The role of central banks in managing financial crises in times of rapid change

Good afternoon. I would like to start by thanking the Central Bank of the Republic of Austria, SUERF, the Joint Vienna Institute and the Yale Program on Financial Stability for inviting me to deliver this year’s SUERF Annual Lecture. It is a pleasure and an honour to be here, and I am grateful to Governor Kocher for the kind introduction.

I am particularly pleased that the conference has brought together such a distinguished group of central bankers, supervisors, market participants and academics. These exchanges are especially valuable today, as many of the issues we face lie at the intersection of monetary policy, fiscal policy, financial markets and regulation. 

My lecture today is about the role of central banks in financial crises. More specifically, developments in fiscal policy, financial markets and innovation are rapidly shifting the landscape around us. In that context, the central question is deceptively simple: how will these shifts affect the role central banks can and should play in future crises?

Lessons learned from recent crises 

Let me start with an overview of what we have learned from the crises of the last two decades.

We have learned once more that central banks play a unique and critical role in managing financial crises. Most recently, we observed this during the Covid-19 pandemic, and before that during the Great Financial Crisis (GFC) and the euro area sovereign debt crisis. Standing and emergency liquidity facilities have provided crucial support to curb pressures at financial institutions and stem fire sales. Emergency lending and asset purchases also helped to restore functioning in core funding and bond markets, supporting the flow of credit to households and businesses. In each case, timely central bank action was key to preventing a shock from spiralling into a threat to financial and price stability. 

We have also learned that managing expectations through credible commitments is a key element of a successful crisis response. In several cases, the mere announcement of future central bank asset purchases was enough to restore market functioning. This announcement effect was on display for central bank purchase programmes in government bonds, commercial paper, asset-backed securities or local government bonds.¹  Several central bank lending operations also succeeded at supporting credit supply while being used only lightly by banks.²  In those cases, credible central bank commitments, rather than outright use of the central bank balance sheet, helped to restart markets. 

At the same time, each success in managing recent crises revealed challenges and limitations. 

One challenge that became apparent is how to design asset purchase programmes. Only a few major central banks purchased assets during the GFC. These programmes were used more extensively during the pandemic.³  Government bond purchases addressed market dysfunction and prevented a broader tightening in financial conditions. However, uncertainty about the nature and duration of the shocks complicated their calibration and communication. Some interventions initially designed to address temporary market dysfunction also morphed into longer-term stimulus programmes.⁴  Blurred lines between objectives made it more challenging to adjust and unwind purchases as market functioning improved.  

Another challenge that emerged, first for a few central banks during the GFC, and then more broadly during the pandemic, was how to design lending interventions beyond the markets and intermediaries that usually have access to central bank liquidity. During the pandemic, a larger number of central banks directed interventions at borrowers, or at markets or non-banks critical to the supply of credit to these borrowers. This expansion of the traditional lender of last resort role reflected the nature of the Covid-19 crisis, which originated in the real economy before spilling into the financial sector, and the growing weight of non-bank financing of the real economy. However, these interventions involved greater credit risk exposure and delicate issues around relationships with government interventions.⁵  

How developments could intensify challenges and limitations in future crises

Let me now turn to the future. We can be certain that the next crisis will not look like the last, but we are not starting from a blank slate. As in previous episodes, central banks will play a central role, and their interventions will follow the very same public interest mandate objectives. However, different aspects of that landscape could shape central banks’ response to future crises. I will focus on three: high government debt, the increasing footprint of non-bank financial institutions (NBFIs), and digital innovation.

High government debt

High government debt might complicate central banks’ role in crisis management. Public debt is near post-war highs in many economies, deficits remain large, and fiscal pressures from ageing and public investment are set to persist. This makes sovereign bond markets more exposed to shifts in investors’ assessment of fiscal risk. 

For central banks, diagnosing the cause of large price movements in sovereign bonds and the appropriate response may be challenging in this context. A rise in sovereign yields may reflect a reassessment of fiscal sustainability or temporary market dysfunction, for instance caused by forced deleveraging. If market dysfunction threatens financial stability or monetary transmission, central banks need to intervene. But when debt is high and public financing needs are large, even a well-designed operation can be interpreted through a fiscal lens. 

