This blog was published on 14 September 2021.
Intended audience: CROs, CFOs, Treasurers, Heads of Basel implementation programmes, risk professionals with an interest in Basel 3.1’s impact on the banking sector.
Time to read: 5-7 minutes
At a glance: The implementation of Basel 3.1  will require significant effort from all banks .
- Banks that currently use the standardised approaches (standardised banks) will have to invest in their risk weighted asset (RWA) calculation infrastructure to implement changes to the standardised approaches for credit, market, operational, and credit valuation adjustment (CVA) risks.
- Banks with permissions to use internal models for capital purposes (IM banks) will also have to calculate the new standardised approach RWAs for all exposures in all risk categories, as well as implementing changes to the IM approaches. For IM banks that are constrained by the output floor after Basel 3.1 is fully implemented, standardised approach RWAs will determine their overall capital requirements.
- Some IM banks will be calculating standardised RWAs on a routine or ongoing basis for some portfolios or exposure classes for the first time and some still have much to do to prepare for the requirement to calculate standardised approaches for all exposures.
- This is an area of growing concern for prudential supervisors, who want to see banks do more in the run-up to Basel 3.1 implementation. Because standardised RWAs could be the determinant of capital requirements for IM banks, supervisors will expect IM banks to demonstrate that their standardised RWAs are subject to the same standards of accuracy, oversight and governance that supervisors expect of their IM RWAs.
Further reading: For more analysis from us on the dynamics of the standardised output floor, see our blog from 2020 here.
Basel 3.1 encompasses changes to both standardised and internal ratings-based (IRB) calculations for credit risk, changes to standardised and modelled approaches to market risk, a new operational risk framework and amendments to CVA, as well as changes to external reporting, a new real estate exposure class and the need for IM banks to calculate standardised RWAs for all exposures due to the implementation of an output floor. Although the implementation challenges differ between standardised and IM banks, there are some common themes. In this blog we will look at three common themes: managing increased complexity; sources and uses of data; and managing supervisory expectations.
Managing increased complexity
Increased complexity arises from the imposition of new approaches (e.g. the Standardised Measurement Approach for operational risk) that require new data and calculations; changes to risk sensitivity (increased risk sensitivity for some standardised asset classes, decreased risk sensitivity for many modelled approaches); and, for IM banks, the imposition of the output floor.
Key challenges will include: understanding the increased complexity, granularity and risk sensitivity of new approaches; how these will affect portfolios and business decisions; and how to manage a more volatile RWA and capital position.
For example: under the current standardised approach, all performing exposures that meet the characteristics for the retail asset class (other than mortgages) receive a risk weight of 75%. Under the new standardised approach, customers with limit-based facilities (credit cards and overdrafts) may receive a risk weight of 45% if they meet the classification of “transactor” – a customer that has repaid their facility in full for 12 consecutive months at each scheduled repayment date, or overdrafts without any drawdowns over the previous 12 months. Retail customers who do not meet the transactor definition will receive a risk weight of 75%. Transactors that cease to repay in full every month or who use their overdraft will shift into the 75% risk weight bucket, and can only be re-classified as a transactor once they have repaid in full or not drawn their overdraft for 12 consecutive months. This increased granularity and risk-sensitivity will allow standardised banks to hold capital more commensurate with the risk in their portfolios, but will lead to greater RWA volatility as customers move between the 45% and 75% risk weight buckets. Standardised banks will face systems and data challenges in tracking this payment behaviour, if they wish to take advantage of the lower risk weight.
Of the many changes and challenges IM banks will face, principal among them will be implementing the output floor and incorporating it into capital management processes.
IM banks will need to ensure they can calculate both modelled and standardised RWAs and then incorporate the calculation of the output floor into portfolio planning and stress testing processes. Capital planning is already a process that requires assessing the constraints of RWAs, liquidity and leverage; adding in the output floor constraint will further increase the complexity of this process. IM banks will need to ensure that board members and senior management understand the implications of this increased complexity.
An example of this increased complexity arises in the Real Estate exposure class. IM banks will need to classify their real estate exposures into either materially dependent or not materially dependent groupings for residential and commercial exposures, according to the extent to which payment of the underlying loan depends on income generated by the property that secures the exposure. While the dependent/not dependent status will not necessarily affect modelled RWAs, the standardised RWAs that drive the calculation of the output floor can be significantly different for dependent vs non-dependent exposures, so this will need to be factored into strategic decisions on the real estate portfolio.
