Bern Took Back the July Premium, and the Business Did Not Flinch

Results vs. Consensus

Q3 2026 Scorecard

The per-share consensus deserves a caveat. At the quarter's diluted share count, the $2.71bn net profit consensus works out to about $0.86 a share, not $0.91, so the EPS "miss" partly reflects estimates that had not been brought up to date. The net profit comparison is the one that matters, and it was in line: the beat in revenue and pre-tax profit was absorbed by tax. Legal provisions recorded in Non-core and Legacy are not expected to produce a tax benefit, which lifted the rate to 24.7%.

Year-Over-Year Comparisons

The reported comparison flatters and the underlying comparison understates, for the same reason running in opposite directions. Reported pre-tax profit grew 26.8% largely because integration expense and purchase-price effects in operating expenses fell by $871m. Underlying pre-tax profit grew only 13.6% because the year-ago quarter carried $668m of net litigation releases inside general and administrative expenses, the item we flagged when we initiated. This quarter swung to $336m of net litigation expense, which explains most of the 21.8% increase in that line.

Net interest income rose 21.9% while income from instruments at fair value fell 13.0%. The two lines move together in UBS's trading businesses, and the combined total was flat at $5,460m against $5,483m. Within that total, Group Items lost $422m against $153m a year ago on hedge and own-debt valuation effects from the sharp rise in US dollar rates. Personnel expenses fell 6.5% on a smaller workforce, even as financial-adviser compensation rose 13% on higher revenues.

Quarter-Over-Quarter Comparisons

Sequentially, underlying revenues were flat and underlying costs fell, so underlying pre-tax profit rose 4.9%. The legal items explain all of that and more: the second quarter carried $30m of net litigation releases and no settlement gain. On a clean basis, underlying pre-tax profit fell 3.7% from $3,857m to $3,716m. That is a fair outcome for a quarter in which the Investment Bank came off a record second quarter and Asia Pacific transaction activity cooled.

Nine Months 2026

The nine-month totals reconcile to the three quarters: net profit of $8,530m is $3,040m plus $2,800m plus $2,690m, revenues of $41,293m are $14,243m plus $13,700m plus $13,350m, and diluted EPS of $2.66 is $0.94 plus $0.87 plus $0.85. Revenues grew 10.3% on flat operating expenses, and the tax rate normalized from 7.8% to 22.3%. Net profit still grew 29.9%. A nine-month underlying return on CET1 capital of 16.7%, against a 2026 exit-rate target of around 15%, is the operating case delivered.

Quality of Beat

• Revenue: Good, and narrower than the headline. Reported revenues beat consensus by 3.2%, but $697m of the total was a legal settlement gain. Ex legal items, underlying revenues grew 4.0%. Global Wealth Management grew 7.6% on recurring fees and net interest income, and the Investment Bank grew 6.4% on prime brokerage and equities. Those offset a 21% decline in Global Banking and a 7.5% decline in Asset Management, where the O'Connor sale is still in the comparatives.
• Margins: Real improvement. The underlying cost/income ratio of 68.7% is below the sub-70% exit-rate target even before the legal adjustment, and 69.8% after it. Costs ex legal items fell 3.4% year over year as the integration savings arrived. Headcount fell 2% in the quarter to 110,000, 30% below the 2022 baseline.
• EPS: Held back by tax. A 24.7% effective rate against a full-year guide of about 23% cost about $61m, or two cents a share, relative to the guided rate. The rest of the gap to the $0.91 per-share consensus reflects stale estimates rather than weaker operations.
• Flows: The quantity is excellent and the quality is still unproven. Net new assets of $55.1bn were the largest of our coverage, and every region contributed, but only $12.9bn arrived as fee-generating assets. The roughly $42bn gap is mostly IPO and liquidity-event proceeds sitting in custody. Those assets are an opportunity and not yet revenue.
• Capital: Generated and then deliberately released. CET1 capital rose $0.7bn and risk-weighted assets fell $8bn, which took the ratio to 14.7%. Management then told us it will spend the excess above about 14%, including an accrual for 2027 buybacks, before year end.

Segment Performance

Currency note, as in prior quarters. The table is in US dollars, consistent with the filing. Personal & Corporate Banking is managed and guided in Swiss francs. In francs, underlying pre-tax profit rose 22% to CHF 813m and revenues rose 4%, so the 19.3% dollar growth is close to the local figure this quarter.

Global Wealth Management

Underlying pre-tax profit of $1,972m rose 11.2%, and 25% excluding the roughly $200m of net litigation releases in the year-ago quarter, which included a $284m release on resolving a French cross-border matter. Underlying revenues rose 7.6% to $6,898m. Recurring net fee income grew 9% to $3.8bn on higher fee-generating assets, underlying net interest income grew 8% to $1,755m on deposit margins and loan volumes, and underlying transaction-based income grew 4% to $1.3bn. Within transaction revenues, Asia Pacific grew only 2% against an exceptional prior-year quarter, while the Americas grew 11%. Underlying costs rose 6%, but fell 4% excluding variable compensation, litigation and currency. The underlying return on attributed equity was 23.0%.

Regional net new assets are as disclosed and are rounded, so they do not sum exactly to the divisional total. On management's figures excluding litigation, the regions grew pre-tax profit as follows: the Americas 28% at a 16% margin on record revenues, Asia Pacific 19% at a 45% margin, EMEA 39% at a 37% margin with 13 points of positive operating leverage, and Swiss wealth 34%. Net new loans were $5.8bn, $4.2bn of them in the Americas, and net new deposits were $1.8bn.

