Every few years a bank fails, the coverage explains that deposits are insured up to $250,000, and a great many people work out that they are fine because they hold less than that. Understanding how does FDIC insurance work properly means noticing that the sentence has a third clause, and that the third clause is the one nobody reads.
The standard amount is $250,000 per depositor, per insured bank, for each account ownership category. Two of those limits are obvious. The category is where the real arithmetic happens, and it can multiply the coverage several times over at a single bank.
The other thing worth knowing is where the money comes from, because it does not come from the federal budget.
The formula, and the clause people skip
The insurance is automatic. Nobody applies, nobody pays a premium, and there is no form to complete — coverage attaches the moment a deposit account is opened at an insured bank, and it cannot be declined, bought in larger quantities, or lost through inattention.
Within a single ownership category at a single bank, everything is added together into one balance. As the FDIC’s own guidance puts it, the agency adds together all of the deposit accounts you hold in the same ownership category at the same bank regardless of the deposit type, whether that is a current account, a savings account or a certificate of deposit.
So a current account, a savings account and three certificates of deposit at the same bank, all in one name, are not five separate allowances. They are one pot with one $250,000 ceiling. Different branches of the same bank do not help either, because the limit attaches to the institution rather than to the building.
What an ownership category actually is
Categories are legal structures for who owns the money, and each one carries its own separate limit at the same institution.
| Category | How the limit works |
|---|---|
| Single accounts | $250,000 for all single-name deposits combined |
| Joint accounts | $250,000 per co-owner, so $500,000 for two |
| Certain retirement accounts | Separate $250,000, not counted with single accounts |
| Trust accounts | Coverage scales with the number of eligible beneficiaries |
| Business accounts | Separate from the owner’s personal deposits |
The categories are defined in federal regulation rather than by the bank, and the definitions are technical. A joint account only counts as joint if both owners have equal withdrawal rights and have signed the signature card.
The practical effect is that a couple at one bank can be covered well beyond $250,000 without doing anything unusual, simply because their single accounts, their joint account and their retirement accounts sit in three different categories. It also means two people with identical balances can end up with very different coverage, decided by paperwork neither of them remembers signing.
What is not covered, including things sold in the same branch
This is where the misunderstanding gets expensive, because banks sell products that are not deposits, sometimes at the same desk.
The FDIC’s published list of what it does not insure is explicit:
- Stocks, bonds and mutual funds
- Crypto assets
- Life insurance policies and annuities
- Municipal securities
- Safe deposit boxes and their contents
- US Treasury bills, bonds and notes
The last entry needs its footnote. Treasuries are not FDIC-insured, but the agency notes they are backed by the full faith and credit of the US government, which is a stronger guarantee rather than a weaker one.
The others carry no such consolation. A mutual fund bought through a bank branch is a market investment that can lose value like any other, and the insurance does not apply to it at all — the counter it was sold across makes no difference to the legal position, however much it suggests otherwise.
Safe deposit boxes surprise people most. The bank rents you a locked space in its vault and insures nothing whatsoever inside it, which is why box contents are normally covered, if at all, under a household insurance policy.
Where the money comes from
Here is the part almost no explainer mentions, and it changes how the whole arrangement should be read. The FDIC receives no funding from the federal budget. It assesses premiums on each member bank and accumulates them in a pool called the Deposit Insurance Fund, and that fund is what pays insured depositors when an institution fails.
So the insurance is not a government subsidy in the ordinary sense. It is a compulsory mutual insurance scheme, paid for by the banks themselves, with the government running the scheme, setting its rules and standing behind it if the fund is ever exhausted.
The cost is passed on, of course. Assessments are an operating expense like any other, and they end up inside the spread between what a bank pays on deposits and what it charges on loans.
Depositors therefore fund their own protection indirectly, through slightly worse rates, rather than through taxation — which is a meaningful distinction when the arrangement gets described as a bailout.
What actually happens when a bank fails
The image most people carry is of the agency writing cheques to a queue of customers. That is the less common outcome.
The usual method is a purchase and assumption. Another bank agrees to take on the deposits as liabilities and buy some or all of the failed bank’s loans, and the deposits simply move.
Done well, this is invisible. A bank closes on a Friday and reopens on Monday under a different name, with accounts intact and cards still working, and most customers experience an administrative change rather than a loss.
The direct payout is the fallback when no buyer can be found. The agency then pays insured depositors the full amount of their insured deposits, and anything above the limit becomes a claim on the failed bank’s estate, paid partially if at all.
The weekend timing is not accidental. Resolutions are scheduled so the transfer happens while branches are shut and the systems can be switched over, which is why bank failures are announced on Friday evenings with striking regularity.
The limit is a number from 2008
The $250,000 figure has a specific and recent history. The limit stood at $100,000 for years, and was raised temporarily on 3 October 2008 in the middle of the financial crisis.
The temporary increase was extended, then made permanent by the Dodd-Frank Act in July 2010, and applied retroactively to the start of 2008.
What has not happened since is any adjustment for prices. The figure is nominal and fixed, so its purchasing power has been quietly falling for the whole time it has existed.
A limit set in 2008 protects meaningfully less in real terms today, and how much less depends entirely on how inflation is measured — which is itself a contested question rather than a settled one.
Nothing in the law indexes the amount. Raising it requires Congress to act, which historically has happened during crises rather than in advance of them.
The record, and what it proves
The claim made for the system is a strong one, and it is accurate: since the agency began in 1933, no depositor has ever lost a penny of FDIC-insured funds. That is a remarkable run, and it is the reason bank runs stopped being a routine feature of American life.
Read the sentence carefully, though. It is a statement about insured funds, and it says nothing about money held above the limit or in uninsured products.
Uninsured depositors in failed banks have taken losses, and in some resolutions they have been made whole by discretionary decisions taken at the time rather than by any entitlement. The guarantee covers what it covers, and nothing beyond it is promised in advance.
Why the scheme exists at all
Deposit insurance is not really about compensating people after a failure. It is about preventing the failure from happening in the first place.
A bank holds a small fraction of its deposits in cash and lends the rest. That model is sound while depositors stay put, and collapses immediately if enough of them arrive at once, which means a rumour is sufficient to destroy a solvent institution and being wrong about the rumour costs the depositor nothing.
Insurance removes the incentive to be first in the queue. If your money is protected whatever happens, there is no reason to join a run — and if nobody joins, the run does not occur.
The insurance payout is the visible product. The absence of panic is the actual one, and it is the reason the scheme is judged by how rarely it is needed.
What the rules mean in practice
Three details do most of the work in any real situation, and none of them is the headline number.
The first is whether the institution is FDIC-insured at all, which the agency lets anyone verify directly through its own bank lookup tool. Some financial firms that hold money are not banks.
The second is which ownership category each account falls into, since the categories are defined in federal regulation with conditions attached, and an account that looks joint is not necessarily joint for insurance purposes.
The third is that cash held for emergencies is exactly the money most exposed to this, which is a point worth holding alongside any thinking about how big an emergency fund should be.
The system is unusually well documented, and the agency publishes the rules, a plain-language brochure and a coverage calculator. It is one of the rare cases where the official source is also the clearest one available, and where the bank marketing material adds nothing the regulator has not already said better.
