Deposit stability: what should policymakers look at?

Remarks by Mr Fernando Restoy, Chair, Financial Stability Institute, at the FSI-IADI event on "Deposits in the spotlight - Do we need to rethink the treatment of deposits?", webinar, 25-26 October 2023.1

BIS, FSI speech  | 
27 October 2023


Hello and welcome to the second session of the FSI-IADI webinar on the treatment of deposits. As always, it's a pleasure to cooperate with IADI. We organise at least one joint event annually, and we always try to focus on a topical theme relating to bank failure management. Following the events of March, the topic for this year's event suggested itself: the treatment of bank deposits, and in particular uninsured deposits, within the policy framework and the financial safety net.

I am referring, of course, to the failure and resolution in early March of two US regional banks – Signature Bank and Silicon Valley Bank (SVB) – followed rapidly by the near failure of Credit Suisse and its state-sponsored acquisition by UBS.

The recent events

I am sure you are all aware of the narrative, of how these banks failed and the measures taken by the authorities to manage those failures, so I won't reiterate them in detail.

While those banks were quite different, their failure followed a broadly common pattern. Although all failing banks satisfied minimum solvency requirements, market concerns about their viability provoked sharp corrections in equity prices. In a few cases, investors identified an unsustainable exposure to interest rate risks, even when existing regulation allowed banks not to fully reflect the negative impact of the recent sequence of interest rate hikes on asset economic values in their balance sheets and reported capital.

Market corrections affected weak banks globally and triggered deposit outflows in a number of banks in different jurisdictions. However, for some US banks, they provoked massive deposit runs of an unprecedented speed and intensity. Those banks that were more significantly affected were characterised by large holdings of uninsured deposits. The figures are worth emphasising. Uninsured deposits represented 89% of the total deposit base in Signature bank and around 94% in SVB.

In all cases, most withdrawals took place electronically, so there were no pictures of queues outside bank branches. But social media replaced the traditional iconography of a bank run, and confidence was rapidly lost.

Effective interventions by authorities stabilised the situation in both the United States and Switzerland. But it is worth noting that, in each case, while no depositors suffered losses, others were not protected. In the US, shareholders and unsecured debtholders of Signature Bank and SVB will suffer losses. In Switzerland, the acquisition of Credit Suisse was accompanied by the contractual writedown of all its outstanding Additional Tier 1 instruments, amounting to an extinction of CHF 16 billion in liabilities from its balance sheet.

So from one perspective, the turmoil showed that crisis management frameworks were effective. Markets have since stabilised and there was no "domino effect" with further bank failures. Nevertheless, these events have prompted a range of policy questions about the adequacy of the prudential regulatory and supervisory framework and the credibility of resolution regimes. For example, the FSB report published earlier this month2 discusses the preliminary lessons learnt for resolution from the 2023 bank failures.

But let me share a few remarks on the topic for today's session – what lessons for deposit stability and depositor protection we might take away from these events.

The quest for deposit stability

Arguably, a sufficient degree of deposit stickiness is a necessary condition not only for financial stability but, more fundamentally, for the sustainability of the commercial bank business model.

Indeed, much of the policy intervention in banking has been traditionally geared towards minimising the probability and impact of bank runs on the financial system. This is the very reason why deposit insurance was created. In addition, the establishment of prudential regulation and supervision for banks – which was actually born at the same time as deposit insurance in the mid-19th century in the US – is meant to reduce the probability of liquidity or solvency crises and, therefore, to strengthen the trust of clients and other stakeholders.3 More recently, bank failure regimes have been put in place to facilitate the orderly resolution of banking crises, minimise their cost and maximise the preservation of the critical functions of failing banks (including access to client deposits).

Therefore, if we consider that the massive outflows of deposits that took place in a matter of hours in those cases in March indicate a structural reduction in the expected degree of deposit stickiness, this has implications for us as regulators and policymakers. Specifically, some reflections on the role of those aspects of the policy framework that aim to protect banks' stability in going concern and an orderly resolution in gone concern are of course warranted. Let me briefly review some of those aspects.

On deposit insurance

Logically, the first and most direct instrument for protecting deposit stability is deposit insurance. The scope of deposit protection clearly affects the volume of runnable liabilities in the form of uninsured deposits. Of course, deposit insurance coverage should be, and indeed is, reviewed regularly. In the US, for example, the maximum protected amount is now 100 times larger than it was when the FDIC was created in 1933.

Some authorities are currently re-evaluating the scope and level of coverage offered by the deposit insurer in their jurisdictions, to assess whether wider coverage or a higher limit would offer significantly greater benefits in terms of financial stability.

