UBS's CHF 3 Billion Credit Suisse Rescue and the AT1 Wipeout

A 167-year-old bank that ran out of trust, not capital

At the close of trading on Friday, March 17, 2023, Credit Suisse Group AG was worth about CHF 7.4 billion on the Swiss exchange. Its balance sheet said something very different. The bank had closed 2022 with CHF 45.1 billion of shareholders' equity, a common equity tier 1 ratio of 14.1%, and a public statement from its own regulator four days before the rescue that it met every capital and liquidity requirement imposed on a systemically important Swiss bank. By Sunday evening it had been sold to UBS for CHF 3 billion, roughly $3.25 billion, in shares.

Nothing in Credit Suisse's regulatory capital had failed. What failed was the willingness of its clients to leave money there. That distinction is the hinge of the entire case, because a bank that is short of capital and a bank that is short of depositors call for completely different legal machinery, and Switzerland reached for machinery designed for neither.

The CHF 16 billion that paid for the weekend

The most consequential number in the transaction was not the price. Late on Sunday, March 19, the Swiss Financial Market Supervisory Authority, FINMA, ordered Credit Suisse to write approximately CHF 16 billion of Additional Tier 1 bonds down to zero, permanently. Shareholders, who sit below bondholders in every textbook ordering of a capital structure, kept something. The bondholders above them received nothing.

That inversion, executed under an emergency ordinance the Swiss Federal Council enacted the same evening, is what turned a rescue into an argument that is still running in the Swiss courts more than three years later. This study follows three threads to the end: what actually drained Credit Suisse and how fast; what the Swiss state put behind UBS to make a buyer say yes; and whether the price UBS paid was a market clearing level, a transfer of value engineered by the state, or both at once. The facts come from the parties' own releases, the regulators' filings, the Swiss parliament's inquiry, and the court judgment that later declared the central act of the rescue unlawful.

Spygate and the end of the Thiam era

Credit Suisse did not fail in a week. It spent roughly a decade converting a franchise built since 1856 into a series of headlines, and the pattern that mattered was not the size of any single loss but the repetition. The 2019 surveillance affair, universally known in Zurich as spygate, was the moment the damage moved from the trading floor to the executive suite. The bank hired private investigators to follow Iqbal Khan, a departing wealth management head who had defected to UBS, and the operation unraveled publicly on a Zurich street.

FINMA's later investigation found that the bank had planned seven observations between 2016 and 2019, most of which were carried out, and imposed new supervisory measures on the bank. Chief operating officer Pierre-Olivier Bouee left, and in February 2020 chief executive Tidjane Thiam resigned, saying he had known nothing about the surveillance. The financial cost was trivial. The governance signal was not: a bank that spied on its own bankers was a bank whose internal controls could no longer be assumed to work.

Archegos and Greensill: two collapses, one month

March 2021 delivered the proof. Within weeks, Credit Suisse was forced to freeze about $10 billion of supply chain finance funds when Greensill Capital collapsed after losing its insurance cover, funds the bank had marketed to clients as low risk precisely because the underlying credit was said to be insured. Days later, the implosion of Archegos Capital Management cost Credit Suisse roughly $5.5 billion, a loss its own commissioned report attributed to a focus on short-term profit and a failure to restrain a single client's leverage.

Both events shared a mechanism rather than a market. In each case the bank had underwritten concentrated exposure it did not monitor, to a counterparty it did not fully understand, in a business line that generated modest fees. Rivals took losses on Archegos too, but none took a loss of that size, and none was simultaneously explaining a $10 billion fund freeze to its own private banking clients. For a wealth manager, that is the worst possible combination: the people whose money you lost are the people you need to keep.

The Swiss parliament's own inquiry later put the aggregate in a single line. Between 2010 and 2022, the commission found, Credit Suisse booked losses of CHF 33.7 billion while paying out CHF 31.7 billion in bonuses. The bank did not merely lose money; it distributed compensation as though it had not.

October 2022: the quarter the depositors left

The run that killed Credit Suisse began five months before the rescue weekend and had nothing to do with an American bank failure. In early October 2022, social media speculation about the bank's solvency circulated ahead of a strategy announcement, and clients acted on it. Net asset outflows in the fourth quarter of 2022 reached CHF 110.5 billion, against CHF 1.6 billion in the same quarter a year earlier, with roughly two thirds of the flight concentrated in October alone. Customer deposits fell by about CHF 138 billion over the quarter.

