The De-Rating Clause Fires: HSBC Upgraded to Outperform as a 2% Revenue Miss Takes the Stock to 1.8x Tangible Book

Results vs. Consensus

The third quarter is an interim management statement under UK convention: a finance-director presentation, an earnings release rather than a full set of accounts, and a fourteen-questioner call with no chief executive on the line. The comparison that matters is against the consensus HSBC compiles and publishes itself, seventeen analysts as of 16 October, because that is the book the London and Hong Kong markets trade against. On that book the quarter delivered a revenue miss and a credit beat of almost equal size.

3Q 2026 Scorecard

Two mechanical notes. First, each ADS represents five ordinary shares, so $0.42 of earnings per share excluding material notable items is $2.10 per ADS. That matches the company-compiled $0.42 and misses the higher aggregator figure of $2.22; the company book is the convention and the better guide. Second, the lines that missed and the lines that beat sort cleanly by type. Every revenue line tied to interest rates or wealth sentiment missed, and the two lines tied to credit quality and client trading beat. That pattern matters more for the thesis than the 0.7% at the pre-tax line.

Year-Over-Year Comparison

As in the second quarter, the reported column overstates the year and the clean column is the one to hold. The year-ago quarter carried $1.4bn of legal provisions, the Madoff and Herald claim in Luxembourg and a historical trading matter, so reported pre-tax profit rises 32% against a constant-currency, ex-notable rise of 10%. Earnings per share excluding material notable items rose 17%, faster than profit, because the Hang Seng minority no longer takes a share of the earnings. That is a structural gain, and it is the clearest evidence that the January privatisation is paying in the per-share numbers.

Quarter-Over-Quarter Comparison

The sequential column is where the quarter disappoints. Underlying revenue fell $344M from the second quarter, underlying profit fell $318M, and the return excluding notable items dropped 130 basis points to 18.2%. Banking NII was flat to the million in a quarter that had rate rises, structural hedge reinvestment and an extra day. Against that, the share count finally fell, by 53M shares as the $1bn buyback completed, and tangible book per share rose for the first time this year.

Nine-Month Comparison

Nine months is the better frame for a bank whose quarters move with notable items and seasonality. Underlying revenue grew 5.2%, underlying profit 7.2% and earnings per share excluding material notable items 14%, on a return of 18.8% against a 17%-plus target. Target-basis costs grew 2.0%, which is above the full-year guide of around 1% and is the arithmetic problem for the fourth quarter that the cost section takes up below.

Revenue. Reported revenue of $18,620M missed the company-compiled $18,934M by $314M, and the miss sat almost entirely in two lines. Banking NII of $11,656M was $247M short of $11,903M: the Street had modelled sequential growth off rate rises and hedge reinvestment, and instead time-deposit margin pressure in Hong Kong, lower Markets Treasury income and a $0.1bn adverse one-off cancelled both. Wealth fee income of $2,688M was $121M short of $2,809M. Against those, wholesale transaction banking beat by $73M and the corporate bank's debt and equity markets income grew 70% year over year. Revenue excluding notable items and strategic transactions grew 5.8%, so the franchise is still growing at the pace of the 2028 target; it simply did not accelerate the way the second quarter suggested it would.

Credit and costs. The $832M charge equals 31 basis points annualised, the lowest quarterly figure since our coverage began, against 41 in both the prior and year-ago quarters. Hong Kong commercial real estate produced what management called "a very low net charge," and the UK and Mexico both improved. On costs, target-basis expenses grew 1.4% year over year and reported costs were in line with consensus. The quarter did not cause the cost problem; the nine-month 2.0% growth rate did, by leaving the fourth quarter to fall year over year if the full-year guide is to hold.

Returns and earnings per share. Return on tangible equity excluding notable items of 18.2% matched consensus and sits above the 17%-plus target for a third consecutive quarter. Earnings per share excluding material notable items of $0.42 grew 17% year over year. The per-share result held up better than the revenue result for two reasons we regard as durable: the Hang Seng minority is gone, and the buyback has started to shrink the share count. The weak spot is tangible book per share, $0.11 below consensus at $9.41, partly because rising yields reduced the value of the bond portfolio held at fair value.

Segment Performance

HSBC's chief operating decision maker reads the business segments in constant currency, and the table below follows that convention. Revenue excluding notable items differs from reported revenue only in the corporate bank, the wealth and premier business and the Corporate Centre. Returns are nine-month annualised figures because the release discloses segment return on tangible equity on a year-to-date basis. Management told the call that each of the four businesses earned better than 17% in the quarter itself.

The table has two halves. Three businesses grew revenue between 6% and 10% and grew profit faster than revenue. The fourth, wealth and premier banking, shrank on both lines. Corporate Centre is the arithmetic bridge: the year-ago quarter carried $1.1bn of the Madoff and Herald legal provision there, which is why its profit swings by $1.1bn and why group reported profit rises 32% against a constant-currency, ex-notable rise of 10%.

Revenue by Income Line

The management view of reported revenue separates the three engines HSBC sells. It also shows what the strategic transactions removed from the comparison, which matters more this quarter than in any quarter we have covered.

Strategic transactions took $218M out of the year-on-year revenue comparison, and $124M of that sat in wealth. Remove it and wealth fee income grows roughly 5%. Remove the roughly $0.1bn of one-off insurance gains in the year-ago quarter, which the finance director flagged at the time, and the underlying rate is closer to 9% or 10%. That is a respectable number, and it is less than half the 21% the same line printed in the second quarter.

Banking NII by Legal Entity

The entity split locates the sequential stall more precisely than the call did. The Asian bank and the UK ring-fenced bank both grew. The decline came in the non-ring-fenced bank, which carries the corporate bank's European and markets business. At group level the release attributes the flat sequential outcome to lower Markets Treasury income and a $0.1bn adverse one-off, offset by hedge reinvestment and day count. The Asian bank's $64M of sequential growth, in a quarter when Hong Kong alone added $21bn of deposits, is the line that carries the time-deposit story.

Hong Kong

Profit before tax rose 8% to $2,565M on revenue up 6%, and the nine-month return on tangible equity excluding notable items reached 43.4% against 33.9% a year ago. Banking NII rose 8% year over year on deposit growth, offset in part by what the release calls "lower average margins on time deposits due to market competition." Fee income barely moved (up 1%), but wealth fee income inside it grew 7%. The credit charge fell 18% to $248M on lower Hong Kong commercial real estate charges. Lending grew $7bn sequentially, mainly in mortgages, against a housing market management described as up 14% year over year.

"We are benefiting from being full owners of the 2 iconic banks, from the growth of Hong Kong, and from our focus on protecting our leading market position." — Manveen Kaur, Group Chief Financial Officer

Assessment: This is still the best business in the group, and the third quarter shows the cost of defending it. Deposit growth of $21bn in a quarter is share defence after a $13.7bn privatisation, and management said plainly that it is growing deposits faster than it did a few years ago. The return on tangible equity can absorb that. The banking NII growth rate cannot absorb it without slowing, which is exactly what the sequential figure shows.

