The fintech hiding inside a 128-year-old North Carolina bank

When Silicon Valley Bank collapsed over a weekend in March 2023, the conventional wisdom said it was the end of an era. The bank that financed half of US venture capital was gone. Founders panicked. The Fed scrambled. And on the following Monday morning, a 125-year-old family-run bank from Raleigh, North Carolina walked into the carcass and walked out with $72 billion in loans and $56 billion in deposits at a $16.45 billion bargain price.

Think about that for a second. A bank founded in 1898, run by the same family (the Holdings) for four generations, picked up the most important tech-lending franchise in America at what amounted to a fire sale. The FDIC took the loss. The Holding family took the upside. And what they got was not really a bank in the traditional sense at all. They got a fintech business that the market still hasn't figured out how to value.

Here is what the market still misunderstands about the SVB deal. SVB was not a regional bank that happened to lend to tech. It was a vertically integrated financial operating system for the venture economy. It held the deposits of the funds. It held the deposits of the portfolio companies. It provided the venture debt that bridged the rounds. It ran the FX desk for cross-border SaaS revenue. It was the plumbing (yes, I know, but here it actually fits) underneath roughly half of US venture-backed software. When the bank failed, the assets did not. The relationships did not. The data did not. They just got handed to someone willing to keep the lights on.

The Raleigh family-run bank kept the lights on. And then they did something more interesting. They restructured the SVB book to keep originating venture debt, signed a five-year loss-share agreement with the FDIC that effectively socializes the downside on $60 billion of acquired loans, and started cross-selling treasury management products that the old SVB never had bandwidth to push. The result is a balance sheet that earns mid-teens returns on tangible equity while carrying federal insurance on its biggest risk pool.

Step back to the broader fintech picture in mid-2026. The venture funding cycle is turning. AI infrastructure financing has exploded, and most of it is going through private credit and venture debt rather than public markets. CoreWeave's debt stack. Anthropic's compute commitments. The Lambda Labs of the world. Every one of those deals needs a depository institution willing to warehouse the loans, manage the FX, run the treasury for the founders, and lend against ARR contracts that traditional credit committees do not understand. There are maybe four institutions in the country that can do this at scale. One of them is the old SVB platform, now sitting inside a 128-year-old North Carolina holding company that the Holding family still controls through a dual-class share structure.

What makes this interesting from a fintech perspective is that the public market still treats this company as a regional bank. It trades on price-to-tangible-book like a sleepy community lender. The peer comp set on every sell-side model is other Southeast regionals. Nobody is benchmarking it against the actual business it now runs, which is the largest venture banking franchise in the world bolted onto a sticky low-cost deposit base.

The chain reaction from here is the part that matters. AI capex requires financing. Financing requires warehouses. Warehouses require deposits. Deposits require relationships. Relationships require the institutional memory of having banked these founders through three previous cycles. That memory now sits on the balance sheet of a Raleigh holding company that nobody in the fintech conversation thinks about. The second-order beneficiary of the AI capex boom is not Nvidia. It is the institution that finances the people building on top of Nvidia.

And the third-order trade is that as private credit continues to eat the syndicated loan market, the banks that have direct origination relationships with venture-backed borrowers will get repriced. Goldman is trying to build this. JPMorgan is trying to buy this. The Raleigh family-run bank already has it, paid sixteen billion dollars for it, and is collecting fees on it while the rest of the market argues about whether private credit is in a bubble.

The company is First Citizens BancShares.