That is the context in which every March 2023 event has to be read. A bank with a stable funding base can survive a bad news cycle. Credit Suisse entered March with a funding base that had been leaking since autumn, a CHF 7.3 billion annual loss just reported, and a restructuring plan that depended on clients believing in a turnaround they had already voted against with their money.

March 14: an annual report that could not vouch for itself

The sequence began with a filing problem. Credit Suisse delayed its 2022 annual report after a late call from the US Securities and Exchange Commission about prior-period cash flow statements, and when the document finally appeared on March 14, 2023, it disclosed material weaknesses in the bank's internal control over financial reporting. Management concluded that its internal controls were not effective, and the report confirmed that the outflows of late 2022 had not reversed.

The timing was catastrophic. Silicon Valley Bank had been closed by US regulators three days earlier, and markets were already scanning for the next institution whose funding could not survive scrutiny. A European bank admitting that it could not vouch for its own reporting controls, in the same week, was the worst available disclosure at the worst available moment. Understanding why that sequence terrified counterparties requires reading a bank's balance sheet the way a funding provider does, which is a distinct discipline from ordinary corporate analysis and is why valuing a bank works differently from valuing any other business.

March 15: one sentence from Riyadh

The next day, the Saudi National Bank, Credit Suisse's largest shareholder with a stake just under 10%, was asked by Bloomberg whether it would inject more capital. Its chairman, Ammar Al Khudairy, said it could not, for a reason that was entirely mundane: going above 10% would trigger a regulatory threshold the bank did not want to cross. The market did not hear a regulatory technicality. It heard the largest shareholder declining to fund the bank.

Credit Suisse shares fell about 24% that day and the cost of insuring its senior debt against default blew past 1,000 basis points. FINMA and the Swiss National Bank issued a joint statement the same evening confirming that Credit Suisse met the capital and liquidity requirements for systemically important banks and that the SNB would provide liquidity if needed. That statement is worth holding onto, because four days later Switzerland would rely on the opposite proposition to justify writing off CHF 16 billion of the bank's capital instruments.

Emergency liquidity assistance (ELA)

ELA is a central bank's lender-of-last-resort facility for a solvent but illiquid bank, extended against collateral the bank pledges in advance. It is not a subsidy and not a capital injection: it converts illiquid assets into cash so the bank can meet withdrawals. Its binding constraint is collateral. A bank that has not pre-positioned enough eligible assets with its central bank cannot borrow, no matter how solvent it is, which is precisely the gap the Swiss authorities discovered at Credit Suisse.

March 16: CHF 50 billion that bought a single day

In the early hours of March 16, Credit Suisse announced it would borrow up to CHF 50 billion from the Swiss National Bank under a covered loan facility and a short-term liquidity facility, and simultaneously offered to buy back up to CHF 3 billion of its own debt. The intent was to demonstrate overwhelming liquidity and stop the withdrawals.

It did not work, for the same reason such announcements rarely do. A CHF 50 billion emergency facility is not read as strength; it is read as confirmation that the bank needs CHF 50 billion. The stock rallied briefly and then resumed falling, and by the weekend the outflows had accelerated again. Credit Suisse would later disclose that first-quarter 2023 net outflows reached CHF 61.2 billion, with customer deposits down a further CHF 67 billion, and that the withdrawals were most acute in the days immediately before and after the merger announcement.

The parallel with an earlier weekend rescue is exact on this point. When the Federal Reserve extended an emergency facility to Bear Stearns on a Friday in March 2008, the market treated the loan as evidence rather than reassurance, and the firm was sold within 48 hours; the same reflex is visible in the government-orchestrated fire sale of Bear Stearns to JPMorgan. By Friday, March 17, the Swiss authorities had concluded that Credit Suisse would not be able to open on Monday as an independent institution.

From CHF 1 billion to CHF 3 billion

The Swiss federal cabinet convened for a crisis meeting on Saturday, March 18. The only buyer with a balance sheet large enough, a Swiss license, and the ability to diligence a peer's book in 36 hours was UBS, and both sides knew it. UBS's opening position, reported on Sunday, valued Credit Suisse at around $1 billion, a number the Credit Suisse board rejected outright as an insult to shareholders who had put CHF 4 billion of fresh equity in four months earlier.