Corporate & Institutional Banking

The standout of the quarter. Profit before tax rose 34% to $3,605M, helped by the non-recurrence of a year-ago legal provision, and revenue excluding notable items rose 9%. Fee and other income grew 18% to $3,607M. Debt and equity markets grew 70% to $726M on equity derivatives and prime finance, and wholesale transaction banking grew 12% to $2,442M, led by foreign exchange at 19%. Investment banking grew 13% despite the wind-down of M&A and equity capital markets in the UK, Europe and the US. Wholesale credit and lending fell 40% to $90M, and the credit charge rose $54M because the year-ago quarter carried net releases in Asia.

"Foreign exchange had a particularly strong quarter, up 19% as we saw client demand in a period of higher market volatility." — Manveen Kaur, Group Chief Financial Officer

Assessment: In May we flagged the 26% first-quarter decline in debt and equity markets as the line to watch. The interim disclosure did not break it out. This quarter answers the question: the line grew 70% in the quarter and 17% over nine months, which implies a roughly flat first half. The caution is mix: foreign exchange and markets income are the most volatility-dependent revenue HSBC has, and a quarter in which they lead is a quarter that flatters the segment's run rate. Transaction banking fees up 11% at group level is the durable part.

International Wealth & Premier Banking

The weak segment. Profit before tax fell 14% to $1,125M, of which $91M was strategic transactions. Revenue excluding notable items fell 4%: banking NII fell 3% on lower Asian rates and disposals, and wealth fee income fell 2% to $1,904M after a $122M drag from strategic transactions and the non-recurrence of year-ago insurance gains. Investment distribution and private banking both grew on client activity. Costs rose 3% on continued wealth investment. The nine-month return of 26.1% against 19.7% a year ago says the business is still highly profitable; the quarter says its growth has stalled on the reported line.

"Private Banking grew 16%. We are really pleased our strategic focus is paying off. This quarter, we saw growth in mandates and higher trading activity. In Asset Management, AUM growth was offset by margin compression." — Manveen Kaur, Group Chief Financial Officer

Assessment: Most of the decline is perimeter, and the perimeter is shrinking on purpose: UK life, French life, Singapore life, Malta and Australia retail all leave this segment. That makes the reported line a poor guide for two more years. The issue we cannot explain away is the insurance line, where the year-ago one-off and disposals dominate a comparison that grew 21% on higher contractual service margin release only one quarter ago. One quarter of a flat line is not a trend. Two would be.

UK

Profit before tax rose 19% to $1,842M on revenue up 6%, the cleanest print in the group. Banking NII rose 7% on structural hedge reinvestment and balance growth. The credit charge fell a third to $182M on lower stage 3 charges in commercial banking, and costs were flat. Commercial lending grew 13% year over year. The headwind is on the liability side: management described UK retail deposit competition as increasingly intense, with fixed-term rates in the market often above the policy rate, and said the UK book was held stable by choice.

"In the U.K., we are particularly pleased with the 13% commercial lending growth year-on-year. You will hear more from our U.K. team in November, including how our leading position with U.K. companies provides us with great opportunities." — Manveen Kaur, Group Chief Financial Officer

Assessment: The UK is where the structural hedge works without complications. HSBC has the lowest loan-to-deposit ratio among the large UK banks, so it can decline to price for deposits it does not need, and it did exactly that this quarter. The November investor event is the next disclosure point, and the item to look for is a deposit-beta assumption rather than a lending target.

Corporate Centre

Profit before tax of $495M against a $614M loss a year ago, the swing almost entirely the $1.1bn legal provision in the base. Underlying revenue fell to a $194M loss on valuation losses in financial instruments and structural foreign exchange hedges. Associates contributed $652M, up 11%, on Saudi Awwal Bank and Bank of Communications. Markets Treasury revenue allocated out to the businesses fell to $423M from $606M a year ago, which explains a large part of the "Other" fee decline in the income-line table.

Assessment: Nothing here changes the thesis, but the Markets Treasury figure deserves a line in the model. A $183M year-over-year decline in treasury income allocated to the businesses is a quiet headwind that shows up as lower "other" income in every segment rather than as a single line anyone asks about.

Key KPIs

Two rows deserve a comment before the topic sections. The first is the composition of net new money: $41bn is the highest quarterly figure since the new definition was adopted this year, but $14.5bn of it is a one-time Indian government deposit scheme that closed at the end of August, and the investment half slowed to $17bn from $22bn. The second is tangible book. It rose only five cents in a quarter that earned 41 cents a share, because the dividend, the buyback and other reserve movements, including rate-driven valuation losses on the bond book, absorbed most of the profit.

Key Topics & Management Commentary

Overall Management Tone: Steady and unhurried, and noticeably more defensive than in August. The second-quarter call spent its time explaining upside; this one spent most of its time explaining a deposit mix, and management answered with history and ranges rather than with numbers. The finance director ran the call alone, kept every guidance line unchanged, and was most fluent on capital and credit and least specific on the two questions the market cared about most: the size of the time-deposit drag and the level of 2027 costs.

1. Hong Kong Deposits Migrate to Time Deposits, and Banking NII Stalls

The defining disclosure of the quarter. As Hong Kong rates rose, customers moved balances from current and savings accounts into time deposits, and the time-deposit share of the Hong Kong mix rose two points to 38%, mostly late in the quarter. Competing banks with higher loan-to-deposit ratios bid for the same money as Hong Kong lending recovered. HSBC chose to match the market rather than cede share, because it now owns all of Hang Seng and wants to grow with Hong Kong on both sides of the balance sheet. The result was $21bn of Hong Kong deposit growth in the quarter and almost no growth in group banking NII.

"In terms of emerging trends, in Hong Kong, we saw increased time deposit margin pressure this quarter. Most of our strong growth has been in time deposits. In the U.K., the deposit market has become increasingly competitive." — Manveen Kaur, Group Chief Financial Officer

Management framed the shift as a familiar rate-cycle behaviour rather than a structural break. The time-deposit share peaked at around 40% in the last high-rate period, against an industry average near 56%, and management said the deposit growth remains net-interest-income positive and incremental, even though the marginal contribution is smaller. Interest rate sensitivity to HIBOR has been reduced through hedging, so a further rise in HIBOR helps less than it used to but still helps.

"Now in terms of HIBOR sensitivity, yes, we have reduced it. But directionally, when HIBOR is higher, it is always positive for us. Interest rate increases are positive for banking NII. What really shifts or changes is the rate to which that incremental increase to banking NII happens." — Manveen Kaur, Group Chief Financial Officer

Assessment: The precedent supports management. The time-deposit share has moved between the low 30s and 40% across the last cycle, HSBC has always run below the market, and 38% is inside that historical band. What changed is the strategy, not the behaviour: in the last cycle HSBC gave up share to protect deposit cost, and this time it is paying for share because it has just spent $13.7bn on Hang Seng. That choice is defensible on returns, since Hong Kong still earns more than 40% on tangible equity, but it means banking NII growth in Hong Kong now tracks balance growth less a margin drag, not balance growth plus rate leverage. We model 2027 banking NII growth at 3% to 4%, below the roughly 6% that a question on the call attributed to consensus.

