Basel III and the Output Floor – where are we now?

In response to the perceived weaknesses in the banking industry highlighted during the global financial crisis, the Basel Committee for Banking Supervision ("Basel Committee") introduced the Basel III standards in November 2010 (the "Original Basel III Standards"), followed by a number of subsequent amendments, with the final stages of Basel III being published in December 2017 ("Basel III"),¹ in an attempt, among other things, to reduce the excessive variability of risk-weighted assets ("RWAs") and to increase the comparability between the utilisation of supervisor-approved internal models and the standardised approach.

One of the more far-reaching introductions in December 2017 was the implementation of an output floor, which is imposed as an internal rating model floor. In the context of securitisation positions, the output floor would cap the minimum risk weighting under an internal model such as the Securitisation Internal Ratings Based Approach ("SEC-IRBA") to 72.5% (following a transitional period) of the RWA under a counterfactual standardised approach ("SEC-SA"). This means that banks' regulatory capital gains from the use of internal models such as SEC-IRBA cannot exceed 27.5% of what would have been required to be held in the context of a standardised approach.

The European Banking Authority ("EBA"), in its Basel III full implementation impact report found that European banks' Tier 1 capital requirements would increase by 15% as a result of the Basel III changes, with the output floor being responsible for a 7.1% overall raise, with a slightly higher percentage for globally systemically important institutions ("G-SIIs") than for ordinary banks.

The output floor is expected to have a different impact depending on the asset class, so banks may find that they are more exposed to the output floor depending on their  asset portfolios. For example, residential mortgages and residential mortgage-backed securitisations ("RMBS") are on average assigned a very low internal model risk weighting of approximately 10%, whereas the standard risk weights rank from 20% - 70% (at a corresponding output floor of 14.5% - 50.75%), meaning the output floor may have a significant effect on this asset class. Banks that specialise with this type of product may therefore be disproportionately affected by the output floor, and this impact on a relatively low-risk lending activity has been criticised.²

What can securitisation do to help?

Amongst a heavily regulated capital landscape and a squeeze on profitability, banks and their shareholders will be eyeing up possibilities to minimise their capital burden. The two broad solutions for banks is to:

1. 1. 1. Increase their regulatory capital base

This can be done through replenishment of their eligible equity capital (through Common Equity Tier 1, or "CET1") or the raising of additional subordinated debt capital (through Additional Tier 1 or Tier 2 Capital).

However, increasing regulatory capital can sometimes be expensive (particularly in respect of subordinated debt and quasi-equity instruments), and may have additional limitations for the bank, such as dilution risk for shareholders and pressure for buy-back programmes. In addition, recent regulatory impacts, such as the writing off and subordination of AT1 convertible bonds to below CET1 level in the UBS acquisition of Credit Suisse may have a dampening effect on the AT1 bond market, or indeed a re-pricing exercise which may mean that such instruments become much more expensive for banks to raise especially amid the current volatile market conditions, potentially making other forms of risk transfer more attractive.  

In the Basel Committee's monitoring report published in February 2023, it was noted that EU banks' capital ratios fell to pre-pandemic levels following a record high at the end of 2021, to a CET1 ratio of 12.7% (against a minimum required amount of 6%). As expected, CET1 continued to be the predominant form of regulatory capital used by banks (equating to around 74.5% of total capital held), with Additional Tier 2 capital being the second highest at 15.5%. However, the total level of CET1 as a percentage of total capital has been on the decline, with banks starting to utilise the Additional Tiers of capital instead of tapping their shareholders for additional equity.³

Holding high amounts of CET1 may not only have adverse consequences of shareholders as such capital could otherwise be deployed in income-generating investments, but also a broader negative impact for the wider economy, such as reducing the  availability of credit more generally. This was particularly visible during COVID-times, where although banks were generally seen to be well capitalised, the involvement of governments and central banks became necessary to encourage lending into the economy.     

1. 1. 2. Decrease their RWAs

Banks may decrease their RWAs by either reducing their total exposures (i.e. removing assets from their balance sheet or through significant risk transfer) or reducing the risk weight density of its assets by enhancing the creditworthiness of its assets (i.e. turning to "safer" assets). This article will deal predominantly with the former exercise, although a diversification of product and/or business lines as part of the latter exercise, as well as enhanced credit portfolio management, may also mitigate some of the harsher capital requirements imposed on banks.

