Welcome to the Treasury Committee on Wednesday 22 October 2025. We are here to discuss the appointment of Professor Stephen Blyth to the Financial Policy Committee as an external member. This is a confirmatory hearing with Professor Blyth, who submitted to the Committee his CV and questionnaire, which are about to be published. Welcome to the Committee, Professor Blyth, and thank you very much for your CV and questionnaire—that was very helpful. Before we go any further, I want to know whether you had any contact with the Bank about your questionnaire when you filled it in. Did you send it to them, or did they have any feedback?

They sent me the questionnaire, which I filled out and completed. I sent it to the Bank for comments, and they had a few minor comments that they said I could take or leave. There was some minor wordsmithing, I guess, and then I submitted it.

Was it to change your writing style, or was it anything of substance?

No, it is definitely my writing style—hopefully you can see that it is my writing style. I wrote it all from scratch without using AI.

So no plagiarising by the professor. I am sure your former students will be pleased to know that. You have given us a lot of information in the questionnaire, and you talked about the particular skills that you bring. You said: “I believe that I can bring a distinctive combination of financial expertise, international market insight, analytical skills and leadership experience…which provides a strong fit for the needs of the committee.” By that, do you mean you are bringing something that you do not think the committee has already?

First of all, thank you for having me here. I appreciate having the time to speak to you. I should point out that I have a related—not a conflict, exactly—declaration. Dame Harriett and I have met a number of times through my role as former principal of Lady Margaret Hall and Dame Harriett’s role as an alumna of Lady Margaret Hall. I wanted to mention that we have met before. I think I can bring useful and complementary experience to the FPC. My knowledge and view of the role that the FPC plays in promoting financial stability is that it brings experience across a range of backgrounds, including economists, academics and practitioners. I think I can bring valuable experience from having been a global investor—a risk taker across markets—and as someone who understands participants in markets and their behaviour, and who understands transmission vectors, the co-holder risk and the utility function that investors and traders have, which is often orthogonal to financial stability. There are other members on the FPC—you will have met Jon Hall, who has also been risk taker in fixed-income markets—but I think that element of my experience is particularly valuable. Also, having sat in the seat at the Harvard Management Company—especially as CEO, although I was there for 10 years—I have been able to see a range of financial market participants internationally and across different types of institution, not only investment banks, endowments and foundations, but sovereign wealth funds, pension funds, insurance companies, hedge funds, private credit, private equity funds and so on. I think that knowledge base is particularly useful in understanding the interlinkages of the financial markets and how that might transmit to UK banks. In addition, with my statistical training, and having not only been an academic but used that statistical training in practice, I think I can bring a good balance in respect of the pragmatic use of technology. Something that I have learned not only from my academic training but in practice is that there are limitations to financial models and to mathematical models, whatever they are. Actually having the knowledge of those models is important to begin to understand their limitations. That is something I can usefully bring to the committee.

Could you give an example of how that might play out? Can you think of something that might come before the committee for which you could use that particular skill?

If we look at the global financial crisis, one of the causes, plausibly, was overconfidence in mathematical models, and in particular overconfidence in mathematical models that, for instance, priced the risk of pooled mortgages. Essentially, it was about saying, “Okay, there is some decorrelation between housing markets on the east coast and on the west coast and therefore, if you put that in a model, it will come out with the potential credit risk of this pool of mortgages.” That was proven to be incorrect in the financial crisis. There was overconfidence in that model. This is not like saying, “All models are wrong”—that is a famous statistical aphorism; all models have flaws—but understanding where they are being overly relied on is very important. In that case they were overly relied on because these tranches were rated AAA, so people were able to borrow against them and then, when they were wiped out, that had really damaging effects. Similarly, if we now think, “Are there risks of that?”, my view is that it is not clear where that necessarily lies in financial markets currently. That is obviously a really interesting question. Are the tranching and packaging of private credit loans being mispriced in a similar way to the tranching and packaging of residential mortgages in the US? That is a valid question. I think having both the experience and the technical knowledge to say, “Okay, these models look sophisticated”—they are sophisticated; there have been triumphs of quantitative advances; this is progress—“and yet there are limitations,” is somewhere I could bring some insight. If I may, I will just add one thing that I mentioned in the questionnaire on AI. To me, this is an echo of, “No one really knows where AI will fail.” Actually, it is not that no one “really” knows—no one knows where AI will fail. I think that if you talk to experts in the field—I am not a technical expert—they will agree. The complexity is extreme, and the high-dimensional input that will cause AI models to freak out, for want of a better term, is not well understood. That is increasingly better understood, but it is not well understood. So when we have a system that relies on that type of model, we need to be thinking about overconfidence, overreliance and so on.

