The 60% rule: how much cash is safe in one institution

This is you if

Wise, Mercury, or Revolut already closed on you once, or sent the proof-of-operations email that made your stomach drop. Maybe you watched a friend run a $200K/mo business into a locked screen and realized your own setup is identical. Maybe you have already counted the days of payroll you could cover if that email arrived tomorrow, and the number scared you. And somewhere in your stack, more than 60% of your cash sits behind a single institution’s risk algorithm, one you know the support chat for by heart.

You know the feeling now: one email away from zero. That knowledge is only useful once it becomes architecture instead of something you carry around as background dread. If a chartered bank has already frozen you rather than a fintech, the complaint ladder is the separate, immediate move; the rules below are what stop the next freeze from being a crisis.

The backup doctrine

Three ideas do the work here, and they only work together.

Real banks. Chartered institutions, not fintechs. Fintechs optimize for onboarding speed, and the same speed that gets you an account in ten minutes is the speed at which their automated de-risking closes one. Real banks are slower to open and much slower to close, which is exactly the tradeoff you want for the money your payroll depends on.

Correct sequence. ITIN before accounts. Entity before processing. Each piece collected in the order that makes the next piece easier instead of harder. The pieces themselves are public. The order is not, and that is the part that actually determines whether an application goes through clean or gets marked.

Redundancy by design. Two to three real banks, plus backup processing, plus a written playbook for the day any single institution flinches. No off switch anywhere in the stack that one algorithm can reach alone.

The five redundancy rules. Free.

These are the rules that any real fix to a freeze eventually enforces. You can implement all five yourself starting today, whether or not you ever hire anyone to help with the rest.

1. The 60% rule. Never let a single institution hold more than 60% of your reachable cash. Concentration is the freeze multiplier: it is the difference between an annoyance and a payroll crisis, and it is the first number worth checking today.
2. Fintechs are rails, not vaults. Keep at least one chartered bank, and keep your operating buffer sitting there. Fintechs are fine for moving money around. They are not fine for storing the money your payroll depends on existing tomorrow.
3. Keep a second processor warm. A backup processor with even a token volume already running through it is a toggle you can flip in a day. A backup processor you have never onboarded is a three-week emergency project, and you will not have three weeks when you need it.
4. Payroll must be able to clear from two places. If you cannot run payroll from a second institution this Friday, you do not have redundancy, you have a diagram someone drew once. Test it quarterly, not in theory.
5. Write the break-glass playbook now. Who you call, what you say, what you never say, and which account payroll moves to. Write it before the email arrives, because after it arrives is not when you will be thinking clearly.

Two fintechs is not redundancy

It is tempting to count Wise and Mercury as two separate safety nets. They are not. Two institutions running the same style of automated risk model tend to fail the same way in the same week, which is exactly what happened to one founder who lost both in a single quarter. Redundancy means different failure modes on purpose: chartered banks running alongside processing backups, not two accounts that share the same blind spot.

Can’t I just do this myself

The pieces really are public, so yes, eventually, you can build all five rules on your own. But the wrong order along the way writes rejections into files that the US banking system remembers, and undoing that is slower than doing it right the first time. The five rules above cost nothing to start today. What tends to be harder to get right alone is the sequence behind opening the second and third bank in a way that does not trip the same reviews as the first.

Where you actually stand

Knowing the five rules is not the same as knowing which of them your current setup already fails. Concentration creeps up quietly, a warm backup processor goes cold after a slow quarter, a break-glass playbook that exists only in someone’s head is not a playbook.

Run your current setup through the calculator to see exactly where the single points of failure sit today, free. It takes a few minutes and tells you which of the five rules needs attention first.