Banks may elect to decrease the balance of their RWA through a number of different arrangements, including through portfolio sales and the sale of whole businesses relating to non-core businesses and less valuable portfolios. Although a bank may always sell its assets (which would reduce its regulatory burden), this may have a negative impact on its profitability, particularly in respect of high yield and/or income generating assets, and removes the ability of such bank to extract any ongoing profits from the relevant asset. In addition, there may be strong reasons for retaining assets on a balance sheet to ensure that banks remain "visible" in the market for a particular product line, and have continuous relationships with customers and clients which can be used for cross-selling, thereby increasing shareholders' return on investment.  

A bank may instead elect to transfer the credit risk associated with an underlying exposure, in a method which allows it to reduce its regulatory capital requirement, without it having to sell the asset outright.

The above significant risk transfer can be achieved in a number of ways – including insurance or reinsurance policies or risk transfer agreements. A common method of achieving risk transfer is through either a full capital stack cash securitisation or a synthetic securitisation (the latter of which will give rise to counterparty risk unless suitably collateralised/structured as a CLN).  

Most SRT transactions are in a synthetic format, however ‘full-stack’ true sale securitisations have become more prominent over the past year in the case of non-corporate reference assets such as auto and consumer portfolios, and generally are easier to implement with higher quality assets. In such true sale deals, all tranches are normally sold subject to risk retention requirements (typically vertical with 5% retention of all tranches), though in some cases the senior notes are retained. Broadly, in a full capital stack cash securitisation, an originator bank may transfer (by way of true sale) certain assets on its balance sheet to a special purpose vehicle issuer, enabling the bank to divest itself of the entirety of its exposures without the use of leverage (i.e. by retaining the senior tranche or entering into total return swaps) and with the ability to continue the relationship with the borrowers. Provided the relevant criteria under Article 244 CRR are met (including true sale, application of relevant risk weights, the originator having in place appropriate internal risk management policies, etc), the originator bank may be able to reduce its capital burden for the assets it sells to the issuer and also derecognise the assets from an accounting perspective.

For a synthetic securitisation, the assets would remain on the originator’s balance sheet. The originator would, as the "protection buyer", buy credit protection from the investors (the "protection sellers") through credit derivatives (which may include credit-linked notes) and/or financial guarantees/credit default swaps. In each instance, the credit risk associated with the portfolio can be said to have been "transferred" to the investor / protection seller, and therefore alleviated the originator's RWAs. In return, the protection seller gets compensated for the assumption of risk, typically through the payment of a premium or, in the case of credit-linked notes, coupon on the notes. Significant risk transfer transactions have undergone rapid growth in the EU, as a result of highly sophisticated banks using internal models and also (of late) regional and standardised banks turning to an expanding group of protection sellers (including specialist funds, hedge funds, pension funds and insurers), to help to manage their capital requirements.  

Benefits of Significant Risk Transfer transactions

There are a number of benefits in using significant risk transfer ("SRT") transactions to alleviate capital burdens.  

