Article 44 of the EUSR requires the ESAs to produce a report once every three years on certain aspects of securitisation. The next one is due before the end of 2024, and at the Barcelona Global ABS a couple of weeks ago, two representatives from ESMA and the EBA said that this would include a review of the definition of “securitisation” and proportionality for due diligence and transparency (e.g. carving out private issues).
What about the penal capital treatment for securitisation: the excessive degree of non-neutrality, and the notorious upcoming output floor? Earlier this week, Regulation (EU) 2024/1623 – i.e. CRR III was published in the OJEU, and it mostly becomes effective on 1 January 2025 (with some parts earlier). It implements various stop-gap reforms to counter the adverse consequences of the output floor including, as regards securitisation, by a halving of the “p” (non-neutrality) factor, and it holds out the promise of more thorough reform in the medium term to address excessive non-neutrality.
The output floor
One element of the “Basel IV” (or, sometimes, "Basel 3.1") reforms is the “output floor” - a floor under the capital requirements calculated under the Internal Ratings Based Approach. The Basel Committee had concluded that the, more precise, IRBA might under-estimate risk, despite being based on a bank’s own data and experience in its market, and despite it only being allowed where the model had been approved by the bank’s regulator (seemingly a case of Basel giving with one hand but then wanting to take back with the other much of what it had just given). Hence the idea of limiting the degree of capital reduction that it could produce by reference to the Standardised Approach. So, once fully phased in, the output floor will require that capital requirements under the IRBA may not be less than 72.5% of those calculated using the Standardised Approach, and both the UK and the EU have the same transitional period for the implementation of the output floor: starting at 50% on 1 January 2025 and reaching 72.5% on 1 January 2030.
This has been controversial with the market because the IRB reflects the actual experience of the bank in question, in comparison to the one-size-fits-all categories of the SA, and a floor based on the SA risk weights would require increased amounts of capital to be allocated to exposures. Since capital has a cost, this would reduce the return on the exposure, and could render certain transactions uneconomic. This would be undesirable, and the data do not support the conclusion that this is justified. Regulators have not necessarily seen it this way in the past. For example, the Bank of England’s 30 November 2022 proposal, although it noted that the IRB can "often generate significantly lower risk weights than the SAs for similar exposures", drew the conclusion that the output floor "would support competition" by narrowing the gap between the IRB and the SA. Query whether Adam Smith would have agreed that competition is supported by making a more accurate risk analysis more blurred, and forcing more sophisticated approaches to be less sophisticated.
The output floor can have adverse consequences in different parts of the balance sheet. As regards securitisation, a November 2022 AFME-commissioned 70-page number-crunching analysis, “Impact of the SA Output Floor on the European Securitisation Market” suggested that corporate securitisations by originator banks, both for large corporates and SME portfolios, could be largely eliminated, and the output floor was described by AFME in January 2024 as posing “an existential threat” to synthetic securitisations by banks. It was criticised for not having been soundly rooted in an understanding of the relatively riskiness of different asset classes.
In January 2023, the European Parliament ECON committee approved an amendment to the draft CRR amendment regulation, and this is now contained in the CRR, as amended by this week’s CRR III. It provides that, pending the completion of a comprehensive review by the EBA and ESMA (to be done by 31 December 2026), the ‘p’ factor which applies under the SEC-SA (and which is therefore the one which would apply when and if the output floor kicked in) is halved for the purpose of calculating the output floor: for STS securitisations, from 0.5 to 0.25; and for non-STS securitisations, from 1 to 0.5. This is not a panacea, but it will tend to counteract some of the adverse effects of the output floor (some calculations have it more or less offsetting them).
Non-neutrality
The CRR capital regime for banks’ securitisation exposures is “non-neutral”: if a pool of exposures is securitised, and all the securitisation positions are held by regulated institutions, the combined capital charge for all those pieces will be more than the capital charge that would apply if the pool had not been securitised. This is known as “non-neutrality”, and it creates a capital surcharge on the tranches of the securitisation. The capital requirements are first calculated at the level of the underlying portfolio of securitised assets in accordance with the general framework for credit risk (using KIRB or KSA, as the case may be) and each of those capital inputs is then increased as a result of being multiplied by the "p" factor. Under SEC-IRB, “p” is at least 0.3, and for SEC-SA it is 1. This is an add-on, and so, for example, a “p” of “1” means the capital is doubled.