The global footprint of non-bank financial institutions

The rising NBFI footprint reinforces the challenge of high government debt for central banks’ crisis management. NBFIs are now the largest holders of sovereign debt in advanced economies, and leveraged players such as hedge funds have become important intermediaries in several core government bond markets. Greater NBFI participation can support market liquidity in normal times. At the same time, some NBFIs’ reliance on short-term and market-based funding strategies and high leverage can make the system more sensitive to liquidity shocks.

The tighter link between sovereign bond market functioning, fiscal risk repricing and financial stability gives rise to a new fiscal-financial stability nexus as described in the BIS 2026 Annual Economic Report.⁶  The traditional bank-sovereign nexus still matters, but stress can now spread at a large scale through funding markets, across borders and between banks and non-banks. Government bond liquidity can dry up suddenly when fiscal news interacts with leveraged positions, margin calls and constrained dealer balance sheets. In that setting, fiscal space can shrink well before any limit implied by long-run fundamentals is reached. 

The new role of NBFIs in amplifying stress in government bond markets was highlighted in recent episodes. In the March 2020 dash for cash, even deep government bond markets, including US Treasuries, came under strain as mutual funds and hedge funds were forced to sell government bonds in response to redemptions and margin calls.⁷  During the UK gilt episode of 2022, margin calls on leveraged pension fund strategies forced asset sales. The triggers differed, but the propagation was similar: market liquidity became fragile when many government bondholders needed cash at once. Crowded positions can further amplify stress. If many funds that follow similar benchmarks and execute similar trading strategies face liquidity shocks at the same time, a price decline can produce simultaneous deleveraging. Dealers may then be less willing or able to intermediate the flow, and market depth can disappear when it is most needed.

The increasing role of NBFIs challenges the long-established central bank frameworks for emergency liquidity provision. Access to central bank liquidity facilities is often restricted to depository institutions. Hence, to address stress centred on NBFIs, central banks cannot rely on lending operations but instead need to rely on asset purchases, which as discussed before may be hard to unwind and price. At the same time, expanding central bank access to NBFIs is complicated by the fact that access is generally conditional on institutions being well regulated and supervised. While prudential frameworks for banks are well established, prudential frameworks for non-banks often remain incomplete. As a result, expecting central banks to step in without closing the regulatory gap, by subjecting similar risks to comparable regulation, liquidity safeguards and resolution tools, is highly problematic, as it could create moral hazard, encourage regulatory arbitrage and produce an uneven playing field across institutions. 

Digital innovation

Finally, ongoing digital innovation will also affect how stress builds and how central banks may need to respond. 

Digital innovation in banking and reliance on social media by depositors and investors can accelerate liquidity pressures, and this acceleration could increase the speed at which central banks must intervene. Depositors and investors can now move funds instantly across platforms, and information can spread globally at a much faster speed. The speed of deposit withdrawals observed during the 2023 turmoil was unprecedented, as digital online banking likely intersected with the speed of information dissemination via social media.⁸  Beyond speed, false or partial information can travel through social media, and official messages may arrive after market narratives have already formed. Confidence can become more brittle, making future crises harder to stabilise in their earliest phase.

The growth of stablecoins could also raise questions about their access to central bank liquidity backstops and therefore about commensurate stablecoin regulatory frameworks. Stablecoins do not settle on central bank balance sheets, can deviate from par and depend on redemption terms, reserve assets and the reliability of the ledgers on which they circulate. If stablecoins sell reserve assets rapidly, stress could transmit to money markets, and central banks might need to intervene to avoid a broader tightening in financial conditions. The risk is still modest today, but it could become more relevant if stablecoins grew as payment instruments or as stores of value. 

Finally, widespread use of similar artificial intelligence (AI) models by financial market participants may increase co-movement and speed in market reactions. We have seen a version of this with algorithmic trading, where automated strategies reacted to similar signals and amplified price movements. With AI, the concern could be broader-based. Similar tools may be used not only for trading but also for risk management, liquidity planning and portfolio allocation by a broad range of market participants. This might increase herding behaviour in financial markets and might amplify price movements when shocks hit. 

How policymakers can address these challenges

The challenges I’ve set out call for a multilayered policy response. Central banks have a key role to play, but so do regulators and governments. 