• Freeing up capital. Allows banks to deploy freed-up capital elsewhere (including within the same business line in which the SRT transaction has arisen). The expansion of the simple, transparent and standardised ("STS") designation in 2021 to synthetic securitisations (previously only available for more traditional cash securitisations) also means that originator banks can have preferential capital treatment for any tranches which are retained by them and not placed externally with investors, benefiting standardised banks in particular by reducing the risk weight floor on retained senior tranches and the size of the tranche that needs to be placed for a given amount of capital saving. The expansion was put forward following the EBA report in May 2020 reviewing the feasibility of a framework for STS designation to synthetic securitisations, which broadly recognised the use and development of synthetic securitisations in the industry, as well as the comparably lower rate of defaults in the available performance data.
• Maintaining presence. The fact that the asset remains on the balance sheet of the bank allows the bank to maintain a presence in a regulatory capital-intensive or non-core market as well as continuing their relationship with the relevant customer and/or client, as opposed to, for example, selling the asset outright.
• Larger investor base. The potential universe of investors who are willing to invest in an SRT transaction may be wider than investors who want to buy the asset outright, or indeed investors who are willing to invest equity or AT1 capital into the originating bank. The structural nature of SRT transactions, which may include an issuance of debt securities such as bonds in a full capital stack cash securitisation or credit-linked notes in a synthetic context is beneficial and makes it easy for investors to invest in and trade such securities (if needed). The flexibility of the structure, which can include unfunded credit protection such as financial guarantees (which may be attractive to multilateral entities and/or insurance companies) as well as funded credit protection (which may be attractive to specialist credit hedge funds) broadens the nature and types of investors which banks may have access to. In fact, the increasing involvement of insurance companies and their preference for longer-dated deals has supported the extension of the SRT market beyond the short-dated SME and corporate credit reference portfolios to longer-dated asset classes such as mortgages (albeit this involvement also comes with certain limitations, as discussed below).   
• Customisable asset pools. Investing in SRT transactions may be more attractive to investors than investing in other eligible capital of a bank, such as AT1 bonds. Unlike AT1 bonds, which are essentially funding, broadly, some of the riskiest assets of the bank (which are intended to act as a shock absorber in case of credit loss), investors in SRT transactions may perform due diligence and be selective in identifying exactly the type of credit risk and exactly the type of asset class they are willing to be exposed to. This may allow banks to deal with SRT investors which specialise in and have deep experience in specific asset classes and therefore achieve a better pricing for their trades.  
• Customisable risk appetite. The ability for banks to include tranching in their structures means that they can create different products for investors with different risk appetites (with a corresponding change in risk premium), even in the context of the same reference portfolio. This can be seen in mixed-tranche structures where multiple investors can invest at different levels.  

Has Basel III impeded or paved the way for SRT?

The impact of the output floor and the other changes to the Basel framework may encourage banks to consider SRT as a useful mitigant tool. However, it is also important to consider whether any Basel III developments would, simultaneously (and perhaps unintentionally), also make SRT trades less attractive for banks and investors. We note below a few factors which may be particularly relevant considerations for banks using internal models:

1. Different impact on different asset classes

Risk Control Limited's research report, published in November 2022 and commissioned by the Association for Financial Markets in Europe ("AFME"), reviewed the Pillar 3 reports of approximately 30 banks in Europe to assess the impact of the output floor on different asset classes. The results broadly showed that the asset classes which had the lowest RWA under their internal model relative to the standardised approach would be impacted by output floor the most. For example, residential exposures would be impacted the hardest (even with the lowest output floor in the transition period of 50%), whereas corporate exposures would only have a smaller haircut as a result of the output floor. Although this may not be as relevant for banks which are not using internal models (and therefore not subject to the output floor), banks which use internal models for the specific asset classes which are most impacted by the output floor may prefer to enter into SRT transactions relating to those assets to ensure that they are not penalised by the output floor.   

The differential treatment for each asset class is important to consider in light of the relatively fragmented concentration of asset classes on a jurisdictional basis in the EU: for example, UK, Dutch, Spanish and French RMBSs represented €467bn (or 79%) of all European RMBS deals at the end of 2021. On the other hand, Italy, Belgium and Spain represented 74% of corporate SME exposures, whereas Germany and Italy together represented 73% of the auto securitisation amounts generated.

2. To securitise or not to securitise?

The output floor applies to all RWA calculations. This therefore impacts both originators which hold assets on their balance sheets, but also those banks which are either (i) investing directly in a securitisation position and/or (ii) retaining a securitisation position in respect of which they are themselves an originator. The different treatment for different asset classes must therefore be juxtaposed with the different treatment a particular asset may obtain depending on whether or not it is securitised. In other words, there is not necessarily a correlation between (i) the ratio of RWAs under an internal model for an unsecuritised asset versus that same securitised asset and, likewise, (ii) the ratio of RWAs under the standardised approach for that same unsecuritised asset versus that same securitised asset. This means that the output floor may impact at different times depending on whether or not the asset is securitised, and this may be a factor in a bank deciding whether or not to securitise.