"Net new assets were $55 billion, representing a 4.5% growth rate and bringing year-to-date NNA to $128 billion. Momentum was broad-based across regions led by APAC and the Americas." — Todd Tuckner, Group CFO

On Asia, management described the Chinese consultation on offshore direct investment as strategically promising for the wealth business, though too early to quantify. It expects the close of the transition window for settling historical offshore-trust tax liabilities to have an immaterial effect on the book and a small effect on fourth-quarter flows.

Assessment: The division delivered the flow year it promised a quarter early, and the Americas is now a growth contributor rather than a drag. The quarter-over-quarter dip in profit is Asia Pacific transaction volumes normalizing after an unusually uniform rally, which management had twice told us not to extrapolate. The open question is conversion. A $42bn gap between net new assets and fee-generating assets is large, and the bull case requires a meaningful part of it to move into mandates over the next several quarters.

Personal & Corporate Banking

The Swiss bank had its strongest quarter since the acquisition. Underlying pre-tax profit was CHF 813m, up 22% year over year and 20% sequentially, with an underlying pre-tax return on equity above 18% and 9 points of positive operating leverage. Revenues rose 4%. Underlying net interest income, excluding purchase-price accretion, rose 2% year over year and 3% sequentially on lending growth, at a zero policy rate. Non-interest revenues grew 5% on custody, mandate and card fees. Underlying operating expenses fell 5% to CHF 947m as the platform consolidation paid off. Credit loss expense was CHF 27m, well below the roughly CHF 75m per quarter guide, because model releases on the performing book offset charges on non-performing loans.

"P&C delivered its strongest quarter since the acquisition with pre-tax profit of CHF 813 million, up 22%, and an underlying pre-tax return on equity above 18%." — Todd Tuckner, Group CFO

Assessment: This is the second consecutive quarter of net interest income growth at zero rates, so the bear point we retired in July stays retired. The cost line is the bigger story: a CHF 57m sequential step-down in underlying operating expenses, to CHF 947m, with management saying there is "still upside to see in Q4" before integration ends. A Swiss bank earning an 18% pre-tax return on equity at a zero policy rate is the clearest single argument UBS has that its domestic franchise is not the problem the capital debate makes it out to be.

Investment Bank

Underlying pre-tax profit of $854m rose 8.5% at an 18% pre-tax return on equity. Global Markets had a record third quarter, with highs in Asia Pacific, the Americas and Switzerland. Financing rose 32% on prime brokerage client balances up 33%, and equities as a whole rose 21.5%. Global Banking fell 34.5% on the reported line and 21% underlying, against a record prior-year quarter. A 44% rise in equity capital markets was more than offset by weaker leveraged finance, and advisory fell 44% in a fee pool that management said is concentrated in mega-cap transactions where balance-sheet commitment matters. Risk-weighted assets and the leverage ratio denominator were broadly flat year over year.

Assessment: The normalization we and management both flagged arrived on schedule: from a 23%-plus return in the second quarter to 18%, still above the 15% through-the-cycle ambition. The capital-light design held, with revenues up on flat resources again. The weakness is in banking, where a franchise that will not write large balance-sheet commitments loses share in a market dominated by jumbo deals. The other weakness is the equities growth gap to a peer, which an analyst raised and management declined to address.

Asset Management

Pre-tax profit of $300m rose 6.4% and the cost/income ratio improved to 62%. Revenues fell 7.5% on the O'Connor exit and were broadly stable excluding it, while operating expenses fell 14%, or 11% excluding O'Connor. Net new money was negative $5.1bn, on money-market outflows and institutional rebalancing that followed the completion of the custody platform consolidation. Unified Global Alternatives raised a record $12.5bn of new commitments across the wealth and asset management channels, taking its invested assets to $371bn.

Assessment: Profitability is improving on the cost line and the alternatives platform is clearly working. But net new money has now gone from $17.9bn a year ago to $6.3bn last quarter to negative this quarter, and the explanation was given once without a forward view. We said in July this would be asked again if flows did not recover. They did not.

Non-core and Legacy and Group Items

Non-core and Legacy earned $360m of underlying pre-tax profit, all of it from the $697m net gain on settling legacy legal matters, which more than offset $275m of litigation charges. Excluding both items, the division would have lost about $62m. Operating expenses excluding litigation fell 69% to $91m, cumulative cost reductions against the 2022 baseline passed 90%, and credit and market risk-weighted assets fell to $3bn, already below the year-end ambition of about $4bn. Management cut the 2026 exit-rate cost expectation again, to about $300m from about $400m in July and about $500m in April.

Group Items lost $740m on a reported basis. That figure includes the $346m net loss on repurchasing $9.8bn of high-cost Credit Suisse senior debt and $265m of own-credit and hedging losses on accrual-accounted positions. Excluding the temporary hedging volatility, the underlying loss was $138m, mostly deferred-tax-asset funding.

Assessment: The wind-down is effectively finished and keeps beating its own targets. The debt repurchase is the more interesting decision: a $346m charge now, in exchange for lower funding costs in Non-core and Legacy and a net interest income benefit in the wealth and Swiss businesses later. That is the kind of trade a bank makes when it expects to be in business on the same footprint for a long time.

Key Group KPIs

Key Topics & Management Commentary

Overall Management Tone: Combative on Bern and businesslike on everything else. The CEO used the prepared remarks to criticize the upper-house outcome directly, and in Q&A management declined every invitation to price alternatives or describe contingency plans. On operations the register was the most confident of the coverage period: the return target language moved from "well positioned to outperform" to "increasingly confident," and the integration was described as in its last mile.