The official regulatory justification for non-neutrality in the past has been that securitisations are structured products, and the very fact of the structure creates complications and risks which go over and above the credit risks on the underlying assets, so that institutions need to allocate more capital to them because they are inherently riskier than a simple holding of the underlying assets. These additional risks - known as “agency risk” and “model risk” – were perceived to be significant in pre-GFC transactions. The general perception is that in the post-GFC and post-regulation era, they are much more limited, if not actually non-existent in most modern securitisations, but the regulatory capital regime remains rooted in a previous era.
The EU and UK market can look enviously to the USA, which has not yet implemented Basel 3 and is using a modified version of Basel 2. There, it seems that under the standardised approach, “p” is 0.5 for all securitisations, and, while the USA equivalent of SEC-IRBA does not explicitly have a “p” factor in the formula, according to AFME, “implicitly it is close to 0”. And when we consider the output floor, it is ironic that the capital requirements under the SEC-SA standardised approach were, apparently, heavily based on the loss experience of US residential mortgages and US RMBS: which, for a variety of factors, was much worse than in European jurisdictions.
What CRR III says about non-neutrality
There is a general sense now that the degree of non-neutrality in the CRR is excessive, and counter-productive in the context of securitisation and the wider context of CMU. We have seen several papers from public or official bodies to this effect, including, in April this year, “Developing European Capital Markets to finance the future”, from a committee that includes Christian Noyer, ex-governor of the Bank of France and Robert Ophèle, ex-chairman of the Autorité des Marchés Financiers (the French equivalent of the FCA) which was commissioned by Bruno Le Maire, the French finance minister.
CRR III holds out the promise of medium-term reform. Recital (3) notes that the EC will examine the implementation of the output floor, including its level of application, based on input from the EBA, and “will consult with interested parties to ensure that the various perspectives are appropriately considered”. This is provided for by inserting a new Article 506d into the CRR, which requires the EBA to report to the EC no later than 31 December 2026, and which says that:
“In particular, EBA shall monitor the use of the transitional arrangement referred to in Article 465(13) [i.e. the halving of “p”] and assess the extent to which the application of the output floor to securitisation exposures would affect the capital reduction obtained by originator institutions in transactions for which a significant risk transfer has been recognised, would excessively reduce the risk sensitivity and would affect the economic viability of new securitisation transactions. In such cases of a reduction of risk sensitivities, EBA may consider proposing a downward recalibration of the non-neutrality factors for transactions for which a significant risk transfer has been recognised. EBA shall also assess the appropriateness of the non-neutrality factors under both the SEC-SA and the SEC-IRBA, taking into account the historic credit performance of securitisation transactions in the Union and the reduced model and agency risks of the securitisation framework”;
It then tells us that the EC, taking into account this report and the upcoming report from the FSB (which we were told last year was due “mid-2024”) “shall, where appropriate, submit… a legislative proposal by 31 December 2027”. Allowing for trilogues and the rest of the usual process, we might be looking at, perhaps, late 2029 or 2030 for a permanent change in the law. It may not all be plain sailing of course: as recently as December 2022, the ESA’s “Joint Committee advice on the review of the securitisation prudential framework” did not entirely share the market’s enthusiasm.
In the UK
In the UK, we wait to see whether the PRA, once armed with the FSB’s conclusions, will take the post-Brexit opportunity to introduce a more holistic set of reforms significantly in advance of the EU’s likely timescale. Its October 2023 consultation seemed to indicate a preference for “a targeted and data-based adjustment to the Pillar 1 framework for determining capital requirements for securitisation exposures”, and it supported the idea that the BCBS should conduct a wider review of the Basel capital requirements for securitisation, and “particularly its level of ‘non-neutrality”: meaning the size of the “p” factor.
Conclusion
There is a way to go yet but, still, we seem to have a clear direction of travel towards some degree of liberalisation and, if you think securitisation is a niche financing technique, you might want to get ready to think again.
and then 