Sources and uses of data
Both standardised and IM Banks will have to source new data, make changes to existing data, and develop new internal and external reporting as a result of Basel 3.1. In the UK, the PRA paused a number of Section 166 reviews into regulatory reporting at the start of the Covid-19 pandemic. Many of these have now concluded, and the PRA has recently issued a “Dear CEO” letter  setting out the thematic findings from its reviews. In short, the PRA has made clear its view that many banks do not take seriously enough their obligation to produce accurate regulatory reporting. Banks should anticipate that supervisors will expect all RWA calculations to be subjected to the highest standards of governance, controls, reconciliation and accuracy.
New collateral data, such as “at origination” loan to valuation ratio, will be required for standardised mortgage exposures under Basel 3.1 and may not be easily obtained from existing systems or data sources. Where standardised banks are looking at finding proxies for data that is not readily available, they will need to demonstrate to their supervisors that the proxies are reasonable and that they have added an appropriate margin of conservatism to reflect any uncertainty.
IM Banks will face all the challenges standardised banks face in implementing the revised standardised approach, plus a number of additional challenges particular to modelled approaches. There will be instances where the same data is needed for both IM and standardised calculations, but the data is used/processed differently for each purpose. For example, collateral data is used in models for credit risk under IRB approaches and directly in RWA calculations for the standardised approach. A key difference in managing and using the data is that the eligibility of collateral to reduce RWAs differs between the IRB and standardised approaches. The use of collateral data in standardised calculations also requires collateral management systems that can link collateral to customer, facility and exposure in ways that may not otherwise be used in the running of the business (e.g. allocation by bank subsidiary/legal entity for regulatory reporting).
Managing supervisory expectations
For both standardised and IM banks, it will be important to demonstrate to supervisors, rating agencies and other stakeholders that the new standardised calculations are being undertaken to the same level of rigour and standards of control and oversight as supervisors expect of existing calculations. Implementing a tactical solution to meet the implementation date, without having a fully resourced and budget-approved plan for a business-as-usual solution, to be delivered in short order, will be a high-risk approach.
Supervisors will want to see evidence that standardised banks have implemented the changes to the standardised approaches correctly and that they understand the effects of increased risk sensitivity and volatility across all risk types. This will increase the challenge of capital forecasting and stress testing and could increase significantly the challenges of capital management for standardised banks.
IM banks will need to persuade their supervisors that they are making the appropriate investment in systems, controls and data for their standardised calculation infrastructure.
The need for this investment is clear: IM banks caught by the output floor will have their overall capital requirements determined by their standardised RWA number. Supervisors will want to see evidence that the standardised number is robustly calculated, governed and appropriately conservative. Banks’ boards and shareholders will want to be confident that all available options to recognise permitted RWA reduction are being taken into account. In order to meet both of these expectations, banks will need to have a standardised calculation infrastructure that is robust, well controlled, and meets high standards of data accuracy and completeness.
One consideration for all banks in relation to all the challenges above is change capacity – can they implement the necessary changes through internal systems change processes in time? Given it is now September 2021 and implementation is currently scheduled for 1 January 2023, both time and the capacity of human capital to develop and implement these changes are rapidly dwindling resources.
Addressing standardised implementation challenges early will help banks better implement the revised standardised approaches and calculate a more accurate risk weight based on those approaches. This will flow through to the calculation of the output floor and could help mitigate its capital impact for IM banks. This, for most banks, will be a multi-year task that will require planning and investment. In our view, the banks that implement earliest will have an important competitive advantage over those that leave it too late.
One last point: we remain of the view that implementation of Basel 3.1 by 1 January 2023 looks challenging in some jurisdictions, particularly the EU where legislative packages of this type typically take at least two years to negotiate after the publication of initial proposals, currently due in late October. However; while the onset of the Basel 3.1 framework may ultimately be later than 1 January 2023 in some jurisdictions, we believe where this is the case, the five year transition period for the output floor will be shortened so that the 1 January 2028 target for the full 72.5% output floor calibration is maintained. In addition, we believe that it would not be advisable for banks to delay work on developing and implementing the capability to accurately calculate standardised RWAs, as supervisors and other stakeholders will increasingly expect banks to be able to speak confidently about the impact of Basel 3.1 on both pillar 1 and pillar 2 capital requirements.
 We use “Basel 3.1” to describe the December 2017 BCBS publication Basel III: Finalising post-crisis reforms
 Basel 3.1 will affect a range of financial institutions including banks, building societies, broker dealers and some investment firms. We use “banks” as a catch-all term.
 Dear CEO letter Thematic findings on the reliability of regulatory reporting (bankofengland.co.uk)