Lessons from the 2023 banking turmoil – a new ASC report

At the ESRB’s Advisory Scientific Committee we just released a new report (joint with Vasso Ioannidou, Enrico Perotti, Antonio Sánchez Serrano, Javier Suarez, and Xavier Vives), titled: “Addressing banks’ vulnerability to deposit runs: revisiting the facts, arguments, and policy options”. This paper is not a post-mortem of the bank failures in the US and of Credit Suisse, but – based on these episodes – a broader discussion of bank fragility and possible policy options. Herewith a short summary.

First, deposit runs are a major source of bank fragility and are almost always the result of a combination of weak fundamentals (concerns about the solvency or liquidity position of a bank or the banking system) and strategic considerations by potentially withdrawing depositors when processing the relevant information (in plain English: no one wants to be the last one to get to the cash machine). The runs in spring 2023 unfolded much more rapidly than previous runs, fuelled by the concentration of the depositor base and by the role of social media and the feasibility of making fast (instant) payments and transfers from the accounts of the affected banks, and thus forced supervisors to intervene much more rapidly than in previous crisis situations.

Second, exposure to interest rate risk (critical for the failure of Silicon Valley Bank) is a direct consequence of banks’ involvement in maturity transformation. While there is evidence for a natural interest rate hedge for banks (Drechsler et al., 2021), due to the deposit franchise, individual banks or business models may fail to be effectively hedged against interest rate risk, as the bank failures witnessed in the US in early 2023 illustrate. These failures also show that the franchise and hedging value of deposit funding is vulnerable to any force that leads depositors to withdraw their funds or suddenly requires banks to pay much higher rates for them to roll over their deposits.

The optimal prudential treatment of interest rate risk might therefore interact with banks’ funding structure. If the system can guarantee the stability of a greater fraction of deposits during crises, the minimum capital that ensures that a bank stays solvent declines. Alternatively, imposing a minimum fraction of highly liquid assets provides an immediate buffer to accommodate outflows without having to sell longer term or less liquid assets, but holding more liquidity reduces banks’ expected net interest rate income and may result in a reduced capacity to accumulate loss-absorbing capacity. In sum, the existence of multiple margins along which interest rate risk interacts with banks’ funding structure calls for considering it in conjunction with the safety guarantees and the capital and liquidity positions of each bank rather than with a single dimensional tool.

Finally, we analyse a number of policy options. The first list of categories includes options that could be further considered without major structural changes in the current regulatory and supervisory framework, and might be implemented in the form of adjustments within the margins of discretion of Basel III:

1. Enhancements in the supervision of bank liquidity and funding positions (e.g. increasing the frequency of monitoring of the banks found to be more vulnerable to runs).
2. Amendments to liquidity requirements (particularly, concerning run-off rates of uninsured deposits, which have been calibrated to previous bank run episodes).
3. Amendments to capital requirements (e.g. concerning the level of the requirements or the treatment of interest rate risk in the banking book).
4. Amendments to the pricing of deposit insurance (e.g., making it more risk-based, including concerning depositor concentration).
5. Enhancing going concern recapitalization capacity (e.g., enabling early recapitalisation through the timely conversion of contingent convertible debt) to thus allow earlier intervention and giving supervisors enough time to successfully and effectively turn around a bank, given how difficult resolution (especially bail-in) has proven to be applied.

The second list of categories includes those with policy options that would imply deeper structural transformations of the institutional setup or the banking industry and that either we do not promote as the most desirable or would require further analysis:

1. Narrow banking – an eternal policy proposal that would undermine the intermediation capacity of banks and simply shift risk-taking outside the banking sector
2. Prepositioning of collateral at the central bank, as proposed by Mervyn King, which has the disadvantage that it might interfere with central banks’ monetary policy mandate
3. Tightening convertibility conditions for uninsured deposits (e.g., charges or gates), which would, however, change dramatically the nature of bank deposits as money-like instrument.
4. Extending deposit insurance coverage (e.g., to cover larger balances from all or specific classes of depositors), which would, however, undermine market discipline further.
5. Extending mark-to-market accounting to broader asset classes in the banking book; a gain, this has been a long-standing discussion and currently there is no consensus on the convenience to promote full (or as much as possible) mark-to-market accounting in banking.

In summary, we see the first list of policy options as input into ongoing discussions at regulatory and supervisory authorities in Europe, while the second set of policy options might be the basis for further (currently more academic) discussions.