The Investment Association agrees with the paper’s assessment of the impediments to investors’ ability to invest in this asset class, including:
• the stigma attached to the asset class following the financial crisis. This has limited institutional investors’ willingness to allow their asset managers to invest in these instruments.
• the risk retention requirements that:
o under the CRDIV/CRR and AIFMD do not ease the identification of qualifying instruments and therefore investment in this asset class.
o increase the investment risk to investors in securitisations, as they are being held to account for originators satisfying retention requirements yet have no obvious means to reliably control the behaviour of originators over the life of the transaction. Ideally, the retention requirements should operate as disclosure obligation on originators, rather than (as present in the EU) on investors.
• lack of market data for less developed asset classes e.g. ABS backed by SME loans.
• sovereign ratings caps applied to ABS from certain EU countries that mean that they are treated more harshly than their credit performance would imply. This would:
o limit institutional investors’ - specifically insurers’ - ability to invest in these assets due to the higher capital charges under the standard model that would be applicable under the proposed Solvency II rules; and
o limit asset managers ability to allocate funds to these assets due to mandate and investment guideline restrictions.
The Investment association welcomes the European Commission’s proposed revisions to the draft Solvency II rules reducing the capital charges for high credit rated securitisations backed by certain categories of assets. However, these changes may not be sufficient to encourage insurance companies to invest in these products as the proposed capital charges remain high relative to other fixed income products such as corporate and covered bonds.
In addition, there remain severe cliff effects between the treatment of AAA rated and non-AAA rated securitisations. For insurers to invest in these products a graduated approach to risk is required; or better still, a removal of such hard-edged usages of credit ratings in regulatory definitions (in accordance with the wider EU policy of addressing over-reliance upon credit ratings).
The Investment Association’s members agree that synthetic securitisations can introduce undesirable counterparty risks. They are also concerned about their ability to able to take control over the underlying assets in enforcement scenarios. This limitation dilutes the investor protection available in synthetic structures relative to those in true sale securitisations.
However, The Investment Association’s members recognise the importance of synthetic securitisations in certain instances, such as where a true sale securitisation is either uneconomic or not possible due to contractual limitations.
The Investment Association members believe that adequate safeguards should be included in any regulatory framework that admits synthetic securitisations as qualifying securitisations. For this to be achieved it is important to draw a distinction between synthetic securitisations where only the asset-side of the securitisation is synthetic and those where both the asset and liability side of the securitisation are synthetic. It is our members’ view that the former and not the latter should be considered to be a qualifying securitisation subject to the safeguards set out below at a minimum (in addition to the criteria that qualifying true sale securitisations will have to satisfy).
• Where only the asset side of the securitisation is synthetic, the securitisation is exposed to two separate asset pools, the funding collateral and the reference collateral.
• Regulatory eligibility criteria should be applied to the funding collateral in a qualifying synthetic securitisation, with the aim of ensuring that the bankruptcy-remoteness of the securitisation from the originator is not impaired via the reference collateral (e.g. senior unsecured bonds issued by the originating bank would not be eligible as funding collateral).
• The funding collateral should not consist of instruments with credit risk that is significantly correlated with the credit risk of the reference collateral (e.g. senior unsecured bonds issued an EU bank used as reference collateral in a securitisation from another EU bank).
• The originator must certify that:
o the reference collateral will only be securitised once and will not be securitised in a true sale transaction in addition to the synthetic transaction, nor will it be securitised in multiple synthetic transactions.
o that the reference collateral is not referenced in any other credit instrument, securitisation or not.
o the ability of the securitisation to enforce the reference collateral is in no way inferior to that in a corresponding true sales transaction.
• The originator must own the reference collateral over the life of the transaction. In the event of a transferor insolvency, the transaction must enter liquidation.
• The originator must post additional cash collateral in a ring-fenced account to be deposited with the originating bank to be deposited with the originating bank (in a ring-fenced account) or held by a counterparty and pledged to the securitising bank.
• Higher risk retention requirements should be imposed to ensure strong alignment on interest between the originator and the investor.
• Originators should use standardised transaction documents and that the terms of those documents (including amendment) should be publically available following the conclusion of the transaction.
The CRR definition as per Article 178 of the CRR would be most appropriate. This is the definition used in the draft delegated acts under Solvency II (Article 177). This will not only ensure standardisation but would ease investors ability to undertake their due diligence requirements when making a decision to invest in a particular instrument.
