


In March 2023 a large number of startups discovered simultaneously that their entire cash position sat in one institution, largely uninsured, and that they could not move it fast enough.
The lesson was widely misread. It was not that a particular bank was a poor choice — it had served the ecosystem well for decades. The lesson was concentration: everything in one place, above insurance limits, with no prepared alternative.
This guide covers what deposit insurance actually protects, the categories of provider and how they differ, how to structure cash sensibly, what changes as you grow, and the tension between diversification and what your lender may require.
Federal deposit insurance protects deposits up to a standard limit per depositor, per insured bank, per ownership category. That limit is $250,000, and it has not moved.
The arithmetic is immediate: a company that has raised $8 million and holds it in one bank has the overwhelming majority of it uninsured. This is normal, it is not a scandal, and it is the reason structure matters.
Two further points founders regularly get wrong:
Enormous balance sheets and systemic importance. Less familiar with venture-backed companies, slower to onboard, and frequently indifferent to a small startup. Reliable, unexciting, and hard to get good service from at small scale.
Understand venture-backed companies, offer venture debt and credit alongside deposits, and can move quickly. This category consolidated significantly after 2023 but has not disappeared — several institutions expanded into the space.
Excellent product experience, fast onboarding, good integrations with accounting and payroll, and modern treasury features.
The critical distinction: most of these are not themselves banks. They are technology companies that partner with chartered banks to hold deposits. Your money sits at a partner bank; the platform provides the interface and the programme management.
That arrangement is entirely legitimate and widely used. It also means:
The practical response is not to avoid these products — they are genuinely good — but to ask directly: who holds my deposits, at which chartered institution, and what happens to my access if you fail? A provider that answers clearly is telling you something useful; one that does not is telling you something else.
Services that spread a single balance across many partner banks, each holding an amount within the insurance limit, so a large deposit can be substantially insured. Offered by many banks and platforms.
Genuinely useful. Check how quickly funds can be withdrawn, what the yield is net of fees, and how many banks are in the network.
For balances beyond operating needs, cash can sit in government money market funds or short-dated Treasury bills rather than as bank deposits. These are not deposits and are not FDIC insured — they carry different risk, backed by the underlying securities rather than by insurance.
For a company holding a large round, this is frequently the sensible destination for the bulk of the balance.
The single most useful thing you can do is unglamorous and takes an afternoon.
This structure solves most of the 2023 problem regardless of which providers you choose, which is why it deserves more attention than the vendor debate.
Treasury requirements step up at fairly predictable points, and the error is either building institutional infrastructure at seed or still running a single account after a Series B.
Pre-seed and seed. One account is defensible when the balance is below or near the insurance limit, because concentration risk is theoretical at that size. What matters here is the basics: the account is in the company's name, personal and company money never mix, and someone other than the founder can access it if necessary. Open the second relationship once the balance exceeds insured limits, which usually means immediately after the first real round.
Series A. The structure described above becomes genuinely necessary — two institutions, operating separated from reserves, and the bulk of the balance in a sweep network or government instruments. This is also the point at which a written treasury policy is worth the hour it takes, because the board will start asking and because the person managing cash may no longer be the founder.
Series B and beyond. Balances are now large enough that yield is a real number rather than a rounding error, and the treasury question becomes an active management question. Expect a finance hire to own it, laddered maturities matched to the spending plan, and counterparty limits enforced rather than aspirational. If you have taken debt, covenant compliance and reporting now sit alongside the cash management.
International operations. A separate step change whenever it happens. Local accounts in each operating jurisdiction, intercompany funding arrangements, currency exposure on payroll and revenue, and reporting obligations that vary by country. This is the point where treasury stops being an administrative task, and it is worth specialist advice — particularly for companies that have completed a Delaware flip and now run two entities in two currencies.
Approaching an exit or a listing. Acquirers and auditors examine cash management, controls and counterparty exposure. A company with a documented policy, evidenced dual authorisation and clean reconciliations moves through this quickly. One without spends diligence explaining why.
Here is the conflict founders discover at the worst moment.
Bank lenders frequently require, as a covenant, that you maintain your operating accounts — and sometimes substantially all your cash — with them. That is precisely the concentration you are trying to avoid.
The requirement is rational from the lender's perspective: deposits fund lending, visibility into your cash informs their risk, and control over the account matters in a default.
What to do:
Once you hold meaningful cash, write this down and have the board approve it. It takes a page.
Bank failure is dramatic and rare. Payment fraud is undramatic and common, and it targets startups specifically during financings when large wires are expected.
Broadly yes, and millions of businesses do. Understand the structure: your deposits sit at partner banks, your protection depends on records being correct, and the failure of the platform is a distinct risk from the failure of the bank. Ask the questions above and keep a second relationship elsewhere.
Enough for near-term obligations with comfortable margin — many companies use one to three months of operating expenses. The rest belongs in insured sweeps or short-dated government instruments.
Modestly, and never at the cost of liquidity. Startup cash is runway, not an investment portfolio. Government money market funds and short Treasuries capture most of the available yield with minimal risk; anything more complex is optimising the wrong variable.
They did not before 2023 and several now do. A board-approved treasury policy and a documented backup relationship are cheap, and they demonstrate operational competence — which is exactly what diligence is assessing.
Adds complexity — different insurance regimes, FX exposure, tax and reporting obligations. If you have genuine international operations you will need local accounts. If you do not, keeping cash where your expenses are is simpler and cheaper.
Longer than founders expect, which is the entire argument for doing it early. Two to six weeks is common once you account for entity verification, beneficial ownership documentation, and the identity checks required on every signatory. A foreign-owned entity can take considerably longer. An account you begin opening during a crisis will not be ready in time, which is why the backup has to exist and be funded before anything goes wrong.
At least two people, always. A single-signatory account is a genuine operational risk — if that person is unreachable, ill or departs badly, the company cannot pay anyone. Give a second person independent access with appropriate limits, and make sure the board knows who has authority to move money. This is also a control question: the person who initiates payments should not be the only person who can approve them.
Insured deposits are typically available very quickly, often within a business day or two, because the resolution process is designed around exactly that. Uninsured balances take longer and are resolved through the receivership, with recoveries paid over time and historically substantial but not guaranteed. The operational consequence is what matters: payments in flight may fail, direct debits may bounce, and you may not know for several days what you can access. That is the scenario the backup account exists for.
Direct Treasury bills give you a known maturity and no fund fee, at the cost of managing the ladder yourself and being locked in until maturity unless you sell. A government money market fund gives you daily liquidity and diversification for a modest expense ratio, which for most startups is the better trade. The one thing to be clear on either way: neither is a deposit and neither is FDIC insured — the security comes from the underlying government obligations, not from insurance, and your treasury policy should say so explicitly.
The lesson of 2023 was concentration, not vendor selection. Open a second relationship before you need it, split operating cash from reserves, use insured sweeps or government instruments for the bulk of the balance, and negotiate a carve-out if a lender wants all your deposits.
Then write a one-page treasury policy and turn on dual wire authorisation — because the loss you are statistically most likely to suffer is fraud, not a bank failure.
Global Capital Network connects founders with investors, lenders and the financial institutions that serve them. See upcoming events or get in touch.
This article is general information, not financial or legal advice. Deposit insurance rules and provider structures vary and change. Verify current details directly with any provider before relying on them.



.png)




