There is a special category of money that ruins more sleep than any other: sums with a date attached. The house deposit needed next spring, the tuition due in eighteen months, the wedding in two years. It sits there, large and idle, while a voice whispers that it should be 'working'. That voice has cost countless deposits a third of their value at exactly the wrong moment.
The governing rule is horizon, not return: money needed within about five years should not be exposed to assets that can fall by a third in a bad year. Stocks are magnificent over decades and reckless over months — the same asset, judged by different clocks.
Why five years, roughly
Historically, deep stock market declines take years to recover — sometimes a few, occasionally many. A ten-year horizon has almost always been long enough to ride out a crash; a two-year horizon simply is not. The five-year line is not physics, but it marks where the odds of forced selling at a loss become too real to ignore.
The test is brutal and clarifying: if this money fell thirty percent the week before you needed it, what happens? If the answer is 'the house purchase collapses', the money was gambled, not invested — whatever the brochure called it.
The boring toolkit
Short-term money has three worthy homes. High-interest savings accounts: instant access, deposit-protected in most jurisdictions, the default choice. Fixed-term deposits: lock the money until near its date in exchange for a better rate — the lock is a feature, not a bug, when the date is known. Short-dated government securities or money-market funds: the institutional version of the same idea, useful for larger sums where available.
Check two things everywhere: that the institution is covered by your country's deposit-protection scheme and within its limits, and the real after-tax rate. Then stop optimising. The difference between the best and second-best safe rate on a deposit is trivial next to the damage one bad equity year would do.
Inflation, the honest cost
Yes — in safe accounts, inflation may nibble the money's purchasing power. Accept it consciously: for dated money, a small, certain erosion is the insurance premium you pay against a large, uncertain loss. The goal of short-term money is not growth; it is arrival. It must show up on its date, whole.
Growth is the job of long-term money, which lives in a different bucket with different rules. Keeping the buckets separate — by horizon, in separate accounts, with separate expectations — is one of the simplest structures in personal finance, and one of the most protective. Boring, here, is the whole strategy.
Where the five-year figure comes from
The threshold is a rule of thumb rather than a calculated boundary, and it is worth understanding what generates it. Looking at historical rolling periods for broad equity markets, the proportion of periods ending below their starting point falls steadily as the period lengthens. Over one year it is substantial. Over five it is smaller. Over twenty it becomes small in most markets studied.
So five years is not a point at which risk disappears; it is roughly where the probability of a loss falls to a level many people consider acceptable for money that is not essential. For money that absolutely must be available at a specific date, even that residual probability is too high, which argues for a longer threshold or none at all.
The other consideration the simple rule omits is the consequence of being wrong. A house deposit that arrives thirty percent short does not merely delay the purchase; it can end a chain, forfeit a deposit and cost more than the shortfall. Where the consequence of a loss is severe, the horizon rule should be applied more conservatively than the probabilities alone suggest.
The instruments available and what distinguishes them
The toolkit for short-horizon money is deliberately dull and the options differ on three dimensions: access, certainty of return, and protection if the institution fails. Instant-access savings offer full access and a variable rate. Fixed-term deposits offer a known rate in exchange for locking the money away, with penalties for early access.
Short-dated government debt, held directly or through a fund of very short maturities, offers high credit quality with minimal price sensitivity, and money market funds hold a diversified pool of very short-term instruments. These typically track prevailing rates closely and are not deposit-protected in the way a bank account is, which is a meaningful distinction.
Deposit protection schemes exist in most countries and cover balances up to a limit per institution. For anyone holding a substantial short-term sum — a house deposit is the common case — checking that limit and, if necessary, splitting across institutions is a straightforward precaution that costs nothing beyond a second application form.
Matching the instrument to the actual date
The most efficient arrangement is usually not a single account but a ladder matched to when the money is needed. Cash needed within months sits in instant access. Money needed in a year can accept a twelve-month fixed term, which typically pays more. Money needed in three years can accept a longer term or a short-dated bond holding.
This captures a higher blended return than holding everything in instant access, without taking any meaningful risk, and it requires setting up once. The constraint is that the dates need to be genuine, since early access to a fixed term carries a penalty that can erase the advantage.
