Charles Goodhart Interview

26^th June 2025

Andrea Enria and Pedro Duarte Neves in conversation with Charles Goodhart

Andrea Enria: This is the opening contribution for the Forum on Financial Supervision and I couldn’t think of a better person than you, Charles, to set the stage on the key issues we should focus on when discussing financial regulation and supervision, and potential threats to financial stability. Thank you very much for accepting our invitation. 

The first point I would like to raise is the impact of geopolitics on financial regulation and supervision. The rules-based international order that has accompanied the extraordinary increase in globalisation in the last decades is increasingly challenged, and this could well include the mechanisms for international standards setting for financial firms and markets. You wrote the first history of the Basel Committee on Banking Supervision (BCBS). Is this international framework under jeopardy?

Charles Goodhart: At the moment, I would prefer to focus on the internal politics within the US. The reason for that, in my view, is that the international standard setting by the BCBS is actually to the benefit of large American banks. You will remember that it was the British and the Americans that took the initiative to develop the Basel I standards, partly in order to protect themselves against the more competitive Japanese banks, which were then operating with lower capital ratios. There is also the issue that, if the international standard setting by the BCBS were to break down, it would mean that the international business of large American banks could very well face differential regulatory requirements in the many countries where they operate. This would be a very considerable operational burden for them, and in case of a more conflictual situation it could be that the American banks could get discriminated against. So, I think that large American banks will ask for some adjustments in the implementation of the Basel standards, but they will not want to disrupt the present set up.

Although there could be some political hostility towards large banks in the US, I think this will not lead to a departure from international standards, but rather to the adoption of a much less onerous set of regulations for smaller banks, which may trigger pressure for a more differentiated treatment of local banks also in other jurisdictions. This could be justified in light of the difficulty of small banks to deal with administrative costs imposed by rather complex regulations. So in my view the most likely outcome is that international agreements remain in place but apply only to large international banks and to a much lesser extent to the smaller banks. Do you agree?

Andrea Enria: I agree large US banks have an interest to have the Basel standards implemented. I am less sure a relaxation of the standards for mid-size banks is a very good idea after what happened in the spring 2023, when turmoil at US regional banks had major systemic effects. 

I take from your answer that you do not expect financial services to become the battleground for trade wars. 

Charles Goodhart: No, I do not see in papers of members of the administration, such as Steven Miran, nor in statements by President Trump any tendency to necessarily move in that direction and I don't see why it would be in the benefit of the United States to do so.

Pedro Duarte Neves: Moving to a financial stability perspective, how do you see the risks for the preservation of financial stability from the current policy mix (monetary policy, fiscal policy and macroprudential policy)? What are in your view the risks on which supervisors should focus most at this juncture?

Charles Goodhart: I see some very severe risks arising. I have always been of the view that there has been an underestimation of interest rate risk. Interest rate risk is not directly captured in international capital standards and I don't think it is so much covered in the oversight of supervisors. On the opposite, governments have always been very determined to make it as easy and desirable as possible for financial institutions, including the banks, to finance their debt. This is now becoming even more important, as debt to GDP ratios rise and fiscal deficit remains far too high, raising concerns as to their sustainability. Furthermore, demographic reasons are likely to keep the growth rate low and, together with other factors, are contributing to drive inflation, and therefore interest rates, higher. 

While facing enhanced interest rate risk, the danger is that politicians might want to push banks to invest even more in government debt, for all maturities. I think this is one of the reasons why there is so much pressure for exempting government debt from the supplementary leverage ratio in the US. Banks could then get heavily invested in bonds issued by their own governments at a time when the potential for disturbances and sharp increases in yields in the government debt market is becoming ever more apparent. We should ask ourselves whether what happened in the Gilts market liquidity event (the “Truss event”) could only be, using a fairly common phrase, a “canary in the coal mine”. And that of course leads on to the question of what should the central bank do.

If confronted with severe disruption in the government bond market, central banks would actually have to engage in extensive quantitative easing (QE), at a time when inflation is still a significant danger, and they might have to be raising interest rates. I see a potential condition, which I describe as the conflicted central banker, in which the central bank is simultaneously increasing balance sheet via QE in order to keep the government debt market from collapsing and, at the same time, raising interest rates to keep the inflation from going well above the target.

