Nominal Return vs Real Purchasing Power

Two questions that look like one

“Did the investment go up?” and “am I better off?” feel like the same question. They are not, and the gap between them is the whole subject of this article.

The first question is about a nominal amount: the number the account reports, in the currency of the day it reports it. The second is about purchasing power: what that number would actually buy. A balance that rises from 100,000 to 150,000 has unambiguously gone up as a number. Whether the holder is better off depends on what 150,000 buys at the end compared with what 100,000 bought at the start, and nothing about the first fact settles the second.

The tool this article accompanies exists because the two get reported as one. Most calculators answer the first question, print a large number, and stop.

Nominal money and purchasing power are different quantities

It helps to stop thinking of inflation as a thing that happens to money and start thinking of it as a change of unit.

A currency amount is a measurement, and like any measurement it is meaningless without a unit. “50,000” is not an answer until somebody says 50,000 of what — of 2026 money, of 2046 money, of money at some other date. Those are different units. The exchange rate between them is the price level, and inflation is the rate at which that exchange rate moves.

Once inflation is a unit conversion, three things follow immediately.

A translation is not a deduction. Inflation removes nothing from an account. The balance is exactly what it was; the label on the axis changed. This is genuinely different from a fee, which removes money.

Every real figure needs a stated date. “In today’s money” means nothing until “today” is fixed. The Real Return calculator fixes it at the start of the scenario you entered, so every real figure on that page is in the purchasing power of the day the projection begins. It is not the date you happen to be reading, and the tool reads no clock to find out.

Two amounts in different units cannot be added or subtracted. Not because it is bad practice, but because the result is not a quantity. Most of the specific errors in the rest of this article are instances of this one.

Why subtracting inflation from the return is only an approximation

The usual shortcut is to take the return and subtract the inflation rate. Earn 7%, inflation is 3%, call it 4% real.

The shortcut is close at small rates and it is wrong in a knowable direction: it always overstates the real return. Here is why.

Start with 100. A 7% nominal return leaves 107. Meanwhile prices rose 3%, so what cost 100 now costs 103. The question is how much of the new, more expensive basket the 107 buys:

107 / 103 = 1.03883...

That is 3.883% more basket, not 4%. The 4% answer treats the extra return as though it could be bought at last year’s prices. It cannot: the gain arrives at the end of the year, when prices are already higher, so the gain itself has to be deflated too.

The gap is 0.117 percentage points here, which sounds like nothing. Over thirty years compounding, the difference between 4% and 3.883% is about 3.5% of the ending value. It widens as either rate rises: at 12% nominal and 8% inflation the shortcut says 4% and the exact answer is 3.70%.

The exact identity

1 + real return = (1 + nominal return) / (1 + inflation)

real return = (1 + g) / (1 + i) - 1

Read it as a ratio of two multipliers, because that is what it is. 1 + g is how much money you have; 1 + i is how much of the old basket that money now costs. Divide the first by the second and you have how many baskets you ended up with.

The subtraction shortcut is what you get if you expand the division and throw away one term. The discarded term is g × i, which is the interaction between the two, and its size is exactly why the approximation degrades when either rate is large.

Percentage fees slot into the same expression by multiplication, because a fee is also a multiplier on the balance:

real return after percentage fees = ((1 + g) * (1 - F)) / (1 + i) - 1

Both of those are rate identities, and neither is a return on a portfolio that received money over time. The moment contributions arrive on different dates there is no single annual rate that describes the whole result, which is why the calculator reports amounts as its main answer and shows these rates as supporting figures rather than as the headline.

Fees and inflation are different layers

They are frequently reported as one number, usually called something like “total drag”. They should not be.

A fee is a withdrawal. Money leaves the account and is somebody’s revenue. The balance is smaller afterwards. The money is also no longer invested, so it stops earning, which is why the shortfall at the end is normally larger than the sum of the deductions.

Inflation is a revaluation. No money moves. The balance is identical. What changed is the price of everything the balance could be exchanged for.

Two practical consequences follow. First, the order in which you apply them does not matter, because multiplication commutes — but only if you keep them as multipliers rather than subtracting rates. Second, and much more important, the two effects live on different price bases and their money amounts must never be added. Fees deducted are nominal amounts taken on the dates they were taken. An “inflation loss” would be a difference between two figures in start-date money. Adding them produces a total that is not denominated in anything.

