Welcome to the Treasury Select Committee on Tuesday 14 July 2026. We are here today to consider the Bank of England’s financial stability report, in one of our regular sessions with the Governor. I am delighted to welcome Andrew Bailey, the Governor of the Bank of England, who is joined by Randall Kroszner, an external member of the Financial Policy Committee at the Bank; Nathanaël Benjamin, executive director for financial stability strategy and risk at the Bank; and Stephen Blyth, another external member of the Financial Policy Committee. Governor, I will start with you. You told us when we spoke to you in January that the risks of financial instability had increased in 2025, for the obvious reasons—all the geopolitical events that we have all witnessed. Do you think that the risks have increased further since then?

Yes, for a number of reasons, but I will pick out two. One reason is the developments in the world economy, particularly developments in the Gulf and the middle east conflict. I do not think that that directly increases the structural issues that we have identified in the past, but it obviously increases the risk of things crystallising, and I think it would increase the impact were things to crystallise. The second thing, which I have written to you about in the last week, is the big new development, I would say: the news on frontier AI and the cyber-risk consequences of that. Unfortunately, I think it is an unambiguous yes to your question.

We will come to AI in more detail later. On the conflict in the middle east, we keep having a ceasefire that is then not a ceasefire. What is your prediction for how that will impact the long-term situation in the UK economy?

If you had asked me this question a week ago, I would probably have given you a slightly different answer. At the hearing that we had on monetary policy a few weeks ago, I would have said that I thought things were panning out at probably the lower end of the three scenarios that we had painted in the MPR, but I would have added the very strong caveat that it seemed to me that the situation remained unstable and that the ceasefire was fragile. Sadly, that instability has come to pass. I certainly do not feel qualified to judge how long the resumption of hostilities will go on or how it will end, but it underlines that this is going to be an unstable process for the foreseeable future. Energy prices have come down. As of a week ago, they were sitting a bit above where they were before the crisis; they are now somewhat further above. Crude oil prices are well down on their peak. I think we are also seeing continued, fairly soft evidence on the pass-through into UK prices, but we will get another inflation number next week. The one other caveat I would add, which I do not think gets enough coverage, is that although crude oil prices have come down, the prices of refined products have not come down as much. At this point, we are talking about things like gasoline and diesel, which are the end products—those are the things that matter. Among the reasons for that, in the global context, are the story around China and the story around Russia. China has substantially reduced its imports of crude over the last couple of months. That has been interpreted as a softening of the market and a reduced demand story. However, as far as we can tell, it has substantially reduced its exports of refined products. In a sense, it has not necessarily changed its balance, and that is tightening the global market for refined products. The other thing—and this is obviously good news in other respects—is that Russia has had a serious blow to its production capacity because of what Ukraine has done. I make the caveat that we all spend all our time quoting crude oil prices, but you have to be a bit careful, because I do not think the refined product prices have come down as much.

One of the challenges is the damage to the infrastructure in the middle east. Once there is a ceasefire that holds, it will take a minimum of eight months to get things back on track. What modelling is the Bank doing to look at the impacts on the UK economy in the short to medium term?

We talk a lot to people in the industry, and we talk a lot to our counterparts in the Gulf area; I spend a lot of time talking to my counterparts. The story there, if we take oil and gas separately, is that on the oil front there has not been as much physical damage to the production infrastructure, although obviously the blockages in the Gulf are another matter. The message that we get from the Gulf states is that if they can ship the stuff out, they will resume production very rapidly. Until last week, they had resumed production rapidly. They also have quite big plans, which they are already executing, for building new pipelines to get around the strait of Hormuz and get to ports south of the strait. That will obviously take some time. On gas, the story we get from the Qataris, who are the people who matter in that market, is quite precise. If I remember rightly, they say that about 83% of the supply is intact. The 17% that was hit by the Iranians in the attack some months ago will take longer to bring back. They too have quite big plans for increasing their overall output of gas, but again they are talking about 2029 or 2030 for that. They are clearly very determined to get supply back on to the market, but obviously what we are seeing in the strait now is going to impact that.

So there is a rocky period ahead. I will ask some of our other witnesses about that in a moment. We are about to have our seventh Prime Minister in 10 years. We have had a lot of change at the top. We are talking about what is happening internationally, but domestically do you think that those changes and all that upheaval is good for financial stability?

Well, I am not going to comment on what is happening in politics.

It is about the stability point.

Let me say two things, which are important. One, as the last few days have demonstrated, is that the big driver of moves in financial markets is not what happens in the UK; it is the global setting. But the evidence does indicate that the UK situation is supported on our side by monetary stability and financial stability, and on the other side there is the fiscal framework. Obviously, you did see a bit of movement in markets, briefly, on news on the fiscal framework, but that came back quite quickly.

You cannot give private advice to us in public, but we will have a new Prime Minister and a refreshed Government, I suppose you could say, coming in next week. What would your top tips be for making sure that they have the best chance of maintaining financial stability in the UK?

The overall message I would give is that the big issue is growth in the economy. We can do all we can on financial stability; no doubt we will come on to this, but I do actually think that there are signs of a resilient financial system, as we describe in the financial stability report, in two respects. One is that the core banking system is resilient, as I am sure we will come on to. Also, debt levels in the household sector and the corporate sector are not stretched at the moment. That gives us an underlying picture of resilience. What is much more challenging is that we have had low growth in the economy for the best part of 16 or 17 years. This is not a story about any one Government, by the way; it is a general story. It is not political, in that sense, but it is a story about UK growth. I have made a number of speeches recently on the subject and it is important, because it is the critical structural issue. As far as I will go into fiscal policy, which is not far, it is interesting to do a simple, naive counterfactual and imagine that we had had the same potential and actual growth rate for the last 17 years that we had for the preceding period, which was around 2.5% a year. We have been at around 1.5% for the last 17 years. If you add that 1% back in and imagine that we had had exactly the same fiscal policy for every Government and the same shocks of covid, Brexit, Ukraine and the Gulf, where would we be today? I think the public sector debt level would be about 10 percentage points lower. You would not have solved all the problems, but I would argue that it would be a very different debate.

That is very helpful. There has been a strong drive in the UK for deregulation of the banking sector to promote growth. When we meet financial institutions, they are keen to see that, and they constantly stress that we are a long way from 2008. Do you think that that emphasis on deregulation and growth poses a risk for financial stability? That is the tension that is sometimes put to us by other people: that the Bank is holding back growth because you do not want to see deregulation. Is that fair comment?

My overall view is that we will not get growth if we do not have financial stability. We can point to episodes in history that support that view. Financial stability and monetary stability are key underpinnings of growth. Does that mean that every one of our rules in financial regulation is perfectly formed? No. Would that ever be the case? No, because the world always moves on. Do we support a process of review? Yes. I do differ from some of the arguments that are put. Let us take the banking system, because it is important. The key to the banking system supporting the economy is to have a system that is earning returns that more than cover its cost of capital, which means that businesses can invest their returns to support lending into the economy. That is the best system we can have. After a long period of rebuilding capital in the banking system post the financial crisis, we are now in that situation: they earn more than their return on capital, as you can see in the market numbers and in the price-to-book ratios. It therefore seems to me that they have a choice as to whether they reinvest earnings into their own business or pay them back to shareholders. That is a choice that should be made in the market. We were not in that position before. The reason that I push back is that when you read some of the debates, it is as if we have a fixed amount of bank capital and must therefore lighten regulation to allow it to be spread around further, but it is not fixed at all. A healthy banking system will earn its returns, and they can be reinvested into the economy.

They will have heard that. Mr Kroszner, what does a Warsh-led Fed mean for financial stability globally and in the UK? Do you think that there will be any changes as a result?