Could I just build on that? I do recognise the relevance of your experience then, but one issue will have been, perhaps, the lack of rules around transparency and the way the market was familiar with what was underneath. It is generally said that there is not a complete repeat, but what lessons can practically be applied from that? You have just mentioned, and one of my colleagues will come on to ask about, AI, but it is a completely different phenomenon. The same issue about transparency and lack of understanding of the market seems to be relevant again, so what is the remedy or insight leading to an action that could be taken to deal with this potential risk?

That is an extremely relevant question. During and after the financial crisis, I wrote and spoke quite widely about some of the failings of quantitative models and the use of quantitative models. I really said that the most important thing is to ensure that judgment, which is necessarily based on experience, is integrated in any decision framework, so that one does not rely formulaically on the outputs of models. There is a natural question, “Well, what is good judgment?” That is a great question, but if you think about financial decision making, people need to be aware of the uncertainties in their modelling framework. Financial investing is, of course, full of uncertainty, but one of the uncertainties is the lack of confidence—the error bars, if you like—in any output of any model. It is a case of being really honest about that. There is a tendency for everybody—well, not everybody, but there is a broad tendency in financial markets to become overconfident over time as markets are relatively quiet or growing. One could easily say, for instance, there is overconfidence in the private credit market, because over the last six or seven years that has been a very profitable market to be in, things have gone pretty well and there hasn’t been too much going wrong. People just generally take it for granted, so there is less focus on really re-examining assumptions and outputs. I think that one real benefit of the Financial Policy Committee, and why I am excited about contributing to it if I can, is that it is not a countercyclical body, but a body that wants to continue re-examining areas that participants in financial markets probably would not re-examine on their own—to ask the questions. There are headlines at the moment about the default of First Brands: is that a big fire alarm? Is that idiosyncratic or systemic? No one will know for sure, but that is exactly the right question to ask. A body like the FPC, which in my experience does not exist in the same way in the US, is valuable in having that questioning function.

Basically, you are saying that your previous experience shows you complacencies, gaps and an understanding of what is missing from models that can be used in the same way and applied to some of the phenomena that exist today.

Complacency is somewhat pejorative; it starts off as overconfidence and overreliance, and I think those morph into complacency. There is also, really importantly, to the Chair’s point about what I could bring to the committee, an understanding of the mindset of a hedge fund investor, a risk taker and a swap dealer, which is to be able to take the most risk possible when they feel they have an edge, without really thinking about the overall system. A hedge fund trader has no incentive for financial stability; the incentive is to generate risk-adjusted returns. I have sat in that position as a proprietary trader at an investment bank. I then also ran the internal platform at Harvard, which was essentially an internal hedge fund: it was about 40 people, with the equivalent of about $5 billion or $6 billion in capital. We were looking to generate superior investment returns by taking risk where we felt it was mispriced. We obviously wanted to find mispriced risk, but the most important thing was being able to take that risk as freely as possible. Of course that would be the goal. All participants in the market will naturally want to be as unconstrained as possible, and that is not necessarily consistent with the financial stability of the system as a whole. It is about bringing that perspective. I remember sitting there as an investor thinking, “I do not want to have a haircut on any of my Treasury repo, because that means I can do less Treasury repo when I am doing the basis trade.” That might not be consistent with financial stability, and just being able to call that out would be useful.

Poacher turned gamekeeper.

Yes.

You talked about doing an analysis of the crash after the crash, but you left Deutsche Bank in 2006. Did you see it coming?

No. My team and I were fortunate in that we thought that risk was underpriced going into the financial crisis, so the fixed-income portfolio that I was running at Harvard was relatively successful. That set me on a career that ended up with running the endowment. However, the endowment as a whole was exposed to large amounts of equity risk going into the financial crisis, and it lost over 25%. So to say that we were prepared would be wrong; no, I did not see it coming.

You talked very eloquently about models, and it sounds like that is a very useful skill to bring to the FPC, but were you thinking like that when you were at Deutsche Bank? If lots of people had been able to predict the crash, we would have been in a different place, so you are not the only one who did not see it coming. But how come you did not see it coming, given what you have told us about your skillset and modelling?

It is very difficult to predict financial markets and to disentangle a lot of the complexity and interlinkages in markets. What happened in the mortgage market in the US was clearly identified by some people. People were aware that it was a bubble, although it was not something that I saw; it was not something that I traded. Ex post, one can see that the overconfidence in models was high. One thing that I bring is that there are certain things that make me feel anxious, and that made me feel anxious before the financial crisis—for example, where illiquid assets are marked and cannot be sold after some liquidity event, and people say, “Okay, well, that’s fine. We’ll just keep them marked at the same level.” That was happening before the financial crisis with the Bear Stearns hedge funds. People said, “Okay, we can’t sell these, but we will just keep them as is and take them on to balance sheet.” Things like that are flags. When things like that come up, it is helpful for me, with my experience, and others with that experience to point out that this is not necessarily a great sign. For example, if there are private credit loans—which are within the focus of the Bank and others—that are paying payment-in-kind, or PIK, interest and the companies are no longer able to afford the interest, that is not a strict default, but it is a pretty worrying sign. It likely means that there are undercurrents that are worse than is being transmitted by the first-order pricing.

