A framework for synthetic STS securitisation - A snapshot of recent developments
In May 2020, the European Banking Authority (EBA) published its long-awaited report (EBA/OP/2020/07) (EBA/OP/2020/07) on the feasibility of extending the framework for simple, transparent and standardised (STS) securitisations to synthetic securitisations. In this report, the EBA proposes such an extension and suggests criteria to be considered when labelling a synthetic securitisation as STS. Moreover, the EBA indicates pros and cons of favourable capital treatment for such transactions. The report is limited to balance sheet synthetic securitisations, which make up the majority of the synthetic transactions on the European market since the 2008-2009 financial crisis.
Appealing not only to securitisation enthusiasts, this report is one of the most interesting files currently before European policy makers and financial market participants. A framework for synthetic STS securitisation would likely stimulate synthetic securitisation, which is an essential tool for banks in managing their credit risk, balance sheets and capital requirements. By freeing up capital, synthetic securitisation might assist in enhancing banks’ capacity for lending to the wider economy and could, thereby, play a key role in addressing the consequences of the Covid-19 crisis. In addition, it may offer opportunities for accessing additional funding sources for implementing the EU's sustainability agenda.
Most recently, the High Level Forum on the Capital Markets Union (CMU) has supported these ideas in its report on a new vision for Europe’s capital markets published in June 2020. The High Level Forum recommends expanding the scope of STS synthetic securitisations and to apply the same regulatory treatment to traditional and synthetic securitisation. This constitutes a cornerstone in a broader attempt of scaling up securitisation in the EU, as a central element of functioning capital markets and a key feature of the post-pandemic EU economy.
This briefing discusses the key issues arising from the EBA report as the principal basis of further legislative action.
When the STS framework was introduced in the Regulation (EU) 2017/2402 (Securitisation Regulation) in late 2017, it was restricted to traditional ABCP and non-ABCP securitisations. Synthetic securitisations – where the transfer of risk is achieved by the use of credit derivatives or guarantees, while the exposures being securitised remain exposures of the originator (Article 2(10) of the Securitisation Regulation) – are excluded. This is because in order to qualify as "STS" inter alia, a true sale of the relevant assets or an assignment or transfer with the same legal effect is required (Article 20(1) of the Securitisation Regulation).
However, Article 45 of the Securitisation Regulation mandates the EBA, in cooperation with the European Securities and Markets Association and the European Insurance and Occupational Pensions Authority, to provide a report on the feasibility of a framework for STS synthetic securitisation, limited to balance sheet securitisation. EBA issued its report in response to this mandate.
The report follows on a discussion paper of September 2019 and takes into account comments received from the industry during a two-month consultation period. It also builds on previous EBA publications, such as the discussion paper on the significant risk transfer (SRT) in securitisation of September 2017 and the report on synthetic securitisation of December 2015. In the latter, the EBA had first set out potential STS criteria for synthetic securitisations and had proposed differentiated capital treatment for certain SME synthetic securitisations, a recommendation which was adopted in Article 270 of the Regulation (EU) 575/2013 (CRR) – as amended by the Regulation (EU) 2017/2401 – at the end of 2017.
Key take-aways from the EBA report
Criteria for synthetic STS securitisation
In its new report, following a comprehensive analysis, the EBA recommends introducing a legal framework for synthetic STS securitisation based on defined criteria. Among these are requirements on simplicity, standardisation and transparency similar to those for traditional (non-ABCP) securitisation, in an attempt to ensure as much consistency across the “STS” label as possible. In addition, the report includes requirements which are specific for synthetic securitisation.
Most of these STS criteria are not substantially different from those put forward in the 2019 discussion paper, with a number of refinements made, largely to reflect responses submitted during the consultation period. In the following, we highlight key changes in comparison to the discussion paper, and other remarkable points.
- Balance sheet synthetic securitisation (criterion 1): In response to requests from stakeholders, for a securitisation to be considered as “balance sheet”, the final report does not require the protection buyer to qualify as an EU-regulated undertaking, as defined under points (a) to (g) of Article 2(12) of the Securitisation Regulation (as did the discussion paper), but as an EU-regulated entity subject to an authorisation or licensing regime. This essentially allows for entities which are not prudentially regulated, especially non-bank lenders, to be a protection buyer. Moreover, according to the report, the underlying exposures should be held on the balance sheet of a member of the corporate group of the protection buyer, not necessarily the protection buyer itself (as the discussion paper suggested). This amendment might be beneficial for several protection buyers, especially in the insurance sector.