For example, the average RWA for exposures to SMEs was 53.8% (for the internal model) and 85% (for the standardised approach). Under the current proposed EU rules, this would mean that the output floor would only bite during its transitional period when it reaches the 65% floor (i.e. 65% of 85% being 55.3%). In contrast, if the same exposures were securitised, the RWA (for an STS transaction) would be 59.5% (for the internal model) and 127.5% (for the standardised approach), meaning the output floor would immediately bite (i.e. 50% of 127.5% being 63.8%). This may therefore indicate that a bank may prefer to be exposed to this asset on its balance sheet (rather than indirectly through a securitisation). Conversely, if we look at the example of a retail consumer exposure, because of the relatively low SEC-SA approach under the securitisation standardised approach (of 150%) compared to the internal model, the output floor would only be breached following the full transition period, and even then would only result in the SEC-IRBA RWAs being increased from 106.3% to 108.7%, whereas if such assets remained on the balance sheet it would immediately breach the output floor and, following full implementation of the output floor to 72.5%, would mean that the internal model RWA of 39% would have to increase to 54.4% to comply with the output floor.

The difference in the various models therefore can lead to different results not only for the type of asset class, but also on the impact of the output floor depending on the SEC-IRBA and SEC-SA approach. This is largely as a result of the imposition of the supervisory parameter "p" (covered below), rather than as a result of sound methodological analysis on the underlying credit quality of the assets. Banks may therefore adopt different approaches for different business lines. It is unclear whether this was an intended consequence of the on-balance sheet / securitisation dichotomy of RWAs, but it may unfortunately have a disincentivising effect on SRT trades for certain assets (for reasons that may not necessarily be linked to credit performance). 

3. Output floor application to senior tranches of SRT trades

Although the above analysis may be a relevant indication of some trends, it does not necessarily capture the totality of RWA considerations for banks based on asset class. Often (although not always the case), a bank pursing an SRT transaction may securitise a portfolio of loans and retain a tranche (or multiple tranches), adding a layer of complexity to the general trends highlighted above.

The RWAs would then be subject to a number of other variables, including tranche detachment and attachment points, as well as the ability for banks to sell or buy protection on certain tranches to or from external investors and thereby remove such tranche from their own RWA calculations.

For illustration purposes, a bank may enter into a balance sheet synthetic securitisation with three "tranches":  

1. A "junior" tranche, which is the tranche which suffers the first loss (with a tranche attachment point of 0%). This may either be used as the risk retention piece by the originator or, as is the case in some transactions, may be sold to external investors. Depending on the capital charge of the reference portfolio in question, it is likely that the junior tranche will have a very high RWA given its status, and therefore it will be reduced as much as possible;
2. A "mezzanine" tranche, which is the tranche sold to external investors, or the "protection seller". The mezzanine tranche would absorb losses after the "junior tranche" or, if no junior tranche exists, at the same time as the covered percentage (i.e. a vertical slice). Provided the SRT requirements are complied with, the originating bank will not be required to hold RWA in respect of the mezzanine tranche, as the credit risk would, for regulatory purposes, be transferred to the investors; and
3. A "senior" tranche, which can either be retained by the originator, or sold to external investors as well – this may sometimes be structured as a financial guarantee from a multilateral or supranational entity or potentially (potentially on a collateralised basis depending on its credit rating) even an insurance company. The senior tranche normally represents the largest of the tranches.

In the below example, the senior tranche would have a detachment point of 100% and an attachment point equivalent to the mezzanine tranche detachment point. If the capital charge, K_IRB or K_SA, is equal to or above the mezzanine tranche attachment point, the RWA for the junior tranche (both under SEC-IRBA and SEC-SA) would be 1,250%.

Given the mezzanine tranche will not attract RWA (i.e. because the credit risk is shifted to the external investor), it is the thick senior tranche (provided it is retained) that is likely to be of most interest to originating banks when assessing the potential benefits of the SRT trade. The capital floor applied by the bank for the reference portfolio had it not been securitised (the "pool RWA") is compared against the capital floor applicable to the retained tranches (on a weighted average basis, as the junior and senior tranches will have vastly different RWAs) (the "securitisation RWA"). Where the relevant SRT tests are satisfied, an originating bank is permitted to use the securitisation RWAs instead of the pool RWAs. This may be beneficial for a bank, as they may be able to rely on the improved tranche attachment and detachment points of the senior tranche to minimise their RWAs. As the largest retained tranche (and considering the junior tranche will have a very unattractive RWA), the capital treatment of the senior tranche will likely determine whether or not the synthetic securitisation as a whole provides sufficient capital benefit to the originating bank. Given the differences in the US and the EU's implementation of Basel III, and the fact that SRT trades in the US are much less common, the imposition of the output floor to senior retained tranches of SRT trades will most acutely be felt by EU banks. Despite the underlying objective of SRT trades (i.e. to achieve capital relief), banks may paradoxically find themselves in a scenario that the overall capital position is counter-productive as a result of the higher burden of the securitisation RWAs compared to the original pool RWAs.

