How to Choose an Inflation Assumption for a Financial Goal

Meera is asked for a number she does not have

Meera is trying to work out what her daughter's degree will cost when the time comes. The planner in front of her wants one figure: the rate at which she expects the fees to rise. She has no idea. She picks something that sounds sensible, the calculator produces a total, and that total then governs every contribution she makes for the next several years.

The uncomfortable part is that the number she guessed matters more than almost anything else on the page. Over a long horizon it compounds, so a modest change in the assumption moves the target by a great deal — and unlike the return assumption, which at least has a market attached to it, this one feels like it was invented. In a sense it was.

This page is about making that guess honestly: what the evidence can tell her, what it cannot, and what to do with the part that remains genuinely unknown.

What Indian inflation has actually done

Start with the one thing that is not a guess. We hold the official combined consumer price index — the national basket, urban and rural together — from January 2013 to July 2026. Across that whole stretch prices rose at 5.1% a year.

If the story ended there, Meera would have her number and this page would be short. It does not, because that single figure conceals almost everything worth knowing. Look instead at the 151 months in the record for which we can measure the change over the preceding twelve, and the picture is far less settled.

Twelve-month inflation was	Rate
The lowest month in the record	0.0%
Lower than four months in five	3.4%
The middle of the record	4.9%
Higher than four months in five	6.3%
The highest month in the record	8.5%

The middle of that record sits at 4.9%, close enough to the long-run average. But the range runs from 0.0% in October 2025 — a year in which prices, on this measure, barely moved at all — up to 8.5% in January 2014. Within a single twelve-year stretch, the rate a planner would have called normal was at one point almost nothing and at another high enough to double a cost in under a decade. 24.5% of those months sat at six per cent or above, and 30.5% of them sat below four.

That is the finding, and it is not the one most planning tools imply. The average is real, and it is also a poor description of any particular year Meera will actually live through.

Why the headline rate is not her rate

There is a second problem, and for Meera it is the larger of the two.

The index above measures a national basket — food, fuel, housing, clothing, transport, everything, weighted by what a representative household spends. Her goal is not a representative household. It is one specific course at one kind of institution, and the price of that has its own path, driven by staff costs, capacity and what parents are willing to pay. Education, healthcare, urban rent and anything imported have all at various times moved differently from the headline, in both directions.

We hold no separate index for school fees or medical treatment, and this article will not invent one. What can be said honestly is the direction of the warning: a goal concentrated in a single category should not assume the national average describes it, because the national average is an average precisely of things that move differently. If Meera has the actual fee schedule for the last several years from the institutions she is considering, that is better evidence than any published index, and it is evidence she can obtain by asking.

Where no such record exists, the honest position is that the central estimate comes from the national series and the uncertainty around it is wider than the national series suggests.

Higher is not automatically safer

The instinct on being told all this is to be conservative — to assume a high rate and be pleasantly surprised. That instinct is half right and worth being careful with.

A higher assumption produces a larger target, which produces a larger required contribution. If that contribution is affordable, the caution costs nothing and buys a margin. If it is not, one of two things happens: the plan gets abandoned as unrealistic, or the number gets quietly ignored while the contributions stay where they were. Neither leaves Meera better off than a central estimate she actually funds.

An assumption is only prudent if the plan built on it survives contact with the household's cash flow. A rate chosen for safety and then disregarded is worse than a central one taken seriously, because it also destroys the reader's confidence in the rest of the plan.

The useful form of caution is not a higher single rate. It is running the goal at more than one rate and knowing in advance which lever moves if the high case turns out to be the real one.

Doing it as a range

The practical method takes about ten minutes and replaces the guess with something Meera can defend.

Calculate the future cost three times: at a low rate, a central rate and a high one. For a goal tracking general prices, the spread in the table above is a reasonable place to draw those from — the middle for the central case, and something near each end for the others. For a goal in a category known to run hotter, shift all three up rather than only the top one, because the whole distribution has moved, not just its tail.

Then look at what the three answers do to the contribution required. This is the step that matters, and it usually produces one of two reactions. Either the three contributions are close enough that the choice of rate barely matters, in which case Meera can stop worrying about it and get on with funding the goal. Or they are far apart, in which case she has learned something specific: this goal is sensitive to inflation, and it needs reviewing more often than the others.

Finally, decide now what gives if the high case arrives. Education has real levers — a different institution, a longer degree paid across more years, a contribution from the child, a loan for the final portion. Retirement has fewer. Writing that down while the question is still hypothetical is the entire benefit, because the alternative is discovering it at the point of payment.

Keep the two rates in the same units

One error is common enough to be worth its own warning, and it is easy to make in a spreadsheet.

If the goal has been inflated to a future rupee amount, then the return assumption applied to the contributions must also be a nominal one — the number an account statement would show. If instead Meera is working in today's purchasing power, the return must be a real rate, net of inflation. Mixing them is silent: the spreadsheet produces a plausible answer, and it is wrong in whichever direction the mismatch runs.

The check is simple. Ask whether the answer on the screen is in rupees she could spend today or rupees she will spend in a decade, and confirm that every rate on the page is in the same world.

Reviewing rather than forecasting

The last thing to say about the assumption is that it does not have to be right, because it does not have to stand for the whole period.

Meera's estimate is a planning input for the next year or two, at which point she will have something the estimate never had: actual prices. Fees will have been published. The gap between what she assumed and what happened is information, and revising the target on that basis is not an admission of error — it is the plan working as intended. A goal reviewed annually against real prices converges on the truth regardless of where it started. A goal set once from a confident guess and never revisited does not.

So the review matters more than the rate. Set the rate carefully, run the range, fund the central case, and then check the actual cost every year and move the target when it moves.

What to take away

Prices in India rose at 5.1% a year across the record we hold, and that average conceals a range wide enough that planning to it alone would be misleading. Meera's own goal is in a category the national index does not describe, and we have no index that does, so her best evidence is the fee history she can ask for.

Use a central assumption you can actually fund, test the goal at a low and a high rate to find out whether it is sensitive, keep nominal and real quantities apart, and decide in advance which part of the goal changes if the high case arrives. Then review it against real prices once a year, which is the only step in this whole process that replaces a guess with a fact.