2. "At Least $46bn" Now Requires Only $11.45bn in the Fourth Quarter

Banking NII of $11,656M brings the nine-month total to $34,552M, so the "at least $46bn" guide needs only $11,448M in the fourth quarter. That is $208M below the third-quarter print, and the guide was reiterated rather than raised for the first time this year. The company-compiled consensus going into the print had the full year at $46.93bn, which required roughly $12.1bn in the fourth quarter on the Street's own third-quarter estimate. That number now has to come down.

"We benefited from a couple of rate hikes, although I would say that these were in line with expectations a few months ago. As you would expect, our structural hedge is being reinvested at higher yields." — Manveen Kaur, Group Chief Financial Officer

The sequential bridge offsets four items. Hedge reinvestment and day count each added roughly $0.1bn. A $0.1bn adverse one-off and lower Markets Treasury income took away the rest, and the release names time-deposit margin pressure as the drag inside the Hong Kong business. Net interest margin fell 5 basis points to 1.57% because interest-earning assets grew faster than income. The finance director also repeated a framing she has used before: rates of 3% to 4% are the "sweet spot" for banks, and above that range net interest income still grows, but more slowly.

Assessment: The guide is safe and the consensus is not. We carried $46.0bn to $46.5bn in August, below the Street, precisely because management had twice called its own guide a floor with buffer. A fourth quarter at the third-quarter run rate gives $46.2bn, and our range narrows to $46.1bn to $46.3bn. The Street's $46.9bn implied a second-half acceleration that the deposit mix has now ruled out, and the roughly $0.7bn reset is a large part of why the stock fell more than a 0.7% profit miss would justify.

3. Credit: 31 Basis Points, and the 45 Basis Point Guide Held

The $832M charge equals 31 basis points annualised, the lowest in our coverage and well below the 41 basis points of both comparison quarters. The nine-month charge is 41 basis points, which includes the $0.5bn fraud-related securitisation exposure first charged in the first quarter and the Middle East allowances. The full-year guide of around 45 basis points was held. Taken literally, it requires a fourth-quarter charge of roughly $1.5bn, about 56 basis points annualised.

"Our third quarter ECL charge was $0.8 billion, equivalent to an annualized charge of 31 basis points as a percentage of loans and advances. In Hong Kong commercial real estate, we had a very low net charge in the quarter. Overall, performance remains broadly stable." — Manveen Kaur, Group Chief Financial Officer

Management said it sees nothing of concern for the fourth quarter and explained the unchanged guide as procedure: the books stay open until the end of January, so it does not cut a credit guide early. It also said that "around 45" can be lower. The medium-term planning range of 30 to 40 basis points was reiterated, and management declined to guide 2027.

Assessment: This is the strongest part of the print and the one we discount most. A guide that implies 56 basis points in a quarter management describes as benign is buffer, and we model 40 to 42 basis points for the year, which puts the fourth quarter near 40. The caution is that 31 basis points is not a run rate. The release itself says the Hong Kong commercial real estate book "remains sensitive to further interest rate increases," and rates are rising. The credit tailwind that rescued this quarter's profit is the least repeatable line in the income statement.

4. Hong Kong Commercial Real Estate: Charges Fade, Collateral Keeps Weakening

The Hong Kong commercial real estate book shrank again, to $28.9bn from $29.3bn in June and $30.6bn in December. Credit-impaired balances fell to $5,593M from $5,989M, and sub-standard plus credit-impaired exposures fell to $8.1bn from $8.5bn, 95% of them secured. The net charge in the quarter was very low. The allowance nonetheless rose to $1,405M from $1,331M, and the weighted average loan-to-value on credit-impaired exposures rose to 83% from 82% in June and 71% in December.

"So yes, we are growing our deposit base more than we had a few years ago. But on the other side, we have good lending growth starting for the first time in Hong Kong. So that's the difference in context, aligned with a strong trajectory for GDP growth as well as from a credit risk perspective, normalization on Hong Kong commercial real estate side." — Manveen Kaur, Group Chief Financial Officer

Assessment: In February we told readers to watch collateral rather than migration, and the two have now separated completely. Migration has stopped: impaired balances are falling. Collateral is still weakening: the average loan-to-value on the impaired book has moved twelve points in nine months. The book's total allowance now equals about 25% of the impaired balance, against 17% at December, which is the right direction, and at 83% loan-to-value a secured loan still has collateral above its balance. The risk is concentrated in non-prime segments, which the release says still face higher vacancy, valuation pressure and weak investor demand. We keep this contained, with the loan-to-value ratio as the measure.

5. Wealth: $41bn of Net New Money and Flat Fees

The two wealth numbers point in opposite directions. Net new money was $41bn, of which $38bn came through the Asian bank, taking the nine-month figure to $105bn. Wealth fee and other income was flat at $2,688M. The flow figure includes $14.5bn of deposits from India's foreign currency non-resident deposit scheme, which closed at the end of August, partly offset by deposit run-off in the Australian retail business being exited. The fee figure lost $124M to disposals and roughly $0.1bn to a year-ago insurance one-off.

"Wealth net new money at $38 billion in Asia this quarter brings our year-to-date flow to more than $95 billion. We set this out on Slide 24. Our Wealth business continues to perform." — Manveen Kaur, Group Chief Financial Officer

Inside the fee line, investment distribution grew 7% on mutual fund and structured product sales, private banking grew 16% on mandates and trading, asset management saw asset growth offset by margin compression, and insurance fell on the year-ago one-off and disposals. Management declined to give a wealth growth rate and said quarterly product mix will vary.

"We do not give whether it's double-digit, high, low, or high-single targets or guidance for the year. We didn't do this year either." — Manveen Kaur, Group Chief Financial Officer

Assessment: The flow evidence is strong even after the one-off. Excluding the $14.5bn scheme, net new money was roughly $26.5bn, comfortably above the $20bn-a-quarter downgrade trigger we set in August, and net new invested assets of $17bn are the cleaner flow measure. The fee evidence is weaker than management allowed. Underlying growth near 9% to 10% is half the second quarter's rate, and investment distribution at 7% compares with 26% in the second quarter. Some of that is market mood: equity transaction volumes fell in Hong Kong. Wealth is the pillar the multiple pays for, and this was its weakest quarter in our coverage on the fee line.

6. The Corporate Bank Carries the Quarter

Corporate and institutional banking produced $3,605M of pre-tax profit, up 34%, and its fee lines were the ones that beat consensus. Fee and other income rose 18%: debt and equity markets 70%, wholesale transaction banking 12% with foreign exchange up 19%, and investment banking 13%. Group trade fees grew 5% and payments 4%. Securities services fee income grew 3% on new mandates and market levels.

"While this slide focuses on fees, I would point you to the data back. Wholesale transaction banking revenues, which give you the full picture of our world-leading trade and payments franchises, are up $0.8 billion and 12% year-on-year." — Manveen Kaur, Group Chief Financial Officer

Assessment: Network banking is doing what the bull case says it does in a volatile market: clients facing a Middle East war and rising rates bring their foreign exchange, hedging and trade business to a bank with HSBC's network. The quality is mixed. Trade and payments growth of 4% to 5% is the durable core, and foreign exchange and markets income at 19% and 70% are the cyclical top. For the model, we keep the corporate bank's 2027 fee growth at mid-single digits and treat this quarter's markets income as a high point rather than a new base.

