M-RCBG Associate Working Paper No. 229
Regimes for Lender of Last Resort Assistance to Illiquid Monetary Institutions: Lessons in the Wake of Credit Suisse
Sir Paul Tucker
Introduction
This report has been commissioned as part of the efforts of the Swiss government to learn lessons for its banking regime from the failure and rescue of the Credit Suisse group in March 2023. The mandate was not to investigate, with complete access to private information, the precise circumstances and timeline of the bank's failure and rescue. Rather, it addresses the design and application of lender of last resort (LOLR) policies and operations in general and for Switzerland, and aims to do so in ways that might be useful to those who know why they acted and communicated as they did.
From the outside, the salient features of Credit Suisse's demise are as follows. The group appears, on published numbers, to have been comfortably solvent. Instead, the group seems to have fallen apart for essentially reputational reasons, causing customers to cease using it, which involved their transferring deposits away in ways that spiraled into a full-blown liquidity crisis. The idiosyncrasy of the firm’s malaise fits with the lack of contagion to other parts of the financial system. For some reason(s), the authorities did not arrest the decline in time, or provide funding to bridge to other conceivable solutions. Instead, the group was sold to UBS, assisted by backing from the state and the write-off of certain kinds of bond. It can seem —- again, I stress, from the outside —- as though the Swiss regime for banking stability was deeply flawed, inadequately implemented, or constrained by operational choices made years before.
Other speculations have focused, more narrowly, on an apparent lack of eligible collateral against which the Swiss National Bank (SNB) could have tried to facilitate other possible solutions. There is hint of this in what the Swiss have termed "ELA+" which involved the central bank lending to CS not, in the usual way, against specific assets but unsecured with, thanks to an emergency government ordinance, preferred rights in a bankruptcy. That is quite something. But the improvisation points to more concrete questions. Did CS simply not have much good collateral (putting the assertions about solvency into doubt)? Or was it a matter of the central bank being unable or, differently, unwilling to lend against some types of good collateral? And why was any collateral-shortage problem not identified and addressed years before? While, again, it is not my mandate to dig into exactly what happened, it is possible to identify weaknesses in the design and operation of the Swiss banking regime that bear on those questions.
Without getting into the innards of CS affair, then, what is discussed is relevant to whether different policies for lending of last resort could have helped in several ways. First, a different LOLR crisis-management regime might have enabled an alternative solution. These include the SNB being better able either to help fund an orderly run down (with public sale of parts of the business and perhaps a private auction of others), or to support a special resolution with liquidity if, despite appearances, the group was in fact fundamentally insolvent. Given the nature of this exercise, it is impossible to reach a definitive view on that. Notably, I cannot judge whether the unravelling of the business model precluded other courses once collapse engulfed the authorities. But a better LOLR regime would, second, have changed the incentives for planning, preparation, and pre-emption --- at the SNB as well as at the Financial Market Supervisory Authority (FINMA). Even if none of the circulating speculations about the handling of the crisis hits the mark, those broader LOLR-regime lessons are relevant for the years before policymakers found themselves concluding, in the heat of events, that their best option was to have the whole CS group transferred to UBS.