How Big Should Your Emergency Fund Actually Be?

"Three to six months of expenses" is the most repeated rule in personal finance, and it is nearly useless as stated. Three months of which expenses? And why does a tenured government employee with no dependants get the same advice as a freelance contractor supporting three people?

The size of your emergency fund should fall out of two things: how much a bad month actually costs you, and how likely a bad month is. Let's calculate both.

Step one: find your bare-bones monthly number

Your emergency fund does not need to cover your current lifestyle. It needs to cover your life with the discretionary parts switched off.

If you have not yet mapped your spending, build a basic budget first — this calculation needs those numbers.

List only what you would still be paying if your income stopped tomorrow:

• Rent or mortgage
• Utilities and phone
• Groceries (not restaurants)
• Insurance premiums
• Minimum debt payments
• Transport to look for work
• Childcare or care costs you cannot pause
• Medication and essential health costs

Deliberately exclude: restaurants, subscriptions you could cancel, holidays, gym, new clothes, gifts, and any savings or investing contributions.

For most households this bare-bones number lands somewhere between 60% and 75% of normal monthly spending. That gap matters enormously — it means the fund you need is meaningfully smaller than the one implied by "six months of expenses," which people usually interpret as six months of total spending.

If your normal spending is $4,000 and your bare-bones number is $2,700, then six months is $16,200, not $24,000. That difference is a year of saving for many people.

Step two: choose your multiplier honestly

Now decide how many months of that bare-bones number you need. Start at three and adjust:

Add them up. A salaried software engineer with a working partner and no children might land at two to three months. A self-employed graphic designer who is the sole earner for a family of four might land at eight or nine. Both are correct answers.

Step three: build it in stages, not all at once

An eight-month target is demoralising when you have nothing saved. So do not aim at it directly.

Stage 1 — $500 to $1,000. This is the anti-credit-card tier. Its entire purpose is that a flat tyre, a broken phone, or an unexpected medical bill does not become debt. Most people can reach this in one to three months by pausing every optional expense. Do this before anything else, including extra debt payments.

Stage 2 — one month of bare-bones costs. This changes your relationship with your employer and landlord. You are no longer one bad week from a crisis.

Stage 3 — your full calculated target. Build this slowly, alongside other goals. There is no prize for reaching it quickly, and money sitting in cash is not earning much.

Splitting it this way matters psychologically. Stage 1 is achievable now, and finishing it produces the momentum that carries the rest.

Where to actually keep it

Three properties, in priority order: accessible within a few days, principal that cannot fall, and only then interest.

Reasonable homes for it:

• A high-yield savings account at a separate institution from your daily bank
• An easy-access or notice savings account
• A money market fund, if your platform settles quickly

Poor homes for it, despite the temptation:

• The stock market. Emergencies correlate with recessions. The month you lose your job is disproportionately likely to be a month the market is down 20%.
• Your daily current account. Not because of interest, but because you will spend it. Friction is a feature.
• Anything with a withdrawal penalty or a fixed term. Access is the entire point.
• Crypto. Volatile and, depending on where you are, may take days to convert.

The "separate institution" detail is underrated. Money in the same app as your spending account is one tap away. Money at a different bank requires a transfer that takes a day, and that day is usually enough to reconsider.

What actually counts as an emergency

The fund only works if you are strict about this, and most people are not. A usable test: is it unexpected, necessary, and urgent? All three, not two.

• Car breaks down and you need it for work — yes, all three.
• Boiler fails in winter — yes.
• Job loss — yes.
• Medical bill — yes.
• Christmas — expected. That is a sinking fund, not an emergency.
• Annual insurance renewal — expected.
• A flight sale to somewhere you have always wanted to go — none of the three.
• Replacing a working laptop with a faster one — not urgent.

The failure mode is not one big wrong withdrawal. It is a slow drift where "emergency" quietly expands to mean "unbudgeted," and the fund never gets above $800.

Emergency fund first, or pay off debt first?

The genuinely correct answer depends on the interest rate, but the practical sequence that works for most people is:

1. Build the $500–$1,000 starter tier first, even with 22% credit card debt. Without it, the next surprise goes straight back onto the card and you never escape.
2. Then attack high-interest debt aggressively — anything above roughly 8–10% beats the guaranteed return of extra cash savings.
3. Then return to building the full fund.

The exception: if your debt is low-interest — a subsidised student loan, a 0% promotional balance you will clear in time, a mortgage — build the full emergency fund first. Paying an extra $200 against a 3% loan while having no cash buffer is a bad trade.

Refill it without guilt

You will use it. That is not a failure — that is the fund doing its job. A fund that has never been touched is not necessarily a well-managed fund; it may just be an uneventful few years.

When you spend from it, treat refilling as the next month's top priority, ahead of investing and ahead of extra debt payments. Then go back to normal.

The point of this money is not to grow. It is to make a bad month boring — which is exactly why it belongs in cash rather than in the market, a distinction covered in saving vs. investing.