Huntington Bancshares Q3 2026 Earnings: $2.8B Revenue Up 31%, EPS Down 20% on Cadence Deal

Huntington's third quarter is the acquisition quarter in full: net interest income up 37% to $2.07 billion, assets up 37% to $287.8 billion, deposits up 36% — all of it downstream of the Cadence (and Veritex) deals. But net income applicable to common shares grew only 13%, and diluted EPS fell 20% to $0.33, because the share count grew 38% to pay for it. This is a quarter about what bank M&A giveth and what it taketh away.

The Headline Numbers​

Every figure below comes from the Q3 earnings release and the Quarterly Financial Supplement. No 10-Q was filed as of this writing, so the 8-K exhibits are the primary source. The Beancount fence models the quarter, so the headline table that gates it is Q3.

Read the table top to bottom and watch growth decay: revenue +31%, net income +15%, EPS −20%. Each step down has a named cause. Revenue-to-income decay is expenses: noninterest expense grew 42%, faster than revenue, including $117M of acquisition-related notable items (plus an $8M equity-conversion adjustment on the income side). Income-to-EPS decay is the share count: average diluted shares rose from 1,485M to 2,047M year over year (+38%), all of it acquisition currency, and preferred dividends took another $41M off the top. The bank earned more money and each share got less of it — the central fact of the quarter, and the lens for everything below.

Revenue Deep Dive​

Net interest income is three-quarters of revenue, so start there: $2,069M, up $563M (+37%) on average earning assets up $62.5B (+32%), "primarily due to a $52.5B, or 39%, increase in average loans and leases driven by loans acquired in the Cadence and Veritex acquisitions, in addition to organic loan growth." The margin side cooperated too: NIM widened 12 basis points to 3.25% as deposit costs fell 10 basis points to 1.90% while loan yields slipped 12 basis points to 5.84% — liabilities repricing faster than assets, the favorable side of the rate cycle for a deposit-funded bank.

The loan book underneath is being actively reshaped, not just grown. Linked-quarter, average loans fell $806M (−0.4%): commercial real estate down $854M (−4%) and consumer down, offset by commercial and industrial up $553M (+1%). CRE de-risking while C&I grows is deliberate mix management — shrinking the scary book, feeding the relationship book — and it continued even as the acquired portfolios integrated. Deposits tell the steadier story: +$413M linked-quarter (+0.2%), with money-market and interest-bearing demand inflows offsetting time-deposit runoff, and noninterest-bearing demand up $146M. A bank growing deposits organically while cutting deposit costs is winning both sides of the franchise.

Noninterest income (+15% YoY to $722M) is the mixed line: payments and cash management revenue grew 21% year over year to $211M — the fee franchise working — but the total fell 8% linked-quarter "largely due to lower capital markets fees and unfavorable equity valuation adjustments." Lumpy, market-sensitive, and the reason fee income cannot be modeled as a straight line. The adjusted (non-GAAP) noninterest income of $730M strips the $8M conversion adjustment; the underlying message is the same: fees grew with the acquisitions and wobbled with the markets.

The Margin Story​

NIM is the standout: +12bps year over year and +4bps linked-quarter to 3.25%, with the release guiding nothing but the trajectory speaking — deposit-cost relief (1.90%, −10bps) outpacing asset-yield decay. In a year when many regionals watched margins compress, Huntington expanded its, partly because acquired deposits repriced onto its lower-cost base.

Everything else in the table is the acquisition's cost showing through. The GAAP efficiency ratio worsened five points year over year to 63.5% (the release's headline 61.2% is an FTE/non-GAAP variant; directionally identical): expenses grew 42% against revenue's 31%, and even excluding the $117M of notable acquisition costs, adjusted expenses grew roughly in line with the acquired revenue — synergies are promised ("achieving our targeted cost and revenue synergies," per Steinour) but not yet in the ratio. ROA fell to 1.01% and ROTCE to 14.6% from 17.8% for the same reason: the denominator (assets, equity, shares) arrived on day one, while the earnings power phases in over quarters. This is the standard J-curve of a big bank deal. The question is how long the J takes.

The One Big Question: did Huntington buy growth at the expense of per-share value?​

Put the two sides of the Cadence/Veritex deals on one scale. What shareholders got: NII +37%, a $288B balance sheet (from $210B), deposits +36%, entry into Texas and the Southeast, and a wider NIM. What they paid: 38% more shares, EPS down 20% year over year ($0.41 → $0.33), tangible book value per share down 2% linked-quarter to $9.47 (rates-driven AOCI, but down is down), and a CET1 ratio of 10.1% that is merely adequate rather than fortress-like. The deals are accretive to every aggregate and dilutive to every per-share number that matters. That is not a verdict — most large bank acquisitions look exactly like this in year one — but it sets the falsifiable terms on which this one will be judged:

1. EPS must grow through the dilution. Adjusted EPS (ex-notables) was $0.38 this quarter vs $0.33 GAAP; the GAAP number needs to inflect upward over the next two quarters as the $117M/quarter of acquisition costs burns off and synergies land. A third straight flat $0.33 would say the earnings power is not phasing in.
2. Tangible book must rebuild. $9.47 is down linked-quarter on AOCI; the combination of retained earnings (dividend is $0.155, well covered) and the $100M/quarter buyback pace ($409M year-to-date) should grind it higher absent further rate damage. TBV flat-or-down again next quarter says capital is leaking somewhere.
3. Credit must stay boring. This is the tripwire with teeth: net charge-offs rose to 0.31% (from 0.22%), nonperforming assets to 0.90% (from 0.60%), and nonaccrual loans more than doubled year over year to $1,697M — while the allowance coverage of loans actually fell to 1.68% and ACL coverage of NPAs fell from 312% to 195%. An acquiring bank stretching coverage thinner while NPAs climb is the exact pattern to watch; any Q4 reserve build would confirm the market's quiet worry.