Kevin Warsh and I went through our confirmation hearing together to join the Fed, and we sat next to each other for years during the global financial crisis. He has been battle-tested in financial stability, and he is very aware of and sensitive to those issues. No one can guarantee that there will not be a financial crisis, but he is aware of and very sensitive to those issues. There is a lot of change in financial regulation in the US and there is a new proposal related to the Basel III endgame, which I think is providing much more effective regulation. It is very important not to confuse effective regulation with deregulation. We want to make sure that regulations are effective and that they reach the financial stability goals, but also that banks can intermediate, that they can actually do lending and that they can do what they are supposed to do in the economy. I think that that is the key motivation for what is going at the Fed now, in thinking about financial regulation. From the macro perspective, Kevin will be very aware of those issues and will not want to do anything, in the time that he is at the Fed, that exposes it to a financial crisis. But of course sometimes events overtake what one wants. Certainly, when I was at the Fed I had studied a lot about the 1930s, as had Ben Bernanke; I do not think that any of us thought we would be in a ’30s-like crisis, but having studied that period was very helpful. I think that Kevin, as well as others who have been through that, will make sure that we are best prepared and will be able to respond, to try to minimise the damage if there is some sort of shock that causes a crisis.

Thank you. That is very helpful, from someone who is so close to Mr Warsh.

Let us look at some other risks that we are facing at the moment. We are in the middle of the third wave of extreme weather this year, so clearly climate change is front and centre for all of us in our thinking about how the country and the world are developing. Would you say that climate change is now a bigger risk that you had previously estimated, Governor? I think when you came to us before, you said that it was one of a number of risks, but that it had slightly receded as one of your primary risks. Has the recent focus on climate change changed your view on that at all?

Well, I was making a relative point, not a point about the absolute level of climate risk. Going back to the Chair’s first question, given what we have seen on frontier AI over the last few months, for instance, I have to make that point again. I am not a climate scientist myself, so I will not venture an opinion on what we can conclude from what we have seen in the last few weeks, but obviously Europe has had a very strong period of very hot weather and it reminds us that there are very clear risks from climate, yes.

Does the climate change that we are seeing have an impact on financial institutions in a way that we should be taking notice of?

The channels in which climate change affect financial stability have not changed; they remain very present. There is a risk to banks, to insurers and to core markets from climate change, and that has not changed. Part of our analysis is to gradually embed that into the mainstream of how we test financial stability.

Do we still have strong teams of staff at the Bank looking at those issues and understanding the analysis that you are talking about? Last time we spoke, I think the Governor accepted that there were now fewer staff working on those issues and that that might indicate a lesser emphasis on them, which was perhaps a bit concerning.

We have to allocate our resources to reflect the relative risks. For instance, AI has come up the league table very rapidly.

On AI, what assessment has the Bank done on the needs around water, particularly internationally? In countries like Malaysia, it is such a big part of the economy now, but the physical needs around water, if not handled correctly, could have an impact on economic stability. What assessment has the Bank done on that?

I am not sure that we are the experts on that particularly, to be honest.

There is one aspect—

He may be!

It’s good to have an expert in your organisation.

I am not an expert either, but there is one aspect, and it is related to AI, in which the availability of water supply affects developments. We have all heard about the very large increases in the infrastructure of data centres that are being planned to develop AI capabilities. Those large data centres will require water supply to cool down the production line. The needs for water are large, and one of the questions is whether the water infrastructure will be sufficient to meet the needs for those developments at the pace that is currently embedded in the expectations of the markets.

What assessment has the Bank of England done on the demand on the grid from AI in the UK, and on what impact that might have on financial stability? If certain infrastructure projects do not go ahead, but AI is demanding that—

Yes. Can I make two points? It comes back to this question of growth. I made this point about lower growth over the last 17 years or so, and I subscribe to the view that technology is part of this, because I think, in the long run, technology is one of the drivers of growth. Globally, we have probably been somewhat between technology cycles; we had the internet/ICT cycle, and I think the next cycle is likely to be some combination of AI and robotics. I therefore think the question about grid capacity and data centre capacity is relevant here in terms of supporting growth. The other point I would make—I have to be honest, not being an expert in this field, I had not appreciated this until going around the country and talking to people who are closer to it—is that there is also a question of what is called access to clean power supply. I am using “clean” in a slightly different context from the environmental point here; it is actually about the quality of the power supply. In other words, it is about having a very high-quality and very continuous power supply. Some people have made the point to me that, on their own, renewables do not currently give that clean power supply. They can be a big part of it, but they do not completely solve that problem. You need a baseload of other power to make sure that you have that clean power supply. That is true for data centres, so there does need to be thinking about how we ensure that we have that mix of power supply.

Turning to the decision to ease the leverage ratio by 0.2 percentage points, I want to try to understand the range of views that existed on that, because it appears to have been a split decision. The quote is: “Most judged that the benefits of the proposed changes outweighed the costs.” It would be useful to understand that range of views and how meaningful that range is in terms of decisions. Let me start with you, Governor.

I am happy to answer that. Just to put it into context, the announcements we have made on the leverage ratio are, in my view, really about cleaning up some of the issues around that leverage ratio. There was a particular issue where we had made some changes back in 2019 that I think had created an anomaly, which was particularly affecting the large domestic banks. That is important, because they are the people who make the highest proportion of loans into the UK economy, and it had left them facing a higher leverage requirement than was, I think, justified. So cleaning that point up was important. We have also increased the buffer usability in the leverage ratio, which is important because it is exactly consistent with how we operate the risk-weighted ratio. One thing I would say is that, of course, the risk-weighted regime and the leverage regime are not additive; they are parallel. That is important. We made some changes to the risk-weighted regime, which we talked about in December, and these are changes in the parallel regime. For me, those were all sensible things to do. The challenge we have—and this is why there are some quite important words in the announcement, which I will come to—is that we have an issue with leverage in another context, which is leverage in financial markets. I think we have talked about this in previous hearings, particularly in the context of Government debt markets. In the period since the last hearing, and in recent months, we have seen a growth in leverage in equity markets, so hedge fund leverage in equity markets and leverage in exchange-traded funds. There is further growth in that world. It is all leverage, but it is a different context for leverage. Those two things are connected, in the sense that the banking system is the ultimate provider of liquidity and therefore there are exposures between these parts of the system. The question that we rightly debated at length in the committee was: if we think we need to do something in that market leverage regime, is it better to do it in the banking world or in the market world? Is it better to tackle this market leverage question in the market regulation world, if you like, or in the banking regulation world? My personal view is that it is better to do it in the market world, but we have to land that, because we have put out some ideas and got mixed responses, which I am not surprised about. The final thing I will say is that you will see that we have included some language that actually means we are going to do a bit of further work before we finalise these changes, which is exactly to this point. It is particularly to say, “Do we think these changes might actually increase the market leverage risk?” Frankly, the additional question—it is a question we already have—will be, if so, what are we going to do about it in the market leverage world, or do we do something in the banking leverage world? I think they are sensible changes to make in their own rights, but we have got this outstanding question of what we do about market leverage.

Who can help me understand a slightly complementary view to that among the witnesses this morning?

The aim of the change to the leverage-ratio framework was to fix this idiosyncrasy that was the result of an increase in the counter-cyclical buffer in 2019. There was an offset in the risk-weighted stack of capital. There was an increase in the counter-cyclical leverage buffer. There was not an offset in the leverage stack. That was a discrepancy, and that disproportionately affected the domestic systemically important banks. There was unanimity on the committee that that was something to fix, and that is what has been achieved by this package. There is also a broad view that we want to be getting the overall balance of the risk-weighted capital requirements and the leveraged capital requirements correct. The other view is that, in addition to fixing the discrepancy for the D-SIBs, which makes a lot of sense and will support lending into the UK economy, there is also a slight reduction in leverage ratio requirements for some other institutions, and that could potentially lead to further provision of leverage into markets. That is the question about cost and benefit. To the Governor’s point, I agree that there are mechanisms within the markets, in particular haircuts or margining on repo, that would be very effective for providing stability in markets that can be in some ways arbitrarily levered if you have no haircut on a repo transaction. The leverage ratio will not fix that if you have essentially no haircuts on a repo; that is arbitrary leverage. That was the balance of the view.