7. Capital: Below the Range, a Basel 3.1 Cushion, and a Second $1bn Buyback

The CET1 ratio fell to 13.9% from 14.1%. Regulatory profits generated 90 basis points before the dividend accrual, a net 40 basis points after it. Risk-weighted asset growth took 30 basis points, rate-driven fair-value losses on hold-to-collect-and-sell bonds took 20 and the second-quarter buyback took 10. Management then gave its first estimate of the Basel 3.1 day-one impact, a benefit of 40 to 50 basis points from 1 January 2027, and announced a further buyback of up to $1bn to be completed by the full-year results.

"As always, we judge our capital position on a current and forward-looking basis. That positions us to deliver a further up to $1 billion buyback today after continuing to provide the resources to support our growth." — Manveen Kaur, Group Chief Financial Officer

The Basel benefit comes mainly from lower risk weights on sovereign exposures, small business lending and the foundation approach to large corporates. Management said a related reduction in its firm-specific Pillar 2A requirement would also help but by less. It described the 14% to 14.5% range as a guide rather than a hard floor and said buybacks can continue at its lower end.

Assessment: On a pro forma basis the bank is back inside its range, at roughly 14.3% to 14.4% once the Basel benefit lands, and the benefit is equivalent to roughly $4bn of capital at today's risk-weighted assets. That is the most valuable new number in the release for a shareholder, because buybacks are the residual and the residual just got larger. The pre-print expectation we saw had the buyback halving to $500M on the same capital pressures; $1bn was the better outcome. The negative is sensitivity: a 20 basis point hit from bond marks in one quarter is large, and the finance director's reply that "what goes off in September comes back also in October" is a statement about the yield curve, not a hedge.

8. Costs: 1.4% in the Quarter, "a Rounding Thing," and a 2027 Consensus Near 5%

Target-basis costs grew 1.4% year over year in the quarter and 2.0% over nine months, against a full-year target of around 1% on a $33.9bn baseline. Hitting 1% requires fourth-quarter target-basis costs near $8.77bn, roughly 1.7% below the $8.9bn of the fourth quarter of 2025. Management again said it may add performance-related pay if strong performance continues, and this time sized the effect more precisely.

"What it does in total terms is that it becomes a rounding thing. If you're without that VP uplift, you're running down to a 1% cost increase. And if you add to that VP, because of rounding, you may round up to the next percentage point." — Manveen Kaur, Group Chief Financial Officer

On 2027, the question came with a number: consensus at roughly 5% cost growth, or roughly 7% adjusting for disposals. Management did not give guidance and did not dispute the figure. It described the build as inflation and investment on a baseline that already includes simplification savings, plus redeployment of costs freed by disposals into growth businesses, partly offset by Hang Seng synergies in 2027 and 2028.

Assessment: The 2026 answer is now as precise as it will get: about 1% without extra variable pay and up to about 2% with it, which means the second half lands flat only if no extra variable pay is added. That puts the August upgrade criterion on course for a narrow miss, and we grade it in the trigger review below. The 2027 picture has moved since August. Consensus was about 3% then and is about 5% now, above our own 3.5% to 4.5%, so the Street has already marked the number up past the level we set as a downgrade trigger. Management's description of the build does not tell us on which side of 4% the guide will land.

9. Disposals: $2.2bn of Revenue Leaving Through 2028

Management set out the revenue cost of the exit programme by year: $2.2bn cumulatively, split evenly between banking NII and fees. The 2026 effect is $0.7bn, mostly fees. A further $0.9bn leaves in 2027, more balanced toward banking NII, and the final $0.6bn in 2028, mostly banking NII. In the quarter, Australia's $25.6bn retail loan portfolio moved to held for sale, Uruguay completed, and a strategic review of the retail and domestic mid-market business in Türkiye began.

"In this year, i.e., in 2026, the impact was $700 million, but more skewed towards fee. Next year, it will be a $900 million incremental, which will be more balanced towards banking NII. And as you go further out into 2028, the final $600 million will be more skewed towards banking NII." — Manveen Kaur, Group Chief Financial Officer

Assessment: This is the most useful modelling disclosure of the call and it explains part of the 2027 cost debate. A $0.9bn revenue headwind in 2027 against a 5% revenue growth target for 2028 means the continuing businesses must grow faster than the headline. Management's answer is that the redeployed costs will earn back the lost revenue "and more very quickly," with a short lag. We give partial credit: the four businesses all earn above 17%, so redeployment into them should be accretive, but the lag lands in the same year as the cost step-up.

10. China's Latest Guidance and Hong Kong Customer Acquisition

A new document from China's National Development and Reform Commission, published the Friday before the results, set out further rules on outbound direct investment and appeared to require some individual investors to register investments with provincial governments. Separately, Hong Kong new-to-bank customer numbers fell back from the elevated levels of May and June. Management linked the two months of high acquisition to competitors pausing onboarding while they upgraded compliance, and said acquisition has returned to a normal level of about 100,000 a month across HSBC and Hang Seng.

"So firstly, our monthly customer acquisitions, if you look at an average for the first half of this year or even last year, have stayed pretty steady at about 100,000 per month. So that continues." — Manveen Kaur, Group Chief Financial Officer

Management said it has seen no change in product mix or customer behaviour attributable to the mainland rules, pointed to the reaffirmed role of Hong Kong in China's new five-year plans, and said Hang Seng's account opening from mainland customers has more than doubled since it adopted HSBC's digital onboarding.

Assessment: In August we wrote that the risk from China's cross-border wealth rules expresses itself through enforcement rather than through the text. The new document is a step along that path: registration with provincial governments is an enforcement mechanism. Observed data still support management, since net new money from Asia was $38bn and acquisition is at the pre-May norm. We keep this at low probability and high consequence and note that the market priced it on the day: the bank's sensitivity to Beijing's oversight of wealth moving abroad was one of the reasons cited for the share fall.

11. Simplification at $1.9bn, and Hang Seng Under One Management Team

The organisational simplification programme has actioned about $1.9bn of the $2.0bn annualised savings target, for $1.6bn of the $1.8bn restructuring envelope. The savings reduced the income statement by $0.6bn in 2025 and $0.8bn in the first nine months of 2026. Full-time equivalent staff are down 4,390 since December, to 204,330. On Hang Seng, management has announced the combined management team and is building common technology platforms.

"And our continued discipline saw us action $0.2 billion of simplification saves, taking us to $1.9 billion of the $2 billion we set out at the half year." — Manveen Kaur, Group Chief Financial Officer

Assessment: Both programmes are on schedule, and both are now in their less visible phase. The simplification savings are actioned decisions, so the remaining $0.1bn by year end is close to certain, and the August payback arithmetic still holds: about ten months on the first tranche and about sixteen on the last. The Hang Seng synergies remain absent from the numbers, with $0.5bn of targeted synergies against $0.6bn of restructuring cost through 2028. The combined management team is the step that makes the cost synergies executable, and 2027 is the first year they can appear.

Guidance & Outlook

Nothing moved down and nothing moved up. The quarter's two new numbers, the Basel 3.1 benefit and the disposal phasing, are disclosures rather than guidance changes, and both help the model more than the headline. The reiteration of the net interest income guide is the line the market read most closely, because the second quarter had raised it and the third-quarter consensus had already priced another step up.