The price moved through Sunday to more than $2 billion and settled at CHF 3 billion, but the negotiation that mattered was never about the headline. UBS was negotiating for indemnity, not for a discount. It sought protection against losses in Credit Suisse's non-core portfolio, against litigation it had not underwritten, and against a liquidity call it could not fund alone. Every franc of price concession UBS extracted was small next to the guarantees it obtained, and the state was the counterparty on all of them.

Article 10a: two shareholder votes that never happened

An all-share merger of two listed Swiss banks would ordinarily require approval from the shareholders of both. Neither vote took place. The Federal Council had issued an ordinance on additional liquidity assistance on March 16 and amended it on March 19, and the amended text included a new Article 10a, headed as a deviation from the Merger Act, which allowed the merger to be implemented without shareholder approval on either side.

Emergency ordinance (Notverordnung)

Under Articles 184 and 185 of the Swiss constitution, the Federal Council may issue ordinances directly, without parliament, to safeguard the country's interests or address a serious disturbance to public order. Such an ordinance has the force of law immediately and is time-limited. Switzerland used this power to write a legal framework specific to one transaction over a single weekend, which is why the resulting deal was legally certain on Monday morning and legally contested for years afterward.

The practical effect was to hand UBS something no acquirer normally gets in a distressed situation: complete deal certainty at signing. There was no vote to lose, no shareholder litigation that could unwind the exchange ratio, and no window for an interloper. UBS chairman Colm Kelleher was explicit at the announcement about the asymmetry.

Credit Suisse shareholders got a meeting, not a vote. At the bank's final annual general meeting on April 4, 2023, held with police security in the hall, chairman Axel Lehmann faced roughly 2,000 shareholders who had been given no say in the disposal of their company.

What the state put behind UBS

The support package was layered, and the layering is the analysis. The Swiss National Bank announced up to CHF 200 billion of new liquidity assistance on March 19: CHF 100 billion of ELA+, secured by preferential rights in bankruptcy but with no federal guarantee, and a further CHF 100 billion public liquidity backstop, secured by preferential rights and additionally guaranteed by the Swiss Confederation. That sat on top of the CHF 50 billion already drawn from March 16.

Separately, the Confederation signed a loss protection agreement covering CHF 9 billion of losses on a designated portfolio of Credit Suisse non-core assets, but only after UBS had absorbed the first CHF 5 billion itself. SNB chairman Thomas Jordan later told a central banking audience that the peak liquidity actually provided across all facilities reached CHF 168 billion in three currencies, an amount he described as without precedent for a single bank.

The headline CHF 200 billion figure in the SNB release covers only the two new facilities announced on March 19; adding the CHF 50 billion already in place gives the roughly CHF 250 billion of committed liquidity capacity against which CHF 168 billion was actually drawn at peak. Swiss finance minister Karin Keller-Sutter nonetheless drew a hard line on how the package should be described.

That framing was contested from the first hour, and it did not survive contact with the Swiss parliament. In an extraordinary session in April 2023, the National Council voted 102 to 71 to reject the CHF 109 billion of federal guarantees, a rebuke that was symbolic only because a parliamentary finance delegation had already approved the commitments and the money ran through the central bank rather than through a budget line.

Master the mechanics behind a deal answer: practice 1,000+ technical questions on M&A, financing, and accretion/dilution, download our iOS app for the full toolkit.

Why FINMA said the trigger had fired

Additional Tier 1 instruments exist to absorb losses while a bank is still alive. Credit Suisse had twelve publicly issued AT1 instruments outstanding, sold between December 2013 and June 2022 in dollars, francs and Singapore dollars, with coupons ranging from 3.000% to 9.750% and a combined nominal value of roughly CHF 16 billion. The Federal Administrative Court would later put the figure at CHF 16.5 billion.

Additional Tier 1 (AT1) capital

AT1 instruments, often called contingent convertibles or CoCos, are perpetual bonds that count as regulatory capital because they either convert into equity or are written down when a defined trigger fires. They pay high coupons precisely because the investor is underwriting that trigger. European AT1s typically convert into shares; Swiss AT1s of this vintage were written down permanently instead, a design difference that mattered enormously in March 2023. The instrument family sits alongside more familiar hybrids, and the general mechanics of bonds that convert into equity are the right starting point for understanding it.

FINMA's stated basis was contractual. The instruments provided that they would be written down completely in a viability event, which the documentation defined to include the granting of extraordinary government support. Credit Suisse received exactly that on March 19 in the form of liquidity assistance loans secured by a federal default guarantee.