As noted above any framework for simple, standard and transparent securitisation should seek to ensure consistency with regulation that impacts securitisation.
As such it is proposed that, as set out under the draft Solvency II Delegated Acts, a securitisation should qualify as a simple, standard and transparent securitisation if:
(i) It is listed in a regulated market of a country which is a member state of the EEA or the OECD, or is admitted into trading in an organised trading venue providing for an active and sizable market for outright sales which has the following features:
(ii) Historical evidence of market breadth and debt as proven by low-bid ask spreads, high trading volume and a large number of market participants; and
(iii) The presence of robust market infrastructure.
In the interest of transparency, the jurisdictions governing all securitised assets, all transaction documents and all financial instruments issued or entered into should be disclosed.
We do not believe that there is a need for simple, standard and transparent securitisations to have voting rights conferred only to the most senior credit tranches in a securitisation. Voting rights are a factor that is considered actively when purchasing/pricing different securitisation tranches.
This would act as a disincentive to investors interested in investing in the junior tranches of a transaction as they would not only be taking first loss (relative to the senior tranches) but the protections afforded to them by the ability to vote on material changes (subject to the terms and conditions) would be taken away. This may negatively impact the marketability of these transactions.
We welcome a requirement for disclosure of a liability-side cashflow waterfall model both before the pricing of the securitisation and on an ongoing basis.
In addition, the sponsor or originator should be required to issue a draft prospectus (red herring) which should be made available to potential investors during the roadshow phase of the transaction. The draft prospectus should contain all terms and conditions of the transactions, including priorities of payment and possible changes to the waterfall following trigger breaches. This will provide investors with an opportunity to raise any inconsistencies within the cash flow model and the transaction documents prior to making the decision to invest. The practice of distributing a draft prospectus to potential investors already exists in the high yield corporate bond space.
Comprehensive investor reports must be available at the same frequency as that of the coupon of the highest-frequency coupon paying bond at inception or at a minimum should be available quarterly. Investor reports must be available not later than two weeks following the end of the reporting period covered. Loan tapes and liability-side cashflow waterfall updates must follow the same rules. All three types of documents much cover the exact same period and must be consistent between each other.
All transaction documents should be made publically available, at a minimum from the date of transaction close onward.
The Investment Association agrees that granularity provides benefits, particularly in reliable modelling of portfolio cashflows. However, we are uncertain as to the extent to which granularity can be linked to lower credit risk. Nonetheless, we would welcome a definition of granularity that can be consistently across regulatory frameworks, for securitisations, covered bonds and other collateralised instruments.
As noted in the introduction, the criteria for identifying simple, standard and transparent securitisations should seek to reduce the major non-credit related risks. However, the aim should not be to seek to remove credit risk. For example, the criteria that all assets should be self-liquidating goes too far by disallowing certain assets on the basis of their credit characteristics. This criterion would appear to exclude all interest-only or balloon loans. This could have the unintended consequence of impacting lending behaviour by banks and therefore limiting consumer choice.
We agree that any collateral concentration has an effect of the risk profile of securitisations. However it is not clear why there is a requirement for a pool of exposures to be securitised should be such that the largest aggregated exposure to a single obligor does not exceed 1% of the value of the aggregate outstanding value. Whilst 1% threshold may be applicable for certain asset classes such as residential mortgages, consumer auto receivables and consumer debt securitisations, most other asset classes would have difficulty in fulfilling this criterion. This includes some SME securitisations and a large proportion of commercial auto lease securitisations. Any thresholds that the EBA seeks to include in the criteria should be subject to an impact assessment to ensure that they do not stifle the revival of the securitisation market.
Further, to avoid cliff effects, a securitisation classified as qualifying at inception should remain qualifying over its lifetime, implying that an increase in concentration over a securitisation’s lifetime should not result in it becoming non-qualifying.
The Investment Association is broadly supportive of the proposed criteria. We would make the following points:
• Criterion 1: Currently ABCP are excluded. While we agree that they have differing credit dynamics to term securitisations, they could benefit from a set of SST criteria at a later date.
• Criterion 2: The Investment Association broadly supports this criterion, yet feels that this area should be covered under the managed funds regulatory framework.
