Speech
Many thanks for having me along. It is great to be here at ISDA. I’ll cover three broad things on Basel 3.1 today. First, where do we stand in our implementation in the UK. Second, a reminder of what we’re trying to achieve with the new FRTB rules. And third, something on the areas in FRTB where we might need to look a little closer in due course. I’ll finish by broadening a little and looking beyond FRTB, although many would say there is nothing else in life.
First, on where things stand. Basel 3.1 remains a key part of the post crisis reforms. Risk measurement, fundamentally, needs to be accurate. And we had lost confidence in this post GFC, including because of the variability across firms in risk weights that comes from using modelled approaches. That concern speaks to getting Basel 3.1 done as soon as possible.
But if Basel 3.1 feels to you like the project that comes quite close to never ending, believe me when I say that ‘I sympathize’. It is in everyone’s interests now that its implementation is finished as soon as possible, so that we can all move on, with the benefits of the new regime safely tucked away. We have the vast majority of our rules in the UK entering force in January 2027, and for those there will not be any further substantive changes to what we have already published. For those, firms should be preparing now.
They will adopt legal force when the Treasury makes the SI that allows us to finalise those rules, and we will publish the final rules in Q1 next year. That will give firms a year to prepare before they need to report to us on the new regime, and allow us to undertake the necessary Pillar 2a reset that we promised, to make the Basel 3.1 approach complete and avoid any double counting with the more accurate risk weights generated by Pillar 1. As I’ve said a few times now, this approach to avoiding double counting has been a fundamental principle for us since the start on Basel 3.1, with no need to hold capital twice against the same risk.
The only part of the regime in our proposals entering force after January 2027 is the modelled part of the market risk rules – the FRTB – which enters force in January 2028. That is to allow us to finalise those in light of any developments in other jurisdictions, including the US and the EU. The FRTB applies to truly global and cross border business, and so there is real value in having consistency across major jurisdictions. My sense is that both the US and EU are making good progress in their plans.
That brings me to my second topic of the FRTB and why we have it. The big point is FRTB was called fundamental for a reason. I would argue there were three main GFC related issues it was trying to fix:
- There was a trading book / banking book boundary which was too subjective, leaving the door open to an uneven treatment across firms and for risks to end up in a part of the capital framework that isn’t designed to deal with them. That has been addressed in the FRTB with a much clearer boundary.
- The existing internal model approach missed important risks, for example the different levels of liquidity of trading book positions, and was a bit of a smashing together of elements that have been added to the rulebook over time as and when we spotted gaps in the framework. This has been made more coherent now in the FRTB.
- There was not a standardised approach that acts as a risk sensitive alternative to modelling, meaning big banks had little choice but to use models. And that means supervisors couldn’t realistically challenge firms, and ultimately stop firms from using modelled approaches, even if the models weren’t performing well. The FRTB greatly improves the standardised approach.
So we wanted to fix those issues by having a better-defined boundary between the trading book and the banking book, have a new standardised approach that proves to be a credible alternative to modelling, and having built a credible SA, set a firmer and higher bar for modelling to improve standards. This includes recognising that some risks can’t be modelled well and therefore need capital add-ons. That, of course, is an approach we have been taking with UK firms for some time now, and so is not a particularly new element for the UK, and less of a big change here.
All of this adds up to being an important shift. A key goal – implicit in the new framework – is that modelling should be perfectly achievable for good models, but banks should not just assume they can always use models regardless of whether those models are good, and up to scratch.
That brings me to my third topic of how well does the FRTB measure up to those aims?
Well, it’s a little difficult to fully explore this given it isn’t in force yet. But firms do know the direction of travel, and have been working on preparing, which does give us some information.
But perhaps the place to start with this topic is a brief description of how we got to where we are today. Post GFC, we had identified a range of high-level issues, as I’ve just described, that needed fixing. The Basel Committee set the appropriate working group, chaired by my excellent colleague Derek Nesbitt, to work in identifying fixes. Perhaps more so than in other parts of the rulebook, it appears quite possible in the market risk world to be able to identify changes and fixes that look good on paper and should work well, but generate a different outcome in practice. The early efforts of this Basel group were criticised for being misplaced in practice by industry, so Derek’s group made very extensive use of industry expertise in designing the new rules. Meaning that the new FRTB had a very heavy dose of industry input in its design, and industry wisdom.