The calculator therefore publishes two separate decompositions and never one. The first is entirely in future currency: what the fee-free reference would have reached, what was deducted, what the compounding of those deductions cost, and what remains. The second is entirely in start-date currency: the after-fee balance divided by the price-level factor, and the comparison against what was paid in. No line of either is added to a line of the other.

Why nominal contributions cannot be compared with a real ending value

This is the error the tool was built to prevent, and it is the most seductive one because both figures are honest on their own.

Suppose 50,000 at the start and 500 a month for twenty years. The nominal total paid in is 170,000. Under the calculator’s published example the after-fee balance reaches about 342,890 in future money, which is about 209,256 in start-date money.

Now do the tempting subtraction: 209,256 minus 170,000 is 39,256. That number is worthless. The 209,256 is measured in start-date money. The 170,000 is a sum of 240 separate amounts paid at 240 different price levels, most of them far weaker than start-date money. Subtracting the second from the first is subtracting kilometres from hours.

The correct comparison values every contribution at its own date and then adds them, which puts the whole basis in one unit:

real value of one contribution = C / (1 + i) ^ (t / 12)
contribution basis             = S + sum over every contribution

A 500 payment made in month 240 at 2.5% inflation was worth about 305 in start-date terms. One made in month 12 was worth about 489. The starting 50,000 was made at time zero, so it counts in full. On the published example the basis comes to about 144,602, not 170,000, and the real gain is about 64,654.

Note that the wrong answer here was smaller than the right one. The direction is not the point and does not generalise: with a shorter horizon, a larger starting amount or lower inflation, the wrong subtraction can flatter the result instead. The number is not too high or too low. It is not a number.

What contribution timing actually changes

Timing changes two things at once, and the second one is easy to miss.

The first is ordinary: a contribution added at the beginning of a period is invested for one period longer than the same contribution added at the end, so a beginning-of-period schedule produces a larger nominal balance whenever the return is positive.

The second is that timing moves the purchasing-power date of every payment. An earlier contribution is not just invested longer; it was made when money was worth more, so it enters the contribution basis at a higher real value. Both the ending value and the thing it is measured against move, and they move in the same direction.

That is why the calculator asks for a timing convention instead of quietly picking one, and why the answer is stated near the result rather than buried in a methodology note.

Fixed contributions against inflation-indexed ones

The model contributes a fixed nominal amount. A 500 payment in year one and a 500 payment in year twenty are the same number and not the same commitment: by year twenty, 500 is a smaller share of what an average price level buys.

Many real plans are not like that. Contributions tied to a salary, or explicitly escalated each year, rise in nominal terms and hold their real value better. Such a plan pays in more real money, so it should reach a higher real ending value — and its contribution basis is higher too. Both sides of the comparison move again.

The tool does not model indexed contributions in version 1. That is a stated boundary rather than a claim that fixed contributions are normal or advisable. If your plan escalates, the projection understates both the ending value and the basis, and the real gain it reports is not the real gain you should expect.

One inflation number is still an assumption

Everything above is arithmetic. The inflation figure fed into it is not.

The calculator takes one constant annual rate for the whole horizon. Real price levels do not move at a constant rate, and the sequence matters for anyone paying money in over time: high inflation early erodes the contributions that had the longest run, while the same average arriving late erodes a larger balance for a shorter time. One constant rate cannot express either case.

UBWHY publishes no inflation figure, prefills none, offers no low, base or high band, and fetches nothing from any published index. This is the one tool on the site whose central input has an obvious public data source, and taking it would do two unacceptable things at once: introduce a network request into a calculator that promises to send nothing, and turn a reader’s own assumption into an implied UBWHY forecast. The number is yours or it does not exist.

That is also why the tool offers a second inflation field. It is not a scenario generator. It runs the same nominal path against a second assumption you choose, so you can see how much of the answer depends on the number rather than on the investment.

A general price index is not your inflation

A general price level is an average across a defined basket. Yours is an average across what you actually buy, and the two coincide only by accident.

Nothing exotic is needed to make them diverge. A household that owns its home outright is barely touched by rents. One that spends heavily on energy in a year when energy moves sharply is affected far more than an average suggests. Somebody whose spending is mostly one category is exposed to that category, not to a weighted mean of everything.

