Return on investment sounds like one number, but it has two honest forms: the total gain over the whole period, and the annualized rate that compounds. This calculator computes both, includes the income most people forget, and shows exactly how far the popular "just divide by years" shortcut overstates performance.
Put $10,000 to work, collect $600 of income over three years, and sell for $12,500. Total proceeds are $13,100, net profit is $3,100, and the total ROI is 31.0%. That is the number that goes into the Christmas letter. The next question is the one that causes arguments: what was the annual return?
The street answer divides 31% by 3 and reports 10.33% a year. The correct answer is the compound annual growth rate, 9.42%. Both describe the same investment, and the street answer is a full nine-tenths of a percentage point richer every year. Nothing was manipulated; the division simply answers a question nobody asked. It spreads the profit evenly across the years, as if each year delivered a slice of the gain and the money sat idle between slices. Compounding does not work that way, and the mismatch compounds right along with it.
Hold the same 31% total ROI and stretch the holding period. The divided figure drifts down in a straight line; the compounded figure falls faster and keeps falling:
| Years held for the same 31% total | ROI ÷ years | True CAGR |
|---|---|---|
| 1 year | 31.00% | 31.00% |
| 3 years (default) | 10.33% | 9.42% |
| 5 years | 6.20% | 5.55% |
| 10 years | 3.10% | 2.74% |
Over one year the two answers coincide, which is why one-year performance claims cause so little trouble. Over a decade the divided figure overstates the true rate by more than a tenth of its own size. Anyone quoting a three-decade real estate story as "we made 200%, so about 6.7% a year" is running this same trick, knowingly or not.
The most common DIY mistake is computing ROI on price alone. The default example returns 25% on price alone, $12,500 against $10,000. Add the $600 of collected income and the honest figure is 31%. That six-point gap is not a rounding detail; for assets whose return leans on cash flow, rental property, dividend stocks, bonds, price-only ROI can turn a mediocre investment into a flattering story. The flip side is honesty about timing: this calculator counts income at the end, with no reinvestment, which understates what a disciplined dividend reinvestor actually earned. If you reinvest payouts, enter the reinvested shares' value in the ending value field and leave the income field empty, so the money is counted once.
Variables in plain terms. Ending value is what the position sold for or is worth now. Income is everything the position paid you along the way: dividends, interest, rent, net of nothing. Amount invested is your full cost basis. Years is the holding period, and fractions are fine; eighteen months is 1.5. CAGR takes the total proceeds ratio to the power of one over years, which is the rate that, applied every year with compounding, lands on the same ending wealth.
Proceeds: $12,500 + $600 = $13,100. Net profit: $13,100 − $10,000 = $3,100. Total ROI: 3,100 / 10,000 = 31.0%. CAGR: (13,100 / 10,000)1/3 − 1 = 1.310.3333 − 1 = 9.42%. Divided annual: 31 / 3 = 10.33%. The three numbers agree on the facts and disagree on the story, and only CAGR can be compared against a fund's annualized return without flattering anybody.
Fidelity Investments puts the S&P 500's average annual return at about 10% since the index launched in 1957. Compound 10% for three years and $10,000 becomes $13,310, a total ROI of 33.1%. The default example, 31% over three years, therefore lands just under a boring index fund, despite feeling like a win in isolation. That is the second job of the annualized figure: it turns any investment, a rental duplex, a gold stash, a friend's restaurant deal, into a number you can line up against the market without adjusting for effort.
Three blind spots survive the CAGR correction. First, risk: a smooth 9.4% from Treasury notes and a wild 9.4% from a single stock are not equivalent achievements, and ROI cannot see the difference. Second, endpoints: measure from a crash low to a peak and almost anything shines, which is why the holding period dates should be honest rather than chosen. Third, what comes out the other end: this is a pre-tax, pre-inflation figure. The inflation calculator shows what the nominal gain is worth in constant dollars, and for most taxpayers part of the gain never was theirs to keep.
The compound interest calculator runs the forward version of this math: given a rate, project what money grows into, which is how to sanity-check whether a promised ROI is even internally consistent. The savings goal calculator works backward from a target to the rate or contribution needed. And the FIRE calculator applies realistic long-run return assumptions to a whole portfolio plan, where quoting a divided-instead-of-compounded rate by accident can move the retirement date by years.
One habit worth keeping: whenever someone quotes a return, ask two questions. Over how long? And is that annualized or total? If the answer to the second question is a pause, run the total figure through this calculator and let the CAGR speak.
ROI measures the whole gain over the whole period; CAGR converts it to a yearly rate that assumes compound growth. On the default example, $10,000 growing to $13,100 over three years is a 31.0% total ROI. Dividing by three gives 10.33% a year, but the true annualized rate is 9.42%. The division answer is always higher for any winning investment held more than one year, and the gap widens with the holding period: the same 31% over ten years is 3.10% by division but 2.74% compounded.
Yes. Return on investment means everything you got back, not just the ending price. The default example adds $600 of income collected over three years to a $12,500 ending value, lifting total proceeds to $13,100. Leaving income out would understate ROI by 6 percentage points here. This calculator assumes income arrives at the end and is not reinvested, which is the conservative choice; reinvested dividends would push the compound rate slightly higher.
It depends on the time frame, and that is exactly why the annualized line matters. Fidelity Investments puts the S&P 500 average annual return at about 10% since the index launched in 1957. Ten percent compounded for three years turns $10,000 into $13,310, a 33.1% total ROI. So the default example, 31% over three years, sits just under a plain index fund. Any ROI claim without a holding period attached is untestable against any benchmark.
Division answers a different question. It spreads the total gain evenly across the years, as if you realized a slice of the profit each year and held cash the rest of the time. CAGR answers the question people actually mean: what steady yearly growth rate reproduces the same ending wealth. Compounding makes those two numbers diverge, and the divergence is not symmetric. Gains compound upward, so a winning investment's true annual rate is always below the divided figure; the same 31% total over one, three, five, and ten years is 31.0%, 9.42%, 5.55%, and 2.74% compounded, against 31.0%, 10.33%, 6.20%, and 3.10% by division.
No. ROI treats a dollar of profit the same whether it came from a Treasury note or a speculative stock, which is why it should only compare investments of similar risk. It is a nominal figure: inflation comes out of it, not the formula. Taxes depend on holding period and account type, so an after-tax ROI can differ meaningfully from the pre-tax number this calculator produces. Use the inflation calculator on this site to see what a nominal return is worth in constant dollars.
A negative CAGR is the yearly rate at which the investment shrank; the default inputs dropped to $8,000 with no income over five years would show a negative 4.36% compound rate against a negative 4.00% divided figure, and here the division is the flattering one. A total loss, ending value plus income of zero, has no meaningful CAGR beyond the definition: the investment compounds to negative 100%, everything is gone, and no yearly framing makes it look better.
Same compounding math, different starting question.