We had an excellent discussion that I think got us to a very good outcome. Exactly as the Governor and Stephen have said, the key motivation was to make the leverage ratio more effective and make sure that it is doing what we want it to do, which is to prevent excess leverage in the banking system. I also believe that this is not the right tool to try to deal with market leverage. Things like haircuts or potentially centrally cleared transactions are the best way to try to deal with that. What we were doing was trying to make sure that we had a lot of things that could be released in a crisis. Getting back to the previous question, that is a very effective way to try to make sure that banks can continue to intermediate during a crisis and provide both support in the crisis and more rapid recovery. A 20-basis point change in leverage ratio does not seem to be material to me in changing the ability of banks to provide leverage either within the domestic economy or more broadly. We have much better-targeted instruments. Making much more of the capital stack releasable is very helpful not only in achieving financial stability, which is our main objective, but in providing support for a secondary objective on credit provision and growth, because that is something that will allow the banks potentially to reduce the so-called management buffer that they hold above and beyond the required minima. Of course, they are always going to hold something beyond that—they should not be running right at the edge—but, when there is a lack of predictability and a lack of clarity about what is releasable, how much it has to be rebuilt and how that is going to operate, it leads them to hold a bigger buffer against that. I think the approach that we have taken will help to reduce that, making the leverage ratio more effective without in any way, from my perspective, affecting financial stability or addressing the market risk issues, which we will do through separate means.

Can I come on to the decision about lowering the capital requirement from 14% of risk-weighted assets to 13%? I think there are two broad sets of views. There are those who ask where the benefit goes—to the banks and their shareholders or to the economy as a whole. There is also still significant dissent among many banks over what the actual net effect of this is, because there are so many caveats that the notional 13% is never met because of all the different elements that go into it. People then say that, if you look at the bolder measures taken in the US, we are not really getting there. There is still quite a margin between where we are as a consequence of this change and what is actually happening in other environments. How do you respond to those observations?

I would disagree that the US has taken bolder measures; I do not think it has, actually. The latest package that it has announced is to implement Basel 3.1, which obviously we are doing as well. We have published a number of comparisons. In the latest numbers, we compared leverage ratios, and the leverage ratio is actually higher for international banks. I would disagree that the US has done bolder things, and I would say the same about the euro area—I do not think it has done anything bolder than we have. Let me make two points. First, just to reinforce what Randy said, we know that banks maintain what they call management buffers above our requirements. Randy rightly said that there are reasons for doing that, related to how they run their businesses, and that is understandable. However, what we have done will enable them to have more confidence to reduce those buffers. Certainly, that is our aim. Secondly—we included a diagram in the publication that we made—we would like to see the buffer structure simplified. If we are talking about bold things to do, it is too complicated at the moment. We would like, frankly, a single releasable buffer. The reason I think releasability is important is that banks always worry that, if we do not release buffers and if we say to them, “Look, it is okay to go into your buffers—we won’t get upset,” they always come back and say, “Well, will you?” When conditions change, if you are prepared to take the buffer away, while accepting that there will be a rebuilding process further down the line, that is a stronger message. We would like to see that, and we would like to see it simplified so that there is only one buffer. To do that, we need international agreement. Buffers are a product of the international Basel structure. Sam Woods, when he was with us, made speeches on this subject. We have been leaders in advocating that, and we will keep working to get an international consensus to tackle the buffers. We are not quite there at the moment, but I think that is the bold thing that should be done. We are strong supporters of doing it.

When banks look at their effective situation, they tell us that they never get to that amount because of all those add-ins and overlays that occur. Although I understand the aspiration to simplify and have a single buffer, the reality is that, because of their hesitancy about the implications of not doing this in the right way, they do not, and the net effect is that they end up holding more than they feel they ought to. That is a persistent narrative that they bring to us, and members of this Committee have had that drummed into them at private meetings recently. You present a very smooth and clear version of where you would like to go, but the bottom line is that, operationally, it is not working like that for the banks right now.

Let us take a moment to pick that argument apart. Why do they say that? There are at least three reasons that I can think of. We have already covered one: that they maintain so-called management buffers above what we require. That is their choice, but I recognise that there are things we can do to simplify the structure that might assist them to maintain smaller management buffers—they will always have them, but they could have smaller ones, as Randy touched on. We have made steps in that direction, but as I just said, the big step in that direction would be to move to the single buffer, if we can get international agreement. Secondly, capital requirements will differ from bank to bank, partly because of the business they do, and sometimes, I am afraid, partly because they fail to demonstrate to us the characteristics of running the business that we expect. We do sometimes put add-ons in—that is more in the PRA’s world—

Can you give an example of that?

Yes; if they show that they cannot manage risk properly, our supervisors can put those on.

And that is active—it happens?

Yes, it does.

So we should go back to the banks on that? Nathanaël Benjamin indicated assent.

Yes. That is us doing our job—I am sorry, but they should get their act together.

I look forward to having them.

Mr Benjamin was nodding; I wonder what Mr Kroszner thinks.

It is about good risk management. If they are not doing good risk management, they are not acting appropriately.

Absolutely. They can also get higher capital requirements bank by bank because, for instance, they have greater concentrations of risk in particular activities. That is a reflection of the risk they are taking. Thirdly, and fundamentally, many of them still think that 13% is too high. Coming back to the points I made earlier in response to the Chair’s question, we will have a strong banking system, and they will earn strong returns, if they are well-capitalised, lending into a strong economy and earning the returns. The evidence says that they are in a strong position. There are a lot of complaints about how much they are earning, but as I have said, they have to earn their cost of capital. They are doing that—actually, they are doing somewhat more than that—which is a sign of their resilience. The capital system does support that resilience. They do not like it, and they will never say that to you—the odds of us getting them to say that are such that it is never going to happen—but it is a fact. They are supported by that. If they have weak capital, they will not be supported in that sense.

To them, it feels as though there will be a move forward in one area, but then, for example, Sam Woods did not include the exclusion of unencumbered gilts when looking at the leverage ratio denominator.

But that is not in Basel. Nobody does that—the US does not do that. I know they would like to exclude gilts from the leverage ratio, but that is not in Basel, it would be contrary to Basel and nobody does it.

Mr Kroszner, you were getting animated. Did you want to come in?

This is one of the reasons why we have stress tests, to make sure that they have appropriate risk management. Most banks pass the stress test now. It does not risk setting a barrier or hurdle that is so high it is unattainable and no one can get an A in the class. Most of the banks are now getting an A in that, so they know how to achieve it. It is completely appropriate that, if they are taking undue risks, they should have add-ons and that, if they have processes in place, sufficient capital and appropriate risk management, they are free from that.

So the fact that in reality, major UK institutions have an effective requirement that is closer to 14.25% than to 13% is just a function of them not operationalising risk management appropriately, is it?

That will come down with Basel 3.1. Part of this is that Basel 3.1 will reduce that.

So it is a work in progress.

When we went from 14% to 13%, we were marrying top down and bottom up—the FPC does top down, and the PRA does bottom up. The bottom-up element is coming down, consistent with 14% to 13%. A good part of that is Basel 3.1, which is why the FPC felt it was sensible to match the top down.

Mr Kroszner, can I ask you about what the banks have told us about this apparent gap with the US, which the Governor pushed back on? Some of them have said to us that UK banks now face a 190 basis point capital premium compared with their US competitors. You can dispute the quantum of that, but do you recognise anything like that or any meaningful difference between the US and UK?

There are many different ways to calculate those sorts of comparisons—

And that is how you get out of jail on this subject.

Accounting is very different for what are considered bank assets, so it becomes a challenge to do the international comparisons. We have been trying to undertake a review. What we put into our reports looks at international comparisons, and the UK is a major global financial centre and it has to be competitive. We very much take that into account. I am very cognisant of that and bring it into the discussions, but when we get into the specifics of whether there are 190 basis points or minus 10 basis points, that becomes a bit more of an art than a science.

Do you think we will ever get to a point where it will be apples and apples, you can make a reasonable comparison and we will stop receiving all these representations from the banks?

Nat may want to come in. In December, we published the most detailed comparison. Nat looked at the feedback we have had, which is where they make these points to us.