Finally from me, you told us in your questionnaire that you will retain interests in certain investments. How will you ensure that there is no conflict? Will you be recusing yourself if necessary from any FPC conversations?

On the conflicts on investments, I had to disclose my entire investment portfolio to the Bank, and I went through it with the Bank’s secretary and the compliance team. As you know, I redeemed from a hedge fund run by a former colleague of mine because I thought that there might be a perception of conflict. I sold shares in a UK-regulated entity and have now liquidated the remaining individual securities that I own. There are three private investments that you mentioned. They are all investments in companies run by former students of mine at Harvard—

So this is how professors make their money, is it?

Or lose it, actually. It is not necessarily going particularly well. I have disclosed those to the Bank. Given their likely orthogonality to the UK financial system, they have said that it is okay that I keep those. I would obviously recuse myself if there were any matters relating to them. One is a biodiesel effort in Pakistan run by a former Harvard cricket captain. There are some private investments; I am supporting these ventures.

It sounds interesting. So you and the Bank will both be watching to make sure that, if you need to, you will recuse yourself.

Yes.

Thank you very much indeed.

AI has the potential to deliver huge benefits, but how concerned are you that the financial sector is deploying it without sufficient and adequate understanding on the part of regulators and firms?

That is a really important question, and there are several parts to the answer. The first is that, in terms of employing AI for trading decisions—real-time investment decisions—one of the interesting things about financial markets is that they are time-inhomogeneous. In other words, they change over time. AI is obviously very good at things that do not change over time. It is trained on a static set of data, and is triumphant on things like image recognition, but is much less good on things that change over time. In my experience, the number of investment firms that use AI to make unsupervised trading decisions is relatively low because it is not actually that successful. On other uses of AI, as I said in my questionnaire, AI at some level is simply a regression function from input to output. It is a very, very sophisticated way of going from, “Here is an input”—whether it is a set of data, a text string or an image—to, “Here is an output.” It is extraordinary. The level of capability of AI is extremely positive, in terms of increased productivity and being able to do a lot more with your data and automate a lot of things, so there is a very positive story there. My main concern with AI is that people do not know where it fails. As with anything unsupervised—it is the same even in large language models—the likelihood of AI spitting out something plausible but incorrect is high. People are realising more and more that the plausible output is actually wrong on a lot of things. I think my biggest concern is about where AI will fail, and fail badly. I mentioned in the questionnaire “universal adversarial perturbations”, which scare me. It scares me that you can have an image of a cat and the AI says it is a cat; you change it slightly but it still looks exactly like a cat, and the AI says it is ice cream. That is because the complexity of these high-dimensional models is impossible to understand. You cannot actually understand the causality between the input and output. That is worrying. There is a famous example in speech recognition. You can have a speech translated perfectly, and it sounds exactly the same, but the AI comes out with “FedEx”. It has stuck in my mind that it can be so wrong, so grotesquely different from the input. To me, that is worrying. Broadly, there is an understanding that that is the case, but the balance between the triumph of what it can achieve and the risk of it going badly wrong needs to be focused on.

I reiterate the interest that I have met the professor before, when he was principal of my alma mater. I am very interested in taking up the theme of AI and how you see that combining with the power of quantum computing potentially to undermine the encryption that supports the whole financial sector. What are your thoughts on that?

That is not an area in which I am expert. One thing that brings me some comfort is that a lot of the cutting-edge quantum expertise is in this country, and a lot of it is actually at Oxford; Ian Walmsley has just moved from Imperial back to Oxford to head up the Oxford Quantum Institute. I do not have a technically deep insight into where we are in that phase, but it is important to keep communication open with the experts and leaders in the field. Those leaders are in the UK, so that is the important thing for regulators to be aware of.

We have asked some of the members of the FPC in the past about the potential risks of quantum apocalypse in the financial sector. I am afraid that, so far, we have mainly heard that it is not something they have particular expertise in, so it is good that you know people who can perhaps bring some of that lens to bear on what could be an unknown unknown risk to the financial sector. I also wanted to ask about cryptocurrencies. Do you or have you ever held any cryptocurrency, such as Bitcoin?

No, I have never held or traded any cryptocurrency, either personally or institutionally. My time at Harvard, at the endowment anyway, was pre-Bitcoin. I was asked by several of my students when I was teaching there what I thought of it. I said, “I don’t know. It’s not something I would invest in,” and I have not. So no, I have not.

Do you think it poses any kind of risk to the financial system? In the past, we have been told it is not systemically important.