Fiscal policy

Fiscal discipline is a key prerequisite. Public finances need a credible sustainable path, and fiscal policy should build buffers in good times. Such discipline reduces the risk that rapid public debt issuance disrupts market functioning and compels central banks to intervene. And if central banks determine intervention is still needed, fiscal discipline lowers the risk that interventions are perceived to be driven by fiscal objectives.

Regulation 

Financial regulation also has an important role to play, as it can reduce the need for interventions as well as contain the risks that central banks’ support ex post may breed excessive ex ante risk-taking.

The post-GFC reform agenda has considerably strengthened the banking system. Banks are now much better capitalised, and they hold larger liquidity buffers. But this does not mean we can lean back. As I mentioned in another speech this year, it is important to review regulatory frameworks to make them more efficient while also tackling new emerging risks.⁹  Some measures could reduce the need for central bank emergency interventions, for example improvements in the calibration of deposit outflows to account for faster runs or in the operational readiness of banks to access central banks’ liquidity provision facilities.¹⁰

By contrast, reforms to ensure that NBFIs do not undermine financial stability have been on the policy agenda for several years, but further progress is still needed.

As we argued in our Annual Economic Report this year, a guiding principle for regulatory design should be to pursue “congruent regulation”. This would imply regulatory frameworks that apply similar stringency to financial intermediaries pose similar risks to financial stability, regardless of their legal form or business model. While a useful guiding principle, designing congruent regulatory frameworks is challenging in practice. Gauging contributions to systemic risks is inherently difficult, the more so given the heterogeneity of NBFIs and their business models. That said, a range of activity- and entity-based tools can be deployed to move towards congruent regulatory treatment. Implementation is also challenged by a fragmented institutional architecture that often assigns, even within a jurisdiction, the authority to design regulatory frameworks to multiple authorities in charge of different NBFI sectors. Greater cooperation at the national and international level would be important to overcome this gridlock.

Notwithstanding these challenges, our Annual Economic Report highlighted several regulatory proposals that merit consideration. In the context of the fiscal-financial stability nexus I discussed earlier, measures to address systemic vulnerabilities in government bond markets can be particularly important. One proposal is greater use of central clearing for cash and repo markets. Central clearing can encourage intermediation activity and make the market less fragmented and more resilient. However, central clearing also brings its own financial stability risks, including by increasing the systemic importance of central counterparties. Another proposal is to impose minimum haircuts, which would limit the build-up of leverage. But minimum haircuts should also be applied in a targeted manner. A uniform minimum could inadvertently favour one side of the trade, ie the cash or collateral lender.

Central bank backstop facilities

To enhance the effectiveness of the backstop facilities, central banks are currently considering development of new emergency lending tools or expanding access to existing lending facilities. 

The key overarching consideration for the design of any such facility is to satisfy the backstop principle: namely, that central banks play a role in preventing market dysfunction from harming the real economy. At the same time, these interventions and backstop facilities should not disrupt price discovery or risk management during normal times, ensuring that markets remain self-reliant, reinforced through appropriate macro- and microprudential regulation and supervision.¹¹

Implementing the backstop principle in practice is difficult. For lending facilities aimed at emergency use, the principle implies that the pricing of these facilities should become unattractive as conditions normalise, but it should be sufficiently attractive during stress to avoid stigma that would make firms inefficiently hoard liquidity. A pricing schedule announced in advance can help, but it may be hard to maintain when market conditions move quickly and in an uncertain manner. 

Applying the backstop principle to asset purchases aimed at restoring market functioning is equally challenging. In principle, central banks should buy assets at a penalty price that is below that prevailing in normal times and attractive in times of stress to improve the market dysfunction. In practice, however, separating price movements caused by fundamentals from those driven by market dysfunction can be very challenging. This is particularly true during stress episodes, as dysfunction is often triggered by news about fundamentals, including fiscal shocks.

Another key difficulty is how asset purchases for market functioning interact with monetary policy. When market dysfunction coincides with above-target inflation, upholding the backstop principle is key to avoid blurring lines with monetary policy. The Bank of England’s intervention in response to the liability-driven investment (LDI) crisis offers a blueprint. Strictly limited purchase windows and amounts as well as clear communication and governance mechanisms worked well. But in cases of wider and more persistent shocks, such commitment might not be credible. 