If EPS inflects, TBV rebuilds, and credit stays orderly, Q3 FY2026 reads in hindsight as the trough quarter of a transformative deal — the J-curve bottom. If not, it reads as the quarter Huntington proved it can buy revenue but not earnings per share.

Tracking a $288B bank in plain text​

A bank that grows 37% by acquisition is exactly where double-entry earns its keep: every acquired dollar of loans must appear opposite acquired deposits, equity, or debt, and a plain-text ledger makes the accounting auditable line by line. Our how we model every company guide documents the conventions; here is Huntington's Q3 income statement as booked, signs and all (income is negative, expenses positive, and the check comment proves the transaction sums to zero):

; Check: −2,791 + 123 + 1,772 + 172 + 724 = 0 ✓
2026-09-30 * "Huntington Bancshares Incorporated" "FY2026Q3 Income Statement"
  Income:Revenue                               -2791 MUSD  ; total revenue (net interest income 2,069 + noninterest income 722)
  Expenses:CostOfRevenue                         123 MUSD  ; provision for credit losses
  Expenses:OperatingExpenses                    1772 MUSD  ; total noninterest expense (incl. 117 acquisition-related notable expenses)
  Expenses:IncomeTax                             172 MUSD  ; provision for income taxes
  Equity:Adjustments                             724 MUSD  ; consolidated net income offset (NCI share 3; preferred dividends 41 and EPS 0.33 in the release only)

The balance-sheet number that tells the narrative is goodwill: $5,997M at the end of FY2025, $9,527M today — $3.5 billion of acquisition premium, matched by capital surplus rising $17,244M → $25,101M as 450 million new shares were issued to pay for it. The price of growth, one assertion per line. Explore all six periods below:

The Multi-Year Arc​

*December 2025 quarter-end, per the Q3 supplement's five-quarter table.

Read left to right, Huntington spent four years as a steady compounder — revenue growing ~13% total from FY2022 to FY2025, earnings oscillating in a $1.9–2.2B band, assets marching up ~7% a year — and then spent 2026 becoming a different-sized bank. The Q3 column breaks every trendline: assets +28% in nine months, deposits +27%, shares outstanding +29%. Nothing in FY2021–25 predicts the Q3 balance sheet because the Q3 balance sheet was purchased, not grown. That discontinuity is why the post-2026 Huntington has to be underwritten on deal math (synergies, accretion timetable, credit marks) rather than on the admirable but now-irrelevant organic record. The ledger keeps both eras in one file so the break is visible: five years of compounding, then a step change.

The Verdict: Bull vs. Bear​

Bull Case

• NII compounding with margin expansion is rare. +37% net interest income and +12bps of NIM in the same quarter — most acquirers buy balances and watch margins sag; Huntington is getting paid on both volume and spread as acquired deposits reprice down.
• The deposit franchise is genuinely improving. Organic deposit growth (+$413M linked-quarter) with falling deposit costs (1.90%, −10bps) and rising noninterest-bearing balances — the cheapest funding source growing while time deposits run off is textbook liability management.
• Synergies are guided, not hoped. Management says cost and revenue synergy targets are being achieved; $117M of quarterly acquisition costs burning off over the next few quarters is a visible, time-bounded earnings tailwind worth roughly $0.05 of quarterly EPS.
• Credit costs are still negligible. Provision of $123M is flat year over year and just 4% of revenue — for all the NPA hand-wringing, realized losses (0.31% NCOs) remain well within a moderate-to-low risk profile, and the ACL still covers 195% of NPAs.
• Capital returns continue through the integration. $100M of buybacks this quarter ($409M year-to-date), a steady $0.155 dividend, and CET1 edging up to 10.1% — the bank is returning capital while digesting its largest acquisition, which only works if internal capital generation is strong.

Bear Case

• EPS is down 20% and that is the scoreboard. $0.41 → $0.33 on 38% more shares — shareholders own a meaningfully smaller piece of a bigger bank, and GAAP EPS has now printed flat at $0.33 for two straight quarters with no visible inflection.
• Tangible book is going the wrong way. $9.47, down 2% linked-quarter on AOCI, with $3.5B of new goodwill and intangibles sitting atop it — the per-share value story is deteriorating while the aggregate story shines.
• Credit is inflecting the wrong way. Nonaccrual loans doubled year over year, NPAs rose 50% (0.60% → 0.90%), and allowance coverage of both loans and NPAs thinned simultaneously — individually explainable (acquired marks, mix), collectively the pattern that precedes reserve builds.
• The efficiency ratio broke out. 63.5% GAAP, up five points year over year and the worst in the five-year ledger — scale was supposed to bring leverage, and Q3 shows the opposite; synergy realization is a promise, not a print.
• Concentration risk arrived with the deals. CRE acquired plus Texas/Southeast geographic expansion plus a $191B loan book growing faster than the allowance — the "moderate-to-low risk profile" is now a larger, less seasoned portfolio than the one that earned the label.

Our Take

Wait for the J to curve. Huntington bought real earnings power — the NII and deposit numbers prove the franchise value of what it acquired — but it paid with 38% dilution, and Q3 shows the cost (flat EPS, soft TBV, creeping NPAs) arriving ahead of the benefit (synergies, accretion, reserve releases). Bank deals are judged at month eighteen, not month nine. Hold the EPS, TBV, and NPA tripwires from the Big Question section against Q4: two of three improving means the curve is bending; two deteriorating means the market is right to discount the dilution.