Three Months Is Not a Number. It's a Guess.

The email lands at 9:40 on a Tuesday. Twenty minutes later your laptop stops syncing, and you are sitting in a kitchen that has become very quiet. The first question is not what do I do now. For almost everyone, the first question is how long can I do nothing?

That number — the count of weeks between today and the day your balance turns into a crisis — is the only thing an emergency fund actually buys. Not peace of mind in the abstract. Weeks. And most people have never once sat down and worked out how many weeks they own.

"Three to six months" is a slogan, not a calculation

The standard advice is to keep three to six months of expenses in cash. It is repeated so often that it has stopped sounding like a recommendation and started sounding like a law of nature. The trouble is that it bundles together people whose situations have almost nothing in common.

Consider two readers. The first is a salaried accountant in a mid-sized city whose partner also works, with no children and a skill set that a dozen local firms hire for. The second is a freelance designer, the only earner in the household, paid per project, in an industry where budgets vanish first when the economy tightens. The same sentence is handed to both. For the first, six months of cash is probably over-saving — money sitting idle that could be doing something. For the second, six months may be thin.

An emergency fund is not a badge. It is a number of weeks, and yours is not your neighbour's.

The fix is not a better slogan. It is doing the arithmetic once, properly, with your own two inputs: what your life costs at its minimum, and how long you would plausibly need to cover it.

Step one: find your floor, not your spending

Most people, asked what they spend per month, quote a number that includes a lot of things they would stop doing immediately if income disappeared. That number is the wrong input. What you need is your floor — the cost of keeping the lights on, the roof over your head, and everyone fed, with no discretionary spending at all.

Pull three months of statements and sort every recurring line into two columns:

The gap between the two columns is usually larger than people expect. A household spending $4,200 a month often has a floor closer to $2,900. That difference is not trivial — it is the difference between needing roughly $25,000 of savings and needing roughly $17,000.

One honest note while you sort: the "basic version" of groceries is a real category, not a fantasy. But do not build a floor you could only survive for a fortnight. If the number assumes you never buy coffee, replace a pair of shoes, or take a bus anywhere social for six months, you have built a spreadsheet, not a plan.

Step two: multiply by how long you would actually be out

Here the data helps more than intuition does. In the United States, the median duration of unemployment was around 11 weeks in early 2026, while the average sat near 26 weeks. That gap is the whole story. Most people find something in roughly three months. A meaningful minority are out for half a year or longer, and they drag the average up.

You do not get to know in advance which group you are in. What you can assess is which way your own circumstances lean. Ask: how many employers within commuting distance hire your exact skill? Is your income one stream or two? Does anything — a visa, a mortgage approval, a custody arrangement — depend on you being employed by a specific date? Is your pay steady, or does it arrive in lumps you cannot schedule?

A reasonable way to translate that into a multiplier:

Adjust downward slightly if you have genuine severance entitlement or unemployment insurance that replaces a real share of your pay, and upward if you are self-employed and entitled to neither. This is a starting point, not a verdict.

The number, worked

Take the household above. A floor of $2,900, a single salaried earner with a child, so a multiplier of six: $17,400. Written down like that, it is an intimidating figure, and this is the moment most people close the tab.

Do not. The target is not the point of entry. The first $1,000 to $2,000 does a wildly disproportionate share of the emotional work, because it covers the things that actually happen to people: the transmission, the emergency root canal, the boiler that dies in February. Those are not unemployment — they are the ordinary shocks that push someone onto a credit card and keep them there for two years.

So build it in tiers, and treat each one as finished business:

• Tier 0 — $1,000. Stops small disasters from becoming debt. Weeks, not years, to reach.
• Tier 1 — one month of floor costs. You can now absorb a late invoice or a gap between jobs without panic.
• Tier 2 — your full multiplier. The version that buys you the right to turn down a bad offer.

The first thousand stops the bleeding. The rest buys you the right to say no.

That last point is underrated. People describe emergency funds defensively, but the most valuable thing cash does is widen your options. Someone with four months of runway negotiates differently, leaves a toxic manager sooner, and does not take the first job that clears the mortgage.

Where the money should sit — boring on purpose

Your emergency fund has to pass three tests, in this order: you can reach it within a day or two, it cannot fall in value, and it is not so reachable that it quietly becomes your spending money.

That last test argues for a separate account at a separate institution from your day-to-day banking. Not because anyone is stopping you, but because a balance you do not see when you check your phone is a balance you do not casually spend. Interest matters less here than people think — the yield on savings accounts has varied considerably, and rates as of writing differ widely by country and institution — but it is free money for a few minutes of setup, so take it where a reputable insured account offers it.

Keep it insured and keep it whole. In the US, deposits are covered up to $250,000 per depositor, per insured bank, per ownership category; other countries run their own schemes with different limits, so check which applies to you. And resist the temptation to invest this particular pot. Emergencies are correlated with downturns: the month layoffs sweep your industry is disproportionately likely to be a month markets are falling. A fund designed to be sold at the worst possible moment is not a fund.

Filling it without hating the next year

The mechanism that works is dull and automatic. Set a transfer for the day after payday, not the day before the next one, and size it so you barely notice — an amount you would not have felt as a raise is an amount you will not feel as a loss. Consistency beats intensity here; $200 every month for a year outperforms one heroic $800 push in March followed by nothing.

Windfalls do the heavy lifting. A tax refund, a bonus, a deposit returned, a side project that paid better than expected: split it in half, spend half guilt-free, send half to the fund. People who try to save windfalls entirely tend to save none of them.

Two distinctions worth drawing now. First, sinking funds are not emergency funds. Annual insurance, car registration, December holidays, the family trip — those are predictable, so they get their own pot. Mix them and you will drain the emergency fund every December and call it an emergency. Second, if you carry high-interest debt, the order is usually: build Tier 0, then attack the expensive debt hard, then finish the fund. Paying down a balance charging over 20% beats earning single digits in savings — though if your employer matches retirement contributions, taking that match first is rarely the wrong call.

Decide now what counts as an emergency

The rule is much easier to write while you are calm. A simple, durable test: unexpected, necessary, and urgent — all three. Job loss qualifies. A medical bill qualifies. A failed refrigerator qualifies. A sale, a destination wedding, or a phone that is merely two years old does not, however convincingly you can argue it at 11pm.

Write the rule somewhere you will see it, and add one more line to it: the refill clause. If you spend from the fund, the automatic transfer restarts that same month, even at a smaller amount. The people who never rebuild are almost always the ones who treated the withdrawal as the end of the project rather than a normal part of it.

The short version

• Use your floor — the minimum your life costs — not your current spending.
• Multiply by a number that reflects your actual job risk, not a slogan. Unemployment has a median of about 11 weeks and an average of about 26; your circumstances tell you which end to plan for.
• Build in tiers: $1,000, then one month, then the full figure.
• Keep it liquid, insured, separate, and unglamorous.
• Define "emergency" before you need to, and always restart the transfer afterwards.

None of this makes the Tuesday email pleasant. But there is a real difference between opening it and feeling the floor drop, and opening it and thinking: I have until March. That sentence is the whole product. Everything above is just the arithmetic that gets you there.

This article is general information as of the time of writing, not personalised financial advice. Figures, rates, and insurance limits change and vary by country — check what applies where you live, and speak to a qualified adviser about your own situation.