Blog
How to Calculate Return on Investment (and Why the Simple Version Misleads)
Basic ROI ignores time, which makes a 40% return over ten years look better than a 15% return over one. Here is the version that accounts for it.
ROI is gain divided by cost. It is a useful one-line summary and it has one significant blind spot: it says nothing about how long the return took. A 40% ROI is excellent over one year and poor over ten, and the basic formula cannot tell them apart.
The formulas
The annualised version converts any return into an equivalent yearly rate, which is what makes different holdings comparable. It is the same compounding relationship used for interest, rearranged to solve for the rate.
Basic ROI = ((Final value - Initial cost) / Initial cost) x 100 Annualised ROI = (((Final value / Initial cost)^(1/n)) - 1) x 100 where n = number of years
Worked example: why time changes everything
Investment A: 10,000 becomes 14,000 over ten years. Basic ROI is 40%.
Investment B: 10,000 becomes 11,500 over one year. Basic ROI is 15%.
On the basic figure, A looks better. Annualised, A returns ((14,000/10,000)^(1/10)) - 1 = 3.42% a year, while B returns 15%. B is more than four times better per year.
Any comparison between investments held for different periods needs the annualised figure. Comparing basic ROI across different time horizons is not a comparison at all.
Worked example: a business purchase
A machine costs 12,000 and saves 4,200 a year in outsourced work, with 600 a year in maintenance. Expected life is five years, with a residual value of 1,500.
Net annual benefit: 4,200 - 600 = 3,600. Over five years that is 18,000, plus 1,500 residual, giving 19,500 returned on a 12,000 cost.
Basic ROI: (19,500 - 12,000) / 12,000 = 62.5% over five years. Annualised: ((19,500/12,000)^(1/5)) - 1 = 10.2% a year.
Payback period is a useful companion figure here: 12,000 / 3,600 = 3.3 years before the purchase has repaid itself. ROI tells you the size of the return; payback tells you how long your money is committed.
ROI for different kinds of spending
The marketing row is the one most often done incorrectly. Using revenue rather than gross profit inflates ROI dramatically. A campaign costing 5,000 that generates 20,000 of revenue at a 30% margin returned 6,000 of gross profit, giving a 20% ROI — not the 300% that a revenue-based calculation shows.
| Application | What to use as cost | What to use as return |
|---|---|---|
| Financial investment | Amount invested plus fees | Final value plus any income received |
| Equipment | Purchase price, installation, training | Savings or extra revenue, minus running costs |
| Marketing | Full campaign spend including time | Gross profit from attributable sales, not revenue |
| Property improvement | Materials, labour, permits | Increase in value or rental income |
| Training or a course | Fees plus time at your own rate | Increase in earnings attributable to it |
What ROI leaves out
None of these make ROI useless. They mean it works as a first-pass filter rather than a decision rule, and that a high ROI figure is the start of a question rather than the end of one.
- Risk. Two investments with identical ROI can carry entirely different chances of losing the money.
- Timing of cash flows. Returns arriving early are worth more than the same amount arriving late, which is what net present value addresses and ROI does not.
- Opportunity cost. A 6% return is good in isolation and poor if a comparable-risk alternative paid 9%.
- Inflation. A 5% nominal return with 4% inflation is a 1% real return.
- Tax. After-tax return is what you actually keep, and it varies by account type and jurisdiction.
- Scale. A 200% return on 100 and a 12% return on 200,000 are not comparable in any practical sense.
How ROI figures get flattered
- Comparing returns over different holding periods without annualising.
- Using revenue instead of profit as the return, which overstates the result by the whole cost base.
- Omitting fees, transaction costs and taxes from the cost side.
- Ignoring the value of your own time on projects where it is a substantial input.
- Attributing all of a result to one action when several contributed.
- Applying a past ROI forward as if it were an expected return, which turns a measurement into a forecast it does not support.
Assumptions and limitations
Annualised ROI assumes the return compounded smoothly at a constant rate, which is essentially never true. It is a summary figure describing the average, not a description of what happened year to year, and an investment that fell 30% before recovering shows the same annualised figure as one that rose steadily.
It also assumes you can identify what the return actually was and that it was caused by what you spent. For financial assets this is straightforward. For marketing, training or a business improvement, attribution is genuinely difficult, and an ROI figure built on a confident attribution that cannot be tested is not a reliable number.
Most importantly, a calculated ROI describes something that already happened. Projecting it forward assumes conditions repeat, which is an assumption you are choosing to make rather than one the arithmetic supports. Investment returns in particular are not guaranteed, past performance does not indicate future results, and any projection should be tested at a materially lower return to see whether the decision still holds.
Use ROI with a time horizon
A simple ROI percentage does not tell you how quickly the return was earned. A 20% gain over one year and a 20% gain over five years are not equivalent economic outcomes. When comparing alternatives, keep the measurement period consistent and consider cash-flow timing, recurring costs and the amount of capital tied up.
Frequently asked questions
What is a good ROI?
It depends entirely on the risk, the time period and the alternatives. The only meaningful comparison is against what a similar-risk option would have returned over the same period.
How is ROI different from ROE or IRR?
ROE measures return against equity specifically. IRR accounts for the timing of every cash flow, which makes it more accurate than ROI for projects with uneven returns spread over time.
Can ROI be negative?
Yes. If the final value is below the cost, ROI is negative and represents a loss as a percentage of what was invested.
Should ROI be calculated before or after tax?
After tax gives the figure that reflects what you keep, and it is the right basis for comparing options taxed differently. Be consistent — comparing a pre-tax ROI to an after-tax one is meaningless.
Does a higher ROI always mean a better project?
Not necessarily. Time, risk, cash-flow timing, scale and the reliability of the assumptions can change how useful a headline ROI is.
Conclusion
Calculate basic ROI to size a return, then annualise it before comparing anything held for a different length of time. Use profit rather than revenue, include every cost on the cost side, and treat the result as a measurement of what happened rather than a prediction of what will. Where a decision depends on a projected ROI, run it again at a lower return — if it only works at the optimistic figure, that is the most useful thing the calculation has told you.
Figures in this article are illustrative. Results depend on your own rates, fees, taxes and circumstances, and are for educational and planning purposes rather than financial advice.