What is a sinking fund — and how do you set one up?

Most budget failures are not caused by overspending on groceries. They are caused by a $1,100 car repair, an annual insurance renewal, or December — expenses that arrive irregularly, feel like emergencies, and get paid from the emergency fund or the credit card because no single month's budget had room for them. The awkward truth is that almost none of these are surprises. You do not know which month the car will need brakes, but you know it will need them; you know roughly what the holidays cost; you know the renewal date of your insurance to the day.

A sinking fund is the mechanism that absorbs this class of expense. You name the expense, estimate its cost, divide by the number of months until it is due, and move that amount into savings every month. When the bill arrives, the money is already waiting, and an expense that used to blow up a month becomes a line item like rent. The term is borrowed from corporate finance, where a sinking fund is money a company sets aside on a schedule to retire a bond; the household version retires a future bill instead.

Known-irregular is not the same as unknown

The useful boundary is not between regular and irregular expenses but between known and unknown ones. Rent is known and regular. A brake job is irregular but known — not the date, but the category and, within a range, the cost: every car eventually needs work, and averaged over a few years the amount is fairly stable. Genuinely unknown expenses are rarer than they feel: a job loss, an injury, a roof failing a decade early. How big an emergency fund should be is its own question, but whatever the answer, it only holds if the fund is reserved for that last category.

The test worth applying to any expense that just ambushed you: could I have named this category a year ago? If yes, it belongs in a sinking fund, and the ambush was a planning gap rather than bad luck. Run last year's transactions through that filter and most "emergencies" reclassify themselves. This matters because an emergency fund that doubles as an everything fund never reaches its target — it drains for brakes in March and a wedding in June, and when a real crisis finally arrives it finds the account half-full.

There is also a difference in how spending the money feels, and it is not cosmetic. Drawing down an emergency fund registers as failure even when it shouldn't; spending a sinking fund is the plan completing. People defend sealed accounts and quietly raid open-ended ones. A sinking-fund setup gives every predictable bill somewhere legitimate to land, which is what actually keeps the emergency fund sealed.

The standard categories

The list is personal, but the same categories recur because the same expenses ambush everyone:

• Vehicle repairs and maintenance. The canonical case — near-certain occurrence, unknowable timing. Owners of older cars should budget more per month, not hope harder.
• Insurance paid annually. Home, auto, tenant, life. Insurers usually charge a premium for monthly instalments; a sinking fund lets you pay the cheaper annual bill with money that still left your budget monthly. The fund earns you the discount.
• Gifts and holidays. December is the most predictable expense of the year and still the most common budget wreck. Add birthdays and weddings.
• Travel. Any trip you intend to take is a known expense with a chosen date — the easiest fund to calculate and the easiest to postpone when money is tight.
• Technology replacement. Phones and laptops fail on a roughly known cycle. Funding the replacement over the device's life means the failure costs you nothing but the errand.
• Home and pets. Owners carry a steady drip of maintenance; a common rule of thumb is one percent of the home's value a year. Pets age on a schedule vet bills follow.

Don't adopt the list wholesale. Pull the last twelve months of transactions and find the withdrawals that hurt — your categories are in there, priced. A budget built without them isn't wrong about groceries; it is silent about a meaningful slice of the year's real spending, which is why it keeps failing in months that contain nothing unusual.

The arithmetic

Every sinking fund is one division: the amount you will need, divided by the months until you need it.

For dated expenses, the deadline sets the pace. A $1,440 premium renewing in six months needs $240 a month until renewal — and then, with a full year of runway ahead, drops to $120 a month permanently. The first cycle is always the expensive one, because you are catching up and pre-funding at once. If the catch-up rate doesn't fit, you have two honest options: seed the fund from cash you already hold, or accept a leaner first version of the expense. What you cannot do is skip the months and expect the deadline to move.

For ongoing wear — car repairs, home maintenance, vet bills — there is no due date, so the divisor is simply twelve: estimate the annual cost and contribute a twelfth of it forever. Estimate from your own history where you have it, and round up where you don't. Precision matters less than existence; a car fund holding $800 against an $1,100 repair converts a crisis into a $300 problem.

A full setup, worked

Here is a complete first setup for one household — six funds, mixed types, one monthly total:

The monthly total looks like new spending. It is not — these categories were already costing this household roughly $5,700 a year; the transfers just move the money earlier in time so it leaves in even instalments instead of shocks. What changes is the shape of the year, not the size of it: the surplus you thought you had in quiet months was never real, and the deficit in expensive months never had to happen.

Run your numbersSinking fund→

One account or many

Both structures work; they fail differently.

Separate accounts — one per fund — make the balance the answer. No ledger, no arithmetic: the travel account holds what the travel fund holds. Many banks now offer unlimited sub-accounts or "vaults" for exactly this. The cost is sprawl, and at banks that still price per account, fees that eat small balances.

One account with a ledger concentrates the money and the interest, and adds nothing to your banking. The cost is bookkeeping: a note, a spreadsheet, or a tracked category that says how the single balance divides. Skip the ledger and the balance becomes fog — and fog gets spent, usually by the most optimistic member of the household.

The deciding question is maintenance, not elegance. If you know you will not update a ledger, use sub-accounts; if account sprawl offends you, keep one account and treat the ledger as part of the monthly transfer ritual. Either way, keep the money out of chequing — the small friction of a transfer is a feature, not a flaw — and keep it in a high-interest savings account rather than anything invested. These balances are measured in months, not decades. This is money with an appointment, not money seeking a return.

Keeping it running

Two things go wrong in practice, and both have boring fixes. Underestimates: the repair invoice exceeds the fund. Cover the gap from flexible spending, not the emergency fund, and raise next year's estimate — the fund was still doing its job, just at the wrong size. Overshoot: a fund finishes early or an expense comes in light. Skim the surplus into the next-thinnest fund and move on. Once a year, re-price the recurring funds against what the categories actually cost, retire the finished ones — the trip happened — and add at most one new category, chosen because it surprised you, not because a list suggested it.

If your income is irregular, the even-monthly rhythm is the part to change, not the idea. Fund the categories aggressively in strong months so the weak ones inherit paid-for bills — the mechanics are covered in budgeting on an irregular income, and sinking funds are arguably worth more there than anywhere, since an irregular income has the least capacity to absorb an irregular expense.

One last accounting habit: keep sinking funds out of your savings numbers. These balances are committed spending parked on a schedule, not wealth — counting them inflates what you save each month right up until the month the money does what it was always going to do. Their job is quieter than growth, and just as valuable: they protect the money that stays.