partners with the one which has strong funding abilities, such that credit assets are sourced, serviced and risk-absorbed by the first one (say, Originating Partner), and are housed on the balance sheet of the latter (say, Funding Partner). The Originating Partner takes the credit risk, to a degree sufficient to absorb the expected losses and unexpected losses of the credit assets, continues to service the assets, and eats the entire excess spread, being the difference between the actual portfolio rate of return and the Funding Partner’s expected yield. The Funding Partner puts the loans on its balance sheet, gets only the expected yield, and essentially takes the risk in the Originating Partner, often collateralised by a funding deposit. Thus, the lender has loans with practically no risk, and the originator has risks with no loans.

In regulated transactions between financial entities, similar result could have been obtained with securitisation (originator providing first loss support sufficient to absorb  a certain multiple of expected losses), credit default swaps (In India, credit default swaps are possible only in case of certain debt securities; total return swaps are currently not allowed), etc. This partnership could not have been possible in case of transfer of loan exposures, as the originator is not allowed to provide credit support to the assignee. Such a partnership would also not be strictly possible in case of co-lending. A securitisation transaction, if done with the originator’s risks and rewards of the degree mentioned in our example, would have resulted into on-balance-sheet treatment for the originator.

The sourcing partnership as above has permitted financial entities to achieve what they could not have achieved in a regulated transaction. While regulatory arbitrage is certainly a concern, the larger concern is that the build-up of risks with the Originating Partner goes virtually unseen, as the loans are not on the books of the Originating Partner. The Originating Partner could easily be tempted to abdicate underwriting standards to step up origination volumes, earn credit spreads to make up for the funded guarantee it gives, and pass to the Funding Partner what looks like a risk backed by collateral, but may eventually turn out to be a bunch of subprime loans, made tempting enough for the borrower by features such as interest-only period, or low EMIs in the initial years.

Some granular details

Before proceeding further, let us put some more granular details for the illustrative example:

• The Originating Partner is an NBFC, say, the one with digital lending strengths that originates MSME loans. Typically, these loans are working capital loans, unsecured, and have a tenure of 3 years, ticket size about Rs. 10 lacs. These loans will qualify as priority sector loans (PSL).
• Loan sourcing is done by the Originating Partner, who collects the loan application and forwards the same to the Funding Partner. The Funding Partner does its own credit due diligence (and KYC checks), and decides to approve the credit.
• The Funding Partner is, say, a bank. The bank wants to build an MSME loan book, either for the purpose of meeting its PSL requirements, or simply to diversify its loan book, or simply achieve credit growth.
• The loan carries a rate of interest of, say, 25%. The yield expected by the Funding Partner is, say, 12%.
• The Funding Partner will originate about Rs 100 crores worth loans in a month. The Originating Partner gives a guarantee on a first loss basis to the extent of 10% of this portfolio, and to back up the guarantee, it places a deposit of Rs 10 crores with the Funding Partner.
• The Originating Partner is appointed as a servicer, and gets servicing fee equal to [25% – 12% = 13%].
• Let us assume that the parties did an estimation that the expected losses of this pool will be in the range of 6- 9%.
• As we may notice, the Funding Partner is fully protected for the expected losses. Also the spreads by way of servicing fees are enough to recover the first loss default guarantee provided by the Originating Partner.

Combination of capabilities

Theoretically, differentiated capabilities of financial entities will, as they should, result into alliances. Gone are the days of universal banking. Currently, we have entities with differential strengths. Entities may have strengths in one or more of these – retail deposits, wholesale funding sources, origination abilities in different customer segments, technology strengths, capital, etc.

It will be quite natural for these entities to do a win-win deal by aligning their differential abilities. For instance, if an entity is strong in, say, housing finance assets, and another one is strong in MSME loans, the two would find it much better by each focusing on their strong domains, and exchange their portfolios wrapped by the risk mitigation of the originating entities, if they decided to diversify their books in the asset domain where either of the entity does not have origination strengths.

There are also quite often different strengths on the liability side. There are several depository banks which have a strong franchise with retail investors, and there are periods when their deposit books swell. There are some P2P lenders as well who would have piled up very strong depositor interest. An alliance with origination-strong entities will lead to a win-win proposition for both of them.

