How Big Should Your Emergency Fund Be, Where Should It Sit, and How Long Will It Take?

The emergency fund is the rare piece of money advice nobody argues about, and the rare one most households have never actually built. That gap is not really about discipline. If you do not know how much it should be, where it should live, or how many months it will take, there is no first step to take — so nothing happens. This is the whole thing, with numbers.

What counts as an emergency

An emergency is an expense you could not have seen coming three months ago and cannot postpone. Both halves have to be true. That sounds like hair-splitting, but it is the difference between having the money on the day it matters and discovering the fund was quietly spent in October.

• The boiler, the fridge or the washing machine dies
• A car repair that has to happen before you can drive to work
• An unexpected settlement bill from the energy or water company
• Dental work, or a medical cost your cover does not take
• Losing a job, having your hours cut, or a client that does not pay
• A move you did not plan, with a new deposit attached

What is not an emergency: Christmas, the summer holiday, winter tyres, the annual insurance premium, the phone whose battery has been fading for six months. Those are predictable expenses, and they belong in their own savings pots with a monthly amount. Households that pay for them out of the emergency fund reliably have no emergency fund in December, and January reliably brings a bill.

How much: three stages, not one number

The usual answer is “three months of salary”, and the benchmark is wrong. In an emergency you do not need your income, you need your baseline: everything that keeps running once saving and wants are switched off. Rent or mortgage, utilities, insurance, food, getting to work, and the minimum payments on any credit.

Take a household on 2,400 a month, with a baseline of 1,800. That fixes the stages.

1. Stage 1 — 1,000. Covers the ordinary repair and keeps you out of the overdraft. This stage matters more than any investment, and it is reachable quickly.
2. Stage 2 — one month of baseline, so 1,800 here. From this point an unexpected bill is an annoyance rather than an event.
3. Stage 3 — three months of baseline, so 5,400. That is the target for an employee in a stable job.

Six months, 10,800 here, is the right figure for the self-employed, single earners with children, anyone with a chronic condition, and homeowners, where a roof is part of the risk. Holding much more than six months in cash is rarely worth it: past that, the safety costs you more in returns than it buys in calm.

Where to keep it

The account has to do three things: release the money within a day or two, never fall in nominal value, and be slightly inconvenient to reach from a supermarket queue.

• An instant-access savings account is the standard answer. Available on demand, protected by the deposit guarantee scheme up to the limit per bank per person. The interest rate is a bonus, not the selection criterion — availability is the whole point.
• Not your current account. Money sitting next to everyday money is everyday money within two months.
• Not a notice account or a fixed-term deposit. A fund you cannot reach for ninety days is not an emergency fund.
• Not invested. Shares and funds are the right home for money you will not need for years; an emergency has an unhelpful habit of arriving in the same month the market is down 20%.
• Not a credit card or an overdraft “in reserve”. That is not a fund, it is a debt you have pre-agreed to take on at the worst possible moment.

One practical detail that decides more than the interest rate: keep the fund at a different bank from your current account, without the card in your wallet and without the app on your home screen. A transfer that takes a day is exactly the friction you want between a bad afternoon and 400 of your buffer.

The plan, and how long each stage really takes

Same household, putting aside 200 a month by standing order on the day after payday. Not at the end of the month out of what is left, because what is left is a number nobody has ever managed to predict.

1. Stage 1, 1,000: five months. Bring it forward with anything irregular — a tax refund, a bonus, the deposit from an old flat, something sold.
2. Stage 2, 1,800: another four months, so nine in total.
3. Stage 3, 5,400: eighteen months more, about two and a quarter years from the start.

Two years sounds slow, which is exactly why people quit in month three. Look at the risk curve instead: the first 1,000 removes most of the situations that would otherwise have become expensive debt, and that stage is done in five months. The long part of the road is the safe part.

If 200 is not available, the honest answer is that 50 is fine. Stage 1 then takes twenty months, and it still ends with you having a thousand and a habit. A fund that grows slowly beats a plan that assumed 400 and got abandoned in week six.

Emergency fund or paying off debt first?

The one genuine argument in this topic, and it has a practical answer rather than an ideological one. If you carry credit card debt at 20% or more, saving 3% while paying 20% is a loss on every unit. But going in with no buffer at all is worse, because the next repair goes straight back onto the card and the balance never falls.

• Build stage 1 first — a small buffer, quickly, so the card is not the emergency plan.
• Then throw everything at the expensive debt until it is gone. Nothing you can safely buy pays 20%.
• Then go back and build stages 2 and 3, with the money that was going to the debt.

One exception worth knowing: if your employer matches pension contributions, keep taking the match throughout. A 50% match is a return no debt interest rate beats.

When you spend it — because you will

Using the fund is not a failure. It is the fund working. The failure mode is the month after, when nobody restarts the standing order and the balance stays where the emergency left it.

So make refilling automatic and boring: the standing order never stops, and after any withdrawal you write down the new target date on the calendar. If the fund went from 3,000 to 1,200, at 200 a month you are back at nine months. Knowing that is what stops one bad month becoming a year of drift.

The part that actually decides the number

Everything above turns on one figure: your real baseline. Most people guess it 20 to 30% low, because they think of rent and food and forget insurance, the annual bills and the standing orders that leave without asking. Guessing it low means a fund that runs out in week seven of a three-month problem.

The only way to know it is to have a few months of your own spending written down and split into fixed and variable. That is what our app is for: every expense recorded in seconds — photograph a receipt and the AI reads the amount and date itself — recurring bills entered once with their due dates, and a monthly report that gives you the fixed-cost total as a single line. Set the fund up as a savings goal and the progress bar does the rest.

The emergency fund is not a clever piece of financial engineering. It is the thing that decides whether the next broken washing machine is an afternoon of irritation or eighteen months of credit card interest. Work out your baseline, set the standing order for tomorrow, and let stage 1 arrive before summer.