That is right. There has been a range of feedback coming from different sides. The package that we have now set out brings UK capital requirements into line with requirements in other jurisdictions in a way that is consistent with the analysis we did and with the feedback that we received. The other point to note is that I used to supervise US banks. They often hold internal management buffers further above their regulatory requirement than some of the UK banks, yet that has no detriment to their earnings or their profitability.

We also hear about the international comparisons in the US. The US banks always feel that they are being unfairly gold-plated. It is a natural thing for domestic banks to go to their domestic regulator and say, “Look at the others. They get away with this and we don’t.” I think you would hear a very similar thing if you were sitting in the US Congress.

It sounds like managing teenagers, but perhaps that is a bit flippant.

Can I ask one more question on the net effect on the banks and their shareholders, and the economy as a whole? The narrative they bring to us is that they can get access to more money to lend to grow the economy, yet we see banks doing quite well in these circumstances. Do you have any views on how we should review the performance of banks financially and the impact on the economy, and what they do with the freedom they have?

I come back to what I said at the beginning. I push back at what I sometimes call the lump of capital argument, which is the idea that they have a fixed amount of capital, and if only we would lighten the regulatory regimes, they could lend more, support the economy and everything would be good. There is not a lump of capital. They earn their earnings, and they have a decision to take about how much of their earnings they distribute to their shareholders and how much they retain to grow the business. Obviously, that is a process that they have to go through with their shareholders. I recognise that we went through a period, up until only about two or three years ago—after the financial crisis—in which they were not earning their cost of capital. It was very hard to go to the shareholders and say, “You should leave this money in the business because it is a good thing,” because the shareholders would say, “Well, it is actually not a good thing because you are not earning the cost of capital. I can take it somewhere else and find someone who is.” But that is not true any more; they are more than earning their cost of capital. The best illustration of that is the price to book ratios, which are now on average about 1.7. They were below 1, which is the point at which they are at cost of capital. They are in that position. They are resilient, which is why they pass stress tests. They can make choices now, and the choice they can make is to support lending into the economy with those earnings. To my mind, that is what they should be doing. A resilient system will support that, and it will support the economy.

I am just going to bring in Mr Blyth, although we need to wrap this section up. We could spend forever discussing capital and banks.

We published the feedback from a very broad consultation about this matter, not only from banks but from academics, rating agencies, bank analysts and bank investors. That is a very rich picture, and it is clear that banks will, all things being equal, prefer to have less capital requirements—any investor or risk taker would like that. The arguments we heard from academics and bank analysts in particular is that their capital should be higher and that bringing it down is not the right thing to do. I think this is quite a balanced argument, with rating agencies and bank investors saying, “We look carefully at the capital people are holding and we want to make sure that’s appropriate in order to invest in these banks.” I do not think their view is that capital levels should be going down.

That is very helpful. Chris Coghlan has a brief question.

Governor, you sound slightly sceptical about the banks’ position on this. Of course, prior to 2008, banks were making the exact same argument—“You just need to deregulate and lower capital requirements. It’ll be fine; the economy will go gangbusters”—and in the process they blew up the global economy. To what extent does that history inform your thinking now?

It does a lot, actually. I am realistic enough to know that the memories of the financial crisis are now disappearing into rear-view mirrors. Some of us—all of us, probably—have unfortunately undying memories of it, having had to deal with it, but they are disappearing. You have seen this in history before. You get these cycles, and I think we are seeing it, where people are now saying, “The response to the financial crisis was too great and too strong.” It did not feel like that at the time, I have to tell you, when we were dealing with a system that was extremely weak and having to rebuild it. I get people saying, “But you’ve solved that problem. You should move on.” Yes, in the sense that I think the system is now really resilient—it is rebuilt. But as for this idea that we therefore relax the standards because it is all over—no, that is not the way it works. I’m philosophical; this is a debate we have. I come back to this point: I am not going to defend all our rules as perfect; they’re not. But the central point is that we will not have a strong economy without financial stability.

Thank you very much indeed. I am going to move on, because we could go on forever discussing this, but we do have other topics. On the subject of AI, which we parked at the beginning, Governor, I am going to bring in Bobby Dean.

Governor, I would like to return to something you made a point of highlighting in your opening remarks, which is about the heightened risks around frontier AI. I think critics will say, “I told you so.” When the discussion started about AI a few years ago, it was kind of levelled down as another risk to manage, and now you are saying that it is growing much faster than you anticipated. What is it that you did not foresee previously?

It’s not just us. If you talk to people who run Anthropic or OpenAI, they will tell you that the advance that they have seen in these models in the course of the last months has been much faster than they expected. They draw what are called exponential curves, and they are saying we have now moved into the vertical part of the exponential curve, in terms of the advances of these models, so they, too, were taken by surprised. If you take Mythos, which was the Anthropic model that really kicked this off back in April, they were surprised—I have talked to Dario Amodei about this—by the advances that model made. The reason is that these models are self-learning—they do this recursive learning process. It is the speed and effectiveness, and the ability of these models, that have moved dramatically forward. When they are subjected to the sorts of tests that the AI Security Institute in the UK—which, by the way, is absolutely world leading on this—subjects them to when it tests the models, to say how effectively and quickly they are solving real-world problems that they are given, what they can do and the speed at which they can do it have moved on dramatically.

I think we can all accept that the reality has moved on much more quickly, but we had an inquiry into AI in financial services, and we heard evidence about the potential advances in AI, and how it was felt that the financial services sector was perhaps under-prepared for the potential movements. The fact that it has happened more quickly than anticipated is one thing, but one of the things we heard from lots of regulators at the time was that, actually, our existing risk frameworks are fine, and we can be technology-agnostic. Do you still hold that view?

There are two things here, and they are both related to AI but they need to be kept separate. On frontier, we are talking about cyber specifically. To use a sports analogy for a moment, frontier AI will strengthen the defence and the offence. In defence, if you are us—the banks and firms—you can put frontier AI to work to strengthen things, because it will find the flaws that have been buried in there, sometimes for a long time, and you can fix them. Unfortunately, if the offence gets hold of it, they will do the same thing to attack you. Obviously, the whole thing at the moment is about whether we can make sure that the defence gets strengthened effectively before the offence. That is getting much more challenging. The second thing that I think you are referring to is equally important: how banks use AI. We are not talking about cyber; we are talking about how they use it and the risks that come with that. That is equally important, and your report was very helpful and very good in this respect. We are very actively engaged on that subject too. We are all using AI a lot more—that is a truism—but banks are using it a lot more, too. Where I think it is still more talk than actuality is things like agentic trading systems—setting up your own agents—because there are some very big questions still to be answered about how that world would actually work. For instance, I am not a lawyer, but something strikes me about having an agentic trading system as a bank. To go back to frontier AI for a moment, frontier AI demonstrates that the recursive learning capacity of these systems means that they actually develop their own abilities. As my colleague Sarah Breeden has said—we don’t invent this; we get this from the firms who are developing it—they learn to cheat and they learn to lie, I am sad to say. Where do they learn that from? It is partly inherent in them and partly because they are trained on the data of past experiences. And they cover things up. If you have established an agent, you have to be very clear where the legal liability rests in all this. I think there are things that will need to be resolved before that world really develops.

Mr Benjamin, we might want to continue the Governor’s world cup analogy about attack versus defence. What access do firms currently have to frontier models? They are presumably on the defensive side of the equation; how is their defence?

The financial sector generally has been putting a lot of focus on these things. At the moment, firms are in the process of testing those types of capabilities on their own systems. The big risk that comes with frontier AI development is the speed at which weaknesses in IT estates can be identified, and what that therefore means for patching and the speed at which software upgrades need to be made. That is the work that is happening. Andrew mentioned earlier things that give the edge to the attackers versus things that give the edge to the defenders. One of the key aspects is the speed at which those capabilities become accessible to the attackers. They are typically accessible to the defenders before that, which buys them time to identify the weaknesses and vulnerabilities and to prepare for much faster patching. Another risk is that the moments when software is upgraded are also the moments when there is a greater risk of IT disruptions internally, not because of an attack but because of the effect of the upgrade itself. That work is happening within the banks and within large firms at the moment, to get them prepared for this new state of the world.

