Sinking Funds: The Piece Most Budgets Are Missing

Your car will need replacing. You know roughly when, and roughly what it will cost.

If that is $18,000 in five years, it costs $300 a month starting today — whether or not there is a line in your budget for it.

A budget with no such line is not a budget with a surplus. It is a budget with a liability it has not written down, and the month the liability arrives it will be paid with an emergency fund or a credit card.

A note before you start. This is general education, not financial advice. Figures are computed by this site's own savings-goal engine, which solves the future-value-of-an-annuity equation for the monthly payment. Investment returns are illustrative and not a forecast.

1. What a sinking fund is, and what it is not

A sinking fund is money set aside monthly for a specific, known, irregular expense.

The three words doing the work are specific, known and irregular.

The distinction is not pedantry. It is the reason emergency funds fail.

An emergency fund drained by a car repair, an annual insurance renewal and a Christmas is not an emergency fund that was too small. It is an emergency fund that was doing a job it was never designed for — and the household will conclude, wrongly, that they cannot maintain one.

Naming those expenses and funding them separately is what makes an emergency fund stay full.

2. The arithmetic

A sinking fund has exactly one calculation: cost divided by months.

Those four together are $709 a month.

That number is the point of this article. A household running all four of those expenses — which is an ordinary household, not an extravagant one — is carrying $709 a month of real cost that a monthly budget does not show.

On a $60,000 salary in Ohio, take-home is $4,121 a month. $709 is 17% of it. A 50/30/20 split that does not account for it is overstating available money by nearly a fifth.

3. Where the money comes from

This is the part people get stuck on, and there is no trick to it.

Sinking funds come out of the same take-home as everything else, so adding them means something else gets less. The value is not that they create money; it is that they move a cost from "surprise" to "scheduled."

Where each one belongs depends on what it is:

Replacement and maintenance of something you need — a car, a roof, a boiler — is a need. It belongs in the needs bucket, and it is the honest reason a needs bucket runs at 60% rather than 50% for most households.

Holidays, gifts and travel are wants. They belong in the wants bucket, and funding them monthly is strictly better than funding them in December on a card.

And annual insurance or registration is a need that simply arrives annually. Dividing it by twelve and treating it as a monthly bill is the whole technique.

The practical test: if the expense would happen whether or not you wanted it to, it is a need. If skipping it is genuinely an option, it is a want.

4. Cash or invested?

A sinking fund can earn a return. Whether it should depends almost entirely on the horizon.

Under about two years, do not bother.

On a twelve-month $2,000 goal, a 4% return supplies $36. Three dollars a month. That is not worth the friction of a separate account with a transfer delay, and it is certainly not worth any risk of the balance being lower than $2,000 in month twelve.

Over five years it starts to matter. $1,710 of an $18,000 car fund — 9.5% of the target — arrives from growth rather than from your paycheck. Over eight years, growth covers 15% of a $12,000 roof.

The rule that follows: short horizons want a high-yield savings account; long horizons can tolerate something with more return and more variance.

Two honest caveats. The 4% used above is illustrative, not a rate anyone is promising you, and a fund that can fall in value is not suitable for an expense with a fixed date. And a sinking fund's job is to be there — a car fund that is 15% down in the month the car dies has failed at its only task, however good the long-run arithmetic looked.

5. Which funds a household actually needs

Most people who adopt this idea start with fifteen sinking funds and abandon the system within four months. The tracking cost exceeds the benefit long before the fifteenth one.

Three to six is the range that survives contact with real life.

The four that earn their place in almost every household:

Car. Replacement, plus tyres and the repairs an older car generates. The single largest irregular expense most households face, and the most commonly unfunded.

Home or rental maintenance. For an owner: roof, boiler, appliances, exterior. For a renter this one is smaller but not zero — moving costs, a deposit on the next place, furniture.

Annual bills. Insurance renewals, registration, professional fees, annual subscriptions. Add them up, divide by twelve. This is the easiest one and it removes the most surprises.

Holidays and gifts. Predictable, dated, and the single most common cause of January credit card balances.

Two more that are worth it if they apply to you:

Medical. Deductibles and out-of-pocket maximums are known numbers. If you have an HSA, this fund and that account are the same thing — and it is the most tax-efficient dollar in the system, because a payroll HSA contribution escapes FICA as well as income tax.

Pets. Veterinary costs are irregular, large, and almost never budgeted.

What does not need its own fund: anything under about $200 a year. Fold it into the annual bills line and stop counting.

6. One account or many?

Both work. The trade-off is friction against clarity.

Separate accounts — one per fund — make the balances unambiguous and make it psychologically harder to spend the car money on a holiday. The cost is administration, and most banks make opening six accounts more annoying than it should be.

One account with a spreadsheet tracking what each balance is "for" is simpler to run and requires the discipline to believe the spreadsheet. For most people this is the right answer, because the version you actually maintain beats the version you abandon.

What does not work is keeping sinking funds in your current account. Money that is visible in a checking balance is money that gets spent, and the entire mechanism depends on the balance not feeling available.

One structural point worth making: keep sinking funds separate from the emergency fund even if they are in the same institution. The moment they are one pot, section 1's distinction collapses and the emergency fund starts absorbing predictable costs again.

7. What starting late costs

The monthly figure is the target divided by the months remaining, which means every year you wait raises the payment — and it does not rise gently near the end.

The same $18,000 car:

From ten years out to five, the cost doubles. From five to one, it multiplies by five.

This is the arithmetic that makes "I will start next year" expensive. A year's delay on a five-year horizon takes the payment from $300 to $375 — a 25% increase for twelve months of not starting.

It also explains why a partially funded sinking fund is still worth having. A car fund that reaches $11,000 of an $18,000 target has not failed; it has converted an $18,000 problem into a $7,000 one, and $7,000 on a card at 24% is a very different situation from $18,000.