1. The Upper House Chose 90%, and UBS Has Re-Priced It

On 23 September the Council of States voted 29 to 16 to require systemically important banks to deduct 90% of the carrying value of foreign participations from CET1 capital. The Federal Council's draft had proposed 100%, and the reporting that drove the stock up through June and July had pointed to 70% to 80%. The bill now goes to the National Council, whose economic affairs committee heard the CEO on 26 and 27 October. The full chamber debates it in the December session, and differences between the chambers could take the process into 2027.

The quarterly report quantifies the upper-house version. If enacted as voted, it would require about $16bn of additional CET1 capital at UBS AG, phased in from 2028 at the earliest. That is on top of about $2bn from the ordinance-level measures and the roughly $15bn the acquisition already required, for about $33bn in total since the acquisition, with an annual capital-related cost of about $2.5bn. The equivalent figures for the government's 100% draft, as UBS gave them in April, were about $37bn and about $3bn a year.

"September's outcome in the Upper House is very concerning. We have made every effort to engage and contribute constructively to create a basis for a truly democratic process. Our goal has been and remains to reach a solution that balances financial stability, taxpayer protection, and competitiveness. Unfortunately, the latest proposal on the table also fails to achieve that balance." — Sergio P. Ermotti, Group CEO

Our arithmetic, on the same simple basis we used in April: an extra $18bn on today's $73.1bn of CET1 is about 25% more capital. Spread the same earnings over it and the 16.7% headline return becomes about 13.4%, and the clean return of about 15% becomes about 12%. A seven-year phase-in softens the path but not the end-state.

Assessment: The vote is better than the draft and worse than the price had assumed. The incremental ask fell about 18%, from $22bn to $18bn, but the measure remains a full CET1 deduction, with no role for AT1 instruments, and it passed one chamber comfortably. For a stock that traded at 1.93x tangible book on the expectation of a 70% to 80% outcome, 90% was a negative surprise, and the share price has absorbed it.

2. "Protect the Long-Term Interests of Our Shareholders": the Domicile Question Is Open

In the prepared remarks the CEO acknowledged "a lot of speculation" about how UBS might respond, said the bank would not address all of it, and then chose one point to make explicit.

"We do not intend to address everything, but on one topic, we want to be very clear. For UBS, the top priority is to protect the long-term interests of our shareholders, clients, and employees around the world. In a very competitive environment, it is essential to be able to continue to grow sustainably while remaining strategically committed to our diversified business model and our global and regional footprint, which allow us to create value for all of our stakeholders." — Sergio P. Ermotti, Group CEO

Read alongside the Q&A, the statement is precise about what it commits to: the business model and the global footprint. It does not commit to the domicile. Three analysts approached relocation in different forms: the cost of contingency actions, the trade-off between a capital-heavy investment bank and a Swiss-branded wealth franchise, and the size of any exit tax. Management twice refused to discuss contingency plans until the rules are final. In the third answer the CEO argued that the proposed rules penalize "global reach" rather than any business line.

Assessment: We do not think UBS is about to move, and management was explicit that there is "no rush." But a bank that wanted to rule out relocation would have done so, and this one pointedly did not. The option has value in the negotiation with Bern and possibly for shareholders. It also has large costs that UBS has not quantified, and executing it would be a multi-year project with its own regulatory approvals. For the next two quarters it is leverage, not a plan.

3. The Integration Buffer Comes Off in January

The most valuable piece of capital information on the call came in answer to a planning question. With the integration ending, management will stop holding the buffer above its CET1 target that it kept for integration risk and regulatory uncertainty.

"By the way, as Todd mentioned, we are now also able, at the end of this integration process, to free up the buffer on capital that we kept over the minimum requirements." — Sergio P. Ermotti, Group CEO

The CFO gave the mechanics: CET1 is expected to end 2026 "at or very close to" the target of about 14%, including a potential accrual for the first part of intended 2027 buybacks. At 30 September the ratio was 14.7% on $496bn of risk-weighted assets. Our arithmetic puts the excess above 14% at about $3.5bn before fourth-quarter earnings and dividend accruals. Asked whether this signals a lower CET1 target, the CEO said it did not, and that the bank would build capital out of earnings "if and when" new requirements take effect rather than hold it in advance.

Assessment: This is a material shift in posture. For two years the risk was that UBS would pre-fund the Swiss requirement out of distributions. Management has now said, in effect, that it will not, because the rules phase in from 2028 at the earliest and the bank has the earnings to build capital as they arrive. That supports 2027 distributions directly. It also means the February capital plan is the event that shows how much of the excess goes to shareholders.

4. The Buyback Finishes Four Months Early

UBS repurchased $1.8bn of shares in the quarter, against the floor of at least $1bn committed in July, and intends to buy another $1.2bn in the fourth quarter. That completes the second $3bn programme by the time of the fourth-quarter results, against "at the latest by the end of the second quarter 2027" when it was announced.

"We have repurchased $4.8 billion of shares and accrued $2.8 billion for our dividend year-to-date. We have already achieved the ambition we set early in the integration journey for 2026, with total capital returns now to exceed pre-acquisition levels. Over the next three months, we intend to repurchase $1.2 billion of shares, completing our current program." — Sergio P. Ermotti, Group CEO

On those figures, 2026 buybacks will total $6.0bn, double the $3bn programme announced in February. Pace and size remain conditional on short-term performance, a CET1 ratio of about 14% and "further visibility" on the parliamentary deliberations. That is the same set of conditions as in July.

