Yale Program on Financial Stability Conference on Building Resilience and Managing Financial Crises
It is a great pleasure to speak to you today on a policy area at the forefront of our work at the IMF in helping our members prepare for, and deal with, financial instability. I will provide a snapshot of the progress that has been made and what remains to be done to deal effectively with bank runs and bank failures. I will also explain what we are doing at the IMF to help our membership make further progress in this critical area.
The bank failures in 2023 in the US and Switzerland presented the most significant test since the global financial crisis of the reforms taken collectively to end “too-big-to-fail.” It’s not often that policymakers get to field test plans for dealing with failing systemic banks, let alone one for a global systemically important bank (G-SIB).
In our view, the failures of Credit Suisse in Switzerland and SVB, Signature, and First Republic in the US, showed that while significant progress has been made, further progress is still required to deliver on the too-big-to-fail reform agenda and reduce the risk that taxpayers bail out shareholders and creditors when banks fail.
On the one hand, the actions the authorities took last year successfully avoided deeper financial turmoil. In addition, unlike many of the failures during the global financial crisis, significant losses were shared with the shareholders and some creditors of the failed banks. However, taxpayers were once again on the hook as extensive public support was used to protect more than just the insured depositors of failed banks.
In Switzerland, amid a massive creditor run, the Credit Suisse acquisition was backed by a government guarantee and liquidity facilities nearly equal to a quarter of Swiss economic output. While the public support was ultimately recovered, it entailed very significant contingent fiscal risk, and created a larger, more systemic bank. Indeed, UBS now has the largest ratio of assets to home country GDP of any individual G-SIB.
The use of standing resolution powers to transfer ownership of Credit Suisse, after bailing in shareholders and creditors, rather than relying on emergency legislation to effect a merger, would have fully wiped out the equity of Credit Suisse shareholders and limited the need for public support.
What lessons have we learnt?
Domestic and international authorities have published extensively on the lessons learnt and we share many of the conclusions. The key points I would highlight include:
The importance of intrusive supervision and early intervention. Credit Suisse depositors lost confidence after prolonged governance and risk management failures. The banks which failed in the US pursued risky business strategies and very rapid growth with inadequate risk management. Supervisors in both jurisdictions should have acted faster and been more assertive and conclusive. Policymakers need to empower supervisors with both the ability and the will to act.
Even relatively small banks can prove systemic. A lesson from many past crises, including the US bank failures in 2023, is that you can’t always judge in advance which banking problems will become systemic. In many countries, including the US and Switzerland, we think authorities should do more to be ready for crises affecting their medium-sized banks. Banking supervisory and resolution authorities should ensure that sufficient recovery and resolution planning takes place across the banking sector as a whole. This should include, on a proportional basis, banks that may not be systemic in all circumstances, but that could certainly be systemic in some.
Central banks should be prepared to provide extensive liquidity support during a crisis. Banks should be familiar with the central bank’s operations and facilities and be ready to use them at short notice. Who can access central bank lending is also an important question as liquidity risks have partially moved away from the usual central bank counterparties. While widening the counterparty list could help central banks intervene more broadly in a crisis, it runs the risk of rewarding regulatory arbitrage, giving raise to difficult trade-offs and requiring careful assessment. Central banks may well have to lend against illiquid collateral in a crisis. In that context, prepositioning would help to ensure operational preparedness especially to ascertain the legal claim on the collateral and to calibrate appropriate haircuts. An open question is whether the prepositioning should be voluntary or required, and how much counterparties should preposition if required. The benefits of enhanced lending “fire power” would have to be compared with the cost that prepositioning entails for the banks and the costs to the central bank, including risks to its balance sheet. If propositioning is directly linked with risk (e.g., a percentage of uninsured deposit), the impact on intermediation and the interaction with other prudential regulation would need to be carefully assessed.
