First Citizens BancShares, Inc. FCNCA

Key data

The business

First Citizens is a bank that grows by eating other banks. Run by the Holding family since 1898 out of Raleigh, it has absorbed more than 20 failed or struggling lenders, most in FDIC-assisted deals where it pays less than the assets are worth and books the difference as a gain. The two that reshaped it: CIT Group in 2022, which added equipment finance, factoring and railcar leasing, and the bulk of Silicon Valley Bank in 2023, seized after the run and handed to First Citizens at a steep discount. The money is made the ordinary way, lending out deposits and keeping the spread, plus fee income from wealth management and the commercial-finance units. The moat is twofold: a low-cost deposit base that depositor inertia keeps sticky, and a family that thinks in decades and has proven it can buy distressed banks and integrate them without blowing itself up.

The numbers

The numbers tell two stories at once, an earnings machine humming and a credit book quietly softening.

The five quarters show a business that is not growing its top line, net interest income has drifted from $1.73B to $1.66B, but is earning more per share anyway. The June quarter's $672M profit, up 17% from a year earlier against revenue that actually fell 0.5%, came from lower expenses and a smaller credit provision, not from more banking. Diluted EPS rose 31% year over year, and the gap between 17% profit growth and 31% EPS growth is the share count doing the work.

The annual picture is dominated by 2023, when a $9.8B bargain-purchase gain on Silicon Valley Bank inflated net income to $11.5B and diluted EPS to $785. Strip the gain and operating earnings were closer to $175 a share. What matters is the trajectory since: earnings normalized from an inflated 2024 down to $166 a share in 2025 as the acquired assets rolled off their initial marks, and 1H 2026's $98 keeps the run rate near $200.

The compounding lives in the share count. First Citizens bought back $3.0B of stock in 2025 at an average of $1,926, below today's $2,208, after $1.6B in 2024 and $1.2B in 2022. The diluted share count has fallen from 15.55M in 2022 to 11.54M last quarter, down about 26% in under four years. Tangible book per share now stands at $1,854, and return on tangible common equity runs near 11%, a steady rather than spectacular number. Free cash flow per share compounded about 12% a year over the last three.

Credit is where the story gets its one warning. Noncurrent loans have climbed every year, from 0.75% of loans in mid-2023 to 1.12% by March 2026, while reserves against those loans have fallen from 1.23% to 1.04%. Noncurrents rising as the cushion thins means coverage is compressing. Charge-offs themselves remain contained near 0.30% to 0.45%. The concern raised a week ago, that the leftover Silicon Valley book would keep pushing noncurrents higher, has held rather than reversed.

Management

The family runs it like owners. Frank Holding Jr. is chairman and CEO, and the buyback record, consistently struck below the current price, is the clearest signal of how they read their own value. Deals have been paid for in cash at a discount, not in goodwill-inflating stock, which is why book value compounds instead of bloating. Insider activity over the past year nets to a small sale, about $2.3M, a rounding error against a $25B company, and the eye-catching $10.3M sale paired with a $10.3M purchase by Holding on the same day in May is an option-related swap that nets to nothing, not a signal. The record says one thing: keep buying cheap, keep shrinking the count.

How it fails or surprises you

Credit turns from creep to break. Noncurrent loans have risen three straight years to 1.12% while reserves fell to 1.04%. If the acquired SVB and CIT books deteriorate faster than charge-offs suggest, the bank must rebuild reserves through the income statement, and a single quarter of doubled provisioning would erase the earnings beat that carried the stock. Watch noncurrents against reserves each quarter.

The beat was expenses, not banking. Q2 profit rose 17% on revenue that fell 0.5%, and the gain came from lower operating costs and a light provision, not more lending. That is the number this read explains least well. If provisions normalize upward while net interest income keeps drifting down, the $200 run rate is optimistic and the forward multiple is not as cheap as it looks.

Another bank falls into its lap (right tail). First Citizens sits on excess capital and a proven playbook. A regional-bank stress event, rates cracking a weaker lender, would hand it another discounted portfolio the way SVB did, and the market pays nothing today for a deal not yet announced. The tell would be an FDIC-assisted acquisition or a sharp step-up in the buyback pace.

Closing thoughts

Nothing in the next four quarters settles this one cleanly. The stock is priced roughly fairly, about 12 times earnings and 1.2 times tangible book against an 11% return on that book, which is close to what a decent bank earning decent returns should cost. The buyer betting on more is betting the family keeps repurchasing stock below intrinsic value and lands another cheap deal, and the seller is betting the credit creep becomes a credit problem. The left tail is the thinning reserve, the right tail is the next acquisition, and on this evidence the right tail is the fatter one, because the balance sheet can absorb the credit drift while it waits.

The bet stays what it was: a family that buys banks cheap and shrinks its own share count, turning other people's failures into its own earnings. What breaks it is credit, not growth, so the pair to watch is noncurrent loans against reserves. The day noncurrents push well past 1.2% while reserves keep falling is the day the acquisition math stops flattering the picture. Until then, the shares just keep getting scarcer.