Assessment: This is an acceleration, not the enlargement our trigger specified. We hold to that distinction, but it is a good outcome: management bought faster after the upper-house vote, not slower. The more important number is the 2027 plan in February, which will now include the integration buffer.

5. $55 Billion of Net New Assets, and the $42 Billion Gap

Net new assets of $55.1bn were up 47% on a year ago, and every region was positive. But net new fee-generating assets were $12.9bn, unchanged on the second quarter, and the difference was explained on the call. Large liquidity events, including major US listings and strong Asian issuance, delivered proceeds that clients have placed with UBS but not yet invested. Management also pointed to about $500bn, roughly 10% of assets under management, held as custody or non-contracted assets, as "dry powder."

Assessment: These inflows are genuine, and they come from the IPO-related wealth events that UBS's investment bank and wealth model is designed to capture. The point the CEO made about the investment bank's role in wealth gathering is visible in these flows. But custody assets earn little until they are invested, and a cooler fourth quarter for markets, which management described, can slow conversion. We model the fee benefit gradually and do not treat the $55bn as a new run-rate.

6. The Americas Delivers the Quarter It Was Promising

The Americas added $16.7bn of net new assets, after outflows of $9bn a year earlier and $14bn in the fourth quarter of 2025. Pre-tax profit rose 28% on record revenues, the margin rose more than two points to 16%, and net new loans were $4.2bn, extending a run of positive lending quarters.

"In the Americas, we continue to execute with discipline against our strategic priorities, delivering further improvements in profitability and growth. Pre-tax profit increased by 28%, with a profit margin of 16%, up more than 2 percentage points, supported by record revenues. During the quarter, we added $17 billion of net new assets, driven by strong same-store performance." — Todd Tuckner, Group CFO

With the national bank charter in hand, management expects to pilot the first new banking capabilities from the start of 2027 and to scale them during the year, aiming for higher net interest income and margins. Asked whether a 20% margin is now realistic, the CFO pointed back to the 18% ambition for 2028 and to the fourth-quarter update.

Assessment: The bear point we carried as Materializing in February is resolved on every measure except adviser headcount, which management now says will settle at normal industry net attrition rather than turn positive. A 16% margin a year before the 2027 target, with the charter benefits still to come, makes the 2028 ladder look conservative.

7. The Legal Ledger: a $697 Million Gain Inside Underlying

For the second time in our coverage, legal items moved the underlying result materially. Non-core and Legacy settled certain legacy legal matters for a net gain of $697m, which is booked in other income and included in underlying revenues. Against that, it recorded $275m of litigation charges, and the group recorded $336m of net litigation expense in total.

"This largely reflected continued progress in addressing legacy cases, with net income from the settlement of certain legal matters more than offsetting $275 million of litigation charges." — Todd Tuckner, Group CFO

The release does not identify the matters settled. The provisions in Non-core and Legacy also carry no expected tax benefit, which is why the group tax rate rose to 24.7%.

Assessment: Settling legacy cases for a net gain is good news for the Credit Suisse legal-tail bear point, and it is genuinely cash and capital in the bank. It is not a recurring earnings stream, and it inflates the 16.7% return. We measure the quarter at about 15% clean, which is still on target, and we would like the release to give the clean figure itself next time.

8. Buying Back $9.8 Billion of Credit Suisse Debt

In September UBS repurchased $9.8bn of legacy Credit Suisse senior debt for about $9.9bn, recognizing a net loss of $346m after releasing $241m of purchase-price adjustments. The loss sits in reported results and is excluded from underlying.

"The debt repurchase accelerates the reduction of the funding drag in Non-core and Legacy, while also supporting net interest income in GWM and P&C going forward." — Todd Tuckner, Group CFO

Assessment: A sensible use of liquidity that pays back over time through lower interest expense. Total loss-absorbing capacity fell to $187bn from $194bn, a buffer UBS can afford. The excluded loss is real money, which is one reason we read the reported return on CET1 capital of 14.8% alongside the underlying 16.7%.

9. The Integration Over-Delivers, at a Price

Gross cost savings reached $13.2bn after $0.6bn in the quarter. Management now expects about $13.7bn by year end, above the $13.5bn ambition, after accelerated integration milestones identified further synergies. About 98% of the legacy Credit Suisse applications in scope are no longer in use and 90% are fully decommissioned. UBS also agreed to transfer former Credit Suisse fund administration businesses in Luxembourg and Switzerland to Northern Trust, closing in the second quarter of 2027 with no material profit or loss expected.

"Accelerated delivery of integration milestones has enabled us to identify additional synergies supporting this expected outperformance. Capturing these opportunities entails additional costs." — Todd Tuckner, Group CFO

The additional costs are visible. In July management guided about $750m of integration expense across the second half, split roughly evenly. The third quarter carried $452m, and about $0.5bn is expected in the fourth, for roughly $950m. Cumulative integration expense will be about $15.1bn at constant currency, for a cost-to-achieve multiple of about 1.1x, in line with guidance.

Assessment: Spending about $200m more to secure about $200m more of annual savings is a good trade, and the programme ends on time. It is still a guide that was missed, and the release did not present it that way. The pillar is complete, and the focus for 2027 shifts to whether the cost base stays down once the synergy target is gone.