An important thing to see is how connected cryptocurrencies—Bitcoin and Bitcoin trading, and other tokenised and digital assets—are with the core financial system. My sense is that that connection is not significant, but it is growing. Empirically, a lot of the volatility in crypto, and in crypto exchanges, does not transmit to the ability of the UK financial system to service households and businesses. Of course, the market is evolving, and this is something that needs continued surveillance. My personal sense is that it is not currently a huge systemic issue, but it is growing as a potential.

As a potential to disintermediate a lot of the systemically important members of the financial system?

I am not an expert on the technology, but clearly, the underlying distributed ledger technology is an interesting technology that could improve payment systems, reduce counterparty risk, reduce single points of failure, and reduce settlement risk—settlement risk being one of the major things that financial counterparties worry about all the time. There is some sense that it could be beneficial, and from reading some of the Bank’s material, I think the Bank is interested in exploring it, and rightly so. Some of the more speculative trading in digital currencies, where perhaps the regulatory framework is not as well fleshed out, could cause volatility, but again, it does not feel like a major systemic issue right now.

Bringing to bear your expertise with quant, and your judgment from your experience to date, if we look at the stress tests run by the Bank to analyse where the potential risks might be, can you give us some commentary on whether you think it has been running the right sort of stress tests? Do you think there are stress tests that are missing the risks to the financial sector at the moment? What are your views on stress tests generally?

One of the most useful and impressive things that the FPC and the Bank have done is the system-wide stress tests and system-wide scenario analysis. Certainly, in the previous financial crisis and, I would posit, in any future period of market instability, the most important factor is understanding the transmission risk between entities or the transmission vector, or the co-holder risk—where people with similar trades all act in a similar manner—or where correlations will move in stressed environments. That is very hard to do, stressing a single entity. The system-wide exploratory scenario for public markets, especially related to the gilt market, is really useful. Certainly, there are always different scenarios that one could stress. Having been on the other side, when asked by regulators in the US, “What would happen if…?” we would often think that they should be asking, “What would happen if…?” about something else. From a general perspective, I have been quite impressed by the Bank’s focus on understanding the interlinkages between broadly defined non-bank institutions and the regulated banks.

Would you advocate for a particular stress test that has not been done yet or looked at?

I expect that, if I were on the FPC and looking through both the SWES and our stress tests, I would point out where other things could be asked. We have seen in the regulatory reforms after the global financial crisis that, in things like Basel and so on, there are so many different elements that one needs to look at, whether it is leverage, risk weightings and so on. There are many dimensions one could stress of either an institution or a system. What I can bring is judgment. For instance, I could go, “Which are the ones that are most likely to cause a correlated or a cascading risk?”

Are you an advocate of the Bank of England bringing in a central bank digital currency?

I do not have a strong opinion, one way or the other, on that.

The gilt market is core to both the financial system in the UK and its fiscal policy. I have traded gilts for a long time. My main experience is that the gilt market, despite being a highly developed and high-grade Government bond market, can be very idiosyncratic. When I started trading gilts there was not enough supply in the long end and pension funds did not have enough gilts to buy. Forward rates in sterling gilts were close to zero or negative. Now, we see an almost complete reversal, with long-end risk premia at extreme highs. The steepening, or more particularly the increased risk premia, of gilts relative to swaps has been mentioned in previous financial stability reports and by the Bank. It is matched across other global Government bond markets—it is similar in the US and Japan—so there are some global phenomena that are not atypical to the gilt market. Clearly, the gilt market disruption in 2022, which resulted in leverage in liability-driven investment funds—essentially, leveraged long positions in sterling swaps that were not sufficiently collateralised or capitalised—had a negatively convex impact on the gilt market. That transmitted through to the real economy and mortgage lending. The big question is, “How robust is the gilt market to future disruption like that?” It has rightly been a topic for the Bank. The work that has been done on gilt repo—consultation on gilt repo and ensuring that there is appropriate gilt repo financing—is essential. It is a highly positive thing that dealers feel they can get financing for gilts. As you know, we are at an all-time high in open interest on the gilt future. That is partly because issuance is much bigger, but also because there is a lot more leverage in the system. Understanding the linkages is extremely important. I know that this Committee and the FPC have looked at leverage of the basis trade—holding gilts against being short in the gilt future by leveraged participants. Clearly, that is still a trade. We should examine if there is confidence in the underlying mechanics of the gilt market. That type of activity is entirely appropriate to bring together the mispricing from real money buying futures rather than cash bonds so that it is actually an additive to the UK financial system.

