Where a bank is unable to use any of the SEC-IRBA, SEC-ERBA and/or SEC-IAA in respect of its exposures, it will be required to use the SEC-SA model. There is no requirement to obtain supervisory approval in using the SEC-SA model (provided the bank is able to calculate the relevant inputs required).
The bank would have to incorporate (i) the standard capital charge for the underlying exposures ("KSA"), (ii) the ratio of delinquent underlying exposures to the total underlying exposures in the pool ("W"), (iii) the tranche attachment point A, and (iv) the tranche detachment point D. The supervisory parameter "p" in the context of SEC-SA is set at 1 for securitisation exposures, 1.5 for resecuritisation positions and 0.5 for STS Securitisations. Although Basel III does include capital weights for resecuritisation positions, it must be noted that resecuritisations are not permitted under the EU Securitisation Regulation or the UK Securitisation Regulation.
All of a securitisation SPV's exposures will be treated as an exposure to the "underlying pool" – even when the exposure is to contractual counterparties such as swap counterparties or liquidity facility providers. Banks are able to exclude these exposures from the calculation of KSA if they are able to demonstrate to their national supervisors that these risks have been mitigated (for example, by requiring cash collateral to be posted by a swap counterparty).
The SEC-SA is subject to the following risk weight floors:
- 10% for senior tranches of securitisation positions; and
- 15% for non-senior tranches of securitisation positions.
v. Default position (1,250% risk weighting)
For any scenario where the bank does not have sufficient due diligence to perform the calculations under any of the above models, it must apply a risk weighting of 1,250% (that is, hold eligible capital in a value equivalent to 100% of the securitisation position).
What are some issues with the Basel III proposed framework?
Impact of the Output Floor on SEC-IRBA, SEC-ERBA and SEC-IAA
One of the amendments to the Basel III framework was the addition of an "output floor", which was created to address modelling risk for the internal calculations (SEC-IRBA, SEC-ERBA and SEC-IAA, the "Internal Models"), where it has been argued that banks have sometimes mismodelled or undervalued their assessment of the risk, and that this created a larger than expected divergence between the Internal Models and SEC-SA. Because of the role and discretion granted to national regulators, there was also the additional risk of discrepancies in approved internal models between different countries. The output floor essentially creates a minimum RWA for any of the Internal Models based on the RWA which would have been applicable had the bank used the standardised approach. This essentially pegs the potentially varied Internal Models to a standardised output.
Where banks use one of the Internal Models for their securitisation positions, they must:
- Input 1: calculate the RWA under the Internal Model;
- Input 2: calculate the RWA under the SEC-SA method;
- Input 3: multiply the result in Input 2 by 72.5% (i.e. apply a haircut of 27.5% to the SEC-SA result);
The RWA which the bank will be entitled to use under the Basel III framework will then be the higher of Input 1 and Input 3. This essentially means that the maximum benefit which a bank may gain from the Internal Models (compared to the SEC-SA calculation) is limited to 27.5%. This may be particularly problematic where (i) a bank has a particularly good knowledge of the underlying exposures (for example, because it originated them) and is therefore able to obtain a more realistic and lower RWA than under the strict SEC-SA method; or (ii) there are credit enhancement features in the securitisation which are omitted from calculations for the purposes of the SEC-SA. In these instances, banks may find that even though their Internal Model calculates an accurate depiction of their RWAs, the existence of the output floor may nullify a portion of the benefit. Ironically (in light of the Basel Committee's objectives of encouraging investors to diligence the underlying positions ), this may mean that, at specific thresholds, banks may determine that there is no capital regulatory incentive to undertake further due diligence in respect of an underlying exposure as they would be unable to derive any further benefit from such diligence.
For illustration, in the below table a bank calculated its RWA under the SEC-SA approach to be 100%. It then subsequently ran the Internal Model (for example, the SEC-IRBA), and calculated a more accurate RWA of 60% (due to a number of variables in the transaction which the SEC-SA did not take into account). However, as a result of the operation of the output floor, the bank would have to use an RWA minimum of 72.5%, thereby losing the benefit of the additional 12.5% (i.e. 72.5% - 60%) it was able to obtain under SEC-IRBA.
The European Banking Authority (the "EBA"), on 2 August 2019 in its policy advice in respect of the Basel III reforms suggests that the imposition of the output floor will lead to an increase in banks' capital requirements by an average of 9%. This suggests that, on average, banks are able to use the Internal Models over the SEC-SA approach to get (an average) of a 36.5% lower minimum capital requirement using the Internal Models. The impact of the output floor will also be concentrated on the subset of banks which have the supervisory approval and the internal operational infrastructure to be using the Internal Models, so the increase in RWA for the banks which use the Internal Models may be higher than the 9% which is predicted across the entire EU average. The current EU Banking Package 2021 therefore includes a transitional implementation which starts with a 50% output floor (from 1 January 2023) and, over a period of five years, progressively increases to the 72.5% (from 1 January 2028) minimum required under Basel III.