So it should work well. But as I said, what works on paper, even when designed with industry input, is not guaranteed to be perfect in practice, as annoying as that is.
What do early preparations of firms tell us how well the FRTB achieves the aims I outlined earlier. Well, I think the two of the aims are broadly in good shape. We do have a much better defined and clearer boundary established, and the new standardized approaches are more risk sensitive.
But it’s the other aim around modelling where some questions seem to be lingering.
To be clear, the aims for the modelling component – the new IMA – are good. And many of the most significant improvements, such as moving to a single, new type of model for the market risk component, and the improved way that liquidity is incorporated into it, seem to work well. But the question that is emerging for a small number of parts of the framework, such as the P&L Attribution Test, is have we put the bar for modelling at the right level and is it too complex? As I’ve already argued, despite much of the design of the IMA reflecting firms feedback as the standard was re-worked, that in and of itself is not a guarantee that it will work well in practice. In truth, as is often the case with changes that look good on paper, it needs some data and in practice experience to really assess that.
And there, the struggle has been to get good quality and convincing data. The good news is that is now starting to come through, as firms invest in getting models ready. This investment in getting ready is important, so that we can understand what can be fixed just by a bit of persistence, and what might need to be fixed through possible changes. Meaning it is looking easier to check whether the bar is where we had intended it to be.
And its for that reason that, in the early part of the Summer, we announced we would be implementing our entire Basel 3.1 package in January 2027, as previously announced, but we consulted on postponing the IMA modelled approach until January 2028. In the meantime, the idea is that firms can stay on their current market risk models. This would also allow time for greater clarity to emerge in other jurisdictions on their own implementation in this area as well, as consistency really matters. We will publish our final approach on that consultation with the rest of the final package in Q1 next year, meaning that we are on track for our January 2027 implementation date for the main package.
As with all of our Basel 3.1 package, we have tried to write the FRTB rules in as clear a way as we can. But there is no avoiding that they are quite involved. And no matter how clearly we try and write the rules, there will be aspects that firms want to clarify with us. By this, I don’t mean re-opening. We need a strong focus on getting the package across the line now, rather than continually re-opening. But clarifying is a distinct and very legitimate need. To accommodate this, in the PRA we’ve launched a new series of Banking Policy Roundtables. There are three per year, rotating the organization of them across UK finance, ISDA/AFME, and the building societies association. These particular roundtables have had the aim of grouping clarifying questions, and we’ve just had a very successful one organized by ISDA and AFME covering market risk.
That’s all I have time to cover on market risk. Let me finish with the briefest of tours of the horizon.
The meta headline is that we remain very SCGO focused.
We have been doing a lot this year. Following through on Mansion House with PRA initiatives. For example, to assist firms, through the ‘scale up’ unit in the PRA and the wider concierge service for firms new to the UK. We’ve also made commitments, with KPIs, to make regulatory transactions like authorisations, SMCR applications and model approvals speedier and more agile.
Outside of Mansion House, we’ve launched deeper review of SMCR and modernized a range of MREL related thresholds so the regime keeps pace with the times.
And all of that is on top of earlier initiatives covering Basel 3.1 where I think our implementation of the global standard is very proportionate, our strong and simple regime for small firms, securitization reform, remuneration reform, and many initiatives from my colleagues working on insurance reforms.
My purpose there is not to blind you with a long list. It is to get across a sense that it is a very big focus for us, that we have been and continue to take very seriously.
The big thing I didn’t include in that list, but is very much a big focus right now, and so I’ve pulled it out separately, is the capital refresh project being run by FPC and PRC. I don’t have anything to say on that here given it is separate to our Basel 3.1 work. But we released our analysis and where we have got to in the December Financial Stability Report, today in fact. But I can say it has, and will continue given the next steps, to occupy a lot of bandwidth in the PRA as we look to do the best evidence-based job we can in order to do it justice.
Thank you very much for listening.