10. Book Value Barely Moved, and the Reason Is Rates

Tangible book value per share rose 0.4% to $27.00, even though net profit was $2,690m. Other comprehensive income was a loss of $1,789m: cash-flow hedges lost $1,320m on rising long-term dollar rates, and currency translation lost $718m as the dollar strengthened against the franc. Total comprehensive income attributable to shareholders was only $902m, against $1.8bn spent on repurchases, so equity attributable to shareholders fell $552m. Book value per share rose only because the share count fell.

"Tangible book value per share increased slightly sequentially as profit in the quarter was largely offset by net charges in OCI and share repurchases." — Todd Tuckner, Group CFO

Assessment: Cash-flow hedge losses are excluded from CET1 capital and unwind as the hedges mature, so this is a drag on the book-value denominator rather than on distributable capital. It still matters for valuation: over the four quarters of our coverage, tangible book per share has risen 1.7%, from $26.54 to $27.00. The investment case continues to depend on the multiple more than on book growth.

11. The Fourth-Quarter Tone Turned More Cautious

Management described the fourth-quarter environment so far as "mixed" for investors. Rising correlations, weaker credit conditions, higher bond yields and renewed inflation concerns are pushing clients toward defensive positioning, and elections are adding to the caution. On Asia, the CFO described the third quarter as "a bit of a tale of two cities," with periods of risk-off sentiment after a uniform rally earlier in the year. The Investment Bank expects typical seasonal moderation in markets volumes and a tough comparison against a prior-year gain on disposal.

Assessment: Management was clearly more cautious about the fourth quarter than about the third. Combined with fourth-quarter cost seasonality, including the bank levy, it argues for a softer fourth quarter on revenue. The exit-rate cost/income target leaves little room for slippage, and management did not suggest it was at risk.

Guidance & Outlook

Where this leaves 2026: nine months at a 16.7% underlying return on CET1 capital and a 69.6% underlying cost/income ratio leave the exit-rate targets of about 15% and below 70% comfortably within reach. A softer fourth quarter on revenue and the usual cost seasonality are the risks, and the guide implies management does not see them as threatening either target. Clean of legal items the third quarter was at about 15%, so "outperform" in the fourth quarter will need either revenue resilience or further cost delivery.

Where the Street sits: consensus had $2.71bn of net profit and $3.52bn of reported pre-tax profit and got $2,690m and $3,587m. The two items most likely to move estimates are the extended wealth net interest income outlook, with 2027 growth guided at a similar pace to 2026, and the release of the integration capital buffer, which raises 2027 distribution capacity.

Guidance style: Five consecutive quarters of meeting or beating operational guides, with one exception this quarter: the second-half integration cost came in about $200m above the July guide. Management still declines to quantify anything that depends on parliament, but this quarter it stated the basis on which the February targets will be set. That basis is current regulation, which means the targets will not price the foreign-participations rule.

Analyst Q&A Highlights

What a 75% Outcome Would Cost, and What Management Would Do About It

The call opened with the question the market most wanted answered: with the lower-house committee meeting that week and a 75% variant in circulation, what would that outcome mean for UBS, and how does the board weigh the cost of contingency actions against the long-term effect of higher requirements? Management declined to discuss any number other than the 50/50 structure it has already endorsed publicly.

Q: "Now, what would a 75% scenario mean for UBS? I guess the capital costs of higher requirements goes well beyond the mere billion-dollar charge, but I'm keen to hear your thoughts on this. Related to that, how do you view, and the Board of UBS view, the cost of any contingency actions here versus the potential long-term implications that higher capital requirements may have on your ability to compete as a global business?"
— Antonio Reale, Bank of America

A: "To be honest, I do not think that it is appropriate for us to go into discussing every single numbers that is thrown out. You make a reference to one of that number, and the one that we publicly, myself and our chairman, commented on was the 50/50 proposal. While being still a very expensive outcome considering our current capital position and the true lessons learned of what happened with Credit Suisse, we felt that a 50/50 proposal was balancing the needs of financial stability and reinforcing the current regulatory requirements while allowing us to preserve a competitive position going forward. I will not comment any further on any numbers, as I mentioned."
— Sergio P. Ermotti, Group CEO

Assessment: The refusal is tactically sensible while the committee is drafting, and the cost is that investors must do their own arithmetic on 75%. The answer did confirm one thing: 50% CET1 with 50% AT1 is still the outcome UBS considers acceptable. Because UBS calls even that "a very expensive outcome," 75% CET1 would be more expensive still on its own framing.

Whether the Swiss Domicile Matters More to Wealth Than to the Investment Bank

A questioner set out the tension directly: the capital-intensive investment bank may need the Swiss domicile less, while the capital-light wealth franchise arguably draws value from the Swiss brand. How does UBS weigh that when it considers whether Switzerland remains the right home? The CEO rejected the premise that the businesses can be separated and reframed the regulation as a penalty on global reach.

Q: "Then second, to come back to the domicile point, there is an interesting tension within the group because, on the one hand, businesses such as the IB are relatively capital intensive, where the Swiss domicile may be less strategically important. While on the other hand, you've got more capital-light businesses such as GWM, which arguably derive significant value from that Swiss brand."
— Chris Hallam, Goldman Sachs

A: "If you look at the current proposed regulation, what they in fact they are doing, they are not even penalizing business segments. They are penalizing global reach. ... So being a bank that has a DNA of Swissness, that is rooted in Switzerland, is a big advantage, but it is not the reason why our clients are banking with us."
— Sergio P. Ermotti, Group CEO

Assessment: The most revealing answer of the call. Saying that Swissness "is not the reason why our clients are banking with us" weakens the main argument for staying put, namely that the wealth franchise depends on the Swiss brand. Coming from the CEO in the middle of a parliamentary process, that is a negotiating position, and it is also a view the board could act on if the final rules disappoint.