It is really hard to say. Markets get scarred; they often react and hypercorrect. Is there excessive fear of another dysfunction in the gilt market and a really significant sell-off in the long end, and how much of that is gilt specific? I do not have the answer. After “liberation day” and the tariffs in April, the treasury market started a significant bear steepening with a sell-off in the long end and a cheapening—or increase in the risk premia—of long-end treasuries. Both long-end treasuries and long-end gilts trade at a negative swap spread. Fundamentally, that says something very strong about the cheapness of or the lack of demand for long-end Government paper, but it is the same in the US as in the UK, so it is hard to say. I have sat in the US for the majority of my investing career, and most US asset managers view the treasuries market as by far the most dominant market and the UK market as kind of quirky. Of course, it is not quirky for the UK in any way; it is absolutely central and core. Within the financial system, there is therefore more vulnerability of the gilt market to shocks, which may be not just UK shocks, but may be international shocks, shocks about the independence of central banks or sustainability of Government debt globally, which is of course a global issue. It is very hard to say. There is of course not a counter-argument, but something that many investors will be thinking about: if and when inflation starts to come down, and if and when there is some growth and tax receipts are up, there is less issuance and then these long-end Government bonds look very attractive from an investment point of view, thinking about where terminal short-end rates will be. There will be a point at which there is demand for those long ends from real money.

I think they learned the right lessons. I was just transitioning from the US. Personally, I think the response to that was appropriate and effective. The question is, what has been learned from the fact that that event happened? In some ways—there was analysis of this—pension funds that were hedging their long-dated liabilities by receiving fixed-in swaps in a leveraged form, or buying gilts in a leveraged form off balance sheet, were doing the appropriate thing to hedge their fixed liabilities. There was awareness that in a sell-off their liabilities would be positive for them, that they would lose money on their hedges and that there might be some liquidity mismatch. There was some scenario about, “Okay, if there is a 50 or 100 basis point sell-off, then the following things will happen,” but it is interesting because the sell-off was even bigger, and then we hit a significant negatively convex, negatively gamma event of liquidating gilts to post margin against losses on gilts. There was a death spiral. A really interesting question therefore is, should that have been better understood or better signposted, or a better light shone on it? That was quite an extreme move. I think that lessons have been learned, but we should think about the fact that the knowledge of the leverage and of the behaviour was actually present. That is quite an interesting dynamic.

That is a really difficult question to answer. One can model a number of extreme and very unlikely circumstances, such as the stock market falling by 50%. That would cause massive ripples. Does that mean that not modelling it is a failure? It really needs a sense of judgment to go back to ask that. The Bank’s language is “extreme but plausible”, but of course the worrying things are the extremes that we have not thought about. I am almost less worried about the extreme-but-plausible ones, because they have been looked at and people kind of understand what is happening. I think it is the unknown unknowns—

Like Rumsfeld.

Yes, Rumsfeld, or Maynard Keynes and the irreducible uncertainty. There is irreducible uncertainty in financial markets. There is not only uncertainty about the worry if stocks go up or down, but irreducible uncertainty, which is about things that one really cannot model mathematically.

What do you worry most about then? You were quite forthright on AI, for example, but I do not want to put words in your mouth.

What I worry most about is a cyber-attack.

On the Bank?

On the electricity grid, because I don’t feel comfortable in my own knowledge of cyber-security to think— I feel more comfortable with AI, thinking, “Okay; I know this is a regression function of some sort,” but a cyber-attack and understanding it—I think it is really important to understand where the knowledge is advancing in cyber-security between poacher and gamekeeper. That is the thing. To the point about quantum computing, I feel that the knowledge in the openly discussed forum of higher education and research firms is relatively open, and it is unlikely that there will be a quantum leap that isn’t known to the experts in the field. I don’t know enough about cyber to have that same level of confidence. That is the thing I would worry about.

I will pick up on something that the Committee looked at last week with respect to the capacity around AI. When I was in the Treasury, we designated the cloud as critical infrastructure. We saw an AWS outage this week.

Amazon Web Services.

I am sorry, Chair. It feels as if the Government have done what any Government would responsibly have done, which is to say, “Look, we have a concentration of three cloud providers for 75% of our hosting.” Your explanation of what you are worried about rather demonstrates that, because how do we reassure ourselves that we have the capacity in our oversight mechanisms to deal with phenomena such as market concentration of the provision of critical infrastructure to the markets? Are we even capable of deploying remedies, given how embedded some of those relationships and the market dominance are? I wonder whether you have anything to say about that, because it is an extension of the application of your anxiety about what we do to deal with some of these things. It is one thing to say that the Government have asked the Bank to view it as critical, but are we rapid enough to be able to deal with the nature of the risks that you perceive to be No. 1?

I agree. Step 1, which as you say is the essential step, is understanding and designating critical third parties, I guess—not just providers of technology services but central counterparties and so on that are key providers. But the question you raise, which is very apposite, is whether there is enough expertise and understanding to understand the risks of those critical third parties?

That is why I am asking.

Yes. I would just go back to that. It is a slight pivot, but I feel that what I can bring to the FPC is what the Bank is trying to do in that regard on risk-taking, transmission and model usage and misusage within financial markets. I would think that the Treasury and the Bank are also able to bring expertise to bear in areas regarding stability of cloud provision. This is the expertise that exists in the private sector.