We have been talking about risks in the abstract; could you describe the impact on a standard financial firm or institution if, say, the attackers got there first?

There are two ways in which cyber tends to manifest itself: one is those who want to disrupt systems, and then there is often a blackmail, effectively; and the other is stealing information. Those are the two things that can go on. As Nat was saying, what is being found is that flaws that have existed in very core pieces of software—by the way, financial systems are just a subset of this, obviously; it goes across the whole shooting match, as it were—that have often been there for a very long time can be discovered by these systems, and can be discovered quite quickly. There are people who know much more about this than I do. To go back to the football analogy for a moment, it currently requires a lot for an offensive party to set this stuff up to do this. You don’t go home, get Mythos, plug it in and off you go. It is about rather more than that. We have a second issue. They say that we are now on this vertical exponential curve. These models have a shelf life of only about two or three months and they are then replaced by something even more powerful. If you talk to the developing firms, they say that there will be open-source versions of these things coming on to access, which will mean that it is much harder to restrict access to them. And let’s be honest: where are other regimes in terms of developing their own? Obviously, China is the big case there. They may be some months behind, but they are not that far behind.

I am not a lawyer, so it would be good to—and we are slightly doing this—get the views of the legal world on this, but here is my understanding. Let us go back to basic principal-agent law. If I am the principal and you are my agent, I think that in the traditional sense, I give you a remit to act on my behalf. I am liable for what you do when you dispense my remit. If you go outside my remit and start doing other things, you are liable. That works, of course, when you and I are both legal persona—in other words, you could be held legally liable and so could I. But obviously with an agentic AI model, that does not work, because it is not a legal persona. It is normally the case that if the agent does something outside the remit, the principal is liable where the agent is not a persona—I think. This hasn’t been tested much in the AI world, but let’s say that if that is the case, I think the principals—the banks—will therefore have to think very carefully about how they manage the use of agentic AI. If they are going to be responsible for everything it does, and if there is this issue that it can go off piste, they have to think that through before they start using this stuff.

It opens a whole can of worms.

Those are the limits of my knowledge, I have to be honest with you.

There are third parties involved in this as well—the AI companies and some of the major tech companies. Perhaps we can move on to a slightly separate point about the critical third party regime. Four will be brought into the regime. A concern has been raised about how much oversight you are able to have over these firms when they are quite opaque about the technologies they hold. Mr Benjamin, in your capacity, what is your relationship with these providers? How much do you get to look inside their black boxes to understand what the real risks are?

The regime is coming into place now, and our colleagues on the PRA will be the ones supervising them. So that is going to begin, and we needed the regime to be in place first.

Are there general concerns about working with these providers and how much information they are willing to share about their systems in order to be able to understand them?

If you are referring to the four that have been nominated, no, I wouldn’t say there are concerns at all. We have had relationships and discussions with them just for broader reasons in the past. They are very open to those types of discussion, so I am not worried about that.

I guess my point is more about regulatory capacity. I am sure that they will appear to be very open and willing to co-operate, but does the Bank have the staff, the knowledge and the know-how to be able to properly hold all their systems to account?

Our colleagues in the PRA have entire divisions of people who are specialists in operation resilience and well versed in these technical topics.

I will move on to a slightly different threat that is still related to AI. Mr Blyth, there are growing concerns about the threat of an AI bubble and the impacts that that could have. The easy comparison is that of the dotcom bubble. Are the risks similar or worse?

It is a good comparison in terms of the level of concentration and heightened valuations, especially in equity markets. There are a number of elements here. First is that the continued increase in equity valuations, especially in AI stocks, has been coincident with very strong earnings. However, the current valuations are predicated on a very optimistic future earnings path. It is not clear that it is necessarily a bubble. However, in our and the Bank’s estimation, it is based on quite an optimistic path going forward. If earnings are not realised along that optimistic path, then there could be a significant repricing of AI equities.

On that point, J. P. Morgan said in a report that something like $650 billion-worth of revenue per year would be needed to get a reasonable return. It also remarked on some of the physical limitations in ever being able to achieve that kind of revenue because of water supply, data centres and other physical infrastructure. Is that something that you recognise? Do you think that, as things stand, the returns required are physically impossible to get at?

I would not say that they are physically impossible, but we are very cognisant of the risk. Power and cooling, along with planning and being able to get approval to build data centres, are all risks to a positive earnings story. On the question of financial stability, there are two elements on the radar that are particularly concerning. One is that the funding for a lot of AI expansion is now going forward in debt rather than equity form or from cash flow. You can see from some of the data that the growth in 2026 has been extreme. Obviously, if there is a popping of the bubble, then that debt can have a more impactful downside. Secondly, there is a lot more leverage on the long side in equities. That is both institutional and retail, and both of those can be concerning. Institutionally, a lot of levered hedge funds have been taking long positions in equities, often funded either in derivative form or by prime brokerage. Then, on the retail side, there are a lot of levered ETFs. That means that if there is a fall in equity prices, that could lead to a negative spiral. Market volatility is not per se a financial stability concern. However, when it can be amplified rather than contained, then it becomes one. I think that those risks are growing on the AI side.

It seems that the change towards increasingly leveraged positions has happened very quickly over the last year or so. For example, we got information about Alphabet’s position. Does that make you feel that a potential bursting of a bubble is much closer than it was this time last year?

Again, it is hard to predict. I think that the impact of a bubble bursting would be more significant given that leverage. As you point out, the pace of change regarding leverage is remarkable. The data on the prime brokerage increase is remarkable. The pace of the growth of levered ETFs is also remarkable, although it is not as big as in prime brokerage. However, that pace, and the pace of increase in debt issuance, is extraordinary. There are many things that we will be thinking about. Is this crowding out other debt issuance because there is so much debt issuance coming in AI? Is this now just increasing the leveraged long to the AI story across the financial system, and does that mean greater interlinkages when the bubble bursts? It is hard to say what the answer is regarding the probability of a valuation reset.

Governor, can I ask you about the UK’s specific position in relation to this? Obviously, we do not have so many developed AI firms, but we will still be exposed to any impact. There are two sides to the coin. One is the risks of that bubble bursting and what that means for the UK. However, in the meantime, what is it doing to our growth prospects if so much capital is going in one direction?

As you say, even if there was a bursting of a bubble that is primarily not located in the UK, the market effect is now so large that it clearly would wash over. I think that is right. We ran a stress exercise that looked just at this question, and it did reduce UK growth. On exactly how the effects would run through, you would expect some adjustments in UK macroeconomic policy, which is another way of saying that you might expect some adjustment in interest rates to reflect that. You would have to decide how you would play it through, but yes, it would wash over On the UK growth side, let me start by adding two points to what Stephen was saying because there are two other issues on how this could go wrong. One is that history would tell us that these big general purpose technology innovations take longer to come through into productivity and growth than people think. This one may be a lot quicker because the pace of change in AI has clearly accelerated a lot, but if you go back to something like Edison inventing electricity, it took a lot longer. If you go back to ICT, Bob Solow famously commented about seeing computers everywhere except in the productivity series. That is one way of saying that it takes time for these things to come through. The second thing is that everybody is currently priced to win. Everybody does not win in history. If you can remember the first company that actually got the biggest share of internet browsing, it was not Google; it was Netscape.

That dates us all, Governor—those of us who remember that.

Sorry! In terms of the UK, there is a highly globalised market for these things. The UK does have real advantages in this area. There are real centres of expertise, we have a university system that is a very strong producer of research and output, and we have a very encouraging state of affairs on tech start-ups at the moment—Patrick Vallance was saying to me the other day that he thinks we have the most ever in this country. The persistent question is about the next stage and how you go from start-up to scale-up. That remains a big issue in this country.