- Representations and warranties (criterion 2): In line with the above, the group aspect has also been introduced in the title criterion: according to the report, it is sufficient if any member of the corporate group of the protection buyer (not necessarily itself) has full right, good and valid title to the underlying exposures. Furthermore, following comments from market participants, a knowledge qualifier has been inserted in the representations to be made regarding validity and enforceability of the financing agreements and accuracy of information.
- No active portfolio management (criterion 3): During the consultation, market players had also proposed additional exemptions from the prohibition on active portfolio management. The EBA has included only one of these, ie that an underlying exposure may be removed from the securitisation if it has been disposed of during the ordinary course of the protection buyer’s business, provided that this would not constitute implicit support.
- Underwriting (criterion 7): Further, the EBA has not deleted the prohibition on special purpose entities (SPEs) as obligors of the underlying exposures underwritten, although this had been criticised. This criterion means that, for example, project or acquisition financings where the borrower is an SPE are not eligible for synthetic STS securitisation. This constitutes a difference to the traditional STS regime which does not exclude SPE obligors.
- No embedded maturity transformation (former criterion 13): Notably, the EBA has abandoned the criterion contained in the discussion paper that the repayment of the underlying exposures should not be predominantly reliant on the refinancing of such exposures or the resale value of the assets financed thereby.
- Interest rate and currency risk (criterion 14): One of the commercially most important changes in comparison to the discussion paper is that the EBA has removed the criterion that the protection buyer should bear no currency or interest rate risk in relation to the credit protection it receives. Instead, the transaction documentation should merely describe how such risk will affect payments to the protection buyer and investors and, as far as interest rate risk is concerned, how it will be mitigated. This modification has been widely welcomed because aligns with typical structures of synthetic securitisations where any currency or interest rate risk, to the extent it exists, falls on the protection buyer. To fully remove such risk from the protection buyer would have meant to put it on investors (which many investors might not have accepted and which would have differed from the approach taken in Article 21(2) of the Securitisation Regulation for traditional STS securitisation) or to perfectly hedge such risk (which, if possible, would have been costly and would have added complexity to the structure, in contradiction to what STS seeks to achieve).
- Referenced interest payments (criterion 15): Moreover, in reaction to input received during the consultation, the EBA has revised the criterion on referenced interest payments in relation to the securitisation (which we understand to refer to payments to investors): according to the report, these may be based on either generally used market rates or collateral yield. Yet, the EBA has not followed the wish expressed by stakeholders that no requirement should apply to referenced interest payments in relation to the underlying exposures. Instead, the EBA suggests that these should be based on generally used market rates or generally used sectoral rates reflective of the cost of funds, and that they should not reference complex formulae or derivatives.
- Enforcement and termination (criterion 16): Another amendment in response to stakeholder input is the deletion of mandatory automatic acceleration or enforcement. Rather, the EBA proposes that following an enforcement event in respect of the protection buyer, it should be at the discretion of the protection seller whether to terminate the credit protection or take enforcement action (with further clarifications made in respect of funded credit protection). Besides this, the EBA has introduced another exception to the rule that when a securitisation special purpose entity (SSPE) is used, following any termination or enforcement no cash should be trapped in the SSPE. Accordingly, the SSPE may retain cash necessary to ensure (i) the operational functioning of the SSPE, (ii) the orderly repayment of investors and – which is new – (iii) protection payments in respect of defaulted underlying exposures that are still being worked out. This allows SSPEs to withhold cash to cover potential losses resulting from such work-outs.
- Amortisation of tranches (criterion 17): Furthermore, the EBA has clarified that not only pro rata, but also “hybrid” amortisation is permitted (the latter meaning a combination of pro rata and sequential, or pro rata applying to only some tranches), provided that clearly specified triggers are included to switch to sequential amortisation. Such triggers should relate to the performance of the underlying exposures and should at least include the deterioration in their credit quality below a predetermined threshold. Without such triggers, sequential amortisation should apply. This suggestion seems reasonable, considering that the level of credit protection to cover losses crystallising towards the end of the transaction is higher in the case of (partial) sequential amortisation compared to (pure) pro rata amortisation.