I would just observe that, as someone who was a Minister in the Treasury for several years, I am reticent to say that just because a Minister—any Minister—declares in good faith that something is going to happen, the state and all its apparatus is necessarily capable of delivering it. I would respectfully urge you to continue asking questions laterally of Government Ministers, officials and all elements of supposed regulation. I accept that your answer is around what you are responsible for—market behaviours and so on—but actually your deep, broader experience of this more laterally needs to be applied more laterally too.

That is a very polite way of saying, “Don’t always trust Whitehall to get it right.”

You mentioned, in the context of our discussion about the gilt market, threats to central bank independence. That engages political risk, and the wider political risk around China, Taiwan and so on. Do you think we are under-pricing and under-examining that? Is that something that keeps you awake at night?

It is interesting. I am not sure we are under-examining it. We might be under-pricing it. In some ways, mispricing is not necessarily a risk to financial stability. Changes in information, changes in policy or geopolitical upheaval can often cause market volatility, as has been shown. One could argue that there has been less market volatility than one might have expected, given the level of geopolitical activity. Through a financial stability lens, it is a question of thinking about what the exposures are to these financial assets and how they transmit into the UK financial system. The FPC, in its most recent financial stability report, highlighted the risk of potential overvaluation in AI, geopolitical risk and so on, which is appropriate. One then needs to think about what the exposure of the UK banking system and the UK financial system is to those types of financial assets. The examination needs to include transmission and exposure, rather than market volatility itself. On your point about central bank independence, and particularly how that can affect Government bond markets, it was pretty clear when there was questioning of the independence of the Federal Reserve in setting interest rates in the US—there was a significant steepening of the US Treasury curve, with the front end coming down, because rates had been mandated to go down, and the long end going up, because of less credibility on inflation fighting. As we mentioned on the gilt market, these pressures on global Government debt and the fiscal sustainability question can transmit into financial stability. In some ways, that is a more direct transmission vector than geopolitical turmoil.

We were speaking earlier about unknown unknowns, but I want to go back to a known unknown. You have spoken a bit about the instability in the private credit market. There have been some recent defaults that caused a lot of speculation, and I wondered if I could hear more from you about what lessons you think could be learned from those recent defaults and what the potential impacts on the UK might be.

I know the Governor said yesterday that the key question is whether this is an idiosyncratic default, because companies default, or a canary in the coalmine—basically the first reveal of significant issues in the private credit market that could transmit into the UK banking system. That is an open question. Maybe I could just step back. The evolution or creation of an increased private credit market was based on sensible foundations. There are capital providers, whether pension funds, insurance companies or endowments and so on, looking to invest in credit markets and fixed income and take credit risk appropriately, dealing directly or through private credit funds in private transactions. It is a natural asset class. My concern is that that natural asset class, because of its success, has grown significantly. Anecdotally, and from reading some of the Bank material, there are concerns about the underwriting standards of loans. There are therefore concerns about events like First Brands, which suddenly went bankrupt, and concerns about opacity. Personally, as an institutional investor, I was never particularly keen on private credit as an asset class because of the opacity about its actual valuations. As a private, untraded security, the loan or the security on the loan could essentially be marked at par ad infinitum until either maturity or default. My concern, which perhaps First Brands is a potential indicator of, is that loans that are basically marked at par could default and transmit significantly through to the financial system.

Could I push you on the potential impact on the UK market? I have read the comparisons to what was happening in the noughties, during the build-up to ’08. The narrative comparisons are the complexity of the products and the fact that people have a lack of understanding across the system. I understand the narrative alignment, but I want to understand better the level of risk, and the level of exposure of UK markets to that risk. If it were to go pop, would it be as big as ’08, or would it be much more contained?

That is the key question—the UK banking system’s level of exposure to lending to private credit funds and to lending securitised by a portfolio of those types of loans. I have not looked at the detail of that, but it is something that the Bank is looking at. I know that an analysis of what you asked is planned to happen, to try to get a better understanding of a system-wide exploratory scenario for private credit. That will involve dealing with the major private credit and private equity capital providers, dealing with the investors in those private equity firms, talking to some of the borrowers, and doing an analysis of how that translates. You are absolutely right that the question that needs to be answered is: what is the overall exposure?

You also said how it is a natural and normal market to have emerged, for the reasons you stated, but I have seen some commentary that it has grown to the size it has because of the tightness of regulation on the traditional banking industry. Do you have any comment on that? Do you think that restrictions in one part have led to this situation?