Mr Bailey, I want to go back to your comments about leverage in the gilt markets and the issues there. Earlier this month, the Financial Times reported that new reforms would be announced this month about the management of the gilt markets and in particular the use of haircuts. Are there still plans to announce those reforms this month? What is the timeline for those?

We put out, as we do, a discussion document. We have obviously done a lot of analysis; we ran a system-wide exploratory scenario. From that, we concluded that rather than classic regulation in the financial institutional sense, there were areas where market regulation, particularly in areas such as clearing for gilt repo, would be sensible. It is fair to say that we got a mixed bag of responses. We were not surprised because obviously there is a whole set of different interests at play in this. I was not at all surprised by that. We are currently working out what to make of that and we will come out with proposals.

Part of the responses we had was that clearing of repo is happening in the US. A natural question from us was, “It is happening in the US; is there any reason why it cannot happen in the UK?” That is the question we posed. The feedback was that there were differences in the structure of the UK sovereign debts market that were not present in the US. We are carefully reviewing and reflecting on that feedback. The option of doing nothing on that front is not really an option. It is important. Markets have been much more resilient than they used to be; that is a good thing. We have gone through a lot of shocks, so it is a very positive thing. But that resilience has to be locked into the structure of those markets.

A big point, which does not necessarily sit easily, is that we have seen a fundamental change in the market structure in recent years. By the way, that is the case in other Government debt markets as well—the UK Government is not alone in this respect. It has enabled, I would argue, a lot more Government debt to be issued—let’s be honest. However, that has created a more leveraged and, arguably, fragile market structure as a result. We have got to come to terms with all aspects of that, frankly, because there are some difficult issues in there.

Mr Benjamin, what is the timeline for those future reforms?

We are in the process of evaluating and reflecting on the feedback we received. We will provide an update on our thinking as part of our Q4 financial stability report. That is the next step.

Will that be an update on the thinking rather than an announcement of new measures?

Yes, these things will take time.

Okay. The discussion paper you mentioned earlier, “Enhancing the resilience of the gilt repo market”, published in September 2025, was launched in collaboration with other agencies—such as the DMO, FCA and Treasury—as well as with market participants, because, of course, management of the gilt market falls between those different agencies. Do you plan, in setting out those reforms, to continue working in that collaborative manner? How will you consult those agencies?

We are working together with those colleagues on an ongoing basis. That is not a new thing. Those are issues that we discuss very collaboratively with our colleagues in other agencies.

Will it be a combined decision of those agencies, or will it be a Bank decision?

Ultimately, who has the power to implement specific decisions will depend on the nature of the measures, but those are things that we discuss together.

I hope we can come to a consensus on it.

I want to go to the most controversial of the measures, which is the haircut on gilt repo. Mr Blyth, you mentioned that earlier. Am I right in thinking that you are on the pro side of such a measure?

I think I mentioned in my appointment hearing that, when I was trading a lot of repo and a lot on the buy side, all things being equal, I would rather have no haircut and be able to get unlimited leverage. [Laughter.]

At least he’s honest.

Poacher turned gamekeeper.

I think I would speak for every hedge fund manager. Even a small haircut might increase my costs marginally. However, from a financial stability point of view and in the round, it is very important that we look at having some minimum haircut on gilt repo. It seems to me entirely appropriate. I do not think that would impact the functioning of the market or its ability to be resilient to shocks. Yes, I would be on that side—appropriately calibrated, of course.

What would you say to your former colleagues who responded to your discussion paper by saying that this might backfire because it may raise funding costs and reduce gilt market liquidity?

I think that is a very effective lobbying argument. I would obviously have made the same argument when I was in that seat.

Obviously an inspired appointment.

I do not think we should allow unlimited leverage in the gilt market in order to maintain a certain level of leveraged hedge fund return. If one has slightly higher capital, that might mean hedge fund returns are slightly lower. That does not mean funding costs would go up or that the market would be less liquid. It would just mean leveraged hedge fund returns are slightly lower, and that is probably the right place for the cost to be in the system.

Is the big concern here that there is a lack of quantitative evidence on either side of this argument, as outlined in the discussion paper?

Certainly, as the Governor said, the market structure for gilts has shifted, particularly as defined benefit pension funds have become less of a demand, and also as real-money asset managers and insurance companies and so on have bought their duration in gilt markets through futures. We have a particular structure, not just in the UK market but in the US one, where buyers of cash gilts against the futures are hedge funds, and that is financed by the repo market. So we have a particular case here, but I do not think there is evidence that that would be disrupted in any way by having an appropriately calibrated repo haircut.

On quantitative evidence, we have very deliberately started publishing much more aggregate fact on exposures and positions in core UK markets, so that market players can make better use of that information for their own risk management and be better prepared for shocks when they hit. That complements the system-wide stress test that we did. On your point about the cost of the minimum haircut, part of the feedback we receive—it is a valid point—is that the eventual costs to participants are not based just on one repo being done; they are based on all the trades being done by those players in the round. The repo is usually only one part of a broader set of trades being done, together with some futures or derivatives, for example. It is important to look at the all-in cost, rather than just the cost of a minimum haircut.

On this point, Mr Bailey, the chief economist at Handelsbanken has written a note describing that the underlying problem here is the loss of a stable investor base in gilts, and that all the issues you now see with leverage is one manifestation of the underlying problem. Is there work being done in the bank or across Government to look at how to increase domestic or retail participation in the UK gilt market?

It is all about pension funds, but there are two parts to this. One is that, because of the profile of the population, we have seen what I call a maturing of the defined benefit pension funds in particular. In other words, their age profile is coming to a point where they are not going to be natural buyers of longer-dated Government debt. That is why the DMO has reduced the duration of its issuance: because that natural source of demand, which was a very good thing because it was sustaining longer maturities and longer duration in the UK Government debt market, is naturally falling away. That is a structural point, so you cannot somehow replace that; it is a function of the population to a great extent. That is the first point. The second point is one we discussed a number of times in the past, which is that—Mr Glen and I did a lot of work on this in the past—we also want to see a greater share of pension fund investment going into the real economy. There are again a number of big moving parts to this story. It has obviously resulted—the UK is not alone in this respect—in a greater share of purchases being done by more price-sensitive investors in the hedge fund world. I would also say that the UK has had a current account deficit for a long time, which means that there is a greater reliance overall—you cannot narrow it down to one market—on overseas investors in that sense.

There are two points. It underlines the points I made early on to the Chair’s question about UK politics. I said that I would not comment on UK politics, but I made the point that the fiscal framework and fiscal rules—this is again not a particular point about this Government; it has been true in previous cases—are an important underpinning of that. You might have seen that in the response of the gilt market to the last Budget, which was favourable, because there is a commitment to, in a sense, stabilising the debt ratio. That is the first point. The fiscal rules are important in that sense. The OBR, as you will probably know, has published its structural long-term fiscal outlook, which is very helpful but not entirely reassuring. It obviously points to a situation that will mean that there are more pressures. I think that we are going against at least three structural headwinds. Again, they are not unique to the UK by any means. One is, as I said, the ageing population, which does not get enough consideration in this respect. It will go on because it is baked in, to use an analogy. The second is another subject that is much in the news, which is the defence dividends. The post-Cold war defence dividend is over and has turned round. The third one, which we discussed earlier, is climate change. If we had one of those at any given moment, we would be saying, “We’ve got this big issue to deal with,” but we have three of them.

I am really glad that you asked that. I think that this is an important first step, but I do not think we have all of the data. We are making a lot of progress by doing something like the SWES. It really is world-leading. Other countries are now doing things like that to try to gather this kind of data. The positive is that the market participants are excited about this because they are also getting data that they did not have. That helps them to understand which trades might be very crowded and how correlation might come in during a crisis or during pressure. That is how we get co-operation with these SWES exercises. So we are not there, but I think this is a very important step in the right direction.

If I can add something, we have had very good collaborative engagement on a voluntary basis from all the major players in the private market ecosystem, many of which are headquartered overseas. Through that engagement, we have obtained what is quite a unique dataset on these exposures for the first time. That is very meaningful progress. Already, we have published in our financial stability report what we found, just as a result of that dataset, without yet applying the shock to it, which is very useful to have. So it is a first.

