Luigi Federico Signorini: Lessons from recent crises

Speech by Mr Luigi Federico Signorini, Senior Deputy Governor of the Bank of Italy, at the BIS-FINMA-SNB Workshop on "Tackling the risks in systemically relevant banks: lessons from March 2023", Basel, 16 May 2024. 

The views expressed in this speech are those of the speaker and not the view of the BIS.

Almost a year ago,¹ the international financial system experienced the most significant episode of banking turmoil since the Global Financial Crisis (GFC). Four mid-sized U.S. banks and Credit Suisse, a global systemically important bank, encountered severe distress more or less simultaneously. Swift and effective supervisory actions, along with extensive public support measures, enabled the U.S. and Swiss authorities to control the situation. More importantly, the international banking system withstood these events with no major consequences, thanks, at least in part, to the strengthened prudential framework developed in Basel over many years after the GFC.

The final Basel III rules marked the conclusion of this regulatory journey. The process involved intense negotiations and some compromises (often reached after difficult discussions), needed to accommodate varying needs and practices. As a member of the Basel Committee for many years, your speaker took part in many of those discussions and contributed in some small way to many of those compromises. By definition, compromises cannot be 100 per cent satisfactory for anyone; I shall come back to certain issues that, in my view, still need fixing. Nonetheless, my opinion is that, on the whole, the outcome of this process was good. Basel III marked decisive progress towards achieving a balanced approach that combines enhanced risk sensitivity with reduced discretionary choices for banks, particularly when using internal models.² The fact that the financial system has since withstood repeated shocks and (as they say) has been part of the solution rather than part of the problem is, in my view, evidence that these reforms have been useful.

Final Basel III implementation

The implementation of the final Basel III standards is advancing in most jurisdictions. It is important to maintain momentum and avoid unjustified delays that could lead to fragmentation and loopholes.

After more than two years of negotiations, the process is nearing completion in the European Union, with the new rules expected to take effect at the start of next year. The CRR3-CRD6 texts incorporate the Basel standards into the EU rulebook, while accounting for certain specific EU characteristics. Most deviations aim to give banks time to adapt to more stringent requirements, without compromising the Basel spirit. These deviations will be gradually phased out.

While the Basel standards primarily target large internationally active banks, the EU applies them to all banks, albeit with some simplifications for smaller and less complex institutions. This approach has been in place since the 2010 Basel III standards were implemented and, I think, is appropriate. The 2023 turmoil showed that threats to financial stability can arise not only from very large banks, but also from smaller intermediaries with specific imbalances, if fears spread on the market that other intermediaries might have similar problems. This phenomenon, which I call 'contagion by analogy', goes beyond the factors currently used to identify systemically important banks, such as size or direct interconnectedness.

I have already stressed the importance of a broadly similar and simultaneous application of the final rules in all major jurisdictions. The U.K. and U.S. are also advancing with their national implementation, with rules expected to apply later in 2025. It is to be welcomed that U.S. regulators are thinking about expanding the range of banks subjected to Basel standards. The debate is ongoing; some argue that revising prudential rules and extending Basel rules to smaller banks would hinder credit availability and economic growth. I find this argument unconvincing.

First, the Basel Committee's evaluation of Basel III reforms from 2011-21 shows that, as banks increased their capital and liquidity reserves,³ and market-based systemic risk measures improved, the banking sector became more resilient and less vulnerable to distress while maintaining its key functions under stress. The analysis found no evidence of adverse effects on banks' capital costs and lending. In fact, the banks most affected by the reforms saw a reduction in their capital costs.

Second, smaller banks matter. The failed U.S. banks were not required to comply with the most rigorous Basel III standards. Despite their assumed 'regional' nature, their crises raised global concerns. While stricter regulatory requirements might not have prevented the crisis (I don't know), at the very least they would have raised a warning about these banks' condition long before their distress. For instance, in the case of SVB, the short-term liquidity requirement would have rung alarm bells in plenty of time.

The way forward

It is challenging to pinpoint the specific causes behind a crisis. For the U.S. regional banks, several factors contributed: unsustainable business models, poor management, too rapid growth, overreliance on uninsured deposits, and concentration in loans and funding. Credit Suisse's crisis had more complex causes, including large losses from the Archegos and Greensill cases, governance weaknesses, and repeated strategic changes that eroded market confidence.

Despite the varying causes, the commonalities reveal potential areas for improving the regulatory framework. In all cases, a rapid loss of client and market confidence occurred. All the banks involved faced substantial liquidity outflows, amplified by social media and facilitated by digital banking.

One key area to think about is therefore liquidity regulation. Let me start with a very basic observation.

Why are banks subject to regulation and supervision? The banking business has two key features: leverage and maturity/liquidity transformation. Both are necessary for banks to perform their function; both, however, entail a degree of fragility and the risk of spillovers and contagion, with potentially destructive effects that go far beyond the shareholders of individual institutions. Bank regulation is there to limit the system's exposure to such risks, without hampering its fundamental function.

This much, I think, is obvious. However, the fact is that, for some reason, international regulatory standards have long concentrated on the first cause of fragility, leverage, by requiring e.g. minimum capital ratios; such requirements have been at the centre of global standards since the start, with the first Basel accord, and have gradually been strengthened over time. The second cause of fragility, maturity transformation (and, relatedly, liquidity), were not covered at all in Basel I or II, and only started to appear in Basel III in the form of liquidity standards and Pillar-2 provisions for the interest rate risk in the banking book (IRRBB). This is strange in a way: it is, after all, common wisdom that bank crises are, more often than not, due to liquidity problems.