FINMA also pointed to Article 5a of the amended emergency ordinance, inserted that evening, which expressly authorized the regulator to order a write-down of Additional Tier 1 capital. On that reading, the trigger was not discretionary; it was a term the investors had bought.

Why every other regulator said it had not

The market's objection was not that AT1s could be written down. It was the sequence. Credit Suisse shareholders received UBS stock; AT1 holders received nothing. The day after the announcement, the Single Resolution Board, the European Banking Authority, and ECB Banking Supervision issued a joint statement that read as a public correction of Switzerland.

The Bank of England issued its own statement the same day reminding investors that AT1 ranks ahead of common equity and that holders should expect to be exposed to losses in the order of their position in that hierarchy. Both statements were, in substance, reassurance to their own bank funding markets that this would not happen in their jurisdictions.

The immediate market damage was severe and short-lived. AT1 spreads blew out above 1,000 basis points in March 2023, then compressed back through their pre-crisis range, reaching roughly 310 basis points by the end of 2024, and European banks issued a record €45 billion of AT1 that year. The asset class survived; the assumption that regulators would always respect the hierarchy did not.

What the write-down actually bought

Follow the capital, and the purpose of the write-down becomes visible. Credit Suisse's first-quarter 2023 accounts recorded a net profit attributable to shareholders of CHF 12.4 billion, driven almost entirely by booking CHF 15 billion of written-down AT1 notes as a gain, and its common equity tier 1 ratio jumped from 14.1% at end-2022 to 20.3% at the end of March. The nominal amount extinguished was around CHF 16 billion; the accounting gain reflected the instruments' carrying value.

Bail-in

A bail-in recapitalizes a failing bank by imposing losses on its own investors, writing down or converting their claims into equity, instead of injecting public money. It is the central design principle of every post-2008 resolution regime, and the point of Additional Tier 1 is to be the first instrument bailed in after common equity. What made March 2023 contentious was not that a bail-in happened but that it happened outside the Swiss resolution framework, under an ordinance written the same evening, and in an order that reversed the market's expectation.

That is the mechanism, stated plainly: the write-down converted subordinated debt into common equity capital inside the entity UBS was about to absorb. UBS was not handed the cash, but it inherited a target whose capital position had been repaired overnight at the expense of one specific creditor class. Whether that is a legitimate use of a loss-absorbing instrument or a state-directed transfer of CHF 16 billion from bondholders to an acquirer is the question that has occupied the Swiss courts since. Loss-absorbing capital exists precisely so that this transfer can happen without public money, and the design logic behind it runs through the Basel framework's capital requirements for banks.

22.48 to one, and what the ratio implied

The consideration was entirely in stock: one UBS share for every 22.48 Credit Suisse shares, an implied CHF 0.76 per Credit Suisse share and CHF 3 billion in total. Credit Suisse had 4,002.2 million shares issued at the end of 2022 and 3,941.3 million outstanding net of treasury, so the ratio implied roughly 178 million new UBS shares, and UBS ultimately issued 178,031,943 of them.

Set against the target's own reported book, the ratio is startling. Credit Suisse ended 2022 with CHF 41.8 billion of tangible shareholders' equity, so UBS paid about 7% of tangible book value. Against the market's own mark on the Friday before, after a week in which the shares had already been savaged, the price was roughly 60% below the CHF 7.4 billion closing capitalization. An all-share structure also mattered: UBS paid in its own currency, preserving cash and capital for the integration, and Credit Suisse holders were left with an interest in whatever value UBS extracted rather than a clean exit.

Negative goodwill: the gain that was not a profit

Because UBS paid far less than the fair value of the net assets it acquired, the difference had to go somewhere. Under IFRS it goes straight to the income statement as a gain on a bargain purchase.

Negative goodwill

When an acquirer pays less than the fair value of the identifiable net assets it acquires, the shortfall is recognized immediately as a gain in profit or loss rather than sitting on the balance sheet. It is the mirror image of ordinary goodwill, and it is rare because sellers do not usually accept less than their assets are worth. It appears almost exclusively in forced transactions. The accounting sits inside the wider treatment of goodwill and intangibles in acquisition accounting.

UBS reported provisional negative goodwill of $28.9 billion for the second quarter of 2023, producing a headline quarterly net profit of about $29 billion against an underlying pre-tax profit, excluding the gain and integration costs, of only $1.1 billion. After refining its acquisition-date fair value estimates, UBS reduced the figure by $1.2 billion to a final $27.7 billion in its 2023 annual report.

