How to build an emergency fund from scratch

An emergency fund is not a savings goal. It is insurance you underwrite yourself, and like all insurance it looks like wasted money right up until the week it is the only thing standing between an ordinary setback and a permanent loss.

That framing matters because it changes how you judge it. Nobody complains that their fire insurance underperformed the market last year. The buffer is held to the same standard: its job is to be exactly the amount you expected, on a day nobody warned you about.

What the fund is actually for

Three specific things, and it is worth being precise, because a pot with a vague purpose gets spent.

• It stops a forced sale. Without cash, an unexpected bill means selling something invested, at whatever price is on offer that week. Bad weeks for markets and bad weeks for households have an unpleasant habit of arriving together.
• It stops expensive borrowing. The alternative to cash on hand is usually credit at a rate that no investment plausibly beats. Avoiding an 18 percent balance is worth more than most portfolios will earn.
• It buys time to think. A household with three months of costs in the bank can take a week to decide what to do about a lost job. A household with nothing has to decide today, and decisions made today are worse.

How big, in months of essentials

The usual advice is three to six months of expenses. It is decent advice ruined by one word: expenses. What you need is months of essential spending, which is a much smaller and much more useful number. In a genuine emergency you are not funding your normal month, you are funding rent, food, utilities, transport, insurance and any minimum debt payments.

Where you sit inside the three to six range is not a matter of temperament. It is a matter of how quickly your income could be replaced and how many people depend on it. A person on a stable salary in a field with constant vacancies can sit near three. Someone self employed, on variable income, in a specialised field, or supporting a household on one wage, belongs nearer six or beyond.

Getting the essential figure right requires actually knowing it, which almost nobody does from memory. That measurement is the whole subject of how to track your spending, and it is the step that comes before this one.

How long it takes to build

This is the part that stops people, so here it is without softening. A six thousand dollar buffer, built at 250 a month, takes 24 months. At 100 a month it takes 60 months. Those are long stretches, and pretending otherwise helps nobody.

Which is why the fund is best built in two stages. The first thousand is the one that changes your life, because it covers the overwhelming majority of ordinary shocks: a tyre, a boiler part, an unplanned trip, an excess on a claim. Reaching it in four to ten months is realistic. The rest of the buffer can then be built more slowly, and in many households it is built alongside the first investment contributions rather than before them.

Where to keep it

Three requirements, in order of importance, and only three.

1. Reachable within about a day. A buffer you cannot get to during a bank holiday weekend is not fully doing its job.
2. The amount cannot move. No market exposure. Not a small amount, not a cautious amount, none. A buffer that can be 12 percent smaller during a crisis is not a buffer, it is a small investment with a misleading name.
3. Slightly inconvenient to spend. A separate account from your everyday one, without a card attached to it if that is possible. The friction is a feature.

Interest comes fourth, and a long way behind. If two accounts satisfy all three conditions, take the higher rate, and do not let a rate difference tempt you into an account that fails the first two. And yes, this money will lose ground to inflation over time. That is the premium you pay for certainty, and it is set out honestly in why saving and investing are different jobs.

What counts as an emergency

A rule that survives contact with real life needs to be applied in the moment, not debated afterwards. Three tests, and something has to pass all three.

• Unexpected. A car service scheduled in the calendar is not an emergency, it is a bill you knew about. Annual and irregular costs deserve their own planning, which is what turning goals into monthly numbers is for.
• Necessary. It has to be something that genuinely must be dealt with, not something that would be better dealt with now.
• Urgent. Waiting has a real cost, financial or otherwise.

A broken boiler in winter passes all three. A tempting flight sale passes none. A worn set of tyres passes two, and honestly should have been in the irregular costs list. The point of writing the tests down in advance is that you will be applying them at a moment when you are stressed and inclined to be creative.

Using it, then refilling it

Using the fund is not a failure. It is the fund working exactly as designed, and treating a withdrawal as a personal defeat is how people end up refusing to use the money and reaching for credit instead.

What matters is the refill, and the useful discipline is to decide the refill rate in the same week you spend it, not later. If the buffer was built at 250 a month, refilling at 250 a month is the default, and any investment contributions pause until the buffer is whole again. That pause is short, it is deliberate, and it protects the thing that protects everything else.

This page describes a general approach and uses round figures for illustration. It does not know your income, your obligations, your job security or the rules where you live, and none of it is a recommendation about any particular account.