- Reference register (criterion 21): Regarding the requirement that the reference register should identify obligors of the underlying exposures – upheld by the EBA despite comments by the industry – we note that caution must be paid in order to ensure compliance with data protection obligations.
Requirements specific to synthetic securitisations
- Credit events (criterion 28): With respect to credit events that have to be included in every credit protection agreement (as a minimum and without prejudice to the parties agreeing on further or stricter credit events), the EBA has deleted the restructuring of the underlying exposure where the credit protection takes the form of a financial guarantee. This makes sense in order to avoid financial guarantees being treated as derivatives in accordance with the relevant accounting standards and, as a consequence, potentially giving rise to mark-to-market volatility.
- Credit protection payments (criteria 29 and 30): Consistent with the extension to group members regarding criteria 1 and 2 above, the EBA has clarified that credit protection payments should be calculated based on the actual realised loss suffered by the originator or the relevant lender (the latter being relevant should the underlying exposures not be held on the balance sheet of the originator, but of a member of its group). However, although market observers had questioned whether these should be mandatory, the EBA has kept the detailed prescriptions for the calculation of and timeframes for credit protection payments, including interim payments. Interim payments should be made, at the latest, six months after a credit event has occurred (if the work-out of the relevant underlying exposure has not been finalised by that time) and should be adjusted by a final credit protection payment once the actual realised loss has been determined. Where upon the scheduled maturity or early termination of the credit protection agreement the work-out is still ongoing, it should be extended for a period of up to two years, following which the final credit protection payment should be made based on an estimate of the loss. Hence, the EBA has maintained the cap of the extension period at two years, despite comments that some assets may have longer work-out periods (eg in real estate finance or project finance). For investors, this cap of course increases legal certainty with respect to the timing of final payments. As a more general note, it has to be seen how all these criteria will be handled in practice, considering that they require credit protection payments to relate to specific underlying exposures, while in some cases they only correspond to risk tranches of the portfolio.
- Credit protection premiums (criterion 31): This criterion has been revised in a helpful way to clarify that the transaction documentation should only describe how the protection fee and any note coupons are calculated in respect of each payment date over the life of the securitisation. The discussion paper had required disclosure of all relevant information used to price the credit protection agreement, which might have included sensitive details usually not revealed in capital markets transactions.
- Early termination events (criterion 33): Further, the EBA added more impediments to those regulatory events which allow for originator calls. Changes in laws or regulations, or the tax or accounting treatment of a transaction, only entitle the originator to terminate the transaction prior to its scheduled maturity if they have a material adverse effect on the amount of capital that the protection buyer is required to hold in connection with the securitisation or the underlying exposures, compared with that anticipated at the time of entering into the transaction. Furthermore, this material adverse effect must have been reasonably unforeseeable at that time. Apart from that, it is now an independent ground for termination if a competent authority determines that the protection buyer (or any of its affiliates) is not or is no longer permitted to recognise SRT in respect of the securitisation (in accordance with Article 245 of the CRR). However, although industry stakeholders had hoped to see this, the EBA has not included changes in the methodology of rating agencies which result in tranches being downgraded as early termination events.
- Excess spread (criterion 34): Probably the most important difference to the discussion paper is that the EBA has finally come round to the view that it is permissible in an STS synthetic securitisation to use excess spread, thereby making the treatment of synthetics more commensurate with true sale securitisations. This seems especially crucial with respect to certain asset classes (such as consumer lending or SMEs). In order to address concerns that excess spread may erode effective SRT, its use is subject to three conditions (to be specified in the transaction documentation): (i) the committed amount of excess spread per payment period should be a fixed percentage of the total outstanding portfolio balance, (ii) the committed amount per year should not exceed the one-year regulatory expected loss on the underlying portfolio, and (iii) a “use it or lose it” approach should apply (ie if not used to cover losses materialising during the relevant payment period, the excess spread should be returned to the originator). The final EBA report on SRT, expected to be published before 2021, will provide further insight.
- Eligible credit protection agreements, counterparties and collateral (criterion 35): Finally, despite strong criticism by the industry, the EBA has largely retained the narrow scope of eligible credit protection arrangements, with only a few reliefs introduced.