Going back to financial stability, I would hope that supervisors would be extremely careful to keep the maturity structure held by banks, to really very, very low risk. As governments are going to want banks to take up as much debt as they possibly can, the supervisors should be worried about interest rate risk. I see the possibility for conflict becoming greater than it has been in the past.

Andrea Enria: It is indeed a regret that international standards do not cover properly interest rate risk, although there is guidance for supervisors and authorities are more alert to this risk following the spring 2023 turmoil. Do you think that besides the interest rate risk, also sovereign risk could become relevant again? We've seen it at work during the euro area sovereign debt crisis, in which the change in valuations of government bonds were adversely affecting the market’s perception of the solvency of banks. Can this now happen in other jurisdictions, and should supervisors do something about that?

Charles Goodhart: The conditions in the euro area are rather special. In countries like the US and the UK I see sovereign risk and interest rate risk as being almost the same thing. It is different than in the euro area because of this separation of fiscal policies, which are at national level, and monetary policy, which is in the remit of the ECB.

Andrea Enria: This also links with what Hyun Song Shin was saying at the lecture in your honour a few days ago, in which he highlighted the structural shift in the last 20 years from private debt to government debt, from bank intermediation to non-bank intermediation and an increasing role of the FX swap market. 

Charles Goodhart: For an old person like me, that is not actually a major structural change: if you take the period from 1939 onwards the banks actually got plenty of government debt in their balance sheets. This changed extremely slowly as fiscal conditions improved and as nominal growth became higher than nominal interest rates. The period in which private debt was the main feature of banks’ balance sheets is not necessarily the historical standard. The holdings of government debt have always been the centrepiece of liquidity and safety of the commercial banks. There is nothing really new about that, it has always been the case.

The problem is when investors begin to feel that government debt, including US Treasury bonds, is becoming riskier. What happens if the public sector debt is seen as potentially risky? What should banks and supervisors do? That is a much more worrying problem.

Pedro Duarte Neves: You mentioned before the demographic trends, which you discussed extensively in your book “The Great Demographic Reversal”. How do you see these demographic changes affecting the financial sector? And how should financial regulators and supervisors adapt to these long term trends – not only demographic trends, but also, for instance, the climate transition?

Charles Goodhart: I would like to emphasize the tendency for the banking sector to move much more into mortgage lending to households and for finance for the corporate sector to move to non-bank financial intermediaries. The relationship between housing and banking is going to be very important over future decades for several reasons. 

As the pressure for higher public sector expenditures goes up - you mentioned climate change, but there is also defense, ageing because of demography and infrastructures – governments will see a progressively declining base for income tax, as the workforce will keep declining as a result of demographic trends and cutbacks to inwards migration will likely persist, driven by populist policies. The answer I give to that conundrum is that we will be forced to move towards taxation on property, particularly taxation on land. If introduced, that would cut back land and housing prices and would cause very considerable financial difficulties in the short run, even if in the long run it would lead to better outcomes.

We had a housing bubble inflating in the run up to the great financial crisis and we addressed it by policies focused on loan-to-value and debt-to-income ratios. But that has not stopped housing prices from going up in many countries, including in the UK. This has particularly impacted the young, less wealthy individuals, who cannot afford to buy a property and tend to stay longer with their parents, which is one of the reasons why marriage and birth rates are so low. Taxation on properties has to go up and will probably need to be focused on the squared meters of properties, as other metrics (e.g., on windows) have not worked well. The interrelationship between banks, taxation and housing is going to be one of the major issues for the next few decades.

Pedro Duarte Neves: You mentioned the increased use of borrower-based measures such as Loan-to-Value and Debt-Service-to-Income limits. How do you assess the contribution of macroprudential policies to the resilience of the financial sector since the great financial crisis, how better are we prepared to deal with future crises?

