保释的可信度（英）

by Alessandro Di Stefano, Yvan Lengwiler and KumarRishabh Monetary and Economic Department June2026 JEL classification: G21, G28, G01, G33, G14, G12 Keywords: bank resolution, capital regulations, bail-incredibility, Credit Suisse BISWorking Papers are written by members of the Monetary and EconomicDepartment of the Bank for International Settlements, and from time to time by othereconomists, and are published by the Bank. The papers are on subjects of topical This publication is available on the BIS website (www.bis.org). ISSN 1020-0959 (print)ISSN 1682-7678 (online) The Credibility of Bail-In Alessandro Di Stefano†The resolution framework for global systemically important banks has been over Abstract a decade in the making.The failure of Credit Suisse (CS) in March 2023 was itsfirst major test. Authorities had a resolution plan in place but chose a different pathamid financial stability concerns. They facilitated a takeover of CS by UBS, backed bypublic guarantees.Additional Tier 1 (AT1) bonds were written down in full; bail-increditors, who would bear losses next under resolution, were left whole.We studyhow this episode reshaped bail-in credibility across Europe.Using bond-level data 1Introduction More than a decade of regulatory reform set out to ensure that when banks fail, privatecreditors, not taxpayers, would bear the losses. A resolution framework has been built to Credit Suisse (CS), one of the thirty global systemically important banks (G-SIBs), wasfailing. The resolution framework prescribes a loss-absorption sequence: equity and AT1bonds first, then bail-in bonds through formal resolution. For CS, such a plan was in place,and later official accounts confirmed it was ready for implementation (FINMA, 2023; SwissFederal Council, 2024). Market pricing of CS debt was consistent with that expectation. We study how this episode reshaped bail-in credibility.Using roughly 700,000 dailybond-level observations from 94 European banks across 22 countries, we examine howinvestors repriced AT1, bail-in, and senior debt in the year following the event. The episode The probability that bail-in debt bears losses has two components. The first is necessity:the bank must deteriorate to the point where the bail-in layer becomes relevant.The when the situation demands action, not how likely the situation is to arise. Yield spreadson bail-in debt price this probability, so they too reflect both components. The hierarchicalstructure of bank capital allows us to tell them apart. We identify changes in credibility Two results emerge.First, AT1 spreads moved in line with jurisdiction-specific reg-ulatory signals.In Switzerland, where authorities wrote down AT1 in full, AT1 spreadsrose by roughly 11 percent. In the Eurozone and the United Kingdom, they fell roughly5 percent below pre-CS levels, consistent with the hierarchy clarifications issued by EU Senior spreads showed no meaningful movement anywhere, confirming that marketsreassessed how losses are distributed in crisis, not whether banks would fail. But a declinein the perceived probability of bail-in can reflect reduced necessity or reduced credibility.Which interpretation fits the CS episode? The necessity interpretation requires that mar- The cross-section of repricing reinforces this reading. If the decline reflected greater pri-vate acquisition capacity, smaller banks, which are easier to acquire, should benefit most. Instead, the differential spread tightening tracks credit ratings, not bank size: lower-ratedbanks saw disproportionately larger declines, consistent with implicit public support risingwith proximity to distress. Credit default swap (CDS) markets corroborate the shift: the The implications extend beyond pricing. If creditors expect public support, their incen-tive to monitor bank-specific risk weakens. Consistent with this, bond yield responses toearnings announcements declined for bail-in and senior bonds by roughly a third and a These findings highlight a structural tension in the resolution framework.Bail-in ismost important precisely when incentives to avoid it are strongest.The first large-scale Structure.The remainder of the paper proceeds as follows.Section 2 reviews relatedliterature. Section 3 describes the post-crisis loss-absorption regime, the CS episode, and 2Related literature and contributions Bail-in credibility strengthened gradually in the decade following the Global Financial Cri-sis. Berndt et al. (2025) show that implied bailout probabilities for U.S. G-SIBs declined, firm rising bail-in risk premia as the framework matured (Cutura, 2021; Lewrick et al.,2019; Cerasi and Galfrascoli, 2023). But this evidence was built on legislative milestonesand smaller domestic failures. The architecture designed specifically for G-SIBs remained Existing work on bail-in risk pricing typically treats loss-absorbing capital as a singlelayer, interpreting spreads as reflecting the expected incidence of losses.But post