Implied fourth quarter. On banking NII, $46bn less the $34,552M nine-month total leaves $11,448M for the fourth quarter, $208M below the third-quarter print. On credit, around 45 basis points for the year implies a fourth-quarter charge of roughly $1.5bn, about 56 basis points annualised, against 31 just printed. On costs, around 1% growth on the $33.9bn baseline implies about $34.2bn, so after $25,472M in nine months the fourth quarter must come in near $8.77bn, roughly 1.7% below the year-ago quarter; with the extra variable pay management flagged, the fourth quarter would land near $8.9bn and flat.

Street position. The company-compiled consensus entering the print carried full-year banking NII of $46.93bn, a full-year return excluding notable items of 18.0% and a year-end CET1 ratio of 14.1%. The return figure is achievable: nine months are at 18.8%. The net interest income figure is about $0.7bn too high on the third-quarter run rate, and the CET1 figure is 20 basis points too high until the Basel benefit arrives in January. On 2027, the question on the call put cost consensus at about 5%, which is above our own range.

Guidance style. Conservative on revenue and credit, demanding on costs, and unchanged by policy. Management has now called its own net interest income guide a floor in three consecutive quarters, held a credit guide that implies a fourth quarter nearly twice the third, and for the first time faced a market that expected the guide to move and was disappointed that it did not. The finance director told the press that the market sometimes expects an upgrade. That is accurate, and it is a choice: a bank that guides to floors will miss a consensus that models through them.

Analyst Q&A Highlights

Whether the Deposit Mix Makes HSBC Negatively Geared to Higher HIBOR

The opening question went straight to the issue that moved the stock. It combined the reduced HIBOR sensitivity on a static balance sheet with the adverse time-deposit mix and asked whether, together, they leave the bank worse off when Hong Kong rates rise. Management rejected the conclusion while conceding the mechanism: higher rates still help, but by less.

Q: "And then sort of conceptually related to that, your rate sensitivity to an upwards move in HIBOR is very limited, assuming a static balance sheet. If you overlay that with negative mix, it suggests you could even be sort of negatively geared to HIBOR going up. I'm just wondering if that's a fair description. It feels a bit different to what we used to, if I can put it that way."
— Guy Stebbings, BNP Paribas

A: "Now in terms of HIBOR sensitivity, yes, we have reduced it. But directionally, when HIBOR is higher, it is always positive for us. Interest rate increases are positive for banking NII. What really shifts or changes is the rate to which that incremental increase to banking NII happens."
— Manveen Kaur, Group Chief Financial Officer

Assessment: The answer defends the sign and concedes the slope. A bank that hedged away much of its HIBOR upside to protect against falling rates in the first half, and then met rising rates with a time-deposit shift, has deliberately traded rate leverage for stability. That trade was right for the first half and costs something now. The practical consequence for the model is that rising Hong Kong rates are no longer a reason to raise the banking NII estimate.

Whether HSBC Has Changed Its Deposit Pricing Strategy in Hong Kong

The sharpest question of the call, asked last. It noted that the Hong Kong time-deposit share is near its previous peak while policy rates are well below their previous peak, and asked directly whether the bank had abandoned the approach it took in the last cycle, when it gave up share to protect deposit cost. Management said yes in substance and explained why.

Q: "However, if you look at Fed fund rate or the rate movement, it is still far from the last peak, right? So my question is that did we change our strategy in terms of deposit competition? Because in the last round, we gave up some market shares to maintain the deposit cost. So is there a change in strategy, number one?"
— Katherine Lei, JPMorgan

A: "So firstly, we are very mindful that now that we have invested in Hong Kong with the privatization of Hang Seng, we want to be fully participating with the growth in Hong Kong. So yes, we are growing our deposit base more than we had a few years ago. But on the other side, we have good lending growth starting for the first time in Hong Kong."
— Manveen Kaur, Group Chief Financial Officer

Assessment: The most important answer on the call, and the most honest. The time-deposit migration is cyclical behaviour; the decision to fund all of it at market rates is a strategy change, tied explicitly to the Hang Seng privatisation. It is defensible because Hong Kong lending is growing again and the segment still earns more than 40% on tangible equity. It also means the margin drag lasts as long as Hong Kong deposit competition does, rather than as long as customer behaviour does.

Whether the 45 Basis Point Credit Guide Is Simply Conservatism

A direct question about the gap between a 31 basis point quarter and a full-year guide that requires a large fourth-quarter step up, followed by a request for a 2027 view and a question on whether consensus captures the disposal effects. Management confirmed the conservatism, explained it as procedure, declined to guide 2027 credit, and used the disposal half of the question to give the phasing of $2.2bn of lost revenue.

Q: "The print in Q3 is a lot lower than expectations, and year-to-date, the run rate implies quite a big step-up in Q4. I think you said in your prepared remarks that you're not seeing anything that concerns you. So is the reiteration of guidance just some conservatism?"
— Benjamin Toms, RBC

A: "And of course, when we give a guidance of 45 basis points around that range, it can very much be lower too. But as a matter of conservatism, given the length of the quarter, we keep the books open right until the end of January. So therefore, that's the only reason why we are not changing with regard to the guidance on the ECL charge."
— Manveen Kaur, Group Chief Financial Officer

Assessment: As close to a guided cut as a bank will give without changing the number. "The only reason" is a strong phrase from a finance director, and we read the full-year charge at 40 to 42 basis points. The same answer also explains why the market did not reward the credit beat: a buffer that management admits to is a beat the market had already partly assumed.

Why the Bank Is Comfortable Running Below Its Capital Range

A two-part capital question: why a 13.9% ratio below the 14% floor is acceptable, and why rising bond yields hit HSBC's capital harder than European peers' through the hold-to-collect-and-sell book. Management treated the range as a guide rather than a floor, pointed to the Basel 3.1 and Pillar 2A benefits, and defended the bond book's size and duration.

Q: "So why are you happy running below your capital target this quarter? I presume the Basel 4 update gave you some of that confidence. So should we expect you to pay down to 14 -- the lower end of your capital target pro forma the Basel benefit at year-end?"
— Robert Noble, Deutsche Numis

A: "We feel very comfortable that at the lower end of that capital range also, we can confidently do share buybacks and nothing changes there. So I wouldn't use that just as a one cliff edge."
— Manveen Kaur, Group Chief Financial Officer

Assessment: A useful clarification for anyone modelling buybacks. If the lower end of the range is not a constraint and Basel 3.1 adds 40 to 50 basis points in January, the capital available for distribution in 2027 is larger than the 13.9% headline suggests. The bond-book answer was less convincing: describing a mark-to-market that reverses when yields fall is not the same as reducing the exposure, and in a rising-rate year the capital ratio will keep absorbing it.

Whether 2027 Cost Consensus Near 5% Is Too High

The cost question, put with numbers: consensus 2026 costs already sit about $300M above the 1% target, and 2027 consensus implies about 5% growth, or about 7% adjusting for disposals. Management declined both halves. It said variable pay is set on the quality and quantum of outperformance rather than on consensus, gave no 2027 number, and described the components of the build.