Resolution plans and regimes need sufficient flexibility. We very much support the conclusion of the Financial Stability Board’s lessons learned report that resolution authorities need to “better operationalize a range of resolution options for different circumstances.” Every bank failure presents different challenges and resolution authorities need to be flexible enough to deal with the actual crisis that presents itself, balancing risks to financial stability with those to taxpayers. Authorities should make sure that they carefully balance rules versus discretion and detailed planning versus optionality in designing their resolution regimes. The rapid sale of Credit Suisse should prompt us to think about what would be needed for the successful sale in resolution of even the largest banking groups, at least in some circumstances.
Strikingly, every one of the cases I mentioned from Spring 2023, involved the transfer of the failing bank’s business lines to an acquiring bank, even where this had not been the focus of prior resolution planning. Two of the US cases also involved the intermediate step of transfer to bridge banks. So, we have timely and high-profile reminders that transfer powers should be a core part of the resolution toolkit and should be duly planned for and readily implementable, including at short notice.
Cooperation and effective implementation of resolution powers across borders is imperative. One notable feature of last year’s bank failures was the degree of international cooperation between regulators and resolution authorities in their handling of these cases. The Swiss authorities worked intensively with international counterparts to prepare for a resolution of Credit Suisse, which would have needed supportive actions from the supervisors and resolution authorities responsible for Credit Suisse’s main foreign operations, including in the US, UK, and EU. SVB’s UK subsidiary was resolved by the Bank of England, ultimately being sold to HSBC, and the FSB report highlights that the UK relied on the deep relationships built over the years with their US counterparts to help implement this. This cooperation seems to have begun earlier and worked a lot better than in similar cases during the global financial crisis, such as the failure of Lehman Brothers.
That experience highlights how global financial stability depends on authorities being able to work together across borders and to build in peacetime the routine contacts and good understanding ex ante of what each authority would be likely to do to make that possible. However, there was a wrinkle in this otherwise positive experience, as highlighted in the Financial Stability Board’s report on the bank failures, which relates to the importance of the US securities markets to most major foreign banks. Credit Suisse and most other major banks have debt securities issued in US dollars and/or under New York law, the holders of which may incur losses in a resolution. As a recent report of the Financial Stability Board highlighted, there remain significant open questions about how disclosure and other US securities legal requirements would be applied in the circumstances where securities issued in the US are envisaged to be converted in a short period, for example, over a resolution weekend. This is an important issue where further work is needed and this is being taken forward by the Financial Stability Board, the Securities and Exchange Commission, and others.
Finally, effective deposit insurance regimes are crucial. Banks typically fail when creditors lose confidence, even before their balance sheet reflects potential losses. Authorities in many countries need to strengthen deposit insurance regimes. New technology like 24/7 payments, mobile banking, and social media have accelerated deposit runs. Last year’s failures followed rapid deposit withdrawals, and deposit insurers and other authorities should be ready and able to act more quickly than many currently can.
IMF staff are working actively to support efforts in member countries to strengthen their supervision, resolution, liquidity assistance, and deposit insurance frameworks including through FSAPs and technical assistance. In the US, we have seen lessons learned reports and policy proposals from many of the US banking authorities, several of which pick up on issues and recommendations that were discussed in the IMF’s assessment of the US financial sector (“FSAP”) in 2020. Our next FSAPs for Switzerland and the Euro Area will be published next year, and as we start work on that we will be taking a close look at the authorities’ and the FSB’s findings and will likely reiterate many of our previous findings, including on strengthening deposit insurance regimes. We are also contributing to policy formulation at the international level, including a recently announced review of the international deposit insurance standard, and by earlier this year hosting with the Financial Stability Board a workshop for policymakers on the use of transfer powers in resolution.
The bottom line is that progress has been made, but there is still further to go in putting an end to too-big-to-fail. Most of the areas where further progress is needed are already well known; last year’s bank failures should provide the impetus for policymakers to cover the remaining ground.