How to Start a Bank: The De Novo Playbook

Thirty-one banking charter applications were filed in 2025, the busiest year in recent memory. Four new banks actually opened.

That gap is the whole story of starting a bank in America right now. The application surge is real, but most of it came from fintechs, payment platforms, and digital-asset firms chasing limited-purpose and trust charters. The true full-service community bank, a de novo in the classic sense, remains rare: roughly ten to fifteen organizing groups a year nationally get one off the ground.

Here is what the journey actually looks like, for founders thinking about it and for everyone who wonders why their town hasn't seen a new bank since 2008.

The money comes first

The FDIC's binding rule is simple to state and expensive to satisfy: a de novo must maintain a tier-1 leverage ratio of at least 8% throughout its first three years. Because the ratio is measured against the bank's own growth projections, the capital has to be there on day one, sized to the balance sheet you promised to have in year three.

In practice that means an organizing group raises twenty to forty million dollars before the bank earns its first dollar of interest income. The raise itself is a filter: it forces a real market study, a credible management team, and directors willing to put their own money in, all of which the regulators will examine anyway.

The timeline nobody shortens

From first organizing meeting to ribbon cutting, plan on 18 to 24 months. The regulatory clock is only part of it.

The pre-filing phase, assembling the board, recruiting a CEO examiners will accept, writing the business plan, and lining up capital commitments, typically takes two or three months of formal preparation after months of informal groundwork. Once the deposit insurance application is filed, the FDIC determines within about 30 days whether it's substantially complete, aims to finish its field investigation within 60 days, and targets a decision within roughly four months of acceptance. Six to eight months from filing to decision is the realistic band.

The approval is not the finish line. It's the starting gun for the hardest phase: building an actual bank in the months before the doors open.

Then comes the buildout. Conditional approval arrives with a list of conditions to satisfy before opening: the capital must fund, key hires must land, policies must exist, and the technology stack must work. Core system, digital banking, BSA program, verification layer, all of it gets selected, contracted, and stood up in this window.

The three-year probation

A new bank doesn't graduate on opening day. The de novo period runs three years, during which the FDIC holds the bank to the specific commitments in its application: the capital levels, annual financial statement audits, fidelity bond coverage, and material adherence to the approved business plan. Deviating from the plan requires regulator sign-off. Exams come more frequently than an established bank would see.

The first exam deserves particular respect. Examiners arrive with the business plan in hand and check whether the programs it promised exist in operation. A BSA officer on the org chart is not a BSA program. A fraud-controls section in the application is not a fraud control. The de novos that stumble early are usually the ones that treated compliance and verification as post-launch projects.

The stack decision that gets underweighted

Most de novo technology conversations orbit the core: Jack Henry, Fiserv, FIS, or CSI, usually as a bundle with digital banking and payments. That decision matters and takes months. But the layer examiners test first, and the one that determines whether the digital channels in the business plan produce deposits or fraud losses, is verification: how the bank confirms that an applicant is real, employed, and creditworthy before booking them.

A de novo has no legacy intake process to defend. It can be the first bank in its market where digital account opening is the most controlled channel, not the most dangerous one.

The established-bank playbook, collect documents, call phone numbers, file PDFs, is the intake model fraud rings industrialized against, and it's why so many older banks turned their digital channels off. A bank built in 2026 gets to skip that inheritance entirely and verify at the source from day one: identity checked against authoritative records, income and employment pulled from payroll systems, funding accounts confirmed by direct connection.

Starting a bank is a long, expensive, heavily supervised project, and that is arguably the point: the charter is valuable because it is hard to get. The founders who do it well treat the constraints as design inputs. The capital rule sizes the ambition, the timeline sequences the build, and the de novo period rewards the bank that opened with its promises already working.