In the HM Treasury's review of the Securitisation Regulation (Report and call for evidence), published on 13 December 2021, multiple respondents flagged the potentially high impact of the output floor for certain securitisations. HM Treasury noted that the output floor would be examined in the PRA's Basel III consultation which was expected to take place in the second half of 2022, but is now expected in the fourth quarter of 2022. UK Finance, the banks' trade body, has highlighted the level of application of the output floor as one of the key concerns with the new Basel III amendments.
In its targeted consultation on, among other things, the functioning of the EU Securitisation Regulation, the Association for Financial Markets in Europe ("AFME") on 30 September 2021 argued that due to the layering of conservative parameters embedded in the calculation under SEC-SA, the output floor would likely have a disproportionate effect on the treatment of securitisations by bank originators. This was likely due to the fact that the discussions in respect of the output floor during the Basel III negotiations were not focussed on the impact of the output floor specifically on securitisation positions, but more widely in the context of risk weighted assets.
The imposition of the output floor is also expected to affect different asset classes within securitisations differently. This is largely based on the difference between the KSA under the SEC-SA model which a bank may rely on (which is subject to a floor) and the KIRB under the SEC-IRBA model. Where the bank has sufficient information in respect of the underlying exposures, it may be able to apply a risk weight which is significantly below the SEC-SA floor for that particular asset class and, where the magnitude of such difference is higher than 27.5%, may find itself subject to an RWA floor. Banks investing in asset classes which have a higher SEC-SA floor (for example, claims on sovereigns, unsecured loans, claims secured by commercial real estate, or retail loans) may be more incentivised to use the SEC-IRBA model given the potential discount can be much higher. However, the Internal Models' effectiveness in these cases may be limited by the output floor.
In the Basel Committee on Banking Supervision's Basel III Monitoring Report dated December 2021, a sample of 84 banks were asked to submit their risk weights under both Internal Model and the SEC-SA. The average weighting showed that the difference between non-STS SEC-IRBA and SEC-SA was 99.5%, meaning the risk weights under the SEC-SA model were, on average, double the amount under the SEC-IRBA method. This would mean that the output floor would operate to significantly increase the risk-weighting used under the SEC-IRBA model. The position for STS Securitisations was much less, at 67.9%. Although there are risk weight floors, it is important to note that under SEC-IRBA, the internally calculated supervisory parameter "p" is subject to a floor of 0.3 (regardless of whether the securitisation is STS or not). Meanwhile, the STS designation does provide benefits from an SEC-SA perspective (where the supervisory parameter for an STS securitisation is half of that of a traditional securitisation). This means that STS transactions may benefit from a lower SEC-SA and, even though a bank may elect to use the SEC-IRBA method for that particular STS transaction, the fact that the corresponding SEC-SA RWA is lower would result in the output floor being higher.
Impact of the supervisory parameter "p"
Both the Internal Models and the SEC-SA include the application of the supervisory parameter "p" (or the "p-factor"), which is intended to reflect the overall level of capital required for the portion of tranches that reside above the securitisation position which absorb losses up to the amount of capital that would be required if the underlying exposures were held directly by the bank (and not securitised). Under SEC-SA, there is a fixed p-factor of 1.50 (for resecuritisations), 1 (for traditional securitisations) and 0.5 (for STS Securitisations). Under the SEC-IRBA, banks may calculate their own supervisory parameter based on the underlying exposures (as mentioned above, this would require knowledge of the KIRB, or the capital charge for the underlying exposures had they not been securitised as well as information in respect of the number of loans, granularity of the loans and the exposure-weighted average loss-given-default of the pool) subject to a floor of 0.30.
The p-factor imposes a premium capital charge on all securitisations (regardless of seniority, maturity, or any other credit enhancing features of the transaction), and such a surcharge has been questioned by a number of market participants. In particular, it is noted that under the Internal Models, the p-factor floor is the same for traditional and STS Securitisations.
Putting aside the RWA floors in respect of the underlying assets (which are applicable equally to holders of securitisation positions exposed to those assets and banks holding those assets directly), one could argue that there are already alternative correction factors which impose an implicit premium on securitisations (including capital floors set out in the hierarchy of models above, to both senior and non-senior tranches). Whether or not an additional floored adjustment is required in the form of the p-factor to address what the Basel Committee perceived as inherent modelling and agency costs associated with a securitisation, and whether or not such p-factor is appropriately calibrated, remains to be seen. Currently, the p-factor has the effect of locking up additional capital just by virtue of the transaction being a securitisation. Although certain banks have questioned the efficacy and appropriateness of the p-factor, regulators and central banks on the whole did not generally see any reason to diverge from the Basel III framework in this respect. On 7 September 2021, the central banks of the Member States of the EU drafted a joint letter to the European Commission affirming that the EU should incorporate the Basel III framework in a "full, timely and consistent" manner.
What next?
