How to build an emergency fund

What it is actually for

An emergency fund converts a shock into an inconvenience. Its job is not to grow — it is to be available on the worst day of your year, without you having to borrow at 20%+ or sell investments after they have fallen.

The evidence that this is a common gap is unusually good. The US Federal Reserve's annual Survey of Household Economics and Decisionmaking asks whether households could cover a $400 emergency expense with cash or its equivalent, and consistently finds a substantial minority could not. Regulators elsewhere ask similar questions and find similar gaps. If you are in that group, this is the highest-return financial move available to you — not because it pays interest, but because it stops you paying it.

Sizing it: use expenses, not income

The common rule is “three to six months”. Three to six months of what is the part that gets skipped. Use essential monthly expenses — the number you would spend in a bad month, not your salary:

• Rent or mortgage payment
• Utilities, phone, internet
• Groceries (not restaurants)
• Transport to work
• Insurance premiums
• Minimum debt payments
• Childcare, medication and any other non-negotiable

For most households that figure is far below take-home pay, which makes the target much less intimidating than “six months of income”.

When three months is not enough

The right multiple depends on how quickly your income could recover, not on a rule:

• Lean towards 3 months if you have stable salaried employment in a field that hires quickly, a second income in the household, no dependants, and good health cover.
• Lean towards 6 months if you are the only earner, have dependants, or work somewhere with long notice-to-rehire cycles.
• Lean towards 9–12 months if you are self-employed, on commission or contract, work in a highly cyclical industry, or have a health condition that could interrupt work.
• Add a separate buffer for the specific large item you can already see coming — an old car, an ageing roof, a visa renewal. That is a sinking fund, not an emergency fund; keeping them apart stops predictable costs from eating your emergency cash.

If your country provides substantial unemployment or sickness benefits, you can reasonably sit at the lower end. If it does not, sit higher. This is one of the places where national context changes the answer more than any personal-finance rule does.

Where to keep it

Three requirements, in order: accessible within days, nominally safe, and separate from your spending account. Growth is a distant fourth.

• A separate savings account — ideally at a different institution, so transferring takes a day and impulse spending takes effort. Check that deposits are covered by your country's deposit guarantee scheme and stay within its limit per institution.
• Not invested. Stocks can be down 30% precisely when you are laid off; the two events are correlated, which is exactly the wrong property for this money.
• Not locked up. A fixed-term deposit with an exit penalty is not an emergency fund. A short ladder of term deposits can work for the portion beyond your first three months.
• Not in a volatile currency if you earn and spend in something else, and not in crypto — the point of this money is that its number does not move.

Getting there faster

1. Automate a transfer for payday, not month-end. What is left at month-end is not a plan.
2. Start with one month, not six. The first month of essentials removes most of the borrow-for-emergencies risk; momentum handles the rest.
3. Route one-off money straight in — tax refunds, bonuses, a sold bike. These build a fund far faster than trimming a monthly budget.
4. Pause investing, not debt payments, while you build the first month if you have to choose. Then reverse it: with a starter buffer in place, high-interest debt outranks a larger fund, because a 22% APR balance costs far more than a savings account pays.
5. Refill it first after each use, before resuming anything else.

Use the 50/30/20 budget split to find the monthly amount you can realistically commit, then divide your target by it to get an honest timeline.

Four ways people get this wrong

• Counting a credit card as the emergency fund. It is a loan that appears at the worst moment and can be reduced or withdrawn by the lender exactly when conditions deteriorate.
• Keeping it in the current account. Money you can see gets spent. Separation is the feature.
• Chasing yield. The difference between a good and mediocre savings rate on a $10,000 fund is a few hundred a year — real, but not worth an access delay or a market risk.
• Never using it. A fund used for a genuine emergency did its job. Guilt about spending it is how people end up borrowing instead.