4.3Quick understanding of the securitisation process:
The essential securitisation process is well known and is explained at length in several texts14. We spend some time to quickly introduce the essential process, which may be skipped by initiated readers. We explain both the agency-backed method and the private label method, though the two are substantively similar. We also add that in India so far, in the RMBS segment, all securitisations have been agency-backed.
In this case, an mortgage originator having accumulated a sufficiently large mortgage pool sells the pool to an agency, say, National Housing Bank (NHB). NHB sets up a special purpose vehicle in form of a trust, with NHB being a trustee thereof. The special purpose vehicle buys the pool and issues several classes of securities, usually designated as either senior or junior securities, or as Class A, class B, class C and so on. Generally the senior securities or Class A will have a high rating, usually AAA. The securities may be in form of either bonds or pass-through certificates – in India, so far the pass-through structure has been used15. The servicing of the mortgage pool is retained by the mortgage originator, who continues to collect the receivables on the sold pool, and remit the proceeds to investors. If there is a residual surplus in the pool, it is typically paid to the junior-most of the securities.
Private label securitisations
The essential process is the same here – with the only difference that instead of an apex agency such as NHB, there is a special purpose vehicle that would typically be created for the transaction. Owing to isolation and non-consolidation legal criteria, this SPV will not be owned or controlled by the originator – hence, there will be an independent trustee to monitor the SPV’s affairs.
In both the cases, the repayment to investors flow from out of the pool and are not guaranteed by either NHB or the trustees.
4.4Why securitisation should reduce weighted average funding costs:
Our arguments as to why securitisation must result into reduced weighted average cost of funding run as under:
We compare two broad forms of funding – on-balance sheet and off-balance sheet, the latter by way of securitisation. We assume that both are fixed income, debt-type sources.
The on-balance sheet funding is characterized by the following:
The rating of the funding method, say, bonds or loans, is driven by the rating of the mortgage originator.
The lender/investor will be concerned with the bankruptcy or default risk of the mortgage originator, which in turn is affected by all the several risks affecting the business of the originator.
There are capital adequacy considerations – therefore, for every on-balance sheet funding raised, there has to be an equity component required.
The off-balance sheet funding or securitisation is characterized by the following:
The rating of the transaction is independent of the rating of the originator – this is because the transaction isolates the assets from the bankruptcy risks of the originator altogether and therefore, ratings are asset-based;
The perceived risks of the investors are based on the rating – the senior class in securitisations will typically carry a AAA rating. Thus, investors invest in mortgage-backed securities at far lower spreads than if the originator were to issue on-balance sheet bonds.
Securitisations typically carry capital relief. The originator is expected to put in first loss support to the transaction, but the extent of the same may be expected to be lower than the element of equity in on-balance sheet financings. This is due to the fact that with isolation of the assets, investors in a securitisation transaction are not affected by other entity-wide risks of the originators’ business. Investors are concerned only with asset-based risks of the isolated pool – therefore, the credit enhancements required to support such risks at a AAA-rating confidence level are considerably lesser.
Thus, the rating independence creates a rating-arbitrage – the transaction gets a better rating than the originator. The isolation of assets creates a leverage-arbitrage – the transaction gets higher effective leverage than an on-balance sheet funding. The combined effect of the two arbitrages is to bring down the cost of funding for the originator.
If it be accepted, based on the above argument, that securitisation will drive down the cost of funding of a mortgage originator, there is no reason as to why securitisation volumes should remain only 0.5% of disbursements. The syndrome of spreads having been compressed in a competitive scenario should only promote securitisation rather than discourage the same.
Therefore, it is quite obvious that the reasons for the lackadaisical securitisation activity in India have to be traced in external inefficiencies. The external inefficiencies have stifled the number of transactions coming to the market. With sparse transactions, rating agencies do not have the benefit of experience of rating survival or migration history, and therefore, they willy-nilly tend to exaggerate credit enhancements which unduly increases the cost of funding. Thus, the a priori argument of reduced cost of funding falls due to market inefficiency, which itself is a creation of inefficiency of the tax and regulatory system.
On the other hand, if the imperfections are removed, a larger number of transactions will enter the market. The cost of funding will be reduced, which in ultimate analysis will also reflect in lower cost of housing finance. There will be lots of lots of other significant benefits. Investors in the fixed income market will get an alternative investment avenue, Various classes of MBS will offer a wide choice to potential investors in terms of risk-return profile, liquidity, maturity, yield etc. and help attract new investors into the long term debt market. Innovations in the US RMBS market in terms of product design have evinced that the RMBS market has potential for creating some very interesting products which are differently sensitive to interest, and therefore, prepayment rates – such as IO strips, inverse floaters, etc.