Although the flexibility of the hierarchy of approaches (including the ability to use the Internal Models) is generally welcomed, the introduction of the output floor has brought renewed focus on the SEC-SA approach, which will now become much more prominent given its role in determining the minimum RWA for any securitisation exposure. And certain criticisms of the SEC-SA approach which formerly would have been dispatched by the ability of banks to elect the SEC-IRBA model may now be tabled again, particularly in respect of the tabular approach to different asset classes which does not always take into account certain qualitative features of the assets which decrease the credit risk associated with such positions. The EBA, in its impact study in 2020, concluded that the imposition of the output floor would result in a weighted average 6.7% increase in the minimum capital which banks must hold. This figure will clearly be much higher for the larger institutions which rely more heavily on SEC-IRBA, and may be closer to 0% for medium-sized or smaller banks which regularly use the SEC-SA.
At the time of writing this article, it is not clear how the EU Banking Package 2021 will progress through the EU legislative process, nor the approach of the UK in implementing the Basel III framework (which is expected to be consulted on during the latter half of 2022). However, given the substantive nature of the legislation and a number of impact studies highlighting the potential capital burden on European banks, it is unlikely that implementation will occur before the current Basel III timeline of 1 January 2023, with some further developments and another potential extension possible.
In October 2021, AFME responded to the EU Banking Package 2021, agreeing that banks have significantly strengthened their resilience and ability to stress test their balance sheets since the financial crisis as a result of the capital requirements, but that a number of impact studies have suggested that further increases may have negative repercussions on the larger banks' ability to free up capital to deploy in the economy. It also called for further adjustments to the output floor, which it saw as having material effect on certain asset classes and business lines. This will inevitably focus the attention of the European Parliament when reviewing the EU Banking Package 21.
It is also worth noting that the above introduction of the output floor and the recalibrated risk-weighted assets for securitisation positions are in addition to a number of other capital and liquidity requirements under Basel III, as well as other regulatory considerations such as large exposures, counterparty and market risk to which banks are already subject to. These changes should not be assessed in isolation, but rather holistically in light of banks' other regulatory obligations which are aimed at improving prudential oversight and minimising systemic risk and which have generally been seen to be successful in ensuring banks' resilience is considerably improved from the pre-financial crisis levels.
Conclusion
The securitisation market plays a key role in the financing of the real economy, as well as a number of issues which are becoming more prominent in global regulators' views – such as the environmental, social and governance ("ESG") factors and the transition to the sustainable economy as envisaged under the European Green Deal and the green securitisation framework being developed in the UK.[6]And, unlike the last financial crisis, the general consensus was that banks were in much more robust financial positions following the COVID-19 pandemic and the unprecedented economic shock that arose as a result. This has allowed them to play an important role in re-energising the economy and deploying capital to provide finance in the capital markets. The securitisation market has generally been seen as innovative and capable of being able to accelerate and drive market developments. The introduction of the output floor should be observed contextually with the other capital and prudential obligations of banks, and should be weighed against the creation of a capital charge which may increase the cost of capital for banks to invest in securitisations (therefore have a dampening effect on originators wishing to use securitisation as a means of funding their businesses) and, paradoxically, removing the incentive for banks to commit to additional due diligence of the underlying assets in a securitisation where they feel the output floor caps their preferential capital treatment. Encouraging investors to participate in securitisations and unlocking excess capital may play a pivotal role in achieving some of the more ambitious capital-intensive objectives of regulators.
The finalised Basel III framework, how it is ultimately incorporated across various jurisdictions and the impact this may have on banks' capital requirements may therefore coincide with an important global crossroad on which governments and regulators are seeking to drive the green initiative to help achieve a net zero goal.[7]
References
[1] This is sometimes referred to as "Basel 3.1", as they are amendments to the (original) Basel III agreed in 2010. Because of the magnitude of some of the changes, they have also been referred to as "Basel IV".
[2] There are a number of different institutions which are the subject of various capital and prudential requirements, including (but not limited to) banks, insurance and reinsurance undertakings (predominantly dealt with under the Solvency II framework in Europe) and collective investment undertakings.
[3] There are multiple tiers of eligible capital under Basel (Tier 1, Tier 2 and Tier 3), which broadly govern the quality and liquidity of assets which the bank must hold in respect of its position. Tier 3 capital is being removed under Basel III. In summary, Tier 1 capital refers to the bank's "core capital" such as equity capital which is able to absorb losses on an ongoing basis without the bank having to cease trading. Tier 2 capital refers to the bank's capital which will only absorb losses in the winding up of the bank (such as certain categories of subordinated debt).
[4] The changes under Basel III has seen the removal of a number of different asset classes which can now no longer be subject to the SEC-IRBA, SEC-ERBA or SEC-IAA models, and will need to be calculated based on the SEC-SA or the default (1,250%) approaches.
[5] https://www.bankofengland.co.uk/-/media/boe/files/prudential-regulation/consultation-paper/2020/cp1420.pdf.
[6] We note that on this point, the EBA has, on 2 May 2022, launched a discussion on the role of environmental risks in the prudential framework.
[7] Although there have been discussions about potential EU green securitisations which (not unlike STS Securitisations) may benefit from a more favourable capital regime, these discussions are currently at their infancy and therefore outside the scope of this article, although we will be actively monitoring this space.