The gain is real accounting and poor economics if read as profit. It measures the gap between what UBS paid and what the assets were judged to be worth on day one; it does not measure cash earned, and the same fair value exercise that produced it also loaded the balance sheet with marks UBS would have to work through. The honest reading is that $27.7 billion is the size of the discount Switzerland made available, not the size of UBS's win.

The costs that came through the door

Against the gain sat obligations UBS had not previously carried. At announcement UBS targeted more than $8 billion of annual run-rate cost reductions by 2027 and warned that restructuring a bank of Credit Suisse's complexity carried real execution risk. The integration ran for well over three years, and by the second quarter of 2026 UBS had delivered $12.6 billion of cumulative gross cost savings against an ambition of about $13.5 billion by the end of that year, with headcount down 28% from the 2022 baseline.

UBS also inherited Credit Suisse's litigation, its non-core assets, and, most durably, its systemic footprint. A bank whose balance sheet is a large multiple of its home country's economy is a political problem as much as a financial one, and that problem is now UBS's alone.

Go deeper before an interview: our 160-page PDF covers deal frameworks, technicals, and behavioral prep end to end, access the IB Interview Guide and pair it with these case studies.

August 11, 2023: UBS hands back the protection

The transaction closed on June 12, 2023, and the state support began unwinding almost immediately. Credit Suisse repaid the public liquidity backstop in full by May 31, 2023, before the deal had even completed. On August 11, 2023, UBS voluntarily terminated both the CHF 9 billion loss protection agreement and the CHF 100 billion public liquidity backstop, and Credit Suisse repaid its ELA+ loan in full.

UBS's stated reason was that having reviewed every asset in the designated portfolio since closing, including severe stress scenarios, and having taken the appropriate fair value adjustments, it concluded the guarantee was not required. The Confederation kept the fees: CHF 40 million for establishing the loss protection agreement, plus commitment and risk premiums of CHF 100.7 million and CHF 60.6 million on the backstop, for total receipts of roughly CHF 200 million.

The bank that is now the country's balance sheet

Legal integration proceeded in stages: UBS AG and Credit Suisse AG merged on May 31, 2024, the Swiss operating entities merged on July 1, 2024, and the global migration of former Credit Suisse client accounts onto UBS infrastructure completed in March 2026, covering around 1.2 million clients. By the second quarter of 2026 UBS reported net profit of $2.8 billion, group invested assets of $7.3 trillion, and an integration on track to be substantially complete by year end.

The strategic result is a single Swiss global bank with no domestic peer, which is the outcome Swiss policy had spent a decade trying to avoid. The regulatory response has been to make UBS carry the risk in capital. Under proposals the Federal Council put to parliament, UBS would have to deduct its investments in foreign subsidiaries in full rather than at 60%, phasing in over seven years from a 65% deduction. UBS quantified the combined effect at roughly $22 billion of additional CET1 capital from the new proposals on top of about $15 billion already required from the Credit Suisse acquisition, some $37 billion in all, and said publicly that it strongly disagrees with a package it called extreme and internationally unaligned.

The PUK's verdict on who failed

Switzerland's parliamentary commission of inquiry reported on December 20, 2024 with more than thirty proposals for reform. Its central finding assigned primary responsibility to Credit Suisse's own leadership over many years rather than to the authorities, and it explicitly concluded that it had found no causal misconduct by the authorities and that they had prevented a global financial crisis in March 2023.

That is not a clean exoneration. The commission criticized FINMA sharply for a 2017 regulatory concession, a so-called filter that relaxed capital requirements at the parent-bank level, describing it as inappropriate even though legally sound, and finding that it obscured the true state of the bank and delayed corrective action. The commission went further on the arithmetic: without the filter, Credit Suisse would have breached its capital requirements, marginally in 2021 and significantly in 2022, which is the strongest available qualification to the fact that the bank formally met every requirement into the final weekend. It also regretted the ineffectiveness of FINMA's supervision and backed giving the regulator power to fine banks and individuals, an authority Swiss law had withheld. The inquiry's implicit judgment is that the crisis was privately created and publicly enabled.

October 1, 2025: an emergency ordinance found wanting

On October 1, 2025, in case B-2334/2023, the Federal Administrative Court revoked FINMA's March 19, 2023 write-down decree. The judgment addressed one lead case out of roughly 360 proceedings brought by about 3,000 complainants, and it went further than most observers had expected.