The consequence is narrow and worth stating precisely. When the calculator says a projection preserves purchasing power, it means purchasing power as measured by the rate you entered. It cannot mean anything else, because it has no information about what you buy. If your own costs rise faster than the figure you entered, the real gain shown is optimistic for you specifically, and no amount of precision in the arithmetic corrects that.

What a positive real projection does not prove

The strongest thing the calculator can say is that under the assumptions entered, ending purchasing power exceeds the inflation-adjusted value of the money contributed. That is an arithmetic result about one path. Here is what it does not establish.

That the assumptions will occur. The return and the inflation rate were typed in. Their being consistent with each other does not make either likely.

That the risk was worth taking. The model has no volatility, no sequence and no probability. Two paths ending at the same real value after taking wildly different risks are indistinguishable to it.

That the money was available. Nothing here models liquidity, lock-ups, exit penalties or the possibility of needing the money on a bad day.

Anything at all about tax. There is no tax input, no tax drag and no after-tax figure. Real tax outcomes depend on jurisdiction, account type, realised against unrealised gains, income type, allowances, timing and personal circumstances. Representing that as one more annual percentage would be a fabrication wearing a decimal point.

That the investment is suitable. Suitability is about a person, not about a projection.

A positive real projection improves the arithmetic case. It does not establish suitability, adequate risk compensation, liquidity, tax treatment, or that the assumptions will occur.

Where the distinction actually bites

Long-dated savings and pension projections. These are the classic case: a forty-year horizon and a large nominal ending figure. At 2.5% inflation the price level roughly doubles over thirty years, so a headline that is not labelled as nominal or real is not interpretable at all.

Anything quoted as a yield. A nominal yield above zero is not a real yield above zero, and a yield paid in an asset whose price is falling can be nominally positive and really negative at the same time. The yield and the change in the asset’s own value are separate facts, and only one of them is usually advertised.

Products with a fee stack. Fees and inflation are two multipliers on the same balance, and the combined effect is worse than either read alone. The companion Investment Fee Drag calculator isolates the fee layer; this one puts the price level on top of it.

Anything with recurring operating costs paid in one currency against rewards in another. Mining and staking arrangements are the obvious examples. Gross output, net output after operating deductions, and net output measured in purchasing power are three different numbers, and the marketing figure is usually the first.

Fixed nominal commitments in general. A promise to receive a fixed amount for a long time is a promise about the number, never about what the number buys.

The formulas, the assumptions and the limits

monthly factors            (1 + g)^(1/12), (1 - F)^(1/12), (1 + i)^(1/12)
price level after m months (1 + i) ^ (m / 12)
real balance               nominal balance / price level
contribution basis         S + sum of C / (1 + i) ^ (t / 12)
real gain                  real ending value - contribution basis
real rate identity         (1 + g) / (1 + i) - 1
break-even, no cash flows  g* = (1 + i) / (1 - F) - 1

What it assumes: constant effective annual return, fee and inflation for the whole horizon; percentage fees behaving as one combined annual charge on the balance; contributions arriving exactly on schedule, fixed in nominal terms and never missed; a fixed fee floored at the balance available, with no unpaid remainder carried forward; and no rounding until display.

What it excludes: tax of every kind, indexed contributions, withdrawals, variable inflation or return paths, sequence risk, volatility, probability, performance fees, transaction costs, currency conversion, and any judgement about whether the figures entered are reasonable.

Where an answer does not exist, the model says so in words instead of printing one. The purchasing-power break-even is reported as unavailable rather than as an invented number when no reliable answer can be found, and no figure anywhere is allowed to reach the page as an infinity or a negative zero.

Run it on your own figures

The Real Return after Fees and Inflation Calculator applies exactly the model above to numbers you enter: what is invested, what you add and how often, for how long, at what assumed return, at what cost, and under what assumed inflation.

It leads with what the projected balance would buy at the start of your scenario, then answers the question that actually decides something: whether that purchasing power is above or below the inflation-adjusted value of everything you paid in. It keeps the nominal and the real figures in separate blocks with separate labels, shows the fee decomposition only in nominal terms, and reports the gross return that would put the real gain at exactly zero.

It runs entirely in your browser. UBWHY does not receive the values you enter, stores none of them, puts none of them in a link, and fetches no inflation figure from anywhere.