Preliminary indications suggest that the case for expanded coverage is not clear-cut. Moderate changes to coverage levels are unlikely to substantially alter the volume of uncovered deposits, which are concentrated with a very limited number of depositors. More substantial changes could have material implications for the size of deposit insurance funds and, by extension, the costs to the banking sector. Moreover, a substantial increase (let alone universal coverage) would raise the usual concerns about moral hazard, redistribution of resources from sound banks to their weaker competitors, and the distortions this could create in the allocation of resources in the financial system.

Whatever changes may be made to coverage levels, I would not expect unlimited coverage, so there will always be the possibility of uninsured deposits.

On the prudential regime

The prudential regime may have a more prominent role to play. Indeed, a bank's failure is not a random event but, mainly, the result of its own vulnerabilities. Undoubtedly, more stringent prudential requirements and oversight could help reduce the likelihood of bank runs – even in the new technological environment.

Some policymakers are considering specific actions to limit the risks of a destabilising run by uninsured deposits. For example, some options would aim to make uninsured deposits a more costly form of funding for banks. That could be achieved by recalibrating the current liquidity requirements. Further-reaching reforms which have been proposed include the introduction of "speed bumps" to slow down the rate at which uninsured deposits may be withdrawn or asking banks to fully collateralise all their uninsured deposits.4 However, such measures would require a very careful cost-benefit analysis as their most immediate effect would be to further restrict – potentially severely – banks' ability to perform their key intermediation function.

Indeed, pending further analysis and evidence on the scale of a possible structural reduction on the degree of deposit stickiness, I am more confident about the role that strengthened supervision could play in preserving banks' stability than I am about the net benefits of introducing much more stringent liquidity or capital requirements. After all, the ultimate cause of the recent bank failures on both sides of the Atlantic was unsustainable business models. Higher capital or liquidity requirements cannot, by themselves, restore banks' ability to remain profitable on a sustainable basis when the business model is fundamentally flawed. Instead, effective supervision is needed to identify and challenge the poor strategic decisions, governance and risk management that lead to weak business models.

On bank failure management

A third relevant piece of policy intervention with implications for deposit stability is the bank crisis management framework. Here, key aspects to improve in many countries include emergency liquidity support and the arrangements for the provision of liquidity for banks in resolution.

Moreover, given the importance of the stability of uninsured deposits for the financial system, bank insolvency regimes should provide a preferential treatment for those deposits relative to less sensitive liabilities. Extending minimum requirements for liabilities which are junior to deposits to any potentially systemic bank – and not only to G-SIBs – would contribute to achieving that goal. This is currently the case in the European Union through the minimum requirement for own funds and eligible liabilities, or MREL, which applies to all banks under the EU resolution regime. In the US, the regulatory agencies have proposed to introduce minimum long-term debt requirements for all banks with balance sheets above $100 billion.5 Furthermore, despite recent developments, a general depositor preference regime, such as that in the US, where insured and uninsured deposits rank pari passu above other non-preferred liabilities, are better able in most circumstances to preserve the trust of the uninsured depositors than those regimes where uninsured deposits are subordinated to insured deposits, as is currently the case in the EU. This argument provides additional support to the recent legislative proposal by the European Commission on crisis management and deposit insurance 6 under which the EU would adopt a general depositor preference rule in insolvency.

Final remarks

Let me finish. I hope my general remarks serve as an introduction to the discussion that is going to take place in today's session. I am sure speakers will consider some of the issues that I have touched upon in more detail, and I think we will have a range of opinions.

1 I am grateful to Rodrigo Coelho and Ruth Walters for helpful comments on an earlier draft.

2 Financial Stability Board, 2023 Bank Failures: Preliminary lessons learnt for resolution, 2023, 2023 Bank Failures: Preliminary lessons learnt for resolution. See also A Carstens, "Some lessons for crisis management from recent bank failures", speech at the High-level meeting on banking supervision of the Association of Supervisors of Banks of the Americas, Basel Committee on Banking Supervision and FSI, 19 October 2023.

3 See F Restoy: "The quest for financial stability", SUERF Policy Note No. 318, August 2023.

4 See Federal Deposit Insurance Corporation (FDIC), Options for deposit insurance reform, May 2023.

5 See Federal Reserve Board and FDIC, "FDIC Board of Directors approves advance notice of proposed rulemaking regarding resolution – related resource requirements for large banking organizations", press release, 18 October 2022.

6 See European Commission, Reform of bank crisis management and deposit insurance framework, 2023.