Bottom line up front: FDIC insurance covers $250,000 per depositor, per insured bank, per ownership category. A funded startup blows past that on day one of a seed round, which means most of your cash is not directly insured unless you deliberately spread it. The fix is not complicated, but it is not automatic either, and the March 2023 SVB collapse proved that “our bank feels safe” is not a treasury policy. This page explains what is actually insured, how startups exceed the limit without realizing it, and the three real ways to extend coverage.

The rule, stated correctly

The number everyone quotes is right but usually explained wrong. The FDIC insures $250,000 per depositor, per insured bank, per ownership category (FDIC). Three qualifiers, and all three matter:

• Per depositor. Your legal entity (the C-corp or LLC) is the depositor. Not you personally, not each of your accounts.
• Per insured bank. Coverage resets at each separately chartered FDIC-insured bank. Two accounts at the same bank do not double your coverage. Two accounts at two different banks do.
• Per ownership category. This is the one founders misread. For a business, your operating entity is a single ownership category. The trust/beneficiary tricks that let an individual stack coverage (POD accounts, joint accounts) mostly do not apply to a corporate depositor. A business entity generally gets one $250K bucket per bank, full stop.

So the honest version for a startup: $250,000 of your corporate cash is insured per bank you hold it at. Everything above that, at that bank, is an uninsured claim against the bank if it fails.

FDIC insurance also only covers deposit products: business checking, savings, money market deposit accounts, and CDs (FDIC). It does not cover investment products. Hold that thought, because it is the trap at the end of this page.

How a startup blows past $250K without trying

You raise a $3M seed. It lands in your business checking account. You now have $2.75M of uninsured deposits sitting at one bank, and nobody sent you a warning.

This is the default state for almost every funded company. Payroll, vendor payments, and runway all live in one operating account because that is operationally simple. The concentration is invisible until the bank is in trouble, at which point it is the only thing that matters.

I lived that week from the inside. In March 2023 I was on the treasury team at a loan servicer, and in the days before SVB collapsed there were whispers in the industry that something bad was coming. SVB had taken a heavy loss selling off securities and was scrambling to raise capital to cover it, and when that capital raise fell apart people got nervous, including our CFO. Here is the detail that made our situation different from the average startup’s: as a servicer we held not just our own corporate cash but hundreds of millions of dollars in investor and borrower cash, principal and interest owed to other people, that we were responsible for reconciling, tracking, and distributing. All of it sat at SVB.

We feared a bank run. We started moving corporate cash to a second bank we had begun building a relationship with, and we started talking to our investor team about options, but the clock was ticking. Thursday night I got a rare overnight message from my boss: clear out all the servicing cash, sweep everything to the investors first thing Friday morning, tee it up tonight so it processes as early as possible. Do not be the one holding the bag when SVB runs out of cash. I set the instructions up that night to move everything out at the earliest possible window.

Friday morning the funds started to move. Before midday, SVB’s cash activity was frozen and we could no longer see movements. It was impossible to tell what had actually gone out, what was stuck in limbo, and what had never moved at all. We went into the weekend not knowing the state of the business, and knowing that FDIC only covers $250,000 per entity, I genuinely thought we might walk into Monday with $250K to cover everything: investor obligations, rent, payroll, all of it. An absolute disaster. The government stepped in over the weekend and backstopped the deposits, and everything we had teed up Friday eventually processed. Over the next couple of days our treasury team moved all servicing activity off SVB to a different institution, lightning fast, and the entities we service for felt only a couple-day delay in their distributions.

The SVB lesson in one line: a bank can be well-run, well-capitalized on paper, and still gap down and get seized inside 48 hours. Diversification of deposits is not paranoia. It is the baseline.

And the lasting lesson for us was permanent, not situational. We diversified off a single institution and never went back. When you are responsible for this much cash, you have to be nimble enough to move off one bank and onto another in the blink of an eye, and you cannot build that relationship on the Thursday night you need it. Extending FDIC coverage protects the dollars. Keeping a second live banking relationship protects your ability to actually move them. A sweep answers the coverage question; it does not, by itself, answer the continuity one.

The three real ways to extend coverage

There are exactly three mechanisms. Everything a startup bank markets is a version of one of these.

1. Open accounts at multiple banks yourself

The manual method. $250K at Bank A, $250K at Bank B, and so on. It works and it is fully within your control, but it is operationally miserable at scale: multiple logins, multiple wire setups, multiple reconciliations, and your controller hates you. Fine for two or three banks. Painful past that.

2. Sweep networks (IntraFi ICS / CDARS)

This is the mechanism doing the heavy lifting behind most startup banking products. A sweep network takes your deposit at your primary bank and distributes it, in sub-$250K chunks, across many other FDIC-insured banks in the network. The largest is IntraFi, which runs ICS (Insured Cash Sweep, for demand accounts like checking and savings) and CDARS (for CDs). IntraFi says more than 3,000 financial institutions participate in its network as of 2026 (IntraFi).

The result: you keep one banking relationship and one login, but your cash is spread so that each slice sits under the $250K limit at a different bank. That is how a provider advertises “$5M” or “$75M” of coverage. It is not one bank insuring $75M. It is your cash cut into 300 pieces across 300 banks, each piece insured to $250K.

