Saving & Investing
Inflation Calculator
See what a sum would need to grow to just to stand still — and what it will actually buy if it does not. Two figures, both true, running in opposite directions.
What the rate does to it
$10,000.00 over 10 years at 3% a year
You would need
$13,439.16
In future dollars — the sum that would buy then what your amount buys today.
It would buy
$7,440.94
In today's money — what $10,000.00 would still buy after 10 years.
Two ways of stating the same erosion: the first is the target to aim at, the second is what happens if you do nothing. Multiply by the inflation factor for one, divide by it for the other.
What the erosion comes to
- Purchasing power lost Purchasing power lost, in today's money — the gap between what the amount buys now and what it buys then.
- $2,559.06
- As a share of the amount Not 30% — the rate compounds on a shrinking base, so the cumulative loss is always less than the rate times the years.
- 25.59%
Nothing is lost — at this rate over this period, purchasing power does not move.
What you entered
- Amount today
- $10,000.00
- Assumed rate
- 3%
At an assumed 3% a year over 10 years, $10,000.00 loses 25.59% of its purchasing power. Two figures follow from that, and they are reciprocals rather than alternatives: you would need $13,439.16 in future dollars to buy what $10,000.00 buys today, and the $10,000.00 itself would buy about $7,440.94 worth of today's goods by then. The cumulative loss compounds on a shrinking base, so it is less than the rate multiplied by the years — 25.59% rather than 30%. The rate is your assumption. This page applies the single constant rate you entered for the whole period; it does not use any published price index, and nothing here is a forecast of what inflation will actually do.
Why the loss is not the rate times the years
Fixed figures, independent of the controls above: $10,000 at an assumed 3% a year. The last two columns are the point — the gap between what compounding actually costs and what multiplying suggests widens the longer the period runs.
| Years | $10,000 buys | Actually lost | Rate × years |
|---|---|---|---|
| 1 | $9,708.74 | 2.91% | 3% |
| 5 | $8,626.09 | 13.74% | 15% |
| 10 | $7,440.94 | 25.59% | 30% |
| 20 | $5,536.76 | 44.63% | 60% |
| 30 | $4,119.87 | 58.8% | 90% |
What this calculates
This applies one constant rate of inflation to a sum over a period you choose, and reports the result in both of the directions it can run. The first is the amount you would need at the end of the period to buy what your sum buys today — the figure to set a savings target against. The second is what your sum itself would buy by then, expressed in today's money, which is the figure that shows what doing nothing costs. It also reports the purchasing power lost in dollars and as a percentage. Those two percentages are the ones most often misread: the cumulative loss compounds on a shrinking base, so it is always smaller than the rate multiplied by the years.
How it works
Inflation factor f = (1 + r)^n
where r = annual rate ÷ 100
n = years
Then:
amount needed later = amount × f
what it will buy = amount ÷ f
purchasing power lost = amount − (amount ÷ f)
percentage lost = (1 − 1 ÷ f) × 100
The two money figures are reciprocal operations on the same
factor, which is why one rises as the other falls and why
they are never equal unless the rate or the period is zero.
Note the percentage: it is 1 − 1/f, NOT r × n. At 3% over
ten years f = 1.343916, so the loss is 25.59% rather than
the 30% that multiplying suggests. The rate applies each
year to what is left after the previous year, so the base
it acts on keeps shrinking.Inflation is a fall in what money buys, and it is easier to reason about as two separate questions than as one. Ask "how many dollars would I need later" and you get one number. Ask "what will my dollars buy later" and you get another. Both describe the same rate acting over the same period, and they move in opposite directions.
Take $10,000 at an assumed 3% for ten years. To buy then what $10,000 buys today you would need $13,439.16. Meanwhile the $10,000 itself, left untouched, would buy about $7,440.94 worth of today's goods. The first figure is a target; the second is a diagnosis. They are reciprocals of one another — multiply by the inflation factor to get one, divide by it to get the other — and averaging them, or mistaking one for the other, is the single most common error made with this calculation.
The percentage is where the arithmetic surprises people. Three per cent for ten years feels like thirty per cent, and the page reports 25.59%. Nothing is broken. Each year's inflation applies to what purchasing power remains after the year before, so the base being eroded gets smaller every year and the cumulative total falls short of the simple multiplication. The gap widens as the horizon lengthens: over twenty years the naive figure is 60% and the real one is 44.63%; over thirty it is 90% against 58.80%. The longer the period, the more the intuitive method overstates the loss.
The horizon does most of the work, though, and it works in the other direction on the money. At 3%, ten years costs $2,559.06 of a $10,000 sum's purchasing power, twenty years costs $4,463.24, and thirty costs $5,880.13. Rate changes bite hard too: holding the period at ten years, 2% leaves $8,203.48 of purchasing power, 3% leaves $7,440.94, 4% leaves $6,755.64, and 6% leaves $5,583.95. A difference of a few points in an assumption you cannot verify changes the answer by thousands of dollars, which is the honest reason to run this calculation at several rates rather than one.