That is a really interesting topic. Has the regulation of the UK and international banking system post the financial crisis meant that risk has shifted elsewhere? Probably, yes. Is that appropriate? That is more of an open question. The fundamentals of regulation after the financial crisis—Dodd-Frank and afterwards—are totally solid. Banks should not have been that levered. Should there be central clearing? Yes. Should there be haircuts on repo? Yes. Should there be initial margining? Yes. If you asked any banker or trader, in private, whether all that was appropriate, they would say, “Absolutely.” There is also the fact that the banking system has a liability stream, the deposits of UK retail, that is very different from the liability stream of insurance companies, foundations and endowments that are investing for the long term. It would be entirely appropriate that the risk regulation of UK banks is different from the regulation of other bodies. I am sure everybody in the industry would agree with that. There is a question about whether the risk is being taken appropriately and in the appropriate places. The second question is whether there is enough knowledge of the risk that has been taken in the non-bank financial system. The answer to the latter question is no, by virtue of the fact that these are unregulated entities that can be offshore, and participants in the UK’s globally open economy. It is a valid question that I do not think has a straightforward answer one way or the other. Having sat in a risk-taking seat, if I were at a bank, I would say that I have too much regulation—I would much rather have less regulation. That is not pejorative; there is just an incentive to say that, and I have been on both sides. One has to look at it—and I think the FPC does, from the material I have seen—and say, yes, there are risks in the non-bank sector that we do not have as much clarity on because they are not regulated like the UK banking system, but that is okay because they are different entities: they are not taking deposits from UK consumers, and they are not offering mortgages to UK consumers.

Our concern and that of the FPC would be whether that could, although it might not look like another direct impact, undermine the stability of the financial system. This is my final question. Let us say that the FPC does the analysis and concludes that something needs to be done that will require global co-operation. At the moment, we have an Administration in the US that is probably looking in the other direction, towards deregulation, so what confidence do you have that the FPC would be able to fix the problem if it assesses that there is one?

Yes, it is a real challenge. That is a question of what I would call shared regulatory endeavour, which is regulation of the global financial system. The UK is an open system, so it is impacted on by the global financial system, and many, many people who never set foot in the UK trade gilts. Among many other things, that endeavour is about a sense of international co-operation, so yes, in some ways, the most important thing about geopolitical risk for financial stability is a lack of co-operation on regulatory openness, discussion, consistency and coherence. That is a risk. I think that the Bank—and the UK—is in a relatively strong position, because it has a very good reputation internationally as a thoughtful regulator. I am not talking for the Bank, but that is from my experience in the US, talking to people at the Fed and people who have been Fed governors. The Bank is pragmatic and flexible. In some ways, because the financial system is so important to the UK economy, and there is a lot of sophistication in the financial system and a lot of international firms in the UK, the UK regulatory regime is really thought about and challenged, and so it is respected. People can disagree with it, but there is real understanding that the Bank and its different arms are trying to do sensible things. In some ways, it is not that this is soft power; rather, the fact is that the Governor is the chair of the Financial Stability Board and there is a real openness in the UK authorities to share knowledge. That is probably the best way forward, because politically it would seem unlikely that a raft of heavy new regulation will come to the US in the near term—that is obviously the case. The question is, how can one keep a level of what I described as shared regulatory endeavour? The optimist in me can see that people will say publicly, “Okay, we need less regulation. This has gone too far after the financial crisis”, but I think in private everybody in the system knows that the regulation after the financial crisis—its core elements—was absolutely necessary, and that financial stability and actually economic growth are dependent on having confidence in the financial markets. This is not a linear thing—more regulation, less regulation—but is about having the appropriate framework to enable enterprise and to have confidence in the system to absorb shocks. It is a multidimensional problem, but I think that international co-operation is really important, particularly so for the UK.

In the questionnaire, you said that stress testing for climate-related risks, improving data and modelling capabilities, and so on are all important roles. You had a focus on tools and on climate stress, but the Bank has only conducted one climate stress test to date. There has also been quite a lot of chat in the Financial Times about the Bank cutting back its work on climate change risk. Are you persuaded that the Bank has the balance right on assessing climate risk?

My sense is that the Bank has been focused on actually understanding how climate change risk could impact on the financial system. In some ways, as I have said on various occasions, the Financial Policy Committee’s role is to identify where major risks in the market could be and to try to understand how that might feed through into the financial system in the UK. Clearly, the transition risks—I prefer to call them the financial transformation risks—of changing policy, changing consumer sentiment and changing technology could have significant impacts on the workings of the financial markets. In some sense, just having that front and centre I see as being a key part of the FPC. What are the jump risks—for want of a better word—that could happen because of climate change? In my view, that seems to be appropriately on the dashboard for the Bank.

The Chancellor has made extensive remarks on how UK regulation has become overly risk averse. Can you point to any examples where you think that the Financial Policy Committee has been overly risk averse?

Regulation is a multidimensional and complex field, and the regulation of financial markets has many elements to it. I do not think anybody would claim that any regulatory system cannot be improved and does not need review. The absence of a significant financial crisis since the global financial crisis and the empirical activity in the economy do not mean that there has been too much regulation. I do not necessarily view it as too much or too little. My view of the FPC is that it has been thoughtful about identifying where risk can transmit into the real economy. Where it has put in limits, for instance, in its mortgage lending limits—

That has changed.