During the Global Financial Crisis, financial regulators learnt a lesson that permitting entities to originate assets and transfer the same without skin in the game may lead to adverse selection, abdication of underwriting standards, etc. Hence, minimum risk retention requirements were laid down. Also, capital regulations have always stipulated capital to the extent of first loss risk.

Therefore, as a matter of principle, if there is (a) capital provisioning, and (b) risk retention by the originator, there should not be regulatory concern.

In our example above, there is risk retention by the originator, as the originator is retaining a first loss risk equal to 10%, as also collateralising the same. The originator may also be providing regulatory capital equal to Rs. 10 crore, being 10% of the pool originated[1]. Hence, there is a case to say that this partnership is answering regulatory concerns.

Transfer of loan exposures by another name

In India, there is yet another regulatory concern. Transfer of loan exposures in India is not allowed to be credit enhanced by the transferor.

In the sourcing partnership discussed above, the loan was sourced or referred by the Originating Partner. The parties may contend that the credit decision is entirely that of the Funding Partner, and from the very first day, the loan was put on the books of the Funding Partner. Hence, the case is not one of assignment or loan transfer, and hence, not falling within the regulatory ambit of the TLE Directions.

However, that is what is precisely called regulatory arbitrage, rather than regulatory lapse. There is a credit asset in the books of the Funding Partner, in form of the MSME loan, but is transparently clear that the Funding Partner is leaning on the funded first loss default guarantee of the Originating Partner. Theoretically, the Funding Partner is doing a credit DD, but really, the Funding Partner couldn’t care less for the customer’s risk, as neither originator, nor servicing, nor the risk is with the Funding Partner. Once again, it may be argued that the guarantee that the Originating Partner has given is only 10% of the total portfolio, and there is still a 90% exposure on the ultimate borrowers. However, given the expected loss rates, that 10% will effectively cover 6-9% expected losses, and to an extent unexpected losses too.

In reality, the Originating Partner has a synthetic exposure in the underlying pool of loans, with the risk as well as rewards on the loans. The Funding Partner has an exposure in the Originating Partner, backed by funded collateral. Bereft of technicality, the transaction is no different from a credit enhanced assignment of the loan pool.

The risks of abdicated lending standards

The risk is not merely one of regulatory arbitrage. The risk is one of subprime lending, the temptation for which is almost irresistible in our example. The Funding Partner is fully collateralised for the expected losses, and therefore, is not much concerned about the credit risk of the pool. The Originating Partner is exposed to the extent of 10%, and makes that money from the excess spread. The real risk of the Funding Partner is the risk of losses exceeding 10%, and that of the Originating Partner is the excess spreads will not be sufficient to recover the first loss default guarantee amount.

Therefore, both parties will be inclined to diversify the pool by adding further loans in the pool. And the easiest and fastest way of originating further loans is by lowering the credit bar. There may be interest-only loans, negative amortisation loans or similar features added, besides lowering down the bureau scores etc. It is worth noting that during the run up to the GFC, the fastest selling subprime loan was IO loans and negative amortisation loans.

Some of the NBFCs relaxing their underwriting standards will soon have contagion effect on the market – others will have to do, to retain their market shares. And soon, the market will have a subprime bubble building up.

Is the risk real?

There are several factors that support that the above apprehension is not mere cynicism. The growing cult of unsecured lending in the name of so-called working capital loans may clearly be hinting at the deprecation of underwriting norms.

Anecdotally, it may not be difficult to establish that most of these loans get originated by NBFCs and shifted to banks. The good-old direct assignment does not permit credit enhancements –therefore, NBFCs may have found co-lending, and even better, sourcing partnerships, is the way going forward.

Way out

Synthetic lending exposures, where the risks and rewards of the loans are with the Originating Lender, should be put on the books of the Originating Lender, and should be allowed to be put off the books following standard principles of de-recognition. If the loans are brought on the books of the Originating Lender, that brings the loan into the system – there is a full system of origination checks, internal audit, board room surveillance, risk committee oversight, regulatory capital, etc. that would mitigate the risks. However, the current arrangement is completely off-the-balance-sheet, except by way of the fair value of the guarantee. In true sense, the arrangement is not a mere guarantee, as it is both risks and rewards, and therefore, is a synthetic position on the underlying loans.

---

[1] In case of first loss position in a structured risk transfer, the capital requirement is equal to the first loss position itself, and not [capital requirement X risk retained]. In our example, the capital requirement of the origination is not 15% of Rs 10 crores; rather, it will be Rs 10 crores.