Q: "And then the second one is, if I look at consensus costs for next year, 2027, it's looking like somewhere around 5% growth. And if I adjust for the disposals, it's probably more like 7%. And that feels like a really very punchy number, particularly after a 1% growth this year or an underlying 1%. So I'm just -- am I missing something?"
— Edward Firth, KBW

A: "So firstly, we are not giving any guidance on costs for next year. As of now, of course, for the full year, we will have that update."
— Manveen Kaur, Group Chief Financial Officer

Assessment: The second quarter in a row in which a 2027 cost consensus was put to management and left alone, and this time the number is higher. In August a 3% consensus drew a refusal; now a 5% consensus draws the same refusal. A management team that thought 5% was badly wrong had every reason to say so on a day the stock fell 6%. We read the silence as a guide that will not start with a 3. Consensus has already moved to about 5%, so the question for the stock is no longer whether 2027 costs grow faster than 4% but whether they grow faster than 5%. The trigger review below sets out how we treat the threshold we published in August.

Whether the Revenue From Redeployed Costs Will Exceed the $2.2bn Lost to Disposals

A precise question about the missing line on the disposal slide: the slide shows the revenue lost and the costs freed, but not the revenue the freed costs will earn once they are reinvested. Management said the reinvested costs should earn back the lost revenue and more, quickly, with a short lag.

Q: "I guess the one thing missing from that slide is the revenue benefit from that cost reallocation. So could I just, I guess, ask whether you think the marginal return on reallocating that cost base will generate more or notably more than the $2.2 billion of revenue attrition that you've earmarked?"
— Andrew Coombs, Citi

A: "And you can see from our 4 global businesses, given where they stand on the return on tangible equity, I feel very confident that we will make that revenue and more very quickly. So that should be an upside. However, I'll remind you that there will inevitably be a very short time lag quarter-on-quarter in terms of how the revenue that has been reduced in one space gets picked up in another deployment."
— Manveen Kaur, Group Chief Financial Officer

Assessment: A confident answer without a number, which is the pattern of the whole call. The logic is sound: costs moved from sub-scale retail into businesses earning 18% to 43% on tangible equity should produce more revenue per dollar. But "very quickly" with "a very short time lag" is not a forecast, and the 2027 model has to carry $0.9bn of lost revenue with certainty against reinvestment revenue that is only promised.

Whether $41bn of Net New Money Is Repeatable

A question on the quality of the quarter's best number: how much of the Asian inflow came from the one-time Indian deposit scheme, and whether the flow is sustainable without it. Management confirmed the scheme closed in August and will not recur, and argued that the rest of the flow reflects broad, multi-product relationships.

Q: "And then just looking at your net new money, which was very strong and very encouraging this quarter. Is that number sustainable? Or does that have some part coming from India that perhaps is just a onetime uplift that we are seeing, especially in the Asia portion?"
— Melissa Kuang, Goldman Sachs

A: "And going forward, of course, we will continue to have the benefit of this FCNR for the coming 3 to 5 years, and most of our book is at the 5-year end, but that FCNR uplift has now closed. It closed at the end of August. So there will be nothing new."
— Manveen Kaur, Group Chief Financial Officer

Assessment: The honest answer, and the one that should reset fourth-quarter expectations. Excluding the $14.5bn scheme, group net new money was about $26.5bn and Asian net new money about $23.5bn, the latter in line with the 8% to 9% annualised run rate management cited in August. The deposits stay for three to five years, which is useful funding, but the flow figure will fall back in the fourth quarter and should not be read as a deterioration when it does.

Whether China's New Guidance Changes the Wealth Story

A question linking the lower Hong Kong new-to-bank figures to the mainland's latest document on outbound investment, which appeared to add a registration requirement for some individual investors. Management treated the May and June peak as an anomaly caused by competitors pausing onboarding, and the new document as further guidance under existing rules.

Q: "We had this NDRC document from China on Friday. It is mostly about ODI that would take the point, but it does seem to mention individual investors having to register investments with provincial governments. So are you comfortable with that document doesn't really change the story a whole lot from May and June, and it's really mostly about ODI?"
— Alastair Warr, Autonomous

A: "And I would just reiterate, all the guidance we are seeing coming is actually further guidance to continue to comply with existing Chinese regulatory requirements. So that's really the stance with which we would take. We actually welcome that when there is a level playing field and clarity, because we don't want to lower our compliance standards to attract more new customers."
— Manveen Kaur, Group Chief Financial Officer

Assessment: The answer is consistent with the August answer and with the data, and it does not address the registration point the question raised. A requirement for individuals to register outbound investments is exactly the kind of enforcement mechanism that can slow cross-border wealth flows without changing a single rule HSBC must comply with. Management's "level playing field" framing is fair, and it is also the argument of a bank that benefits most when competitors are slowed. We keep this on the watch list.

What They're NOT Saying

1. The size of the Hong Kong time-deposit drag: the mix shift moved the stock, and the call gave a direction, a historical range and a reassurance that the effect is still positive at the margin. It gave no basis-point figure for the margin pressure and no sensitivity of banking NII to each point of mix. A model has to guess the one variable that explains the quarter.
2. Any 2027 cost number: for the second consecutive quarter, a published consensus was put to management and left standing. In August that consensus was about 3%; it is now about 5%. Declining to correct a number that has moved two points higher is information in itself.
3. What happened to the $300M Middle East reserve: in August management set a policy of two clear quarters of stability before releasing reserves. The third quarter was the second such quarter. The nine-month commentary still describes Middle East allowances, and neither the release nor the call said whether the reserve was held, used or released.
4. Updated private-credit exposure: the roughly $6bn total and roughly $3bn securitisation-financing figures from the first quarter have not been refreshed in two releases. The nine-month commentary now sizes the single fraud-related securitisation exposure at $0.5bn, against $0.4bn when it was first charged. Whether that is rounding or a top-up on the same name was not said.
5. Quarterly returns by business: management said each business earned more than 17% annualised in the quarter, but the release shows only nine-month segment returns. After a 14% profit decline in wealth and premier banking, the quarter's return for that business cannot be checked.
6. The group insurance contractual service margin: the release answers one August question by disclosing that $1.4bn of contractual service margin moves to held for sale with the Singapore life business, about 9% of the $15.2bn balance last disclosed in the first quarter. It does not update the group balance, so the trend in the accrual that underpins insurance revenue cannot be tracked through a quarter in which insurance income fell.
7. The scale of the Türkiye review: a strategic review of the retail and domestic mid-market business began this quarter, adding to the fifteen exits announced since 2025. The release gives no revenue, cost or capital figure for the business under review.