When market stress coincides with negative news about the economic outlook, for instance as seen in March 2020, designing asset purchases involves distinct challenges. Calibrating the amount, duration and structure of asset purchases can be difficult. A key challenge is to bolster market confidence without overstimulating the economy in the longer term.

Several options exist to mitigate this risk. Central banks should try to better distinguish programmes for market functioning from those for monetary stimulus, so each can be designed to accomplish these specific goals and be communicated as such. Carefully designing programmes ex ante can increase central banks’ ability to respond quickly in crises while lowering risks of confusion about programme objectives and other side effects.¹²  For example, quantitative easing programmes can be defined in a more state-contingent manner, by clarifying exit conditions, adding a degree of reversibility in purchase targets and more clearly outlining the potential costs and benefits of such measures.¹³

Strong global cooperation

As a representative of the BIS, I cannot end without mentioning the importance of global cooperation. The risks I have discussed are common to many countries, and their crystallisation can spill across borders. Cooperation is essential in three dimensions.

First, global cooperation on standard-setting must continue, including for the regulation of banks, NBFIs and digital money. Closing data gaps, in particular for NBFIs, is an important prerequisite for coherent regulation and effective oversight. 

Second, cooperation is a key part of crisis management infrastructure. In the context of the risks related to internationally active NBFIs’ significant reliance on FX swaps, central bank swap lines remain critical to stabilise the global financial system at times of acute distress. 

Beyond that, the exchange of information and data and coordination on announcements can all play a key role in making crisis management more effective.

The BIS has a unique role to play in all three dimensions. It offers a global analytical anchor, integrating microprudential and macroprudential perspectives. And it provides the convening benefits of standard-setting bodies and global forums to help foster dialogue, cooperation and capacity-building across jurisdictions.

References

Bank for International Settlements (BIS) (2024): “Monetary policy in the 21st century: lessons learned and challenges ahead”, Annual Economic Report 2024, Chapter II.

——— (2025): “Financial conditions in a changing global financial system”, Annual Economic Report 2025, Chapter II.

——— (2026): “High public debt and shifting financial markets: challenges for central banks”, Annual Economic Report 2026, Chapter II.

Basel Committee for Banking Supervision (BCBS) (2023): Report on the 2023 banking turmoil, October.

Boyarchenko, N, A Kovner and O Shachar (2022): “It’s what you say and what you buy: a holistic evaluation of corporate credit facilities”, Journal of Financial Economics, vol 144, no 3.

Chavaz, M, D Elliott and W Monroe (2025): “A public-private partnership? Central bank funding and credit supply”, Bank of England Staff Working Papers, no 1,161.

Chavaz, M, A Schrimpf and F Smets (forthcoming): “Central banks as buyers of last resort”, BIS Quarterly Review.

Chavaz, M and F Smets (2025): “Quantitative easing at 25: time for a makeover?”, in D Aikman and R Barwell (eds), The UK Monetary Policy Roundtable, National Institute of Economic Research.

Coelho, R, M Drehmann, R Walters and D Murphy (2024): “Navigating liquidity stress: operational readiness for central bank support”, FSI Insights on policy implementation, no 64.

Coelho, R and F Restoy (2025): “Rethinking banks’ liquidity requirements”, FSI Briefs, no 25.

Committee on the Global Financial System (CGFS) (2023): “Central bank asset purchases in response to the Covid-19 crisis”, CGFS Papers, no 68. 

Eren, E and P Wooldridge (2021): “Non-bank financial institutions and the functioning of government bond markets”, BIS Papers, no 119.

Hernández de Cos, P (2026): “Streamlining financial regulation while safeguarding stability and tackling new risks”,  speech at the International Center for Monetary and Banking Studies, Geneva, 4 March.

Hernández de Cos, P, K Forbes and T Tombe (2024): “External comments on the review of the Bank of Canada’s exceptional policy actions during the pandemic”, 17 December. 

Markets Committee (2022): “Market dysfunction and central bank tools”, Markets Committee Papers.

Minoiu, C, R Zarutskie and A Zlate (2026): “Motivating banks to lend? Credit spillover effects of the Main Street Lending Program”, Journal of Monetary Economics, vol 158, 103897. 

Schrimpf, A, V Sushko and H S Shin (2020): “Leverage and margin spirals in fixed income markets during the Covid-19 crisis”, BIS Bulletin, no 2.