Preferred Bank: A Frozen Balance Sheet and a Tax Code for a Rival
Deposits bleed to Treasuries, nonaccruals bleed to bankruptcy court — and a bank that spent two years defusing its rate exposure suddenly likes it again
PFBC · Earnings Call · 2026-10-23
The quarter in one sentence
Preferred Bank earned $32.9 million, or $2.73 a share, in the September quarter — a number that looks fine until you notice it is the fourth consecutive quarter where the interesting news lives somewhere other than the income statement. Deposits rose $77 million (1.2% linked-quarter), loans fell $14 million after a large credit was moved to held-for-sale, and the reported net interest margin of 3.49% is a mirage: strip out a $2.1 million recovery of interest income and it is 3.60%. “This quarter net interest margin was reported at 3.49%. Without this $2.1 million adjustment, it would have been 3.60%.” — Li Yu, Chairman and CEO · 2026-10-23
What actually changed this quarter is the story management tells about why deposits are hard to keep. And on credit, what changed is the calendar.
The tax code as a competitor bank
For two quarters running, Li Yu has explained deposit disintermediation by pointing at something outside banking. In July it was the equity market: “We see many, many customers are investing their excess cash into the stock market today.” — Li Yu, Chief Executive Officer (CEO) · 2026-07-22 This time the money isn't chasing AI stocks. It's chasing an after-tax yield that no California-chartered bank can match.
Because the Treasury nature is being not taxable in California income tax, which is a 13% top rate. So a 5% Treasury become 5.65% effective yield. So we see many of the customer take off after their CD renewed, they are moving to Treasury.
That is a genuinely company-specific lens on what is otherwise a sector-wide problem. The state income-tax arithmetic — and the effective yield it manufactures — is why "the deposit market is very competitive" lands differently for a Los Angeles commercial bank than for a Minnesota one. The spot rate evidence backs him up: total deposit cost was 3.06% in September against a 3.58% spot margin, and $1.7 billion of CDs mature in Q4 at a 3.81% weighted average with roll-on expected between 3.75% and 4%. That is not a roll-off tailwind; it's a treadmill.
The rest is consensus. Scan this week's reporters and the same words recur across a dozen regional banks — broker deposits shrinking at CNOB, FULT and TCBI, deposit betas debated at AMAL, BOKF, FITB and WAL, retail deposits repricing everywhere. Preferred Bank's deposit squeeze is a sector tide, not a company wound. The tax-exemption footnote is the only part that isn't.
The asset-sensitivity pivot nobody flagged
The most quietly consequential line in the prepared remarks: “If the Fed fund is going to adjust higher, then it will further benefit our net interest margin as Preferred Bank is an asset-sensitive bank.” — Li Yu, Chairman and CEO · 2026-10-23
Read that against the company's own record. In April, management was telling investors PFBC had engineered itself away from that exposure: “within the next few rate moves, Preferred Bank is near neutral in asset sensitivity” — Li Yu, Chairman and CEO · 2026-04-22 — a deliberate two-year project of pushing fixed-rate loans up to roughly 25% of the book and tying corporate deposits directly to Fed funds. In October 2025 the CFO described the bank as not as asset sensitive as we have been in the past.
Six months later, asset sensitivity is a feature again. That is not a contradiction so much as a regime call: the analysts on the call were openly assuming “maybe another rate hike in December and maybe one in March” — Matthew Clark, Analyst · 2026-10-23, and the market's own curated themes for the quarter are dominated by the Fed rate hike narrative. A bank that spent two years defusing its rate exposure now gets paid for having kept some — a interest rate environment bet dressed up as a balance-sheet fact. The catch is loan beta: the CFO conceded the up-cycle deposit beta will be “Definitely slower, Matthew” — Edward Czajka, Chief Financial Officer · 2026-10-23 than the 69% enjoyed on the way down, and management deliberately sacrificed deposit growth to shave 3 basis points off funding costs.
Credit: the courts are now the workout department
Nothing improved here, and the timeline got worse. Classified loans sit at $177.6 million including $11.5 million of OREO, with only about $20 million of additional substandard credit still on accrual. The problem is velocity. The bankruptcy court — the single highest-ranked keyword in the company's quarter — has become the binding constraint on resolution: “Hopefully by Q4, there will be a couple of large loans will be disposed. However, it totally depends on the court decision.” — Wellington Chen, President and Chief Operating Officer · 2026-10-23
Compare that to the promises. In January, management expected “the majority of a good portion of all resolved sometime within two quarters.” — Li Yu, Chairman and CEO · 2026-01-22 The CEO now says certain credits scheduled for this quarter are stuck in the court proceedings and pushed into the third and fourth quarter of next year — a four-to-six-quarter slip on collateral that management insists is well covered. That is the real story of the print: not a loss, but a frozen balance sheet, with roughly $160 million of nonaccruals generating no interest while the legal system runs its course.
What stopped being asked
Equally telling is the silence. In January, analysts pressed on M&A appetite and share repurchases; in July, capital priorities came up again. This quarter: not one question on buybacks, M&A, or capital deployment. The conversation narrowed to deposits, credit and the expense line — a defensive triad. That is what a franchise looks like when it is in hold-and-defend mode, even with expenses genuinely well managed (Q4 guided to $23–24 million, helped by lower legal fees and a smaller incentive accrual) and a loan pipeline management calls very solid for the fourth quarter.
The setup
Preferred Bank is not broken. It is earning a mid-teens return on a book it has kept deliberately short, and it is asset-sensitive into a market that is pricing more hikes. But the two things that would normally re-rate a small-cap bank — clean credit resolution and visible deposit gathering — are both being decided by third parties: a bankruptcy judge in one case, and the California tax code in the other. Until either moves, the stock is a bet on process rather than performance, and the honest read of this call is that management's 2027 optimism rests on forces it does not control.