Two practical conclusions:

Start with a realistic horizon, not an optimistic one. If the car is eight years old, do not budget as though it has ten years left. The horizon that matters is when it will actually need replacing, and people systematically overestimate it.

And revise the horizon rather than the target when things change. If the car survives two years longer than expected, you now have a larger fund and a shorter remaining horizon — which is the pleasant version of this arithmetic and the reason to start early.

8. The difference between this and an emergency fund, in numbers

Section 1 drew the distinction in words. Here it is in dollars, because the sizing methods produce very different answers.

An emergency fund is sized on essential monthly spending:

A sinking fund is sized on the cost of one specific thing, and the two numbers have no relationship to each other.

Which produces the practical failure this article opened with. A household with a $7,200 emergency fund feels well prepared. Then the car dies. $18,000 leaves, and they now have neither a car fund nor an emergency fund — and the emergency fund will take another eighteen months to rebuild at $400 a month, during which any actual emergency is uncovered.

Whereas the same household running a car sinking fund at $300 a month has $18,000 for the car and $7,200 still sitting there. Same total saved. Completely different outcome, entirely because the money was labelled.

Two things that table also shows:

Emergency funds take a long time. Eighteen months for three months of runway at $2,400 of essential spending. That is the normal case, not a slow one, and knowing it prevents the abandonment that comes from expecting it to happen in a season.

And they are sized on spending, not income, which is why the emergency fund article starts there. A household that cut its spending has cut its emergency fund target at the same time — the target moves with the thing it is protecting.

9. How to start from nothing

Four steps, in order, and the first one is the only difficult one.

Write down every non-monthly expense from the last twelve months. Bank statements, not memory — memory systematically omits the annual ones, which are the ones this exercise exists to catch.

Add the ones you know are coming but have not had yet. The car you will replace. The boiler that is fifteen years old. A cost you can see coming is a sinking fund, not a surprise.

Divide each by the months until it is due, and total the column. That total is your real monthly cost of living, and it is higher than the number you have been using.

Then start with the largest one only. Funding one sinking fund properly beats funding six at a third of the required rate, because a fund that is 30% full when the expense arrives has not prevented the credit card — it has only made it smaller.

And accept the first year is the hard one. A car fund started three years before the car dies needs $500 a month, not $300. Starting late costs more, which is an argument for starting now rather than an argument against starting.

10. Six funds, one household, one total

A worked example, so the idea has a number attached rather than a principle.

A household with a five-year-old car, a rented flat, one pet and an ordinary December:

$917 a month. On a $60,000 Ohio take-home of $4,121, that is 22% of everything they receive.

Three reactions are all reasonable, and only one of them is wrong.

"That is too much, I will fund the top three." Correct. Car replacement, annual bills and holidays are $584, cover the largest and most predictable costs, and are achievable. Partial coverage is the normal state of this system.

"That explains where my money goes." Also correct, and it is the main value of doing the exercise even if you fund none of them. A household that knows it carries $917 a month of irregular cost makes different decisions from one that thinks it has a $900 surplus.

"That cannot be right, nobody spends that." This is the wrong one, and it is the most common. Everybody with a car, a home and a family spends this. The only question is whether it is spread across twelve months or concentrated into the four months something goes wrong — and the second version costs more, because it is financed.

Frequently asked questions

What is a sinking fund? Money saved monthly toward a specific, known, irregular expense — a car replacement, an annual insurance bill, a roof. It is not an emergency fund, which is for things you cannot foresee.

Why not just use my emergency fund? Because then it is not there for an emergency. An emergency fund drained by predictable costs is not too small; it is doing the wrong job. Separating them is what lets the emergency fund stay full.

How much does a car sinking fund cost? An $18,000 replacement in five years is $300 a month in cash, or $272 a month if the fund earns 4% — in which case growth supplies $1,710 of the target.

Should I invest my sinking funds? Depends on the horizon. Under about two years, no: a 4% return on a twelve-month $2,000 goal supplies $36, which is not worth the risk of being short on the due date. Over five years it becomes meaningful.

How many sinking funds should I have? Three to six. Most people who start with fifteen abandon the system within months. Car, home or rental maintenance, annual bills and holidays cover the majority of irregular spending for most households.

Where do sinking funds fit in 50/30/20? They have no slot in the plain rule, which is its largest structural gap. Replacement of something you need belongs in the needs bucket; holidays and travel belong in wants. Adding them is the honest reason most households' needs bucket runs above 50%.

Do I need separate bank accounts? Not necessarily. One account with a spreadsheet works if you maintain it. What does not work is leaving the money in your current account, where it reads as available and gets spent.

I am starting late — is it worth it? Yes, and it costs more. A car fund started three years out needs $500 a month rather than $300. That is an argument for starting today, not for skipping it.

What to do next

Write down the twelve months of non-monthly expenses you have already had. That list is the budget line you are missing.

• Emergency fund calculator — sized on essential spending, and kept separate from these.
• 50/30/20 budget calculator — where the funding comes from.
• The 50/30/20 rule and where it breaks — including the gap sinking funds fill.
• Budgeting on an irregular income — when the monthly figure is not stable either.

Every figure on this site is sourced and dated. How we source every number.

Monthly contribution figures are computed by this site's own savings-goal engine, which solves the future-value-of-an-annuity equation for the payment and treats any existing balance as a lump sum growing alongside. The 4% annual return used in the comparison is illustrative only — it is not a forecast, not a rate available from any named product, and a fund invested at a variable return can be worth less than its target on the date it is needed. Take-home figures referenced are computed for tax year 2026 on a single filer taking the standard deduction. This is general education and not financial advice.