UBS Booked $27.7 Billion in Profit on a Deal It Was Pressured Into. The Bill Just Arrived.

Picture a buyer who closes the most consequential acquisition in modern banking and, the same quarter, reports the largest profit of any bank in the world that year. Then read the footnote. Of UBS's roughly $27.8 billion in 2023 net profit, $27.7 billion came from a single line called negative goodwill — a non-cash accounting gain that arrives automatically when you buy something below its book value.³ Strip it out, and the underlying pre-tax profit for the quarter the deal landed was $1.1 billion.⁴ The headline said triumph. The arithmetic said something quieter, and stranger.

The official story is that UBS pulled off the deal of the decade — it snatched a 167-year-old rival for a song and booked a record windfall. UBS won the negotiation. UBS was pressured into a rescue by the Swiss state, paid in its own shares at a price that flattered the math, and the windfall it booked was an accounting artifact resting on a write-off that a Swiss court has since ruled was never legal. The deal of the decade has a tail no one has priced.

What the CHF 3 billion price tag hides

Almost everyone quotes the price as 'CHF 3 billion,' as if UBS wrote a check. It didn't. The deal was all-share: every Credit Suisse holder received 1 UBS share for every 22.48 they owned, worth about CHF 0.76 a share — and the CHF 3 billion is merely the aggregate implied value at announcement.¹ The state did the heavy lifting that made it possible. The Swiss National Bank stood behind it with over CHF 100 billion in liquidity, and the federal government added a loss guarantee of up to CHF 9 billion.¹ This was not an open-market purchase by a confident bidder. It was a state-arranged absorption, negotiated over a weekend, with the buyer's downside underwritten by the public. Kelleher said the quiet part into a microphone: for Credit Suisse, this was an emergency rescue.⁸

Why a 'record profit' was mostly a number, not a result

Negative goodwill is the accounting world's most misleading kind of good news. When you buy a company for less than the fair value of its net assets, the rules require you to book the difference instantly as a gain — a one-time, non-cash entry that says nothing about whether the business will ever earn a franc. UBS booked $28.9 billion of it provisionally, then trimmed it to $27.7 billion a year later after the IFRS 3 measurement-period review revised the fair value of what it had actually bought.³ That revision is the tell. The figure shrank by $1.2 billion the moment UBS looked harder at the assets — which is exactly what you'd expect when the 'profit' is really a snapshot of how distressed the seller was, not how skilled the buyer is. The smaller the price you can get away with, the bigger the gain you must report. The genius of the headline runs on the desperation of the counterparty.

The CHF 16.5 billion that made the math work — and the court that unwound it

Here is the part that turns an accounting curiosity into a live risk. On the same day the deal was struck, the Federal Council passed an Emergency Ordinance — Article 5a — letting FINMA order Credit Suisse's Additional Tier 1 bonds wiped out. FINMA's decree wrote roughly CHF 16.5 billion of AT1 bonds straight to zero, while equity holders, who normally rank below bondholders, walked away with UBS shares worth about $3.25 billion.⁵ That inversion — bondholders zeroed, shareholders paid — is what filled the hole in the balance sheet and helped make the below-book purchase math hang together. Then, on 1 October 2025, Switzerland's Federal Administrative Court ruled the write-off unlawful: the contractual viability event had never actually been triggered because Credit Suisse was still adequately capitalized, Article 5a was unconstitutional, and there was no sufficient legal basis to permanently extinguish bondholders' property rights.⁶ Roughly 3,000 complainants had filed over 360 cases.⁶ The legal floor under the cheapest part of the deal has been kicked out.

Isn't this still obviously a win for UBS?

The fair objection is that none of this stops UBS from being a winner. It acquired a global franchise it could never have built, removed its only domestic-scale rival, and did it with public backstops absorbing the worst-case risk. All true. But 'attractive for UBS shareholders' was the chairman's own framing, and the strength of the position is exactly what makes the hidden cost easy to overlook.⁸ The honest counter to the triumph narrative is not that UBS lost — it is that the deal's economics were never as clean as the record profit implied. A non-cash gain of $27.7 billion is not cash UBS can spend, and the underlying business was earning a fraction of the headline.³⁴ Now layer on a court ruling that the CHF 16.5 billion write-off — the piece that helped the purchase price look generous to the buyer — lacked any legal basis.⁶ The bondholders the deal stepped over are not a closed chapter; they are a contingent liability with a verdict in their favor. The deal of the decade may still be a good deal. It is not the deal the headline sold.

Sergio Ermotti was brought back as CEO in April 2023 for exactly this work — to integrate a failed bank, not to bask in a number — and was paid CHF 14.4 million through December.⁸ That tells you what UBS understood internally that the headline obscured: the real cost of this deal was always going to be the integration and the liabilities, not the price. The negative goodwill was a snapshot of someone else's collapse, dressed as the buyer's brilliance. And the cheapest line on the balance sheet — the CHF 16.5 billion in bonds erased on a Sunday — has turned out to be the most expensive, because a court has now said it should never have happened at all.⁶ The deal of the decade was never bought. It was conscripted — and the bill for the part that made it look like a steal is still being written.