To achieve the optimum output, banks may be incentivised to structure their senior tranche so that their attachment point ensures that they are able to meet the minimum RWA under the SEC-IRBA and/or the capital floor under the output floor (i.e. the lowest possible securitisation RWA under the given approach). Unfortunately, in a lot of cases this would require a significant thickening of the mezzanine tranche (as the senior tranche's attachment point, and thereby the mezzanine tranche's detachment point, would have to move higher up), and the originating bank would have to find new investors to absorb the much higher volumes required and, crucially, may lead to higher premium (or, if structured as a CLN,  coupon payments by the originator.⁴ This may in turn impact the economic feasibility of such trades. Alternatively, the creation of a new "senior mezzanine" tranche in between the senior and mezzanine tranches could bridge the gap – although this may have similar issues in terms of investor base and pricing. The EBA, in its report on SRT in securitisation transactions noted that the cost of credit protection may raise significant concerns on the overall effectiveness of the SRT trade – higher pricing of a thicker mezzanine tranche, particularly if they are comparable or higher than the amount of losses an originator may otherwise be subject to, may lead to the SRT analysis on such a transaction to be undermined. This  requirement for thicker tranching in order to achieve meaningful capital relief for banks through an SRT has led to the development of dual tranche structures, where traditional credit opportunity fund investors continued to participate in the junior mezzanine tranche and the senior mezzanine tranche was placed with (re)insurance companies.  

A potential recalibration to allow for more preferential treatment for retention of senior tranches under SRT transactions, given the acute knowledge of the originating bank in respect of the reference portfolio, may incentivise more SRT trades as a balance sheet management exercise. This may either be structured via a recalibration of the capital floor under SEC-IRBA and SEC-SA, or a potential limitation to the output floor, and the treatment may be reserved for originators (i.e. not extended to third party investors in a securitisation position that were not involved in the origination of the underlying assets).

4. Impact of the "p" factor / supervisory parameter "p"

The supervisory parameter, which is seen by a number of market participants as a premium charge payable for securitisation transactions to address the modelling and agency costs, plays a significant impact in the calculation of RWAs for securitisation positions, and is indeed a key component for the calculation of securitisation RWAs. Under SEC-IRBA, the current floor of the supervisory parameter is 0.3 (regardless of whether the transaction is STS or not). More importantly, the supervisory parameter for SEC-SA is a fixed 1 for traditional securitisations and 0.5 for STS securitisations – this is notwithstanding the standard capital charge (K_SA) of the underlying exposures and/or any credit enhancement through the attachment and detachment points. This arbitrary addition of a parameter which is seen as relatively rigid for the SEC-SA approach now has the dual negative effect of not only making it more expensive for banks using the SEC-SA approach, but also for banks using SEC-IRBA which are caught by the output floor as a result of the conservative calculations under SEC-SA.

The industry (including associations such as AFME) has recommended a slight relaxation of the supervisory parameter, arguing that the modelling and agency risks which the parameter is intended to address is more limited than the impact of the parameter suggests. Industry participants have recommended halving the supervisory parameter for SEC-SA and lowering the floor for SEC-IRBA to 0.1 (for STS) and 0.25 (for non-STS). This may be particularly beneficial for SRT originators: sophisticated banks with internal models will have the requisite information to calculate the inputs required for the SEC-IRBA and may benefit from a lower capital floor and a lower RWA calculation under SEC-IRBA, and even new entrants or banks using SEC-SA will benefit from preferential treatment for the retained tranches. In addition, an argument can be made that the modelling and agency risks associated with normal securitisations (where there may be information asymmetry between the originators and the investors) may not be as pronounced where the originator is itself holding a senior tranche (and therefore not subject to information asymmetry). It remains to be seen whether this is a route which the regulators and legislators will agree to, and whether there will need to be any consequential changes if, for example, a dis-application of the output floor for retained senior tranches of SRT securitisations may create an unfair advantage over banks electing not to use SRT and instead opting to keep the underlying assets on their balance sheet (and thereby being exposed to the full brunt of the output floor).