Market Reaction

• Pre-print setup: the ADR closed at $91.21 on 26 October, up 15.9% year to date against 12.1% for the S&P 500, up 35.7% over twelve months and down 9.6% over the trailing 30 days from $100.95 on 25 September. The 52-week closing range was $66.26 to $107.86. The stock had already fallen 11.6% from $103.21 at our August note, with UK-specific pressure prominent: long gilt yields at their highest since 1998 and reports of a bank tax in the 28 October UK budget.
• Reaction session (27 October, results released before the US open): the ADR opened at $85.55, a 6.2% gap down, traded between $84.89 and $86.57 and closed at $86.01, down 5.7% or $5.20, on 4.0M shares against a 1.8M 30-day average, 2.2 times normal.
• Relative: the S&P 500 rose 0.6%, a 6.3-point relative underperformance. The UK bank ADRs did not follow: Barclays rose 0.85%, Lloyds fell 0.37% and NatWest fell 1.29%. The move was specific to HSBC.
• Primary listings: London closed down 5.42% at 1,291.40p after trading as much as 7.5% lower, while the FTSE 100 fell 0.75%. Hong Kong closed down 5.6% at HK$134.30, its lowest level in almost 20 weeks.
• Following sessions: the ADR closed at $86.37 on 28 October and $86.60 on 29 October, then fell 1.36% to $85.42 on 30 October on a day the UK bank ADRs fell between 3.6% and 9.0%. That leaves the ADR 6.3% below its pre-print close and 17.2% below our August note.
• Peer read: Standard Chartered reported the next day with a 27% profit increase and a raised income target, and its Hong Kong shares rose 8%. Two Asia-focused UK banks reported in the same week, and the market rewarded the raise and sold the reiteration.

A reiteration into a raised bar. The 0.7% pre-tax miss does not explain a 5.7% fall on more than twice normal volume. Three things do. The revenue miss came from the two lines the multiple pays for, banking NII and wealth, while the offset came from the line the market capitalises least, a low credit charge management itself called conservative. The net interest income guide was held flat for the first time this year, against a consensus that had modelled $0.9bn more for the full year than the guide requires. And wealth fee income, which had grown by double digits in each of the two prior quarters, printed flat. Each point alone was a disappointment; together they took away the growth story that justified 2.2 times book in August.

The price had already moved. The setup matters as much as the print. The stock entered the results 11.6% below our August level on UK fiscal and gilt fears, and fell a further 5.7% on the day. Since August the ADR has lost 17.2% while tangible book per ADS has risen from $46.80 to $47.05 and earnings per share excluding material notable items have grown 17% year over year in the latest quarter. The multiple has fallen from 2.21 times to 1.82 times spot tangible book. The market has repriced HSBC from a bank that would beat its guidance to a bank that will deliver it, and at 1.82 times we think that repricing is done.

No second-day fade. In August, a modest first-day decline was followed by a heavier second-day decline on more volume, which we read as a decision taken after the sell side had read the transcript. This time the pattern reversed: the ADR recovered modestly on 28 and 29 October on lower volume, and the 30 October decline came with a broad UK bank sell-off rather than with HSBC-specific news. The selling looks concentrated in the reaction session rather than continuing.

Street Perspective

Debate: Is the Hong Kong Deposit Mix a Cyclical Drag or a Margin Reset?

Bull view: Time deposits at 38% of the Hong Kong mix are inside the low-30s-to-40% band of the last cycle and far below the industry's roughly 56%. The shift is net-interest-income positive at the margin, hedging has stabilised the book, and when rates stop rising the mix normalises. The bank that holds the cheapest funding in Hong Kong is still the best-funded bank in Hong Kong.

Bear view: The behaviour is cyclical but the strategy is not. HSBC has said it will fund Hong Kong deposit growth at market rates to grow with the market after the Hang Seng deal, which means the margin drag persists as long as competitors with higher loan-to-deposit ratios keep bidding. Banking NII was flat in a quarter with rate rises, hedge reinvestment and an extra day, which is what a margin reset looks like.

Our take: The bears are right about the strategy and the bulls are right about the level. We take the strategic point seriously and model 2027 banking NII growth of 3% to 4% rather than the roughly 6% the Street carried. But a bank earning more than 40% on tangible equity in Hong Kong can pay for share and stay highly profitable, and at 1.82 times book the stock no longer needs rate leverage to work.

Debate: Was the Credit Beat Real Earnings or Borrowed Buffer?

Bull view: Thirty-one basis points is the result of a cleaner book: Hong Kong commercial real estate charges have faded, impaired balances are falling, UK and Mexican charges improved, and management said nothing concerns it in the fourth quarter. The 45 basis point guide is procedural conservatism and the full year will land near 40.

Bear view: A single quarter at 31 basis points is not a run rate. Average loan-to-value on impaired Hong Kong property loans has risen to 83% from 71% since December, the release says the portfolio is sensitive to further rate rises in a year of rising rates, and the Middle East reserve has not been released. The credit line paid for the revenue miss, and it will not do so every quarter.

Our take: Both, in proportion. The full-year charge will land below guidance, and we carry 40 to 42 basis points. But we do not capitalise 31: our 2027 assumption stays in the 35 to 40 basis point part of management's planning range, and the loan-to-value trend in Hong Kong is the measure we watch.

Debate: Has the De-Rating Already Priced the 2027 Cost Step-Up?

Bull view: Consensus has moved from about 3% to about 5% for 2027 cost growth since August, and the stock has fallen 17% over the same period. At 1.82 times tangible book the implied cost of equity on the 17% target is about 10.7%, roughly where it sat at our May initiation, before the credit and capital-return questions resolved. Add 40 to 50 basis points of Basel 3.1 capital from January and a dividend yield near 5%, and the cost step-up is in the price.

Bear view: Management declined to dispute a 5% consensus on a day the stock fell 6%, and its description of the build, inflation plus investment plus redeployed disposal costs, sums to something in that range. With $0.9bn of revenue leaving through disposals in 2027 and banking NII growth slowing, 2027 jaws could turn negative, and 2027 earnings consensus has further to fall.

Our take: The bulls, narrowly, and this is the debate that sets the rating. We carry 2027 cost growth at 3.5% to 4.5%, below consensus, but even at 5% the arithmetic holds: on our numbers a 17% return on tangible book that compounds at about 8% a year, bought at 1.82 times, plus a 5% dividend, is a total return of about 13% before any re-rating. The risk the bears describe is real and dated, and we set out below exactly what result in February would reverse this call.

Debate: Is Wealth Slowing or Just Lapping Disposals?

Bull view: Strip out $124M of disposals and a roughly $0.1bn year-ago insurance one-off and wealth fee income grew about 9% to 10%. Private banking grew 16%, net new money was $41bn, and the investment half of the flow was $17bn. A flat reported line in a quarter with weak Hong Kong equity volumes is a strong result.

Bear view: Even the bulls' underlying figure is half the second quarter's rate, investment distribution slowed from 26% to 7%, insurance fell, and asset management margins compressed. The new mainland document adds registration for individual outbound investment. The pillar the multiple pays for just had its weakest fee quarter in our coverage.

Our take: Slowing, from an unsustainable rate, and not breaking. Growth of 9% to 10% underlying is consistent with the bank's 2028 revenue target and with Asian wealth balances still growing at 8% to 9% annualised from net new money excluding the Indian scheme. We hold the pillar at on track but mark it as the one that has to re-accelerate in the fourth quarter.

Model Update & Valuation Framework

Two corrections to our own August numbers belong at the front. Our exit-2026 tangible book of about $10.00 per share was too high: the third quarter added five cents rather than roughly thirty, and we now carry about $9.70. And our 2026 earnings estimate falls about 4%, to $1.71, because the third quarter earned $0.42 against the $0.45 run rate we extrapolated. Both changes reduce value. The price fell further.