Charles Goodhart: I think that the contribution has been considerable. There is a longer-term problem, I call it the “Minsky problem”, which is that it is almost impossible to measure the optimal degree of regulation on capital, liquidity, the capacity to meet mortgages and so on. The reason is that there is always a trade-off between the short run effects, as tighter regulation will slow down lending and growth, and the long term benefits of greater resilience and a lower likelihood of crises. As crises do not occur frequently it becomes extremely difficult to assess what the optimal balance between growth and stability is. After a crisis, the salience of financial crises comes to everybody's mind. Everybody agrees that we must never let that happen again. As the time between the last crisis and the present lengthens, more and more people start wondering whether we got the balance wrong, to the detriment of growth. The pressure to relax regulation and supervision increases, while new risks build up and this creates the condition for a new Minsky moment. The deterioration in fiscal conditions and interest rate risk are the elements that could combine with less demanding regulation and supervision to lead us to the next crisis, although the impact might once again be softened by quantitative easing.

Andrea Enria: Indeed we are seeing clear signals that the US administration intends to move towards deregulation; in the UK the Chancellor has clearly indicated that financial reforms went too far and called for greater attention to growth and competitiveness; in the EU there are discussions of a “simplification agenda”, aimed at reducing the regulatory burden on financial institutions. In the area of monetary policy the time consistency issue has been partly addressed with the independence of central banks. What could help addressing the “Minsky problem” in regulation and supervision? I thought international standards would play as a safeguard, as you set up rigid rules everybody commits to, which require broad international consensus to be changed. This should provide some stability of the regulatory framework across the financial cycle. But it doesn’t seem to work that well. 

Charles Goodhart: You need consensus of at least a small number of really big countries. And if the US refuses to play, that consensus is gone. Also, as you know, the ECB is unique in being based on a Treaty, which cannot be changed without agreement of all Members. That is not true in the rest of the world. In the rest of the world, the independence of the central bank, both in monetary policy and financial stability and supervision, can be overturned by the political authorities if and when they want to do so. There is some protection for central banks in the reaction of markets to an attempt by governments to reduce that independence. This mechanism worked recently in the US, leading President Trump to say that he was no longer thinking or trying to find ways to replace the Chair of the Federal Reserve, Jay Powell.

Andrea Enria: Markets don't seem to react in the same way when the independence of supervisors is questioned. Governments’ pressures for deregulation are generally welcomed by markets.

Charles Goodhart: I agree with you, I think that is true and it is another reason why I feel that a Minsky moment and a future crisis are within eyesight.

Pedro Duarte Neves: Stress tests are used by supervisors as tools to force the banks to prepare for a crisis. Approaches differ somewhat across jurisdictions and a debate is on-going also on the best way to include non-bank financial institutions, which as you said are becoming increasingly active in corporate credit. This is a reason why we included stress testing as one of the special focus topics for the Forum in 2025. You have also done influential work on stress testing, what are your suggestions to improve the effectiveness of these exercises?

Charles Goodhart: Until extremely recently the tendency, which I think was the correct tendency, was to try to extend stress testing to a selection of non-bank financial intermediaries. It was done in the UK, it was suggested that it should be done elsewhere. I do not know whether that will survive the current shift of momentum towards less and simpler regulation.

Banks sought to minimise the effects of regulation on their business, for instance stepping up securitisation to ship a great deal of their activities into a less regulated format. This enhanced the connection with non-bank financial intermediaries. In order to maintain financial stability we are forced to follow the shift of business out of banks into non-bank financial institutions. Whether the tendency to take more account of the potential threat to financial stability arising from non-bank financial institutions will survive in the present atmosphere, I don't know. I think it is uncertain.

Pedro Duarte Neves: All of us recall the famous Goodhart’s law, which basically states that when a measure becomes a target, it ceases to be a good measure. How do you assess the performance of solvency and liquidity ratios or any other regulatory ratio introduced for financial stability purposes under the Goodhart’s law?