5. Can insurers readily access the SRT market?

It must be noted that although the STS designation has also been extended under Solvency II to insurers (which are seen as a potentially large player in the SRT and synthetic securitisation space), capital charges for securitisation under Solvency II remain generally high compared to other traditional  asset classes such as loans or corporate bonds and in particular it has been argued that the STS charges nevertheless remain unreasonably high and are not appropriately justified based on past data, including for non-senior tranches of STS transactions. As mentioned above, the participation of (re)insurers has ensured that thicker tranched deals which are likely to be required continue to be economic for the banks to enter into such deals, in particular as they are often not constrained by the performance return hurdles that funded investors may have, although this has also triggered new funds to be raised targeting lower returns. This therefore may dampen (re)insurers' investment in SRT deals.

6. Synthetic Excess Spread

The recent imposition by the EU of capital requirements for synthetic excess spread ("SES") in SRT securitisations (which is not a feature of Basel III) was primarily focussed on ensuring that excess spread is not too high for the purposes of meeting the commensurate risk transfer tests under the CRR, in particular where excess spread is being used to support the mezzanine and junior tranches on a deal. The EBA did this by requiring the risk weighting of SES as a retained first loss tranche that is risk weighted at 1250% – which on synthetic deals effectively amounts to a full capital charge, and captures lifetime expected losses. The resulting commensurate risk transfer tests then incorporate synthetic excess spread, so banks need to account for the nominal value of the first loss tranche and the retained synthetic excess spread position (on full capital stack transactions, banks achieve market pricing by selling the whole stack and therefore prevent any situation where excess spread could be artificially inflated to support the junior tranches). Banks can elect to implement transactions without SES (and with a corresponding thicker junior tranche) but that may impact economics negatively.

In the EBA's final draft RTS on the determination of the exposure value of synthetic excess spread in synthetic securitisations (published on 25 April 2023), a new helpful derogation was introduced for the calculation of the capitalisation for synthetic excess spread, allowing for the lesser of the actual cash excess spread generated from the underlying assets and one-year expected losses. This deals with one of the concerns around SES, i.e. that it can be used to provide disguised or implicit credit support to a transaction. However, if the SES is less than or equal to the excess cash generated by the underlying assets, there is no risk of such support.

Although this is an aspect which is not utilised in all transactions, SES may provide a further cushion (in lieu of a thickening of the mezzanine tranche) to the senior tranche, by increasing the attachment point of the senior tranche and thereby granting it a more generous (lower) RWA. This will of course have to be weighed up against the higher value RWA in respect of the SES, although this may be a more efficient structure if it proves difficult to find additional mezzanine investors at the right pricing point.

7. Homogeneity RTSs

On 14 February 2023, the EBA published the final report on the drafts regulatory technical standards on the homogeneity of the underlying exposures in STS securitisations (EBA/RTS/2023/01) (the "STS Homogeneity RTS"). The STS Homogeneity RTS maintains the original requirements set out in EBA/RTS/2018/02 relating to the homogeneity requirements for (non-STS) transactions (the "Homogeneity RTS"), which differentiates between SME and non-SME corporate obligors. Neither the STS Homogeneity RTS nor the Homogeneity RTS define SME obligors, although the view is that the distinction between non-SME and SME corporate obligors shall be based on the internal classification and the actual practices of the originating bank. This may create certain issues for smaller originating banks or new entrants which may not have sufficient pool sizes to split these two portfolios, potentially leading to fewer transactions and those with less granular portfolios.

8. Transitional period – any help?

Although the output floor is expected to apply on a transitional basis (in the EU, starting from 50% on 1 January 2025 and ultimately going up to 72.5% by 1 January 2030) to allow for banks to prepare, this may have limited feasibility for SRT transactions. These trades are normally right-sized with a weighted average life which could surpass the entire implementation phase and, given the implementation phase is not expected to be grandfathered on the basis of the issuance of the securitisation, transactions will likely already have to assume an output floor of 72.5% even prior to the 1 January 2025 imposition of the most lenient output floor.  