Basel III thus made progress; but the compromises I mentioned were, in my view, not fully satisfactory in this area.

Specifically, there appears to be room for reconsidering the Net Stable Funding Ratio (NSFR). The current NSFR's one-year horizon means that it cannot fully capture a bank's structural exposure to maturity mismatch. In terms of the NSFR, a 366-day maturity is the same as a 30-year maturity. This is not just a theoretical concern. In the case of SVB, despite extreme maturity imbalance in the bank's balance sheet, the NSFR, had it been applied, would not have detected any anomaly.

One possible way to address this weakness would be an 'NSFR+', based on a more granular maturity classification of assets and liabilities. Many years ago, the Bank of Italy used to have a rough-and-ready maturity transformation rule to prevent extreme cases. One could build on such experiences.

The LCR fared better in the SVB case (as I said, it would have signalled potential distress well in advance)⁴. However, remote banking and the effect of news spreading at lightning speed on social networks have greatly increased the potential volatility of zero-maturity bank liabilities. This might challenge the current treatment of sight deposits. Concentration (an issue that is related to uninsured deposits) is also a concern.

Some further reflection on mandatory pre-positioning of collateral would also be in order.

Additionally, two more issues deserve attention.

First, the IRRBB. The Basel standards already include principles for measuring and managing it under Pillar 2, but these may not be adequate under all circumstances. One solution might be to put the IRRBB under Pillar 1. In my view, however, given the flexibility needed to tackle interest rate risk under changing market and individual conditions, it would be better to strengthen the current Pillar 2 approach, allowing for tailored requirements, but under a more robust and structured framework. The SSM already adopts such an approach.

Second, a reassessment of the rules for deposit insurance would be useful. The inherent trade-off is well-known: a share of uncovered depositors instils market discipline, but their volatile behaviour during crises creates systemic danger. A difficult balance must be struck. Nowadays, as a depositor's flight is just one click of a mouse away, the terms of the trade-off may have changed compared with when the rules were first established. The debate is open.

Effective supervision

Whenever a bank crisis occurs, everybody is reminded that rules need to be supplemented by effective supervision. Last year's episodes highlighted once again the need for supervisors to understand banks' business models thoroughly and to ensure effective governance and risk management in banks. The recent review of the Basel Core Principles Effective Banking Supervision is therefore very timely and welcome.

The matter has been amply discussed in previous interventions in this conference and I shall not pursue it in detail. I concur with other speakers that the use – in a non-mechanical way – of market signals and indicators like a rapid growth of assets is important – and, in fact, already widespread. In general, sharpening supervisors' tools is surely needed.

Let me only sound a note of caution on all this. Even with the best principles and tools, supervisors will never be omniscient or omnipotent. With the benefit of hindsight, one can always find a better course of action that the authorities might have taken in a specific case. However, without perfect foresight, and with no waterproof basis in regulatory requirements, it is not always obvious that the authorities would have been in a position to do beforehand, in an analytically and legally robust way, what post factum appears optimal. Supervision must be continuously improved, by all means; but regulation remains fundamental.

Further issues for the future

Finally, looking beyond microprudential rules for banks, let me mention two further issues that the broader system of financial regulation needs to address-and are, in fact, being actively discussed in international fora.

1. Countercyclical macroprudential rules. Disruptive exogenous shocks (the pandemic was an extreme case) require releasable capital buffers. The Basel Committee has endorsed the concept of using cycle-neutral positive buffers to absorb systemic shocks and avoid procyclicality. Many European countries, including Italy, have introduced this measure with, as far as I can see, minimal costs.
2. Non-bank financial intermediaries (NBFIs): NBFIs' share in financial intermediation, and their interconnections with the banking system, have grown considerably. Credit Suisse's exposure to Archegos and Greensill is just one example of the latter; the Basel Committee's reflections⁵ on it are a valuable starting point for defining better rules for banks. More generally, the FSB's work on NBFIs' impact on financial stability is proceeding, albeit slowly.

To conclude: the 2023 market turmoil has lessons for regulators and supervisors. Banks have proven resilient, but supervisors must remain vigilant. Implementing and updating international standards is crucial to adapting to new challenges. After all, for supervisors there are always 'interesting times' ahead.

---

² The role of internal models has been fundamentally revised. For certain risks, such as the Credit valuation adjustment and Operational risks, rather than questing for perfection, the BCBS simply accepted the idea that models cannot provide meaningful estimates. For other risks, such as Credit risk, the new standards constrain the possibility to achieve capital optimisation by imposing input floors as safeguards against overoptimistic assumptions; moreover, the advanced modelling approach can no longer be used for certain portfolios where data are too few and cannot yield statistically robust estimates. For Market risk, the BCBS fundamentally revised the approach to modelling to make it more consistent with the complexities of modern trading activities. An output floor was set as a safeguard against model risk.

³ The weighted average CET1 ratios of the main global banks included in the Basel monitoring exercise sample almost doubled, from around 7 to around 13 per cent, and the leverage ratio almost halved, from around 3.5 to around 6.5 per cent. The liquidity position improved strongly; the LCR and NSFR were always well above their minima during the same period. See the 'Evaluation of the impact and efficacy of the Basel III reforms' Report, published in December 2022.

⁴ Let me also recall, as the issue of asset valuation is sometimes raised, that in the LCR high-quality liquid assets (HQLA) are marked to market and subject to haircuts.

⁵ See BCBS, 'Newsletter on bank exposures to non-bank financial intermediaries', December 2022.