The court rejected the contractual argument and the statutory one. On the contract, it found that the write-down conditions were not met, because Credit Suisse was adequately capitalized and met its regulatory capital requirements at the time; FINMA had itself approved prospectus wording requiring that customary measures to improve capital adequacy be inadequate or unfeasible, a stricter test than the law demanded, and then ordered a write-down that did not satisfy it. On the statute, it held that Article 26 of the Banking Act and Article 31 of the FINMA Act were too vague to authorize so serious an interference with property rights, and that Article 5a of the emergency ordinance, the provision inserted at 8 p.m. on the evening of March 19, itself exceeded the Federal Council's constitutional emergency powers.

Reporting on the 78-page judgment by the Swiss financial title finews added a detail that reframes the evening: Credit Suisse's own people urged FINMA at 6:24 that evening to abstain from determining that a viability event had occurred, arguing that the bank faced a liquidity crisis rather than a capital shortfall. FINMA issued the order at 10:01 p.m. regardless.

Where the claims stand

Nothing has been paid. FINMA announced on October 15, 2025 that it would appeal to the Federal Supreme Court, and UBS filed its own appeal on October 29. On December 10, 2025 the Supreme Court granted UBS's request for suspensive effect, which means the AT1 instruments remain written down and worthless pending a final ruling; as of September 2026 no substantive judgment has been issued. The Federal Administrative Court has also not yet ruled on the remedy, only on the annulment of the decree.

The parallel actions have mostly fared worse. A New York federal court dismissed a $372 million claim against the Swiss Confederation on September 30, 2025, with Judge Dale Ho holding that Switzerland was immune under the Foreign Sovereign Immunities Act, and the Second Circuit affirmed on July 16, 2026, finding that Switzerland acted as a sovereign rather than a market participant, which closes the US route. One parallel action did land: in a New York ICC arbitration decided on May 8, 2026, a tribunal found that the contractual trigger had never fired for a former Credit Suisse managing director's contingent capital awards and ordered payment, which UBS is contesting in enforcement proceedings. Bondholder groups have also pursued investor-state treaty claims against Switzerland, and Swiss proceedings brought on behalf of institutional and retail holders continue.

What a reversal would actually cost

The exposure is asymmetric and easy to misstate. If the Federal Supreme Court upholds the annulment and a remedy follows, the theoretical maximum is the full CHF 16 billion nominal plus interest, but it would fall on the Swiss Confederation and FINMA rather than on UBS, since UBS acquired a bank whose AT1 had already been extinguished by regulatory order.

The larger cost is to Swiss credibility as a place to issue loss-absorbing capital. A regime whose emergency powers are later found unconstitutional has a harder time promising investors that the next intervention will hold. Some legal scholars have argued the opposite way: writing in the Oxford Business Law Blog, commentators contended that the judgment undermines legislative intent and rewards supervisory forbearance, because regulators facing this level of legal risk will delay acting rather than act decisively in a crisis. That is a real cost too, and it points in the opposite direction from the bondholders' case.

What the evidence has settled

Several disputes are effectively closed. Credit Suisse was destroyed by a funding run rather than a capital shortfall: it met every regulatory capital requirement into the final weekend, and the Federal Administrative Court relied on precisely that fact two and a half years later. The run was not a March phenomenon; it began in October 2022, took CHF 110.5 billion out in one quarter, and never reversed. The cause was a decade of governance failure that the Swiss parliament's own inquiry laid at the door of the bank's leadership, quantified as CHF 33.7 billion of losses against CHF 31.7 billion of bonuses between 2010 and 2022.

The price was not a market price. Two shareholder votes were disapplied by ordinance, there was no other bidder, and there was no time. UBS paid about 7% of tangible book and booked $27.7 billion of negative goodwill, then handed back CHF 9 billion of loss protection within five months of closing and ran the integration to schedule. On the narrow question of whether Switzerland averted a disorderly failure of a globally systemic bank, the answer is yes, and the parliamentary inquiry said so directly.