What this page will not do is tell you what inflation has been or will be. It holds no price index and makes no forecast. The rate is yours, and the result is arithmetic on it. The US Bureau of Labor Statistics publishes the Consumer Price Index for exactly the job of adjusting dollar values between periods, and its own guidance is worth reading before leaning too hard on any single number: the CPI measures average change across a basket of goods for a defined population, and the BLS is explicit that it "does not necessarily measure your own experience with price change." If your spending is concentrated in housing, healthcare or education, a general index is a rough proxy for your position rather than a description of it.
A worked example
$10,000 today, with inflation assumed at 3% a year, looked at over a ten-year horizon.
The inflation factor over ten years at 3% is 1.343916. That single number generates everything else.
Multiply by it and you get $13,439.16 — the amount you would need in ten years' time to buy what $10,000 buys today. This is the figure to set against a savings goal, because a target of $10,000 in ten years is not a target of today's $10,000. It is a target of about 74 cents on the dollar.
Divide by it instead and you get $7,440.94 — what the $10,000 would buy by then, stated in today's money. The difference, $2,559.06, is the purchasing power lost over the decade, and as a percentage of the original sum that is 25.59%.
That 25.59% is the figure most worth pausing on. Three multiplied by ten is thirty, and a reader who does that arithmetic will conclude the page has a bug. It does not. Inflation in year two applies to what was left after year one, and so on down the decade, so the base shrinks each year and the cumulative loss lands below the simple product. Lengthen the horizon and the discrepancy grows: at twenty years the loss is 44.63% rather than 60%, and at thirty it is 58.80% rather than 90%.
Scaling is linear in the amount, which makes the figures easy to carry across. $50,000 at the same 3% over twenty years needs $90,305.56 to keep pace and would itself buy $27,683.79 of today's goods — the same 44.63% loss, applied to five times the sum, for $22,316.21 of lost purchasing power.
Deflation is a legitimate input and inverts every direction. At an assumed −2% over ten years, $1,000 becomes a case where you would need only $817.07 to buy what $1,000 buys today, while the $1,000 itself would buy $1,223.88 of today's goods. Cash gains purchasing power. The engine reports the loss as a negative number and the page relabels it as a gain, because "lost −$223.88" is not a sentence anyone should have to parse.
One last figure, for the reader who wants a feel for long horizons. $100 at an assumed 3% over twenty-four years would need $203.28 to keep pace — roughly a doubling of the nominal sum — while the $100 itself would buy about $49.19 of today's goods. Slightly over half the purchasing power gone, from a rate low enough that it is usually described as price stability.
What this assumes
- The rate you enter is an assumption you are choosing. This page holds no price index, quotes no published series, and makes no forecast of any kind.
- That single rate is applied unchanged across the entire period. Actual inflation varies year to year, sometimes by a great deal.
- Inflation is compounded annually. The factor is (1 + r)^n, so the rate acts on the result of the previous year rather than on the original sum.
- The two money figures are exact reciprocal operations on the same factor, which is why one rises as the other falls.
- The cumulative percentage is 1 − 1/(1+r)^n, not the rate multiplied by the years. It is always the smaller of the two.
- A negative rate is treated as genuine deflation rather than an error, and every label inverts accordingly.
- A zero rate, or a zero period, leaves every figure equal to the amount entered. That is the baseline case, not a missing result.
- Nothing here earns interest or return. This models the erosion of a static sum, not what would happen if it were invested.
What it does not model
- No published price index is used. The Bureau of Labor Statistics maintains the CPI for exactly this purpose and this calculator does not read it, so the result reflects your assumption rather than any measured series.
- The general price level is not your price level. The BLS states plainly that the CPI "does not necessarily measure your own experience with price change", and a household whose spending is concentrated in housing, healthcare or education may have faced a very different rate from the headline one.
- A single constant rate is the largest simplification here. Real inflation clusters — long quiet periods broken by sharp episodes — and a constant-rate model cannot show the difference between a decade of 3% and a decade averaging 3% with a spike in the middle.
- The result is arithmetic, not a forecast. Nobody can tell you what inflation will be over your horizon, and the sensitivity of the answer to that unknowable input is high: at ten years, moving the assumption from 2% to 4% changes what $10,000 buys from $8,203.48 to $6,755.64.
- Taxes, fees and returns are all outside the model. A sum actually held somewhere earns or costs something, and this page deliberately isolates the inflation effect from all of it.
- Annual compounding is assumed. Prices do not move in yearly steps, and a monthly or continuous treatment would give slightly different intermediate figures, though the difference is small beside the uncertainty in the rate itself.
- CPI series are not interchangeable and the BLS warns against assuming they are. Its guidance recommends the US City Average for escalation, notes that local-area indexes carry substantially larger sampling errors, and states that seasonally adjusted data are inappropriate for adjustment agreements. None of that is modelled here because no series is used at all.
- This is not advice about what to do with money. Showing that purchasing power erodes is not a recommendation to invest, and certainly not a recommendation about where.
Questions
Which of the two figures is the answer?