What next?

The Joint Committee of the European Supervisory Authorities (the "Joint Committee"), comprised of the EBA, the European Insurance and Occupational Pensions Authority ("EIOPA") and the European Securities and Markets Authority ("ESMA") in their advice on the review of the securitisation prudential framework on 12 December 2022 did argue in favour of reducing the risk weight floor for originators in so-called "resilient" securitisations from 15% to 12% (for senior non-STS) and from 10% to 7% (for senior STS). Under their recommendation, the lower risk weight floor would only apply on retained tranches and where certain "resilient" eligibility criteria (which would look at factors such as amortisation triggers, counterparty credit risk, tranche thickness – particularly of the mezzanine externally placed tranche, and high granularity of credit exposures) are met. Although this could in theory be applicable to traditional cash securitisations, it more likely to be of relevance and use for the retained senior tranche of SRT transactions. This relaxation of the risk weight floor could bring SRT transactions into a similar capital treatment as covered bonds.   

Unfortunately, the EBA's view was that significant further reductions in regulatory capital may not be prudent and would not, unless accompanied by other measures, have the impact of revitalising the securitisation market. It is important to note that this argument was limited to the traditional "true sale" securitisations – and the Joint Committee clarified that the situation for synthetic securitisations and SRT trades was different, particularly given the predominant incentive of capital optimisation in such structures. The general view of the Joint Committee was that a number of factors in synthetic securitisations, such as the bilateral and "private" relationship with the investor, the generally high level of sophistication of the investors, the ongoing relationship which the originator bank maintains with the underlying obligors, and the primary aim of reducing the bank's credit risk (as opposed to cash securitisations, where a predominant underlying reason is a funding one), distinguished the rationale for creating capital optimisation under the regulatory framework for SRT trades.

Regulators in Europe have generally recognised the importance of SRT (in contrast to perhaps more mixed views on traditional cash securitisations) in the banking space. The very different nature of these types of securitisations justifies a potentially different treatment. The imposition of the arbitrary supervisory parameter p, even in scenarios where there is no information asymmetry, leads to a decrease in efficiency of SRT transactions and, combined with the potential need to thicken the mezzanine tranche, could lead to further decreases in originating banks' appetites given the higher premiums payable for such protection (although we note that this may be partially mitigated through the creation of the senior mezzanine tranche).

Significant arguments have also been made by market participants that the EU regime for securitisations is already a relatively conservative and gold-plated framework when compared with other jurisdictions (in particular the US). Against the backdrop of the relative success of US investment banks, the US has not yet fully adopted Basel III, and still applies the Simplified Supervisory Formula Approach ("SSFA"), which does not include the most recent premium charge (through the supervisory parameter p) applicable to securitisations. For example, under the standardised approach, the supervisory parameter in the US is fixed at 0.5 (which is the equivalent of an STS floor in the EU), with the SSFA having supervisory parameter equivalent inputs of close to 0. This puts EU banks at a competitive disadvantage against their US counterparts. 

The current amendments to the Capital Requirements Regulation⁵ (the draft of which has already been published and which aims to incorporate the majority of the Basel III requirements) is still under review, so it remains to be seen whether the regulators take into account the relatively strong feedback from the industry.

Conclusion

One of the main aims of the EU Capital Markets Union is to enhance the capital markets in order to enable companies (in particular SMEs) to have better access to capital. Incentivising the use of traditional securitisations, which are frequently used by non-bank lenders as well as banking originators, as well as optimising the capital treatment of SRT trades, would unlock a significant amount of capital which can then flow to EU borrowers. SRT trades have been recognised by most European regulators as playing a pivotal role in the ability for banks to risk-share transactions without requiring equity raises or liquidation of assets. An argument can therefore be made that properly targeted amendments to the regulatory framework, which take into account the unique nature of SRT transactions (as compared to other securitisations), could meaningfully assist in the market's recovery and bank's profitability without compromising financial stability, and allow banks to unlock excess capital which can be used to fund the real economy.