What It Would Cost to Leave

A later questioner asked for an order of magnitude on the costs of moving the group's domicile, citing a possible exit tax on hidden reserves and the share of Swiss-booked assets belonging to non-Swiss clients. The CEO refused to engage at all.

Q: "I appreciate I do not comment too much on contingency planning, but just trying to understand really if you had to move the group's domicile out of Switzerland, how to think about some of the costs. Just thinking in terms of tax, whether something like $3 billion-$5 billion of exit tax on hidden reserves is the right kind of order of magnitude and whether that is something you could mitigate."
— Amit Goel, Mediobanca

A: "We are not really commenting on speculations or assumptions or opinions on any of this matter related to a contingency plan. And it is totally inappropriate and everybody can come up with their own opinions, but we will not comment that."
— Sergio P. Ermotti, Group CEO

Assessment: A firm refusal, not a denial. The question put a $3bn to $5bn number on the table, which is in the same range as one to two years of UBS's own estimate of the annual capital cost of the upper-house rule. That comparison is the economic logic behind the domicile debate, and management has chosen to leave it unaddressed.

How the 2027 Plan Treats Rules That Do Not Yet Exist

Asked how UBS allocates capital and investment in its 2027 planning cycle given the possible new rules and their effect on marginal returns, the CEO gave the clearest timetable yet and the basis for the February targets.

Q: "But given that you're going into the 2027 planning cycle already, when you're thinking about how you allocate capital and how you allocate investments across the business, how are you taking into account the new potential capital rules, given what it might mean for marginal returns, in some of those segments?"
— Andrew Coombs, Citi

A: "That is the reason why, as we mentioned before, we are not under any pressure to change anything on our current successful business model. We need to really assess exactly what it is before we take actions that may be then regret moves. I think that you will expect and you will see us presenting at the beginning of the year our refreshed medium-term plans, which will embed the 2028, 2029 outlook, considering that broadly speaking, the current regulation, other than the ordinance, as I said, as of January 1st, 2029, will be the one that is applicable."
— Sergio P. Ermotti, Group CEO

Assessment: Useful and slightly uncomfortable. The February targets will assume current regulation plus the ordinance, so they will show returns before the foreign-participations rule. That is defensible, because the rule is not law, but investors will need to apply their own haircut to whatever 2028 return the company publishes. The CEO also placed the earliest entry into force of the foreign-subsidiary rule in late 2027 or 2028, phased over years.

Whether Releasing the Buffer Means a Lower CET1 Target

The final question asked whether running the ratio from 14.7% down to 14% was the whole message, or whether management was hinting at a target below 14% in the new plan.

Q: "Are you basically just describing the fact that you are running the 14.7% CET1 down to the target of 14% again, or are you maybe hinting at a lower target than 14% in the next business plan?"
— Stefan Stalmann, Autonomous Research

A: "The first item is out at the end of the year, so there is no reason for us to keep excess capital for that reason. We are not going to keep any of that. In respect of the future capital requirements, as I mentioned before, we have plenty of time to build up capital that if and when requirements are coming into effect, we have the earning capacity to do it, and we are not going to just carry unnecessary capital for the next few years. We are not hinting at anything that is going to be then addressed with our Q4 results in respect of the future targets."
— Sergio P. Ermotti, Group CEO

Assessment: The second sentence of the answer matters more than the denial in the last. "We are not going to just carry unnecessary capital for the next few years" means the Swiss requirement will be funded from future earnings as it phases in, not pre-funded from today's distributions. That is the most shareholder-friendly statement on capital that management has made in our coverage.

How Much of the $55 Billion Will Ever Pay Fees

A questioner noted that fee-generating assets lagged net new assets by about $42bn and asked how sticky the remainder is and how much will convert into mandates. The answer identified the source and the opportunity, but put no timetable on conversion.

Q: "Then the second question, just in terms of net new money, I saw obviously very strong net new assets. Although, fee-generating assets a little bit lower. Just curious of the other $42 billion, how sticky it is and how much of that you expect to convert into mandates or fee-generating assets over time."
— Amit Goel, Mediobanca

A: "We are already benefiting from sizable inflows associated with these large liquidity events. That is what is contributing to the roughly $40 billion, by which, as you point out, NNA exceeded NNFGA this quarter. As Sergio mentioned as well, we have about 10% of our AUM, or about $500 billion is in other client assets or non-contracted or custodied assets as well. So all of these assets represent substantial potential to grow recurring fees and advisory and transaction revenues as we help clients diversify and invest over time."
— Todd Tuckner, Group CFO

Assessment: An honest answer that supports a cautious model. Liquidity-event proceeds are by nature lumpy, and clients who have just sold stock tend to diversify slowly. We credit the flows for franchise strength and give them little near-term revenue weight.

When US Adviser Numbers Normalize, and When the AT1 Appeal Is Decided

The question that has appeared in every quarter of our coverage was asked again, this time with a direct request to define "normal" for adviser headcount. The same questioner asked for any visibility on the Credit Suisse AT1 appeal. Both answers moved the story forward.

