How to Estimate the Future Cost of a Financial Goal

Rohit is planning against the wrong number

Rohit and his wife have decided to buy a flat in about four years. They have looked at what the flats they like cost, they have worked out what deposit that implies, and they have set up a monthly amount to get there.

The plan has a hole in it, and it is not a small one. The figure they are saving towards is the price of a flat today. In four years the flats they like will not cost what they cost today, and if prices rise at all over that period, Rohit will arrive with a deposit sized for a purchase that is no longer available. He will have saved diligently for four years and still be short.

This is the most ordinary planning error there is, and it comes from treating a goal as an amount rather than as a thing that has a price on a future date.

The calculation itself

The mechanics are simple, and the simplicity is exactly why the difficulty lies elsewhere.

Take what the goal costs today. Multiply it by one plus the annual rate at which that cost is expected to rise, raised to the number of years until it happens. That is the future cost. The one thing to be careful of is that the rate and the period agree — an annual rate with a number of years, a monthly rate with a number of months — and mixing them produces an answer that is wrong by a factor large enough to be obvious only sometimes.

Two things about the input deserve more care than the formula.

The first is what "costs today" means. The price to use is the price of the specific thing, not a category average. The flats he is actually considering, in the areas he is actually considering, at the size his family will actually need — which may not be the size he needs now. And what comes with it belongs in the figure: registration, stamp duty, the interiors that turn a handover into a home, whatever the building charges before anyone moves in. Those are part of the goal even though they are not part of the quoted price, and leaving them out is a reliable way to be short at the end.

The second is the rate, which is the whole subject of choosing an inflation assumption. The short form is that the national headline figure describes a basket, and property in one city is not that basket.

Goals that arrive in pieces

Rohit's flat is a single payment, which makes it the easy case. A great many goals are not.

An education is several years of fees and several years of living costs, each falling due on its own date. A retirement is decades of monthly spending. A house being built is a schedule of stage payments. In every one of these, compounding the whole amount to a single endpoint is wrong, and wrong in a particular direction: it inflates money that is needed early as though it were needed late, and it overstates the total.

The correct treatment is to inflate each payment to its own date and keep them as a schedule. This is more work and it is worth it twice over — the total is right, and the schedule is the thing that tells Rohit which money needs to be safe soon and which can still be growing. Collapsing it to one number destroys exactly the information the portfolio needs. Planning a child's education in stages works through the case where this matters most.

Why a single number is the wrong output

The strongest reason to be careful here is that the estimate feels far more solid than it is.

A calculator produces a figure to the rupee. It looks like a quotation. It is, in fact, today's price multiplied by a guess, and the guess is compounded — which means the further away the goal, the more of the answer is guess and the less is price. Over four years the effect is modest; over eighteen it dominates.

The way to keep this honest is to produce a range rather than a point. Run the calculation at a low rate, a central one and a high one, and look at what happens to the contribution each implies. If the three contributions are close together, the assumption barely matters and Rohit can stop thinking about it. If they are far apart, he has learned that this goal is unusually sensitive to a number nobody knows, which is a specific and actionable finding: that goal needs reviewing more often, and it needs a lever identified in advance.

Do not let return absorb the uncertainty

There is one manoeuvre to name and refuse, because it is common and it looks reasonable.

When the inflated goal turns out to require an uncomfortable contribution, the temptation is to raise the assumed investment return until the contribution comes back down. The spreadsheet rebalances, the plan looks feasible again, and nothing whatever has changed about what Rohit will actually have in four years.

Cost inflation and investment return are separate quantities and must be estimated separately. If the answer is unaffordable, the levers are the amount, the date, the contribution, or another source of funding. Adjusting the return is not a lever; it is a way of not answering.

Related, and easier to do by accident: keep the units consistent. If the goal has been inflated to a future rupee amount, the return applied to contributions must be a nominal one. If Rohit prefers to work in today's money throughout, the return must be net of inflation. Both approaches are correct and the mixture is silently wrong.

Revisit it, because you will be wrong

The last point is the one that rescues the whole exercise from its own uncertainty.

Rohit does not need his estimate to be right for four years. He needs it to be right enough for about one, at which point he will have something better than an assumption: the actual asking prices of the actual flats. That is a fact, and it replaces the guess.

An estimate reviewed annually against real prices converges on the truth no matter how poor the starting assumption was. An estimate made once, with great confidence, and never revisited does not — and the error compounds silently in the meantime. The review schedule is worth more than the precision of the original number, which is a slightly deflating thing to conclude after a page about how to calculate it, and it is the most useful sentence here.

What to take away

Plan against what the goal will cost on the day it happens, not what it costs now. Use the price of the specific thing including the costs that travel with it, inflate each payment to its own date if there is more than one, and produce a range rather than a single figure.

Keep the cost rate and the return rate separate, and keep nominal and real quantities apart. Then check the estimate against real prices every year and move it when it moves. The first estimate is a planning input; the annual correction is what makes it true.