Here is what that arrangement actually requires, in plain English, from someone who reconciles multi-bank cash movement for a living. The primary institution offering you the big coverage number has to have established relationships and live operations with an entire network of other FDIC-insured banks. On the surface, your full balance looks like it sits at the primary bank. Behind the scenes, that institution is responsible for tracking which network bank actually holds each slice of your money, so that every piece stays under $250K. Run the arithmetic on a $75M ceiling: that is roughly 300 institutions at $250K each, and realistically more than that to handle different states and entity types. Every one of those relationships has to be pre-established, the sweeps have to be operational, and the reconciliation has to be watertight before your money ever shows up.

And that reconciliation is the real failure point. The individual receiving banks have no insight into what is held at the other banks in the network. It is entirely the one primary institution’s job to track where every dollar sits and reconcile it all back to the single balance you can access at any moment. The FDIC has to be able to identify your entity as the beneficial owner at each network bank, and that identification lives in the primary institution’s records. If the primary is not excellent at that tracking, the coverage number is only as good as their bookkeeping. Founders read “$75M FDIC” as a guarantee. It is a records-and-mechanics dependency.

So before I trusted a coverage number with real money, here is what I would ask the provider, in writing:

1. How many institutions do you have relationships with, and how widely can my funds actually be spread?
2. What is the maximum amount held at any single institution?
3. Are there membership requirements anywhere in the network, for example a credit union that requires membership eligibility?
4. Are there any state-level coverage limitations that apply to my entity?
5. Have there been complaints or operational issues with the sweep program?
6. Are there any limits on my cash, or on what I can withdraw at any given point?

If a provider cannot answer those quickly and specifically, the headline number is marketing, not treasury.

3. Treasury / money market funds (NOT FDIC, read this twice)

The third option is to move cash out of deposits entirely and into an investment product: a government money market fund or Treasury bills, usually branded as the provider’s “Treasury” product.

These are not FDIC-insured. A money market fund is a security held at a broker-dealer, not a deposit held at a bank. Its protection, if any, is SIPC coverage, which remains up to $500,000 total per customer, of which up to $250,000 for cash, as of 2026 (SIPC), and SIPC protects you against the brokerage failing, not against the fund losing value.

That is a completely different risk than FDIC insurance. It is not worse in practice for a high-quality government money market fund, but it is different, and conflating the two is the single most common mistake founders make when a bank tells them their cash is “protected.” See our breakdown in best high-yield account for startup runway.

The provider landscape (as of September 2026)

The startup banks compete partly on how much sweep coverage they extend. Current ceilings, all of which move without much announcement:

Two footnotes the marketing pages bury. Mercury’s extended coverage requires balances to actually be in the sweep; a non-sweep balance at a single partner bank is capped at the standard $250K. And Rho’s $75M applies to its savings sweep, not to checking or its treasury product. Also note that Rho’s network includes NCUA-insured credit unions alongside FDIC-insured banks, which is exactly why the membership-requirements diligence question above exists.

Rho’s $75M is the headline for cash-heavy companies, but do not pick a bank on a coverage number alone. A higher ceiling you never approach is worth nothing; the mechanics and the operational fit matter more. See Rho vs Mercury and Mercury vs Brex vs Ramp for the full comparisons.

The practitioner framework: three buckets

Here is the mental model that keeps startups out of trouble, separate from any product.

1. Operating cash (weeks of payroll and AP): sits in checking, needs to be instantly liquid. Accept that some of it is uninsured; keep the amount small and deliberate.
2. Reserve cash (months of runway): this is what belongs in a sweep network so it is FDIC-covered across many banks.
3. Invested cash (the long tail of runway you will not touch for quarters): this is the treasury/money-market bucket, chasing yield, protected by SIPC not FDIC, and sized against your burn.

My honest operator take on how those buckets fill up. In the very beginning, it is all operating cash, and how you burn it should be a sensitive topic. If a dollar is not working toward profitability, question whether it needs to be spent at all, because burn rate and runway are everything early. The next layer is the reserve: something you can tap to survive a down month or a down quarter and keep the engine running while you work back toward profit. Size it against your actual burn, not a round number.

On the third bucket, here is the part the yield-product marketing will not tell you: the best investment for most startups is usually the business itself. Better employees, better efficiency, a shorter path to profitability. Those beat clever yield on idle cash almost every time, and growing top-line revenue is not the same thing as being profitable. So the invested bucket should only ever hold the true long-tail cash you genuinely will not touch for quarters, sized deliberately against burn. If you find yourself optimizing basis points on a big idle pile while the business underneath is not yet profitable, the pile is answering the wrong question.

The startup treasury policy template turns this framework into a one-page document your board will actually approve.

What to do this week

• Find out how much of your cash is currently sitting above $250K at a single bank. For most funded startups the answer is “almost all of it.”
• Move reserve cash into a sweep-network-backed account so it is FDIC-covered across many banks.
• Understand that your treasury/yield product is SIPC, not FDIC, and size it deliberately.
• Write it down as a policy so it survives the next person who touches the bank accounts.

The honest caveat

FDIC rules, SIPC limits, sweep-network ceilings, and each provider’s partner-bank arrangements change, and providers revise the specifics quietly. Every figure here is dated as of the last review. Confirm the live number on the FDIC’s own site and the provider’s disclosure before you move money. This site is a publisher, not a bank and not a financial advisor; nothing here is individualized financial advice. When we earn a commission from a sign-up it never changes what we tell you, because the only asset this site has is being a source you can trust.

Disclosure: some links on this page are affiliate links. See our methodology and disclosures.