Valuation. The ADR closed at $85.42 on 30 October. Each ADR represents five ordinary shares, and tangible net asset value was $9.41 per share at 30 September, so tangible book is $47.05 per ADS and the stock trades at 1.82 times. In August it traded at 2.21 times ($103.21 on $46.80). Holding long-run growth at 3%, 1.82 times implies a cost of equity of about 10.7% on the company's 17% target, against 9.3% in August and 10.6% at our May initiation. Put the other way, at a 10% cost of equity the price now requires a sustainable return of about 15.7%, against 18.4% in August. On earnings, the ADR trades at 10.0 times our revised 2026 estimate of $8.55 per ADS; on the same estimate the August price was 12.1 times, and on August's own $1.78 estimate it was 11.6 times.

The comparison with May needs stating plainly, because the multiple is now where it was when we initiated at Hold. In May, at 1.85 times, we said Hold because credit and capital return were both unresolved and both would be decided in the second quarter. Since then the credit charge has fallen from 52 to 31 basis points, the fraud charge has not recurred, $2bn of buybacks have been completed or announced, and the three strategic reviews open in May have closed. Basel 3.1 adds 40 to 50 basis points of capital that did not exist in May's numbers. Against those, two new risks have appeared, the Hong Kong deposit mix and the 2027 cost line, and tangible book per share is five cents lower than in May. The same multiple now buys a franchise with smaller open questions.

Framework rather than a target. We continue not to publish a price target. Our twelve-month forward tangible book is about $10.15 per share, or $50.75 per ADS: $9.41 today plus about $1.72 of earnings, less about $0.85 of dividends paid and about $0.11 of dilution from $4bn of buybacks above book. On that base, 1.6 times implies $81.20, about 4.9% below the current price; 1.8 times implies $91.35, about 6.9% above; 2.0 times implies $101.50, about 18.8% above; and 2.2 times, the August multiple, implies $111.65, about 30.7% above. Add a dividend yield of about 5.0% on $4.28 per ADS. Total return therefore runs from roughly zero in the bear case to roughly 12% if the multiple simply holds and roughly 24% at 2.0 times.

Why 2.0 times is the base case. A bank that earns 17.5% on tangible equity, grows at 3.5% and is priced at a 10.5% cost of equity is worth 2.0 times tangible book. HSBC has earned 18.8% this year, guides to 17% or better for three years, and has a 50% payout plus a buyback. We do not need the 2.21 times of August, which we called full. A partial recovery to 1.9 to 2.0 times forward book, worth $96 to $102 per ADS, delivers a total return of roughly 18% to 24%, well ahead of what we expect from the index. That is the definition of the rating.

Grading the August Triggers

In August we published the conditions under which the rating would change. We grade them here verbatim before setting new ones.

None of the two-of-three upgrade criteria was met in full, and we do not pretend otherwise. The upgrade rests on the standalone de-rating clause, and its condition is the judgement in this note. We read "without operational deterioration" against what the clause was written to catch: a falling price on a franchise whose profits, returns, credit and capital hold up. Underlying pre-tax profit grew 10%, returns held at 18.2%, credit improved by 10 basis points and the bank generated 90 basis points of capital in the quarter. Revenue growth slowed from 7% to 4.5% and missed consensus, and that is a deceleration we have priced into lower estimates. It is not a deterioration of the franchise. A reader who weights the revenue miss more heavily would keep the Hold, and we set out below the results that would prove that reader right.

New triggers, and every change from August. The August downgrade triggers were written for a Hold rating, where a downgrade meant Underperform. At Outperform, a downgrade means a return to Hold, and we restate the list on that basis.

• 2027 target-basis cost growth guided above 5%. Changed from above 4%. The August threshold was set against a consensus of about 3%; consensus is now about 5%. The new level sits above both consensus and our own 3.5% to 4.5% range, so it fires only on a guide that is worse than what the price now discounts. A guide between 4% and 5% fires nothing, and we say so now rather than in February.
• Fourth-quarter banking NII below $11.45bn, which would leave the full year below the "at least $46bn" floor. New. Our range is $11.55bn to $11.75bn.
• A second sponsor-intermediated or private-credit charge. Carried unchanged.
• The 2026 credit guide raised. Carried unchanged.
• Quarterly net new money below $20bn, excluding one-time schemes such as the Indian deposit scheme. Carried, with the exclusion added. The exclusion makes the test stricter, not looser.
• A re-rating above 2.2 times spot tangible book without a matching improvement in the operating outlook. New. That is the level at which we held the stock at Hold in August.

Conviction would rise on a fourth-quarter buyback of $1.5bn or more, a 2027 cost guide at or below 4%, or a Basel 3.1 benefit at the top of the range with a stated distribution plan. Netting, stated before any of it fires: if February brings a cost guide above 5% together with a buyback of $1.5bn or more, the two offset and the rating stands. Every other downgrade trigger acts alone.

Thesis Scorecard Post-Earnings

The pillars are the ones carried in the standing thesis since initiation, with one bear point added this quarter. Status tags follow the thesis convention: bull pillars run on track, at risk or broken; bear points run contained, emerging or materialising.

Overall: The operating thesis is slightly weaker than in August and the investment case is stronger. Two pillars weakened at the margin, wealth fees and the new deposit-mix bear point, and one bear point de-escalated, the valuation. The weakening is measurable and moderate: lower banking NII and wealth estimates, about 4% off 2026 earnings and $0.30 off exit-2026 tangible book. The de-escalation is larger: the stock fell 17%.

Action: Upgrade to Outperform, conviction 6. Buy the franchise at 1.82 times tangible book with a 5% dividend yield and a buyback that Basel 3.1 should enlarge in 2027. Conviction starts at 6 rather than higher because the call depends on the multiple recovering toward 2.0 times, and because the February cost guide and fourth-quarter net interest income are dated tests that could reverse it.

Bottom Line

In August we held HSBC at Hold with every operating question answered, because the market had paid for the answers in advance: 2.21 times tangible book, a 9.3% implied cost of equity, and a total return that depended on a demanding multiple simply holding. We wrote that a de-rating toward 2.0 times without operational deterioration would change the call. Over the following eleven weeks the ADR fell 11.6% on UK bank-tax and gilt fears, and then 5.7% more on a quarter that missed revenue consensus by 1.7% and matched it on returns.

The quarter was not good. Banking NII stalled as Hong Kong customers moved into time deposits and HSBC chose to pay for share; wealth fees went flat; the net interest income guide was held instead of raised; and the capital ratio dipped below its range. We have cut our estimates for all of it: 2026 earnings about 4% lower, exit-2026 tangible book $0.30 lower and 2027 banking NII growth of 3% to 4%. None of it broke the franchise. Underlying profit grew 10%, the return on tangible equity was 18.2%, credit improved to 31 basis points, $1bn of stock was repurchased and $1bn more announced, and Basel 3.1 will add 40 to 50 basis points of capital in nine weeks.

At $85.42 the stock trades at 1.82 times tangible book and 10 times this year's earnings, the same multiple as at our May initiation but with the credit and capital questions of May resolved. A 5% dividend, about 8% annual growth in tangible book and a partial recovery toward 2.0 times give a total return near 20%. We upgrade to Outperform, and we have said exactly which results in February would send us back.