What remains genuinely open

Three questions are not settled and should not be presented as if they were. The first is the legality of the AT1 write-down, which is under appeal at the Federal Supreme Court with the instruments still at zero. The second is whether the write-down was necessary at all, given the capital position FINMA and the SNB had publicly confirmed four days earlier and the speed with which UBS returned the state's protection. The third is whether Switzerland traded a solvable one-bank problem for a permanent one-bank country, a concern now expressed in a capital package UBS says it strongly opposes, and which parliament has already begun to soften: on August 31, 2026 the Council of States committee backed letting UBS meet half the foreign-subsidiary requirement with AT1 capital rather than common equity, sending the compromise to the full chamber.

The verdict that the evidence supports is narrower than either camp wants. This was a genuine rescue: the alternative on the table that Sunday was the disorderly failure of a globally systemic bank, and no serious analysis disputes what that would have cost. It was also, in the same act, a transfer.

Switzerland paid for certainty with legal instruments it had to write that evening and with CHF 16 billion belonging to a creditor class that had bought a hierarchy the contract did not promise. UBS received the benefit of both without having to bid against anyone. Rescue and steal are not competing descriptions of the March 2023 weekend; they are two accurate accounts of the same set of documents, and the courts are still deciding which one the law will endorse.

1. 1UBS, "UBS to acquire Credit Suisse," media release, March 19, 2023, ubs.com.
2. 2FINMA, "FINMA approves merger of UBS and Credit Suisse," March 19, 2023, finma.ch.
3. 3FINMA, "FINMA provides information about the basis for writing down AT1 capital instruments," March 23, 2023, finma.ch.
4. 4FINMA and the SNB, "FINMA and the SNB issue statement on market uncertainty," March 15, 2023, finma.ch.
5. 5Swiss Federal Department of Finance, "UBS takeover of Credit Suisse," efd.admin.ch.
6. 6UBS, "UBS Group AG voluntarily terminates Loss Protection Agreement and Public Liquidity Backstop," August 11, 2023, ubs.com.
7. 7Thomas Jordan, "The Swiss National Bank's role as lender of last resort in the crisis at Credit Suisse," Bank for International Settlements.
8. 8SRB, EBA and ECB Banking Supervision, joint statement on the announcement of March 19, 2023 by Swiss authorities, ECB Banking Supervision.
9. 9UBS Group AG, Annual Report 2023 and Form 20-F, SEC EDGAR.
10. 10Credit Suisse Group AG, first quarter 2023 earnings release, April 24, 2023, SEC EDGAR.
11. 11Credit Suisse Group AG, fourth quarter and full year 2022 earnings release, February 9, 2023, SEC EDGAR.
12. 12Federal Administrative Court, "Unlawful write-off of AT1 capital instruments," judgment B-2334/2023 of October 1, 2025, bvger.ch.
13. 13FINMA, "FINMA to appeal partial decision of the Federal Administrative Court concerning AT1," October 15, 2025, finma.ch.
14. 14Swiss Parliament, "Lessons from the Credit Suisse crisis, PInC identifies need for action," December 20, 2024, parlament.ch.
15. 15"UBS buys Credit Suisse for $3.2 billion as regulators look to shore up the global banking system," March 19, 2023, CNBC.
16. 16"Credit Suisse to borrow up to nearly $54 billion from Swiss National Bank," March 16, 2023, CNBC.
17. 17"Credit Suisse borrows more than $50 billion from Swiss National Bank after shares crash 30%," March 15, 2023, CNN Business.
18. 18"Credit Suisse shares sink as material weaknesses found in financial reporting," March 14, 2023, CNBC.
19. 19"Credit Suisse chair re-elected after apology at final shareholder meeting," April 4, 2023, CNN Business.
20. 20"UBS shares jump to 2008 highs after profit beat, job cuts announcement," August 31, 2023, CNBC.
21. 21UBS, "UBS statement on regulatory capital announcements made by the Swiss government," April 22, 2026, ubs.com.
22. 22UBS, "UBS successfully completes the client migration in Switzerland," March 18, 2026, ubs.com.
23. 23"The Swiss Job Undone: The Judicial Overturn of the Credit Suisse AT1 CoCos Write-Down," Oxford Business Law Blog.
24. 24"Swiss Court Strikes Down AT1 Bond Write-Off: A Landmark Decision for Bondholders," November 2025, DLA Piper.
25. 25"Word for Word: How the AT1 Judgment Recasts the Credit Suisse Rescue," finews.
26. 26"Credit Suisse rescue rebuked by half of Swiss parliament," April 2023, Associated Press.
27. 27"Morgan Stanley, JPMorgan Advise on UBS Deal for Credit Suisse," March 19, 2023, Bloomberg Law.