Inflation erosion of buying power

Inflation does not remove money from an account. It changes what a unit of that money buys, so the balance is unchanged and the answer to "is this enough" moves anyway. Nothing on a statement records the loss, which is why it has to be worked out separately rather than read off.

The two figures returned here are the same erosion stated from opposite ends. The real value takes an amount held today and asks what it will buy after the years have passed. The equivalent future amount takes the buying power you have today and asks what nominal sum would match it later. One shrinks, the other grows; treating them as interchangeable is the most common mistake in this arithmetic.

The effect compounds, which is why it is so much larger over long horizons than intuition suggests. A rate that seems too small to matter over one year is decisive over the length of a mortgage or a working life, because it applies to a base that has already moved every year.

The rate you supply is a single assumed average. Real inflation varies year to year and, more importantly, varies by what you actually buy — a household whose spending is dominated by housing or energy does not experience the headline figure. This answers a steady-rate question, which is a baseline rather than a description of any particular household.

Every other cost a saver meets announces itself. A fee is deducted, a charge is itemised, a tax is withheld — each leaves a trace in the account it reduces. Inflation leaves none, because it does not act on the account at all. It acts on everything the account might be spent on, and the result is a balance that is arithmetically unchanged and practically diminished. There is nothing to notice, which is precisely why it goes unnoticed.

The erosion is also relentless in a way that intuition handles badly. Each year’s price increase applies to prices that already rose the year before, so the loss accelerates quietly against a base that keeps moving. Over one year that compounding is invisible. Over the length of a mortgage, a career or a retirement, it is the largest single force acting on a long-horizon plan, and it does its work without ever generating a document.

This makes the calculation an act of deliberate accounting rather than of reporting. Nobody will send the figure; it has to be asked for. And because it must be asked for, it is systematically absent from exactly the plans that most need it — the long ones, where a target set in today’s money will be met in money that no longer means the same thing.

The calculation reports the same loss from opposite ends, and the pair is the part readers most often collapse into one. The real value takes an amount held today and asks what it will buy after the years have passed — a figure smaller than the amount. The equivalent future amount takes the buying power the amount has today and asks what nominal sum would match it later — a figure larger than the amount. Neither is a correction of the other; they answer different questions and are used at different moments.

The distinction becomes practical the instant a target is involved. A retirement figure, a deposit for a house, a sum set aside for a child: each is normally chosen in today’s money, because that is the only money anyone has intuitions about. Funding it, however, happens in future money. Setting the target with the equivalent future amount and testing progress with the real value is the discipline that keeps a plan honest at both ends.

The third figure the calculation reports is the gap on the near side: how much buying power the amount loses across the horizon, expressed in today’s money. It is the quantity to quote when the question is “how much of this is going to evaporate” rather than “what will be left” — the same fact, aimed at a different argument.

Faced with two levers — the assumed rate and the number of years — most readers spend their attention on the rate, because it is the one that feels contestable. The arithmetic disagrees. The rate enters once and the years enter as repetitions, so the result is far more sensitive to the horizon than to any plausible disagreement about the rate. Halve the rate and double the years and the erosion is not the same; the naive product is unchanged and the compounding is not.

This has an uncomfortable corollary about how projections get built. Quietly shortening a horizon — planning to a decade because two feels unimaginable — makes a plan look far more comfortable than trimming the inflation assumption ever could. The comfort is entirely an artefact of the shorter horizon, and it disappears the moment the real one is entered.

It also explains why fixed amounts are such reliable traps. A fixed pension, a fixed allowance, a fixed price agreed for a long period, a savings target set once and never revisited: each is a nominal quantity meeting a horizon, and the horizon always wins. Anything fixed for long enough is a shrinking amount, whether or not the document that created it ever says so.

The rate supplied here is a single steady assumption applied to every year alike, and that is a modelling choice rather than a fact about the world. Real inflation varies year to year, sometimes violently, and a steady-rate answer smooths a path that will not be smooth. The smoothing is defensible for planning — over long horizons the average is what dominates — but the output is a baseline, not a forecast, and nothing here tracks any published index.

The deeper limitation is whose inflation is being modelled. A headline measure is an average across a general basket, and no household buys the basket. Spending concentrated in housing, energy, healthcare or education can run persistently ahead of the headline figure for years, which means the honest use of this page is plural: run the published figure, then run one that resembles the life being planned, and treat the pair as the range the plan has to survive.

Held that way, the calculation does the one job it is genuinely good at. It converts a vague unease about the future — that money will not go as far — into two specific figures a plan can be tested against, without pretending to know what any economy will do. That is where this journey ends: with every rate corrected, every amount deflated, and a projection finally ready to be built on numbers that mean what they say.

The engine below arrives pre-filled with the pack’s own declared example: a round sum you would recognise as meaningful today, an ordinary long-run inflation assumption, and a horizon of a couple of decades. Read the two figures as a pair first — the one that shrank and the one that grew — and satisfy yourself that they describe one erosion. Then halve the rate and double the years, and watch the result refuse to stay put.

Read the real value as what the sum still buys, the equivalent future amount as what would be needed then to match it now, and the erosion as the buying power lost along the way. All three are computed live by the verified engine from an assumed rate; this page stores no figures and the calculation forecasts no index.