HSBC's Best-in-Class Deposit Franchise Became Its Loudest Problem

HSBC delivered what on paper is a near-flawless quarter. Revenue rose 5% to $18.7bn, profit before tax climbed 10% ex-notables to $10bn, earnings per share ex-notables were up 17%, annualised return on tangible equity ran at 18.8% year-to-date and the CET1 ratio landed at 13.9% after generating 90 basis points of capital in three months. Management declared guidance unchanged, announced an up-to-$1bn buyback and, in an unusually generous flourish, said it was “on track to meet all our guidance.”

Then the questions came. Ten analysts. Not one asked about trade finance, CIB's standout fee quarter, or the 19% surge in foreign exchange revenue. They asked, in six different phrasings, the same question: what is happening to your deposit base in Hong Kong?

The price of being the best-funded bank in the room

Joe Dickerson of Jefferies framed it better than management did:

In a weird way, you're being penalized for having the best funding position in the market.

He is right, and the mechanism is simple. As Hong Kong rates rose, HSBC's customers did what customers always do: they moved money out of current and savings accounts into time deposits. Group CFO Pam Kaur put the shift at two percentage points, arriving “more increasingly towards the end of the quarter as interest rates got higher,” and volunteered the historical ceiling: “the time deposit proportion can move from sort of low 30s to right up to 40%” — Manveen Kaur, Group CFO · 2026-10-27. Every basis point of that migration raises the cost of the cheapest liability HSBC owns.

What makes this a genuine change rather than a rate-cycle footnote is that HSBC has spent a year telling investors its Hong Kong rate sensitivity was being hedged away. In August, Kaur noted that the bank had “increased our Hong Kong hedge” and taken structural hedge sensitivity down. Today she reaffirmed the hedge but conceded the mix effect runs the other way: “3% to 4% interest rate is at the peak of the sweet spot for banks, even though beyond that, we continue to grow banking NII, albeit at a slower rate” — Manveen Kaur, Group CFO · 2026-10-27. Guy Stebbings of BNP Paribas pushed the logic to its conclusion — if the hedge is neutral and the mix is negative, HSBC could be negatively geared to HIBOR. That is not the franchise investors thought they owned.

Note also, buried inside the deposit section, the sentence that reframes the whole thing: “we are growing our deposit base more than we had a few years ago” — Manveen Kaur, Group CFO · 2026-10-27. Owning Hang Seng outright converts HSBC from a bank that optimised deposit pricing for margin into a bank that defends share in its home market. The deposit ratio is still the envy of the London peer group — the lowest loan-to-deposit ratio of the major UK banks — but a ratio is a stock; the market is now paying for the flow. That trade-off is the single most important thing said on this call, and it was said almost in passing.

There is a second, harder edge. The $38bn of Asia net new money includes $14.5bn from India's FCNR scheme — a deposit inflow that, in Kaur's words, “closed at the end of August. So there will be nothing new.” In the same breath management argues deposit growth remains incrementally NII-positive. Both can be true. But the market needs to decide how much of Q3's deposit story recurs.

What management quietly stopped saying

The most revealing signal in this quarter is a keyword that isn't here. Simplification saves was the company's single highest-momentum theme as recently as the third quarter, and today it is the largest decliner in the entire keyword set — it has fallen out of the top-30 entirely. The topic still appears in the prepared remarks ($1.9bn of the $2bn target actioned), but not one analyst asked about it, and management did not volunteer it.

What replaced it is variable pay, a fresh entrant at the top of the keyword table. Kaur's framing is disarming: the potential uplift “becomes a rounding thing,” taking cost growth from 1% to perhaps the next percentage point. But Ed Firth at KBW did the arithmetic the other way and asked whether consensus's 7% underlying cost growth for 2027 is remotely credible after a 1% year. Management declined to guide. The cost-discipline narrative that Georges Elhedery built through 2025 — the one that justified upgrading saves from $1.5bn to $2bn — has quietly become a story about redeployment and rounding.

Relatedly, strategic transactions arrives as a top-ten theme with genuinely new numbers attached: a $2.2bn cumulative revenue hole from disposals, split $700m in 2026, an incremental $900m in 2027, and $600m in 2028, roughly half banking NII and half fees. “Next year, it will be a $900 million incremental, which will be more balanced towards banking NII” — Manveen Kaur, Group CFO · 2026-10-27. This is the first honest price tag on the “reallocate cost into growth” story: the cost leaves immediately, the revenue takes a quarter or two to arrive, and the disposals give back more NII than fees next year.

The trade HSBC isn't in

The most striking contrast with the wider market sits in lending. The global theme tape shows AI data center as a huge 360-day advancer that has definitively rolled over — it appears among the 90-day and 30-day decliners with 48 negative constituent moves against 6 positive, with AI data center exposure names selling off hard. HSBC, uniquely among large banks, has almost none of it. Asked directly about AI-linked loan demand, Kaur was blunt: “Our data center lending is very small. We are not a big player at all in the U.S. in terms of data center lending.” — Manveen Kaur, Group CFO · 2026-10-27 Loan growth instead came broad-based — $7bn from Hong Kong mortgages, 13% UK commercial lending growth, Asian trade-linked working capital. That looks like a liability in a boom and a virtue in a drawdown, and the tape is now voting for the latter.

Two capital items deserve note. First, the Basel 3.1 day-one benefit of 40-50bps from January 2027, driven by sovereign risk weights, SME exposures and foundational IRB for large corporates — with a footnote on Pillar 2A, which was also among the market's top new themes this quarter and echoed by Santander. Second, the pushback on hold-to-collect-and-sell: “In terms of the structural hedge, the HTCS portion is only 1/6 ... what goes off in September comes back also in October” — Manveen Kaur, Group CFO · 2026-10-27. Fair — but that is a mark-to-market, and it cost 20bps of CET1 in the quarter.

The tell is in the guidance language. In August, with the same “at least $46 billion” banking NII number on the table, Kaur was almost playful: “mathematically, you're quite right with the run rate, you could pretty much get a larger number” — Manveen Kaur, CFO or Senior Financial Officer · 2026-08-04. Today the identical number was reiterated with a warning about deposit mix and a reminder about $900m of disposal drag. Same guidance, different tone — and in a bank trading on a re-rated return profile, tone is the disclosure.