- Firstly, credit protection arrangements may take the form of guarantees meeting the requirements for credit risk mitigation set out in Chapter 4 of Part Three, Title II, of the CRR. In this case, the protection provider should be an entity (i) listed in Article 214(2)(a) to (d) of the CRR and (ii) qualifying for a 0% risk weight in accordance with the standardised approach for credit risk under Chapter 2 of Part Three, Title II, of the CRR (or the guarantee should benefit from a counter-guarantee of any such entity). These – unchanged – requirements essentially exclude all private sector protection sellers, such as commercial banks and insurers.
- Secondly, credit protection arrangements may come in the form of guarantees, credit derivatives and – which is new – credit linked notes (CLN) meeting the requirements for unfunded credit protection and CLN set out in Sub-Section 2 of Section 3, Chapter 4 of Part Three, Title II, of the CRR (as amended by Article 249 of the CRR). In this case, there should be collateral for the protection provider’s obligations in the form of cash or debt securities. The EBA has now allowed for cash collateral to be held not only with third-party banks, but also on deposit with the protection buyer, subject to a minimum credit quality standing requirement (meaning that if such standing is not met anymore, the cash should be transferred to a third-party bank that has the necessary standing or invested in high-quality securities). Further, the EBA has specified that this criterion would be satisfied in the case of the investments coming from CLN issued by the originator (in accordance with Article 218 of the CRR). Considering market practice, which typically involves collateral in one of the forms mentioned above, this approach is welcome news. However, with respect to collateral held in the form of debt securities, the EBA has upheld the requirement that these should be 0% risk-weighted, thereby effectively limiting eligible collateral to certain securities issued by the public sector and international financial institutions.
Preferential regulatory treatment of synthetic STS securitisation
The most heavily criticised aspect of the 2019 discussion paper was the absence of any recommendation for favourable capital treatment for synthetic STS securitisations. Market participants expressed that this was not justified by both the data and the legal analysis and that a prosperous market for synthetic STS securitisation will only develop if an appropriate regulatory benefit is introduced, matching that of traditional STS securitisations.
Traditional STS securitisations are granted preferential treatment in various ways, including from the perspective of investors. Especially, securitisation positions held by credit institutions or Solvency II regulated insurance or reinsurance undertakings benefit from a lower capital charge compared to non-STS securitisation positions (provided that additional requirements are met, as set out in the CRR as amended by Regulation (EU) 2017/2401 and in Delegated Regulation (EU) 2015/35 as amended by Delegated Regulation (EU) 2018/1221). Besides this, STS securitisations may be eligible for inclusion in high quality liquid assets for purposes of the CRR liquidity coverage ratio (under Delegated Regulation (EU) 2015/61 as amended by Delegated Regulation (EU) 2018/1620).
To a limited extent, a capital benefit is also available to synthetic securitisations which meet the traditional STS criteria (with exceptions, such as the true sale requirement). However, this applies only under the circumstances set out in Article 270 of the CRR; in particular, it is restricted to senior positions, certain protection providers and securitisations backed by a pool of exposures to undertakings which largely qualify as SMEs.
Regarding the question of whether differentiated regulatory treatment for synthetic STS securitisations (in general) should be introduced, in its report, the EBA has essentially retained the approach followed in the 2019 discussion paper. The EBA has stopped short of expressing any recommendation, but has only set out pros and cons. (The reason might be that Article 45 of the Securitisation Regulation does not explicitly mandate the EBA to opine on this question.) However, the EBA has made a step forward in setting out how such treatment, if adopted, could look like.
According to the EBA, the evidence and analysis could justify favourable capital treatment, compared with non-STS synthetic securitisations, of synthetic securitisations meeting (i) the STS criteria now suggested by the EBA and (ii) Article 243(2) of the CRR. Such treatment could consist of an adjustment of the prudential floor and of the risk weights as applicable under the STS traditional framework (ie a recalibration under formulae based approaches to include a 50% haircut of the supervisory ‘p’ parameter and a recalibration of the approach based on external ratings to achieve a lowering of risk weights that is consistent with the recalibration of the former approaches).
However, the EBA takes the view that no fully-fledged preferential regulatory treatment should apply.
- First, it should be limited to capital treatment and should not include liquidity treatment.
- Second, such capital treatment (as described above) should be restricted to senior securitisation positions retained by originating credit institutions and should not extend to investing institutions. As originators of balance sheet synthetic securitisations typically place the junior and mezzanine tranches with investors who are usually not banks, we believe that these limitations should not constitute an impediment in most cases.
