What to put aside each month to get there
How much to put aside each month to reach a target by a date, given what you have already and the return you expect. The same annuity formula as compound interest, solved for the one number that is a decision rather than a forecast.
Year by year
| 1 | $4,398.97 | $103.75 | $4,502.72 |
| 2 | $7,797.93 | $350.21 | $8,148.15 |
| 3 | $11,196.90 | $745.19 | $11,942.09 |
| 4 | $14,595.87 | $1,294.74 | $15,890.61 |
| 5 | $17,994.84 | $2,005.16 | $20,000.00 |
An estimate, not financial advice. Figures are illustrative and depend on assumptions listed below. Check anything you plan to act on with a qualified adviser or the provider itself.
The worked example below, drawn. The gap between the lines is what the return contributes rather than you.
How it works
What this works out
The monthly amount that gets you to a target by a date. Compound interest asks what a plan becomes; this asks what the plan has to be, which is the question people actually arrive with.
The table underneath is the year-by-year balance, so you can see when the return starts contributing and when it is still just your own money.
It is one formula, rearranged
The future value of a starting sum plus a level monthly payment is standard:
FV = P x (1+i)^n + M x ((1+i)^n - 1) / i
Everything in it is known except M. Rearranging for the payment:
M = (FV - P x (1+i)^n) x i / ((1+i)^n - 1)
The first term is the gap the contributions have to close: the target, less whatever the starting sum grows into on its own. The second is the annuity factor that turns a lump sum into a level payment.
The schedule is then run forward from that answer with the ordinary loop, not derived from it. That is deliberate — if the rearranged algebra were wrong, the table would end somewhere other than the target and say so. A test asserts exactly that across five rates, five horizons and three starting sums, which makes the table a check on the formula rather than a decoration on it.
The zero-rate branch is not an edge case
At a rate of zero the annuity factor divides by zero. That is not an obscure input — it is money in a current account, or a deliberately pessimistic plan.
The factor ((1+i)^n - 1) / i has the limit n there, so the branch is:
M = (FV - P) / n
The shortfall spread evenly over the months, which is obviously right and is why the branch is written rather than guarded against.
Time beats rate, until it does not
The two levers behave completely differently, and over a short horizon the result is counterintuitive:
| Years | Monthly needed |
|---|---|
| 3 | 494.29 |
| 5 | 283.25 |
| 10 | 125.70 |
Before you read on
You need 20,000 in ten years and you are saving 125.70 a month to get there. If you halve the deadline to five years, what happens to the monthly amount?
It more than doubles, to 283.25. You lose twice over: there are half as many payments, and each one compounds for half as long. The two effects multiply rather than add, which is why the monthly figure always rises faster than the time falls.
Halving the time from ten years to five more than doubles the monthly figure, from 125.70 to 283.25. There are half as many payments and each compounds for half as long, and the two effects multiply.
Over five years the return contributes about 2,005 of the 20,000 — a tenth. Over thirty years at the same rate it would be most of it. Short horizons are about contributions; long ones are about the rate, and the tool’s own table shows which regime you are in.
A worked example
20,000 in five years, starting with 1,000, at a 4% return:
| Save each month | 283.25 |
| Paid in altogether | 17,994.84 |
| Interest does | 2,005.16 |
| Final balance | 20,000.00 |
The second and third figures add to the fourth exactly. “Paid in altogether” includes the 1,000 you started with, which is the same thing the schedule’s “paid in” column means — the two used to differ by exactly that starting sum, which is the kind of discrepancy nobody notices until they add the numbers up.
At a zero return. The same target needs 316.67 a month, because 19,000 spread over 60 months is all there is.
These are the same figures asserted in this tool’s test file, so the page and the formula cannot drift apart without the build going red.
What it does not do
It assumes a constant return, which nothing has. It does not model tax, platform charges or fund fees, each of which comes straight off the rate. It does not adjust the target for inflation — 20,000 in five years will buy less than 20,000 buys now, and that is a separate question this catalogue answers separately. It does not handle variable or one-off contributions, or withdrawals along the way.
For the forward direction — what a given monthly amount becomes — compound interest is the tool.
The formula
The future value of a starting sum plus a level
monthly payment, over n months at monthly rate i:
FV = P x (1+i)^n + M x ((1+i)^n - 1) / i
Solved for the payment:
M = (FV - P x (1+i)^n) x i / ((1+i)^n - 1)
At a rate of zero the annuity factor has the
limit n, so that branch is simply:
M = (FV - P) / n
i = annual rate / 100 / 12
n = years x 12
Contributions land at the end of each month:
an ordinary annuity.
What it assumes
- Contributions are paid at the end of each month, which is what a standing order set up after payday actually does. Paying at the start earns one extra month of growth on every contribution and lowers the required amount slightly.
- The return is constant and compounds monthly. Real returns are not constant, and a fixed deadline is exactly where that matters — a fall in the final year cannot be waited out.
- The rate is nominal and before inflation, tax and charges. A target set in today's money will buy less than you think by the time you reach it, which is the inflation tool's question.
- If the starting sum alone passes the target, the required contribution is clamped to zero rather than reported as a negative. The algebra's negative answer is the amount you could withdraw, which is a different question.
- The monthly figure is rounded to money before it is paid, as a standing order would be, so the final balance lands within a cent of the target rather than exactly on it.
- The figures carry no currency symbol on purpose. The arithmetic is the same in pounds, euros or dollars, and the tool follows whichever currency your locale uses.
Common questions
How much do I need to save each month?
For 20,000 in five years, starting from 1,000 already saved and assuming a 4% return, it is 283.25 a month. Of the 20,000, about 2,005 comes from the return and the rest is yours. That ratio is the useful part — over five years the contributions do nearly all the work and the rate barely matters, which is the opposite of a thirty-year picture.
Why does halving the time more than double the amount?
Because you lose twice over. There are half as many payments, and each one compounds for half as long. The two effects multiply rather than add, so the monthly figure always rises faster than the time falls. Adding time is the most effective lever available and it is usually the least painful one.
Does what I already have really help that much?
More than its size suggests, because it compounds for the entire period while each new contribution compounds only for what is left. In the worked example, 1,000 already saved removes about 18 a month from the requirement — more than the 16.67 a month it would represent if it were just divided across the term.
What return should I assume?
For anything within about five years, a cash savings rate, because a fall at the wrong moment cannot be waited out on a fixed deadline. Longer horizons can justify more, but the figure should be cautious: this tool computes exactly what you ask of it, and an optimistic rate produces a comfortable monthly amount and a shortfall at the end.
What if I cannot afford the answer?
Three levers, and they are not equal. Adding time helps most, because it adds payments and growth together. Lowering the target helps roughly proportionally. Raising the assumed return helps least over short periods, and it is the only one of the three that is a guess rather than a decision you control.
Sources
Method written and checked by Tessalor on Jul 31, 2026.