Is my business account FDIC insured? Pass through coverage, and the part that fails

I rang a provider last month and asked which bank was holding the money. The person on the line was polite, looked for a while, and eventually read me a sentence from their website that had the letters FDIC in it. That was the answer I got: not a bank name but a sentence off a webpage, read aloud to me by somebody who had clearly never been asked the question before and had no obvious place to look for it.

It is not her fault. I want to be clear about that. Do not treat a logo in a footer as an answer, and never accept the word eligible as though it meant insured. The question is not on the script. Almost nobody asks it, and almost nobody asks it because the marketing has been written so that it does not occur to you to ask.

Here is the sentence people are relying on, in the form it usually takes. Funds are held at a partner institution and are eligible for FDIC insurance up to 250,000 dollars.

Every word in that is doing work and 2 of them do something I did not expect until I went looking. I had read that sentence perhaps 50 times before I read it properly, and what I had wrong was simple: I thought pass through coverage was a property of the money. It is a property of the records. That difference is the whole article.

Eligible

Eligible is not the same word as insured. It means the arrangement is capable of qualifying for pass through coverage, which is a real mechanism with a real name, and which depends on a short list being true at the exact moment a bank fails.

The money has to actually be sitting at an insured bank, which sounds like the easy condition and is the one people assume without checking. The records have to show, somewhere the FDIC can reach, exactly whose money is whose. And the relationship has to be disclosed as an agency arrangement rather than the intermediary holding funds as its own. Two conditions out of 3 is not partial credit here, and the FDIC does not grade on a curve when a bank has already closed on a Friday evening.

The second condition is the one that broke in 2024, and it broke in a way I had not modelled and had not seen anybody else model either. A middleware company called Synapse sat between a set of consumer brands and the banks holding the deposits, and it kept the ledger that said which customer owned what. It filed for Chapter 11 in April of that year. When the trustee went looking, the version of the ledger the middleware held and the version the banks held did not agree.

Reported shortfall somewhere between 65 and 96 million dollars. I give a range because the figures published during the bankruptcy moved and I have not seen a final reconciled number, and anybody quoting one exactly is quoting a moment rather than a conclusion.

No bank failed in any of it. Every institution in that chain stayed open and solvent throughout, the insurance fund was never touched, and customers still could not reach their money for months. There is no federal insurance against a bad spreadsheet. I keep coming back to that sentence, because it is the one thing in this subject I did not understand at all until 2024 and it is not written on anybody's marketing page.

The regulator noticed before I did

I found out how load bearing the records are from the FDIC rather than from any provider, and it is worth reading what the agency said in its own words.

In October 2024 it proposed a rule to "strengthen FDIC-insured depository institutions' recordkeeping for custodial deposit accounts with transactional features and preserve beneficial owners' and depositors' entitlement to the protections afforded by Federal deposit insurance". Banks holding those accounts would have to keep records "identifying the beneficial owners of those deposits, the balance attributable to each beneficial owner, and the ownership category in which the deposited funds are held".

Read that as a description of what was missing. A regulator does not propose a rule requiring somebody to know whose money is whose unless there was a case where nobody did. The stated purpose is to let the agency "promptly make deposit insurance determinations" and pay claims "as soon as possible" when a bank holding custodial accounts fails, and the phrase as soon as possible is doing quiet work in a document about money people expect to reach on a Tuesday.

I find it slightly uncomfortable that the clearest description of this risk anywhere sits in a proposed rule that almost no account holder will ever open.

Partner

The other loaded word. A partner institution is a bank you have no relationship with, whose name you may not know, and which did not sign anything with you.

That is not automatically bad. It is how a large share of accounts now work. But it changes what the question is my business account FDIC insured actually means, because the honest answer for a non bank is: your money is probably at an insured bank, and whether you personally get paid depends on a ledger held by a third company.

A short digression about small failures

Since the start of 2023 there have been 11 bank failures, and I counted them out of the FDIC register myself rather than taking anybody's summary. Three were the cluster everyone remembers, First Republic at 212.6 billion dollars in assets, Silicon Valley at 209 billion, Signature at 110.4 billion. The other seven were small enough that nobody outside the trade press noticed: Heartland Tri-State in Elkhart, Kansas at 139 million. Citizens Bank of Sac City, Iowa at 60 million. Republic Bank in Philadelphia at 6 billion. First National Bank of Lindsay, Oklahoma. Pulaski Savings in Chicago at 49 million. Santa Anna National in Texas at 77 million.