Charles Goodhart: Inevitably an institution, or indeed a person, confronted with a regulation or some target, will try to change their behaviour in such a way as to reduce its degree of constraint upon themselves. It was a point raised later on also by the so-called Lucas critique. But I would argue that the law is slightly wider than the Lucas critique, because one of its features is that the authorities themselves have an incentive to try and fix a target and adjust conditions so that the target is achieved. Authorities want to show they have been successful and will actually adjust matters so that they appear to be successful. Lucas only thought of the changes in behaviour of those subject to the regulations, controls or targets, but the behaviour of those who are setting the targets changes as well. The law covers this double action, and this is something that can be very relevant in some cases.

Andrea Enria: For instance, supervisors could become laxer in their supervision of internal models which are used to calculate the risk-weighted assets in the solvency ratio.

Charles Goodhart: Exactly. You want to show that you've been successful. There are times and situations in which the behaviour of those setting the target is also changed. 

Andrea Enria: Indeed, also recently we saw several cases of banks that showed capital and liquidity ratios well above minimum requirements and still entered into very disruptive crisis, as risks were geared up in areas not captured by the requirements and possibly not sufficiently pursued by the supervisors.

Pedro Duarte Neves: For instance interest rate risk was not included in the US stress tests just before the crisis of the regional banks, triggered by the spike in interest rates.

Charles Goodhart: Could I raise one issue with you which we haven’t discussed so far? I understand artificial intelligence (AI) could change the operation and to some extent the cost and burden of regulation very considerably, but this is not a topic that is widely discussed. I see room for AI favouring quasi-automatic, almost real time transmission of information from banks to their supervisors, which in turn can use AI models to perform their tasks, taking action only when something is obviously going wrong.

Andrea Enria: There is a lot of work in supervisory authorities on the implementation of what we call SupTech or RegTech tools. In this first phase this is mainly focused on the application of AI, especially natural language programming (NLP), in order to make the supervisory process more efficient – for instance, speeding up the analysis of application packages sent by banks, summarizing them or extracting the most relevant information. 

As you correctly point out new technologies could significantly streamline the reporting process, enhancing the quality and timeliness of information available to supervisors and reducing the administrative burden for firms.

The next stage will be the development also of supervisory analytics based on AI, which could for instance enable the use of a wide set of stress testing scenarios without raising an unbearable burden for banks.

But I have two concerns. First, budgets of supervisory authorities are currently under very strict constraints, which will limit the ability to invest in AI tools and modernize IT infrastructures as much as needed. Second, even if we were able to set up very effective supervisory monitoring via AI, I remain convinced we would need to maintain sufficient room for supervisor judgement. For instance, I don’t think remote access to information will ever be a perfect substitute for on-site inspections, which enable supervisors to engage in in-depth discussions with bank management and staff, check the quality of data, assessing the internal controls framework, and so on. I don’t think supervision could ever be fully automatized.

Pedro Duarte Neves: AI is already being used for aspects like fraud detection, alerts of possible money laundering activities, identification of bad commercial practices, social media and consumer sentiment monitoring, automation of routine supervisory tasks. This is already state-of-the-art supervision. But I agree with Andrea that an overall assessment of the conditions of a bank will continue to call for supervisory judgment.

Charles Goodhart: I agree with all of that. But when a bank has passed the test and credibly showed to have reliable data, then it might be possible, subject to occasional checks, to reduce the burden and cost of supervision.

Andrea Enria: This is the direction of travel. But it is not a simple project as it may sound. Banks have to report to a variety of authorities – central banks for statistical purposes, prudential supervisors, resolution authorities, conduct regulators, anti-money laundering authorities. To make the process really efficient you would anyway need an agreement to integrate the reporting process across authorities. There is a project in the EU to do exactly that across central banks, prudential supervisors and resolution authorities, building a single dictionary for data requirements, which eventually should enable each authority to fish the data they need out of a single pool of data reported by the banks.

Charles Goodhart: But you need to have an agreement between countries on that dictionary. Within the European Union you can have it, but it is not going to be easy to achieve that at international level.

Andrea Enria: Exactly. I don't think AI can help that much there, it will be very difficult to harmonise data requirements by authorities on a global scale. But I don’t rule out that this might happen in future, we need to remain open minded and hope that new technologies could also lead to a leap forward in collaboration amongst supervisors.