Q: "When would you see advisor numbers normalizing, and what does normal look like? Would that be stable numbers, or would it be slight growth, slight shrinkage? That would be interesting to know. Second question, unrelated, is there any visibility on the litigation from Credit Suisse AT1 holders? Is there any visibility on when you get the decisions on that appeal?"
— Jeremy Sigee, BNP Paribas

A: "No change there as far as we know. Decision not before Q1 of 2027. ... So I think a degree of net attrition is probably the industry norm that you see, as opposed to it being net flat I'd say as a general premise over the long term."
— Todd Tuckner, Group CFO

Assessment: Two useful clarifications. Adviser headcount will not "turn"; management now defines normal as industry-level net attrition, so growth has to come from productivity and same-store flows, which this quarter delivered. And the AT1 appeal now has a timing marker after three quarters without an update, which keeps the legal-tail bear point open into 2027.

Why Equities Grew Half as Fast as a Peer's

A questioner pointed out that UBS's equities revenues grew about 21% while a peer grew more than 40%, and asked whether geography or product mix explained the gap. Management declined to compare, then used the same answer to extend its wealth net interest income outlook into 2027.

Q: "I think you're up 21% year-on-year, versus your peer up 40%+. Just wanted to know what could explain the gap. Is it a geographic mix or product mix that explain the gap?"
— Nicolas Payen, Kepler Cheuvreux

A: "Look, on IB equities, not going to comment on the relative performance, but we're really pleased with the performance of our equities franchise, generated $2 billion of revenues in the quarter, up substantially year-on-year. ... We expect the higher dollar rate environment coupled with net interest margin support from the rollover of our replication portfolios at higher yields to generate similar NII annual growth trends in 2027 as we are seeing in 2026."
— Todd Tuckner, Group CFO

Assessment: The non-answer on equities is the cost of the capital-light model. UBS grows on flat resources, and it will not always keep pace with banks willing to commit more balance sheet in a strong market. We accept that trade-off, but it means UBS can lag in the best markets. The wealth net interest income comment is the more valuable disclosure: growth near this year's pace in 2027 is ahead of what we had modelled.

What They're NOT Saying

1. What 75% or 50/50 would cost. UBS quantified the upper house's 90% to the billion. It declined to quantify the two alternatives now being debated, including the one it supports, so investors cannot price the lower-house outcome from company data.
2. Whether relocation is off the table. The CEO set out a shareholders-first priority, said Swissness is not why clients bank with UBS, and twice declined to discuss contingency plans. He did not say UBS will remain Swiss-domiciled.
3. The size of the 2027 buyback. The buffer release is confirmed and the accrual will take CET1 to about 14%, but no number was given. On our arithmetic the excess above 14% was about $3.5bn at 30 September.
4. 2028 targets, for a third consecutive quarter. They are now promised for February on current regulation plus the ordinance, which means they will not include the foreign-participations rule.
5. The clean return. The release includes a $697m legal gain in underlying results without presenting the return on CET1 capital excluding it. On our arithmetic, it is about 15%.
6. That the integration cost guide was exceeded. Second-half integration expense of about $950m against the July guide of about $750m was presented as the cost of extra synergies, which it is, but not as an overrun.
7. The FRC foreign-exchange versus precious-metals split. Promised in February 2026 and still not provided after three more quarters.
8. A forward view on Asset Management flows. Net new money turned negative at $5.1bn, explained by one-off custody consolidation effects, with no indication of when flows return.

Market Reaction

• Pre-print setup: The ADR closed at $46.79 on 27 October, up 1.0% year to date against an S&P 500 up 12.8%, up 19.7% over twelve months and down 3.8% over thirty days. The 52-week closing range was $36.41 to $55.72. Zurich-listed options implied a move of about 6.2% in either direction through the November expiry.
• Between prints: from the $51.80 close on the second-quarter print day, the ADR fell 9.7% while the S&P 500 rose 5.6%. It first rose to a 52-week closing high of $55.72 on 31 August, 2.07x the then tangible book. It then fell 16.0% to the pre-print close. On 23 September, the day of the upper-house vote, it fell 2.9% to $47.74 against an index down 0.8%. Media speculation about potential suitors followed, and the ADR rose 1.5% on 6.6m shares the next day.
• Reaction session: UBS reports before the US open, so the print day is the reaction day. The ADR opened at $46.46 (-0.7%), traded a $45.93 to $46.74 range and closed at $46.69, down 0.2% or $0.10, with the S&P 500 up 0.3%.
• Volume: 2.6m shares against a 2.1m thirty-day average, or 1.2x normal.
• Following sessions: the ADR rose 3.1% to $48.13 on 29 October with the index down 0.3%, then fell 0.9% to $47.69 on 30 October with the index down 1.8%. Over the three sessions from the pre-print close it gained 1.9%.
• Position versus range: the reaction close is 1.73x the $27.00 tangible book per share, against 1.93x after the second-quarter print. It sits 22.0% above the $38.28 close on the day we initiated coverage.

A quiet reaction to a good quarter, because the stock is now priced off Bern rather than off earnings. The operating numbers were close to expectations at the profit line, and the positive surprises were about capital rather than earnings: an accelerated buyback, a buffer release and an extended net interest income outlook. Those support 2027 distributions, and the market had already de-rated the shares by 16% from their August high, through the upper-house vote. A session that opened lower and closed nearly flat suggests the capital messages offset the slight net profit shortfall and the more cautious fourth-quarter tone.