It has.

Were they overly risk averse before?

I do not have a strong view. I would not say yes or no. Other parts of the Bank do a lot of data gathering across the UK about the availability of credit to businesses and households. My understanding is that that has not shown a lack of availability for businesses. It has also shown that the main hurdle for homebuying is the deposit rather than the availability of mortgages. It is not clear to me that there has necessarily been an overly conservative regulatory regime. That is not an obvious conclusion to me.

So you disagree with the Chancellor?

It is not obvious to me that that is the case. I have not reviewed it as deeply as the Chancellor.

She has talked about “a boot on the neck” of business.

I was in the financial markets during the 1990s. It felt like there was very little regulation, and I was like, “Hey, I can do this trade without any additional capital—how fantastic.” I am now in touch with many people within the financial markets. Sensible regulation breeds confidence, it does not hinder activity. There may be regulations that are not sensible. I have not reviewed that. Mindless form-filling for no purpose would be an example. From my own analysis looking at bank capital requirements: yes, one might want to refresh bank capital requirements, and the countercyclical buffer plays a really important role in that. The countercyclicality of that is incredibly important. But should we review that? Potentially, we should. How about the risk weight? Is Basel III appropriate in how it is risk weighting assets and so on? Those are all things to be reviewed. My personal view—not that of the FPC—is that it does not feel that regulation of the financial services sector is necessarily choking activity.

Of course, there is a risk that politicians like us pass laws that create lots of reporting mechanisms to reassure ourselves that we have done things in the public interest. Much of that reporting activity is expensive to financial institutions and actually superfluous—you recognise that as part of the issue that the Chancellor is probably referring to. A couple of years ago we passed legislation in the Financial Services and Markets Act on competitiveness being an imperative on a pretty bipartisan basis. Could we not reasonably expect you, the FPC and the Bank to be giving more clarity about what we can actually remove and the proportionate risk that we are taking on by doing so? There are lots of reporting obligations on banks that they say are superfluous. They are sometimes reluctant to say it too publicly because they do not want to get on the wrong side of the people who regulate them, but in truth, looking at the global landscape, which you are familiar with, the way that other jurisdictions have been moving recently will make it more difficult to be competitive in London. Is that not reasonable? Do you not feel an imperative to do something about it?

This perhaps speaks to Dame Harriett’s comment. This is not linear—it is not about turning the regulation up or down. It is about thinking sensibly about appropriate regulation. As you rightly say, I have focused in my comments on sensible, appropriate regulation that builds confidence in the system that there are appropriate buffers and appropriate capital, and that things are done sensibly. That is the argument for sensible regulation. I agree that one should remove non-sensible regulation and excessive reporting requirements. I have heard that. I have been on the other side of that, thinking, “It is pointless reporting the following information, which doesn’t”—

Lead to action.

It doesn’t lead to action. This is something that the FPC and the external members can bring. With a lot of reporting from financial institutions, the financial institution will say, “First, it’s not worth while and, secondly, it does not actually reveal anything about what we are doing.” Therefore, it is not worth while to the regulator, which might then be asking for something that is not reflective of the risks being taken or asking for meaningless data that cannot be used. I would be surprised if there are not examples of that. That is a really interesting point for the FPC, whose secondary objective is to support the Government’s objective of growth, subject to financial stability. Personally, I believe that the FPC does have a role to say, “This framework”—or this capital requirement, this leverage limit, this kind of reporting—“is essential to ensure confidence in the market. However, this, this and this does not add anything,” or, “The cost-benefit of this is really not worth while.”

Good.

I have also overemphasised—well, not overemphasised—the voice. Let us think about the overall picture. Financial stability is often not in the best interests of individual traders or firms; it is in the best interests of the country and the system.

I think what we are trying to say is that we all agree that systemic stability is an enduring imperative, but sometimes being competitive in a global market requires saying, “We will not compromise that systemic stability if we remove some of these superfluous reporting things.” The sense I get from talking to people in the markets is that we are sometimes not proactive enough for fear of leading to a dispute over whether it has an impact on systemic stability.

I think regulators and policymakers naturally worry about a hyper-correction swinging back too far the other way. Obviously, there is a concern from the market that we swung too far to significant regulation post crisis. Clearly, some adjustments need to be made and reviewed. The FPC is totally on board with that. Then there is the question of whether we are swinging too much towards deregulation. My view is that deregulation is not linear. It is not more deregulation, higher growth. It is not binary. It has to be the appropriate framework. I personally agree that one should not be saying, “Oh no, we can’t remove the following areas of regulation, which don’t serve any purposes,” because that would be going too far the other way. I totally agree with you on that.

Thank you very much indeed, Professor Blyth, for coming in front of us today. We will now break into private session to deliberate on our report on your appointment. If you want to wait outside, we will hopefully come to you relatively quickly with our views.