Elkhart has fewer than two thousand people. Sac City is about the same. In every case depositors were made whole and the resolution took a weekend, which is the system doing precisely what it was built to do.

I mention them because the ordinary failure is small, local and boring, and because it is the small local boring one that a business with everything in a single institution is most likely to meet.

One number from the failure record changed how I read the rest of it. Of those 11, seven held under 300 million dollars and the median was about 108 million. Small institutions are the normal kind of failure, and they are also where a mid sized local company is most likely to have put everything in one place, because that is exactly the bank that gave it a relationship manager who answers the phone.

The limit, and the word people forget

The FDIC states the limit as "$250,000 per depositor, per FDIC-insured bank, per ownership category". I remembered the first 2 for years and forgot the third, and the third is where most of the confusion lives. The agency gives its own worked example: a revocable trust account "with one owner naming three unique beneficiaries can be insured up to $750,000" at 1 bank.

Ownership category means the legal capacity in which you hold the money. The FDIC looks through a single member LLC to its owner, so your personal balance and the company's balance at the same bank share one limit rather than getting one each. Add a second member and the entity stands on its own, with its own limit. Two accounts at one bank in one category do not double anything. Two separate banks do double it.

Worth doing the arithmetic on a real balance rather than in the abstract. A studio I know holds around 410,000 dollars for about eleven days a month, between a large client settling and payroll going out, after which it drops to roughly 90,000. For those eleven days, 160,000 sits outside the limit. Nothing has ever happened to it. Nothing happening is not the same as being covered, and I keep thinking about that gap more than the size of it warrants.

Sweep, which does the same thing at a chartered bank

One thing I had wrong until fairly recently, and which undercuts the neat division I had drawn between banks and everybody else, is that ordinary chartered banks move your money out of the insured account too. I suspect I am not alone in that, but that is a guess.

A sweep arrangement takes balances above an agreed level at the end of each day and places them somewhere else overnight, into repurchase agreements or a money market fund or across a network of other banks, and brings them back in the morning. Where it goes decides whether it is still a deposit. Into a network of insured banks in slices under the limit, it is. Into repos or a fund it is not, and what protects you then is the structure of that instrument rather than the insurance fund, which is a different kind of promise made by a different party for a different reason and is worth knowing about before the night it matters.

The disclosure for this is usually one line in an account agreement written in the year you opened it. I have asked four business owners what happens to their balance overnight and none of them knew, which is not a criticism of them, because nobody is told and it never comes up until it matters.

Bank failures in 2026 have so far run at 2, both small, and neither involved a sweep going wrong. That is not reassurance. I do not offer it as any. It is the current count and nothing more, and I would not build a treasury policy on 7 months of quiet.

The arithmetic when you exceed the limit

If the balance genuinely needs to sit above 250,000, there are three ordinary answers and none is clever. Split it across institutions, which costs an afternoon of onboarding and nothing per month at most banks I checked. Use a network deposit product that distributes the balance in slices, which large providers offer and which costs a spread you will have to ask about because it is rarely published. Or move the bulk into treasuries, which are not deposits and are not insured because they are a direct claim on the government instead.

A round of two million dollars sitting in one account is nine tenths uninsured on the Monday after it lands, and covering it the naive way takes eight banks. Nobody opens 8 accounts, which is the entire reason the other 2 products exist and also why their pricing is so rarely on a public page that I have started assuming the absence is deliberate.

What to ask, and what it costs to ask

I would ask for the following in writing, and none of it takes longer than the call I described at the top.

Which chartered bank holds the deposit, by name, and I would not accept a group name or a brand. Who maintains the record of my individual balance, and if that is not the bank, what happens to my access if that company stops trading. And is the disclosure written as pass through coverage, naming the conditions, or does it just put the letters near a logo.

A provider that answers all three in one email is telling you something structural about how it thinks. One that cannot name the bank is telling you something too, and I have stopped treating that as an administrative failure and started treating it as the answer.

None of this is advice about your own arrangements, and if the sums are large the question belongs with an accountant or a lawyer who can read your actual account agreement rather than the general case. What I have set out is how the mechanism works and which words in the disclosure are load bearing.

I went back to that provider a week later by email and asked the same question in writing. They named the bank in the reply. It took two days and one sentence, and it had been available the whole time, and I still do not know why it was not simply on the page.