• Criterion 3: It is important to acknowledge that in some contracts a legal ‘true sale’ is not possible. For example, ‘true sale’ does not exist in Germany. Overly prescriptive requirements regarding ‘true sale’ could therefore act as an impediment to the securitisation market (see response to question 2).
• Criterion 13: The Investment Association is wholly supportive of any moves that would ensure that there is a robust trustee in a securitisation transaction.
• Criterion 18: We agree that liability cash flow models are a crucial part of any SST definition. In addition, they should be made available to investors at least one week prior to launch.
• Criterion B: Further justification is needed as to the rationale behind the 1% limit. Any thresholds that the EBA seeks to include in the criteria should be subject to an impact assessment to ensure that they do not stifle the revival of the securitisation market.
• Criterion C: We believe that limiting the criterion to only include exposures to individuals or undertakings resident, domiciled or established in the EEA will serve only to act as a barrier to further expansion of the securitisation market.
There is always a risk that by introducing a qualifying securitisation framework, the market may be divided into compliant and non-compliant securitisations (even within in the same asset class). This may not increase market liquidity in the long run due to difficulty in assessing deals that do not comply.
That said, The Investment Association’s members are supportive of a framework for simple, standard and transparent securitisations would help re-establish investors’ confidence in securitisations and pave the way for a more risk sensitive regulatory framework that can differentiate between different securitisation products.
It is important that as such a framework is developed, securitisations that fall outside the scope of such a framework do not become “non-eligible” securitisations (for example not eligible to be held in an alternative investment fund); therefore precluding investors with the appetite for these instruments from investing in them. This would be of particular concern for transactions which are of the same risk profile as many “qualifying” securitisations but are treated as “non-qualifying” on the basis of one principle e.g. a synthetic transaction.
Further, introducing a qualifying securitisation framework for regulatory purposes that would qualify as high quality and liquid for LCR purposes may incentivise firms to obtain qualifying status for securitisations that they originate and for investors to invest in them. This would also limit the opacity and complexity that has previously been seen in the market.
The European Supervisory Authorities should seek to establish a single definition of securitisation and a single definition of qualifying securitisation. This definition should be harmonised with those used at a global level.
Under such harmonised definitions, a securitisation should be deemed qualifying or not at the transaction level, not at the tranche level. If a securitisation is qualifying, then all exposures to it, including all classes of notes, are qualifying. If a securitisation is not qualifying, then no exposure to it, including none of the classes of notes, are qualifying.
All exposures to a qualifying securitisation should have lower capital requirements than otherwise equivalent exposures to a non-qualifying securitisation.
Regulatory capital requirements should reflect risk in a consistent manner. In particular this should take into account the conditions have to be satisfied with respect to how this risk is assessed. For example: the risk across all exposures to a securitisation cannot exceed the risks present in the underlying asset pool. In fact, because some of these risks flow into exposures other than the issued notes (e.g. rate swaps, basis swaps, currency swaps) and are mitigated via additional protection measures (e.g. liquidity facility, reserve fund), the aggregate risk across all classes of notes in a securitisation is arguably lower than the risk in the underlying assets. Hence, aggregate capital requirements across a securitisation’s note classes should not be higher than the capital requirements for holding the underlying assets directly.
One possibility of calibrating capital requirements across tranches is to use a multiplier as the credit quality decreases. Therefore, the degree to which the capital requirement increases further down the credit spectrum of the tranches should be a function of the expected loss of the particular tranche. Equally a multiplier penalty should be applied to non-qualifying securitisations. For example, if the capital treatment for a given tranche is 2.1%, then the non-qualifying securitisation should be 2.1x[multiplier].
In a regulatory framework assessing capital requirements based on credit rating agency ratings, the only way to avoid having country rating ceilings affect said capital requirement is to request that the rating agencies publish an underlying rating, which ignores the country ceiling, in addition to their final rating. Such underlying ratings would be produced by the rating agencies internally already, as the country rating ceiling is applied late in the process of formulating a rating. In addition, rating agencies routinely produce and sometimes publish underlying ratings in other contexts, such as the BCA (Baseline Credit Assessment) at Moody’s and the SPUR (S&P Underlying Rating) at S&P.
For the sake of simplicity, transparency and consistency, a regulatory framework assessing capital requirements based on ratings should reference the same type of rating across all asset classes.