Both approaches are excessively penalising, making SES economically unviable. The only sensible approach is a one-year deduction on a rolling basis, which is coherent with the design of UIOLI SES calibrated on EL at 1 year. This is the so-called “alternative approach” referred to in the draft RTS.
The full model approach is based on lifetime deduction of SES. As highlighted since the discussions at the time of the Capital Markets Recovery Package, lifetime deduction makes SES uneconomical, i.e. it destroys the economics of the transaction. If retained, originators would have to abandon SES and place thicker non-senior tranches instead.
Article 6 proposes determining the exposure value of a synthetic excess spreads position as the average of losses absorbed by the synthetic excess spread mechanism across the three regulatory loss timing assumptions. Where the future excess spread amounts are variable, Article 4 stipulates that the expected available loss absorption amount is to be used as input to the expected loss allocation amount. No further adjustments will be made to account for uncertainty and the timing of future designation of loss absorption amounts.
This mechanism would disadvantage SES clauses compared to traditional securitisations. Section 3 para 8 (ii) correctly points out that in synthetic securitisations, unlike traditional securitisations, future proceeds from the securitised portfolio will continue to be recorded in the originator’s income statement. However, where SES amounts are contractually linked to this future income (for instance by referencing the outstanding exposure amount, contractual credit spread and subtracting the originator’s funding cost for the retained position), any future loss absorption will only be achieved out of future income. In this scenario the ex-ante capital base will never be impacted by the SES mechanism.
We propose reflecting this in the calculation of the exposure value. This could be achieved by adding the following sentence (either to Article 4(b) or to Article 6):
“Where variable synthetic excess spread is contractually linked to income generated by the securitised portfolio, the amount of synthetic excess spread at any point shall be capped at the amount that has already been reflected in the originator’s income statement.”
As a result of this addition, synthetic excess spread clauses that contractually cannot impact the capital base at inception of the transaction will not lead to exposure values. However, where income has been recorded in the income statement (either intra-period or in trapped mechanisms), the resulting loss absorption potential will be accurately reflected as exposure value, comparable to a traditional securitisation transaction.
In cases where variable synthetic excess spread is not contractually linked to income generated by the securitised portfolio but to other metrics, the wording proposed above would not apply and the exposure value would be determined as outlined in the consultation paper.
As acknowledged by the EBA, the proposed scalar will result in capital deduction way above current levels (based on one-year deduction on a rolling basis). It will therefore make the SES feature uneconomical.
The only sensible prudential treatment of SES is a one-year deduction on a rolling basis for UIOLI calibrated at 1-y EL. Higher capital requirements will destroy the economics of SES.