We have coverage of all the major players; everyone has said yes.

I have been really pleasantly impressed by the amount of engagement. To Randy’s point, the large majority of the capital providers and the asset managers are international, so this really has to be done on a voluntary basis. The fact that there is a lot of willingness from participants is actually a testament to the credibility of the previous SWES and the credibility of the Bank staff in carrying this out. I have spoken to a number of people in the US in particular who are pretty interested in what we are doing. The question of whether people understand the interlinkages between private credit and other markets is still an open question, but everybody in that market wants to understand those interlinkages. I think this exercise is doing a very good job of asking the question, “Are you sure?” I think some of the empirical evidence that we saw with some of the bankruptcies—First Brands and Tricolor—and other events that have resulted in, shall we say, unexpected losses at some of the banks, is sort of re-raising that question and suggesting that perhaps people do not understand those linkages as well as they could. It is not entirely straightforward. I do think this SWES is an important exercise in thinking about the interlinkages between, obviously, the ultimate capital providers, the asset managers, the loans they may get from banks, the ultimate corporates, their relationships with banks and so on. Mapping that out is very useful.

Let me put my other hat on for a moment and speak as chair of the global Financial Stability Board. It is interesting: when you look at the agenda of the FSB over the last five years or so, it has shifted substantially towards the non-bank world, into this world that we are now talking about. Much of our work is now in this world. To your point about data, I think one of the big issues that we still have to solve—and we have a working group in the FSB, which I chair, that is seeking to do this—is to get cross-border data sharing on this whole market-risk issue, because it is not as good as it needs to be. If you go back to the questions about hedge funds in the gilt market and the leverage, it tends to be the same big macro funds that are also leveraged in other government debt markets. You will therefore get the risk of knock-ons across this. If we go back to what we were discussing with Stephen about margining, last year some of the moments where we saw the biggest movements in the gilt market were when we were getting margining requests coming in other markets. When we are talking about initial margining—when we talk about zero margining—the problem with that system is that if you have no constant margining level, the margining from the risk models becomes more pro-cyclical. That is because when volatility does come into the market, margin is demanded. But that is pro-cyclical. We saw some signs last year, particularly during the tariff volatility, of knock-ons between markets going on. The international issue is very important in this respect.

When you talk about risk, do you think that could be a dysfunction in the gilt market completely arising from a sudden correction in equity markets? How far do you think that risk could go? Could it make it crystallise?

Obviously, we have had volatility in the gilt market this year mainly caused by events in the Gulf. The market has been resilient to it; we have had functioning markets throughout. We have had markets that have stood up and been resilient. It is important not to be too negative on that front. But obviously we have also had examples in the past where we have had volatility and one case where we had to intervene. We have developed tools to deal with that because we have to be ready for that.

Do you think you are ready for it?

We certainly have a lot more tools than we used to, and we will use them if we have to. I hope we do not, but we will if we have to.

Mr Benjamin, are you ready for it?

Yes.

That is great confidence. If you sit next to the boss, you have to say that.

There is one point I want to add to the question of this pivot from banks to non-banks—to market-based finance. That has very much been reflected in the type of stress tests that we run. We no longer just run stress tests on banks and insurers. We have started to run those market-wide stress tests, such as the first SWES—the private market SWES. That is because of that shift in the provision of finance across the financial sector. Despite that shift, there are sometimes stories that banks are no longer involved. That is not the case. Banks are still very involved because they are by far the major providers of leverage to the non-bank world. That is the case for private markets, and it is the case for hedge funds. They are still very involved.

The balance between continuing to focus on banks and looking at private markets is a very important one, because I would not want to hypercorrect. All this risk has not gone out of the banking system. There is risk in private markets. It is less easy to get data. We are doing more work on that, but maintaining focus on the banks, which are obviously the deposit holders, is still extremely important.

Ultimately, the risks always go back to the bank balance sheets. No matter where it is in the system, it is going to go back to the banks. That is why we need to understand the interconnections, but we cannot lose sight of the importance of the banks in the system.

Turning to household debt, we have obviously seen an increase in mortgage rates. What is that doing to household resilience?

As you said, there has been an increase by around 75 basis points in mortgage rates since the December FSR. That has affected a number of households. However, the majority of households had already been through some much higher rate hikes. By and large, the household sector in aggregate has been resilient, despite that rise that we saw. A reason for that is that the aggregate level of indebtedness of the household sector is much lower than it used to be. However, for some households, it is clear it has been very challenging, especially for lower income households.

You mentioned a proportion. Just remind us again, what proportion of households would you put in that latter category?

It depends how you reflect it. For example, you have households with a mortgage and households without a mortgage that are renting. The median income of households that are renting is half the median income of households with a mortgage, and those people who rent are most affected by the recent changes. That does not affect the resilience of the banking system in any way, but that does not mean it is not an issue.

It is definitely an issue for policymakers like us. It is a year since you raised the cap on the proportion of mortgages that banks can lend on higher income multiples. How has that gone?

It has worked as intended. We have not raised the aggregate cap. The aggregate cap is the same: no more than 15% for a loan-to-income ratio above 4.5. That aggregate cap has not changed; what has changed is that it is no longer applied at an individual lender level. As a result of the greater headroom that has been released, we have seen an increase in high LTI mortgages. We are now at about 13%, which is higher. That is the system working as intended.

And negative equity—are you more concerned now than you were a year ago?

We have not seen an aggregate rise in the risk of negative equity. The loan-to-value levels have not changed since what was the case before. That indicates that the new households that have a mortgage also have savings buffers to compensate for that, but there has not been an increase in the negative equity rate.

Can I ask about the gap between the 13% and the 15%? What accounts for that? There should be more latitude to do more borrowing, and they don’t.

Well, it has gone up. It used to be lower. It used to be 10%, and before that it was 8%, so it is going up.

I know that. Do you think it will get to 15%?

It might do at some stage, and there are mechanisms in place to bring things back down to 15%. The tool is set up based on the flow of new mortgages granted. That is a means to an end, which is to control the balance of highly indebted households in the stock of mortgages.

What proportion of that increase is first-time buyers?

In aggregate, at that higher level of more than 4.5 LTI, half those households are first-time buyers.

Given that the 15% is across all lenders, and only about six companies control 80% of mortgage lending, do the smaller, more innovative building societies that look to assist first-time buyers get a bit battered around by the changes in the borrowing rules of those bigger companies?

We have also had smaller firms apply for that possibility, so this also supports their ability to extend lending to higher LTI households.

I do not think we have had small building societies saying that they have been squeezed out. I have not picked that up.

What would you say are the risks facing the core banking sector? Is it the geopolitical tensions that we have talked about, the AI bubble, exposure to non-banks or anything else? What are the key things on the radar for that sector?

I think you have covered the big risks to financial stability that we face at the moment. As Stephen was saying, it is market valuations—particularly AI sector valuations—leverage in Government bond markets, growing leverage in equity markets and frontier AI. I would say that those are the big moving parts that we are looking at. Those are the key things that we are focused on with the banks.

The financial stability report talks about banks deploying significant risk transfers. That obviously means taking some of the risk off their balance sheet and putting it into other forms of assets—

Securitisation, yes.

Do you think that is a risk? You have not mentioned it, but it would appear to be an activity that might threaten stability.

I think that falls into the category of, “It’s not large, but it’s growing.” Let us go back to the financial crisis: we saw that happen during the financial crisis and frankly we got bitten a number of times in that era. The reason we got bitten was that we had things that were described as risk transfers that, when push came to shove and the crisis happened, turned out not to be, in the sense that the stuff was not actually off balance sheet; it came back. The PRA looks at every one of these risk transfers, when banks come to it wanting to do one, and it has to look at the question of whether it is robust and whether it is robustly transferred. There is some evidence that, frankly, there is some pushing at the margins of that going on at the moment; that is one reason we highlighted it in the report. Some of that pushing is allowed internationally, I have to say. We have probably held a somewhat firmer line on that front than has always been the case internationally. I will give you an example. When you are securitising and the risk is going off to another party but you still have the economic assets, has your relationship with the counterparty been adequately, in a sense, secured? We tend to ask, “Is it funded?” in the sense, “Do you have pre-positioned assets that will fund it or are you basically going on a promise from the counterparty?” We are seeing a bit of stretching going on, not so much here but elsewhere, and the second of those is growing. It falls into the category of not being the biggest thing, but something that we have to be, and are, very alert to because it is growing.