The common failure is the opposite arrangement: holding everything in instant access at a poor rate because the money might be needed at any time, when in reality only a portion might. Being honest about which portion is genuinely uncertain and which has a known date is worth a surprising amount over several years on a large balance.
The inflation cost, stated honestly
Holding cash over several years carries a real cost that deserves to be stated plainly rather than glossed over. If prices rise faster than the interest earned, the money buys less at the end than it would have at the start, and over a five-year period that erosion can be substantial.
This is a genuine loss and it is not a reason to take market risk with money that has a date attached. The correct framing is that you are paying a known, bounded cost to eliminate an unbounded one. The alternative — accepting a possible thirty percent shortfall to avoid a smaller certain erosion — is a poor trade when the money has a specific purpose.
What is worth doing is minimising the cost rather than accepting it passively. Ensuring the rate is competitive, using the ladder described above, and using any tax-sheltered cash allowance available all reduce the gap between the rate earned and the rate of price increases. In some periods that gap closes entirely; in others it does not, and the cost is simply the price of certainty.
The specific case of a house deposit
House deposits deserve separate treatment because they combine every difficulty: a large sum, an uncertain date, a severe consequence if short, and a target that may itself be moving as property prices change. This last point is the one that generates the most tempting bad argument.
The argument runs that since property prices may rise while you save, keeping the deposit in cash means falling behind, and therefore it should be invested to keep pace. The flaw is that property prices and equity markets are not the same thing and do not move together reliably, so the hedge frequently does not hedge, and a decline arriving in the month you found a house is catastrophic in a way that slow erosion is not.
The more effective responses are unglamorous: increase the saving rate, use any government scheme available for first-time buyers, and be realistic about the timeline. Some countries offer accounts with bonuses or tax advantages specifically for this purpose, and these frequently provide a better risk-adjusted improvement than any investment approach would.
The awkward middle horizon
Money needed in something like five to ten years falls into a genuinely difficult zone, too long for cash to be obviously right and too short for equities to be comfortable. There is no clean answer here and it is more honest to say so than to invent one.
The reasonable approaches involve either a blend, holding some in each and accepting a middle outcome, or a glide path that starts with more market exposure and shifts progressively toward cash as the date approaches. The second is what target-date retirement funds do, and the logic applies equally to any dated goal.
What makes the middle horizon manageable is flexibility about the date. If the goal can be postponed by a couple of years without serious cost, more market exposure is defensible, because a decline can be waited out. If the date is immovable, the horizon is effectively shorter than the calendar suggests and should be treated accordingly. Knowing which situation you are in is the most useful input, and it is one only you can supply. None of this is financial advice.
Reviewing the rate you are actually getting
Savings providers rely on inertia, and the business model is explicit enough to be visible in how products are structured. An attractive introductory rate that reverts after twelve months, a headline rate available only on a new account while existing customers sit on a lower one, a bonus conditional on making no withdrawals.
The consequence is that a rate which was competitive when the account was opened is frequently uncompetitive two years later, without anything having been announced. Nobody writes to tell you that a better version of the same product now exists at the same institution, and in many cases it does.
An annual check, on a fixed date, comparing your rate against what is currently available, takes fifteen minutes and on a substantial short-term balance recovers a meaningful sum every year. Setting a calendar reminder for whenever any introductory period expires is the other half of it, since that is the specific date the rate falls and the one nobody remembers.
Keeping short-term money separate from everything else
A practical failure that undermines all of the above is holding money for a specific dated purpose in the same place as general savings. The balance becomes a single figure, the purpose becomes abstract, and the money is gradually used for other things without any decision having been taken.
Naming the account after its purpose sounds trivial and is measurably effective. Research on mental accounting suggests that money assigned to a labelled purpose is spent differently from identical money in an unlabelled pot, and while that inconsistency is irrational in one sense, it can be used deliberately in your own favour.
The same logic argues for a separate account per goal rather than one large balance covering several. It makes progress toward each visible, it makes any raid on one obvious rather than invisible, and it removes the question of whether the total is sufficient for all of them at once. The cost is a few extra accounts, which most banking apps now make trivial to open.