Finally, the EBA expresses several concerns in relation to the introduction of favourable capital treatment of synthetic STS securitisations. These include a lack of data regarding synthetic deals (which during the post-crisis period were mostly bilateral and almost entirely private) and limited experience with the STS traditional framework so far. The EBA also notes that preferential treatment is not provided for in the international Basel standards and could provide an incentive for overusing synthetic securitisation for a large-scale replacement of regulatory capital by risk mitigation strategies (ie reductions of risk-weighted assets), leading to overleveraging of banks.
Against this background, the EBA suggests that any potential legal framework enabling differentiated regulatory treatment should include a mandate to the EBA to monitor its functioning.
Excursus: Interplay with the new EBA guidelines on the calculation of the weighted average maturity
The paper on STS synthetics came out only days after the final guidelines published by the EBA on the determination of the weighted average maturity (WAM) of contractual payments due under the tranche of a securitisation transaction in accordance with point (a) of Article 257(1) of the CRR (EBA/GL/2020/04). These will apply from 1 September 2020 and will also play a key role in calculating the capital requirements in respect of securitisation positions.
In these calculations, the maturity of a tranche is a relevant parameter for all institutions using the internal ratings-based approach (SEC-IRBA) or the external ratings-based approach (SEC-ERBA). Institutions may choose whether to calculate the tranche maturity based on the final legal maturity of the tranche or the WAM of the contractual payments due under the tranche (as now clarified by the EBA). The final legal maturity approach is usually regarded as more conservative, whereas the WAM approach is more risk-sensitive and, therefore, in some cases more attractive from a commercial perspective.
How the relevant contractual payments are determined for purposes of calculating the WAM (naturally) differs depending on the type of securitisation:
- For traditional securitisations, the contractual payments include those payable (i) to the SSPE by the originator and (ii) to the tranche holders by the SSPE.
- For synthetic securitisations, the contractual payments comprise those payable to the originator by the debtors of the underlying exposures allocated to the reduction of the outstanding amount of the tranche (provided that the documentation is clear enough to allow such allocation). Where tranches are subject to credit protection, the contractual payments of the premia payable by the originator to the protection provider should also be taken into consideration.
On this basis, in line with suggestions made by market players, largely the same WAM calculation methodology applies in synthetic transactions (with respect to payments to the originator) and traditional transactions (with regard to payments to the SSPE in relation to performing exposures). However, prepayments are excluded in the case of synthetic securitisations. Another difference is that where originators have an early termination option, “and the terms and conditions of the transaction contain a positive incentive for the originator” to exercise this option, then the maturity should be the earliest time when such option may be exercised – but only in the case of synthetics.
Although it remains to be seen how all this will be interpreted in practice, for institutions wishing to use the WAM approach it might be worth considering both the criteria for synthetic STS securitisation and the new guidelines when drafting the transaction documentation, in order to have an overview of regulatory consequences.
Conclusion and outlook
Overall, the suggestion to extend the STS framework to synthetics is another step in the direction of finalising the STS regime and, if adopted, would likely constitute a boost for the market, in line with the recommendations by the High Level Forum on the CMU. According to market participants, the proposed framework for synthetic STS securitisation broadly constitutes a reasonable approach, especially in that the criteria closely follows that of traditional STS securitisation. It remains to be seen, however, if too much weight has been placed on standardisation and simplicity to the detriment of flexibility. How the market implements this across classes of transaction participants, sectors and especially assets will be the real test as to whether the right balance has been found. What has been widely welcomed is that the EBA has set out how a preferential regulatory treatment for synthetic STS securitisation could work – and, as it seems, more pros than cons of its introduction – thus cautiously opening the door to the possibility of affording such treatment.
The EBA report is by no means the end of the long and winding road to synthetic STS securitisation. It is now under consideration by the European Commission, which will subsequently report to the European Parliament and the Council, and will assess whether or not to adopt a legislative proposal. Hence, it is in the hands of the European legislator to decide whether to introduce a legal framework for synthetic STS securitisation at all and, if so, whether it should be accompanied by differentiated capital treatment.
We continue to analyse the implications of the report and future developments, including across jurisdictions, and would be delighted to discuss with you.