Do you think there is a big difference between the SRTs and the collateralised debt obligations that were a significant part of the financial crisis in 2008? No matter how good we may be in our approach to regulation, we all know that the CDO problem was most significant in the US jurisdiction and that brought us down as well. Are we covering that risk in the way in which we approach regulation, and are others doing so?

I would say there are a few differences. There are two differences in the fundamentals. First, the banks are stronger now than they used to be. Secondly, the underlying books of loans have been subject to stricter underwriting standards, on average, than was the case at that time. In terms of the structure, the difference from the CDOs that were being traded before is that a significant risk transfer is not something that is then traded; it is a relationship between the bank and the investor. That does not detract from the fact that, if it is unfunded in particular, it can create some counterparty risk to the bank. When we talk about lessons learned from the global financial crisis, banks’ ability to understand the degree to which they have concentrated counterparty exposure is a recurring theme. In the list of risks facing banks that Andrew went through, all of them are channelled through their exposures to the non-bank market-based finance so the importance of understanding counterparty concentration is an important lesson to remember from the crisis.

In our stress tests we do assume those things do not work properly.

Do you think that the banks know enough about what is in the areas that they are transferring risk to in order to have a really good sense of how much risk they are taking?

They have to know how much the capital could be, really—how much risk is actually transferred—for the PRA to actually approve that significant risk transfer. They are approved one by one by the PRA, so it is different.

The Governor talked about how some of the activity appears to be pushing the envelope a little. Is there anything you need to do, proactively, to rein that in a bit and make sure that we are not exposed to an additional risk?

That is happening. Our colleagues at the PRA have been writing to banks to flag the importance of how those are creating that counterparty risk, especially in the case of unfunded transactions.

It will not be long before you are hearing complaints, I am sure.

We have broad shoulders. It goes with the territory.

Please can we explore your attitude—or policy or mindset—to stablecoins? There has been a draft proposal to impose a temporary limit on the ownership, and you have the £40 billion issuance cap. I wonder whether it is viable for anyone to issue with that, so what sort of work has been done on that? Could you also characterise your approach vis-à-vis those of the EU and US—where are we different or similar?

The first thing I would say, to put it in a slightly broader context, is that, with payments, we want to see that we are getting the benefits of digital technology into the payments world, because there are real benefits in terms of smarter contracts, including in fraud and late payments. We are open to how that is done. It can be done in stablecoins. It can be done in tokenising bank deposits, which we are working on with the banks. We could introduce a central bank digital currency at the retail level. Stablecoins are therefore one option. It is important that we have rules that enable them to happen, and that is what we put out our proposals on. On the cap, in our first proposals, we said we would cap at the individual account level. We got a lot of pushback on that, frankly. Our point is that it is only a temporal transitional thing because we have to be careful—if there were to be a big shift of deposits out of the banking system—as to how we would manage the stability of that. We concluded that the more sensible way to cover that off was just an aggregate issuance cap, rather than a per-account cap, because it is much easier for the issuers to manage that. That is why we have come up with the £40 billion. I get a sense from the responses we have had so far that the people thinking about issuing think that is all doable, workable and quite sensible, so we have had a pretty positive reaction. We made an adjustment to the mix of backing assets. We have changed from 60:40, with 60% being high-quality remunerated and 40% being central Bank reserves that we would not remunerate, because our view is that they are not then lending on to create credit in the economy, so they are not performing that function, which is important, and there is a big difference with banks there. We felt, looking at it from the point of view of what liquidity these issuers need to have, that 70:30 was fine and doable. We have also provided for the fact that these issuers will have the ability to have an account at the Bank of England. There will be a repo facility if they need it. It is not a standing repo facility because, again, they are not banks. We felt with all of that in place that 70:30 was doable. Again, the response so far has actually been good. On the question of where we are versus the rest of the world, it is quite funny because we get slightly castigated for being behind, but we are actually ahead of the US. You legislated in Parliament sooner than the US did.

I know.

The second bit of US legislation—the so-called clarity Act—is stuck, as far as we can see. They did the GENIUS Act—brilliantly named—but the clarity Act is sort of stuck. In terms of our rules, some of the US agencies have got their proposals, and some have not, so we are not at all behind in that sense.

The amended proposal makes a lot of sense. The analysis that the Bank has done is that the £40 billion issuer cap is commercially viable. There are issues that we have talked about. There is an issue about the amount of short-end gilts that are available to serve as backing assets. That is the difference between the UK and the US. We are obviously talking with the DMO and others about that, but this is an appropriate change. It is about the pace of adoption of stablecoin. If that goes too quickly there can be disintermediation.

Mr Kroszner, please come in.

On the US, I just want to underscore what the Governor said. We passed the GENIUS Act last year, but there has continued to be an ongoing discussion about remuneration, whether there will be remuneration and in what form it will occur, so we really have not implemented it yet. One of the things that is quite different about the UK regime is that 30% has direct access to the Bank of England, which is a real differentiator because it allows the Bank to monitor on a tick-by-tick basis whether money is flowing out very quickly. We know there will be an at least 30% cushion in case a so-called run occurs. That will be much less likely to have a big disruption in the market, because of monitoring that. That is a very important distinction.

If I could turn to crypto as well, we are told that in the US there are emerging fears that community banks that issue loans to farmers are threatened by the flow of those funds to cryptoassets. Can you make an assessment of whether that could happen over here? What is the potential impact of the growth in cryptoassets in terms of the Bank’s monetary policy and financial policy being rendered less effective in those circumstances? How are you seeing this at the moment?

I would draw a distinction, which I think you are probably pointing to as well, between stablecoins and crypto.

Yes.

I think that is important. Stablecoins purport to be money, therefore, they need to meet the test of money. But as Nat was saying, access to us means that what goes with our holding them to the standard of money is that we also give them what they need from us to be money. The absolute fundamental point about money is that it has assured nominal value. If you are going to use it in payments, you and I know that if we are making a payment, my pound is equal to your pound; we do not need to worry about that. Crypto is not like that; crypto does not have assured value. That is the key difference, so it is a very different animal in that sense.

How do you respond to the challenges of having that destabilising effect, sort of by stealth, of higher volumes of crypto in the financial system?

I do not know precisely what the system is in US community banks. We follow the Basel rules, which are very strict; in fact, many people say they are too strict on the treatment of crypto. But I think it is fair to say that they act to limit banks’ holdings of crypto.

That is right. At the moment, the links between the world of crypto and the core of the financial system are not deep at all. We are monitoring that very closely, but the links are not deep.

I have heard of one or two cases in the US where crypto firms have got banking licences. That is not the case here.

I thank our witnesses very much indeed for their time. We have had an interesting discussion. We have had a long discussion about the capital rules and banks, and whether the balance is right. There has been a lot of discussion about the risks posed by AI and the pace at which that is moving, and the challenge for regulators and firms of keeping up with the risks involved. There is a lot of food for thought there. We also discussed the accumulation of investment in AI and whether this is affecting growth in other sectors, and what might happen if the bubble bursts, among other interesting topics today. It is always an interesting session. I thank our witnesses very much indeed: the Governor of the Bank of England, Andrew Bailey; Nathanaël Benjamin, who is the executive director for financial stability strategy and risk at the Bank; and the two external members of the Financial Policy Committee, Randall Kroszner and Stephen Blyth. The uncorrected transcript of this session will be available on the website in the next couple of days. Thank you to our colleagues at Hansard and thank you to our colleagues at Bow Tie for the broadcasting.