The more telling move came over the following two sessions, when the ADR outperformed a falling index by 3.4 points on 29 October and 0.9 points on 30 October. We would not over-read two days. But the pattern fits a market that had priced the parliamentary risk and was waiting for evidence that UBS would keep returning capital regardless. The release gave that evidence, and the domicile language reminded investors that UBS has options of its own in the negotiation.

Street Perspective

Debate: Is 90% the Ceiling or the Floor?

Bull view: The lower house is already weighing 75% CET1 and a mixed CET1/AT1 structure. Business associations oppose the rule, a referendum is possible, and the phase-in starts no earlier than 2028 and runs for years. The final number is more likely to be below 90% than above it, and the stock has already de-rated as if 90% were final.

Bear view: The government wanted 100%, the upper house voted 90% by a wide margin, and the central bank has called the 90% figure workable. The lower house may soften the number, but the chambers must agree, and the starting point for that reconciliation is now 90% to 100%, not 70% to 80%. A mixed AT1 structure has already been rejected once.

Our take: Neither side can price this well, and we do not think we can either. Our valuation gives equal weight to a softer outcome and to 90%, and the current price sits a little below that midpoint. That is a reasonable price, not a mispriced one, which is why the rating is Hold.

Debate: Is the Domicile Threat Real Leverage or a Bluff?

Bull view: The CEO's language was chosen carefully, the questions were not shut down, and the bank has said its clients do not bank with it because of its passport. A credible exit option strengthens UBS's hand in the National Council, and if the rules end up harsh, a relocation could restore returns for shareholders over time.

Bear view: A relocation would take years, need multiple regulatory approvals, carry an exit tax that one estimate on the call put at $3bn to $5bn before execution costs, and risk the Swiss deposit and client franchise that the capital debate is meant to protect. The political reaction to an explicit threat could also harden the rule rather than soften it.

Our take: It is leverage now and an option later, and both have value. We put no relocation value into our target. If the final rules land at or above 90% with no offsetting concession, the question will move from the Q&A into the strategic plan, and we would reassess then.

Debate: With the Integration Done, Does the Earnings Run-Rate Hold?

Bull view: Clean of legal items, the quarter earned about 15% on CET1 capital at a sub-70% cost/income ratio, with cost savings still arriving, a Swiss bank at an 18% pre-tax return on equity and wealth flows at record levels. Next year adds the integration buffer release, wealth net interest income growth similar to this year's, the US bank charter and a smaller Non-core and Legacy drag.

Bear view: The fourth quarter is starting cautiously, Asia Pacific transaction activity has cooled, the investment bank is back at its through-the-cycle return, and the wealth flows are partly custody assets that do not earn fees. Legal gains will not recur. Without the integration savings narrative, 2027 earnings growth has to come from revenue in a less helpful market.

Our take: The mid-teens return is sustainable under current rules, and that was the operating thesis. The 16.7% headline overstates it slightly and the 13.6% underlying growth understates the improvement. Our concern is not the numerator. The debate that still decides the stock is the denominator, and this quarter's earnings say nothing about it.

Model Update & Valuation Framework

Valuation, and a second admission. In July we said the price already reflected a parliamentary compromise that had not happened, and then marked our target to that price anyway. The diagnosis was right and the target was too generous. The ADR rose to 2.07x tangible book by the end of August, just below the 2.1x level at which we said we would move to Underperform. It then fell 16% by the eve of the print, as the upper house chose 90% rather than the 70% to 80% the price assumed. Our $55 target was too high. At today's price it would imply 17.8% upside, and we do not think that is there.

We now set the target explicitly from two outcomes, rather than from the traded multiple. A softening to 80% or below, or a mixed CET1/AT1 structure, combined with a credible 2027 capital plan, supports about 2.0x our forward tangible book of roughly $28.50, or about $57. The upper-house 90% version enacted as voted takes the sustainable return toward 12% to 13% and argues for 1.40x to 1.50x, or about $40 to $43. Equal weighting gives about $49. We set the 12-month target at $50, about 1.75x forward book, down from $55. From the $46.69 reaction close that implies +7.1%, plus the dividend, a market-like return, and the rating stays Hold. The current price implies about a one-in-three chance of the favourable outcome, a little more cautious than our own weighting but not enough for an upgrade.

What moves the target from here. The National Council debates the bill in December. A lower-house vote at or below 80%, or for a mixed structure, would move our weights toward the favourable case and the target toward the mid-$50s. A lower-house vote at 90% or above would move it toward the low $40s. The February results carry three further inputs: the 2027 capital return plan, the refreshed 2028 and 2029 targets, and any change in domicile language.

Thesis Scorecard: Q2 2026 Signposts Revisited

We set eleven signposts in July. The operating ones were met or beaten. The political one moved against the shares, and two operating items, integration cost and Asset Management flows, went the wrong way.

Standing Thesis Scorecard

Overall: The operating thesis is complete, and this quarter added the one piece that had been missing: an explicit commitment not to pre-fund the Swiss requirement out of distributions. We move Bull 4 to On Track for that reason. The bear point that has always decided the rating moved against the shares in September, and the share price has already adjusted. The thesis is unchanged in structure, stronger on capital return and weaker on regulation than in July.

Action: Hold. Neither published trigger fired: parliament has not concluded, 90% is above our 80% threshold, the buyback was accelerated rather than enlarged, and the shares peaked just below 2.1x tangible book. We cut the target to $50 from $55, set from explicit outcomes rather than the traded multiple, and keep the rating because the expected return from $46.69 is market-like. The December lower-house vote and the February capital plan are the two events that could change it.