Both are, to different questions, and this is the distinction worth getting right before using either. On the default scenario $13,439.16 is what you would need in ten years to buy what $10,000 buys today — use it to set a savings target, since a goal of "$10,000 in ten years" is really a goal of about 74 cents on today's dollar. The other figure, $7,440.94, is what your $10,000 would itself buy by then, stated in today's money — use it to see what leaving the sum alone costs. They are reciprocals of the same inflation factor: multiply to get the first, divide to get the second. The one genuinely wrong thing to do is average them or quote one while meaning the other, because they run in opposite directions.
Why is the loss 25.59% and not the 30% I get from three per cent times ten years?
Because inflation compounds against a shrinking base. Year two's 3% applies to the purchasing power that survived year one, year three's to what survived year two, and so on, so each year erodes a slightly smaller amount than the last and the total lands below the simple product. The formula is 1 − 1/(1+r)^n rather than r × n. The gap widens with the horizon: over twenty years the naive figure is 60% and the actual is 44.63%; over thirty it is 90% against 58.80%. Multiplying the rate by the years always overstates the loss, and the longer the period the more it overstates it. This is one of the few places where the intuitive answer is close enough to the right one that people assume the calculator is broken.
What inflation rate should I use?
There is no rate this page can honestly supply, and the 3% default is a round illustrative figure rather than a claim about any period or place. Because the answer is highly sensitive to a number nobody can know in advance, the useful approach is to run several rates instead of trusting one: at ten years, $10,000 keeps $8,203.48 of purchasing power at 2%, $7,440.94 at 3%, $6,755.64 at 4% and $5,583.95 at 6%. That spread tells you more about your exposure than any single figure. If you want a measured historical series rather than an assumption, the Bureau of Labor Statistics publishes the Consumer Price Index for precisely that purpose, and reading a rate from it is a different exercise from the one this page performs.
Does the CPI describe what inflation costs me personally?
Not exactly, and the BLS is unusually direct about this: the CPI "does not necessarily measure your own experience with price change." It is an average across a basket of more than 200 item categories for a defined population, and the CPI-U covers over 90% of the US population while the CPI-W covers about 30%. Neither covers rural, farm or institutional households. If your spending concentrates in housing, healthcare or education, your own rate may have diverged substantially from the headline number. The BLS also cautions against reaching for local-area indexes as a fix, since they have relatively small sample sizes and are subject to substantially larger sampling errors than the national series. None of that affects the arithmetic on this page, which uses whatever rate you type, but it affects how much confidence any particular rate deserves.
Can I enter a negative rate?
Yes, and it models deflation properly rather than treating it as an error. At an assumed −2% over ten years, $1,000 flips both figures: you would need only $817.07 to buy what $1,000 buys today, and the $1,000 itself would buy $1,223.88 of today's goods. Holding cash gains purchasing power. The page relabels the rows accordingly — the engine returns the loss as a negative number and showing "lost −$223.88" would be gibberish, so it reads as a gain instead. Worth knowing that sustained deflation is historically unusual and tends to accompany conditions in which other things are going wrong, so a long-horizon deflation scenario is an arithmetic exercise rather than a plan.
Why do my figures not change when I set the rate to zero?
Because at a zero rate nothing happens, which is the correct result rather than a failure. Every figure equals the amount you entered, and the two loss rows disappear entirely instead of showing $0.00 — a "zero lost" row implies a loss that happens to round to nothing, which is a different claim. The same applies at a zero timeframe. It is worth looking at deliberately: it is the version of this calculation that assumes nothing, and every figure the page shows above it is the consequence of a rate you chose rather than one anyone has measured for you.
Should I use this to adjust a contract, rent or support payment?
Not on its own. This page applies one assumed constant rate, and an escalation agreement needs a measured index with the terms pinned down. The BLS publishes guidance for exactly that and its recommendations are specific: adopt the US City Average CPI rather than a local series, identify the exact series by population, area, title and base period, name a reference month or annual average, set the adjustment interval and the formula including any cap or floor, and provide for index revisions. It also states that seasonally adjusted data are inappropriate for escalation, partly because they can be revised for up to five years. The BLS neither encourages nor discourages such clauses and will not draft wording or settle disputes about them. For anything with legal force, work from the published series and take proper advice.
Sources
- BLS — Consumer Price Index: Questions and Answers (opens in a new tab)
What the CPI measures and what it excludes: an average of prices across more than 200 item categories in eight major groups, for a defined urban population — CPI-U covering over 90% of the US population and CPI-W about 30%, with rural, farm and institutional households outside both. It also supports this page's statement that a general index is not a description of an individual household's experience, in the BLS's own words that the CPI "does not necessarily measure your own experience with price change". It establishes no rate of inflation, and nothing on this page is derived from it numerically.
- BLS — Using the Consumer Price Index for escalation (opens in a new tab)
The conventional use of a price index to adjust dollar values between periods, and the specific cautions this page repeats: that the US City Average is recommended over local-area series, that local series carry substantially larger sampling errors, and that seasonally adjusted data are inappropriate for escalation because they remain subject to revision for up to five years. It backs the limitation that adjusting a contract requires a measured series rather than the assumed constant rate used here.
- Formula reviewed
- Formula version
- 1
Version 1 means the formula has not changed since this page was published. If it changes, this number moves and the change is described here.