Simple ROI and why it misleads
Return on investment is net profit divided by total cost. Turn $26,500 into $41,000 and you have made $14,500, a 54.72% return.
The problem is that this number has no time dimension. A 54.72% return is excellent over three years and mediocre over fifteen. Comparing two investments by simple ROI without knowing how long each took is close to meaningless, yet it is how returns are most often quoted.
This is why the calculator also shows the annualised figure. Over 36 months, that 54.72% total works out to 15.66% a year — a number you can compare directly against a savings account, an index fund, or another project.
Annualised return is the comparable number
The compound annual growth rate answers a specific question: what steady annual return would have produced this result? The formula is (final ÷ invested)^(1 ÷ years) − 1.
Note that it is not the total return divided by the number of years. Dividing 54.72% by three gives 18.24%, which overstates the true 15.66%. The difference is compounding — each year's return builds on the previous year's, so a lower rate reaches the same endpoint.
Whenever you see a headline return over an unusual period — "up 140% since 2019" — annualise it before reacting. Long periods make ordinary returns look spectacular.
Include every cost
The gap between quoted and actual returns is usually costs. In the default example, ignoring $1,500 of additional costs turns a real 54.72% return into an apparent 64% — a nine-point exaggeration from one omitted line.
For property, the costs that get left out are closing fees, agent commission on sale, property tax, insurance, maintenance, and vacancy periods. Gross rental yield figures routinely ignore most of these; net yield after all costs is often less than half the advertised number.
For securities, it is trading commissions, bid-ask spreads, fund expense ratios, and tax on gains and dividends. For a business, it is your own unpaid time, which is a real cost even though no invoice records it.
A useful discipline: calculate ROI from the total amount that left your account, not the headline purchase price.
What ROI cannot tell you
ROI ignores risk entirely. A 15% annualised return from government bonds and a 15% return from a single speculative stock are not equivalent achievements, but ROI reports them identically. Any comparison of returns without a matching comparison of risk is incomplete.
It also ignores the timing of cash flows. If you invested in instalments rather than all at once, or took distributions along the way, simple ROI will not reflect that correctly. Internal rate of return (IRR) or money-weighted return handles irregular cash flows properly.
And it ignores liquidity and effort. A property returning 12% that consumes weekends is not obviously better than an index fund returning 9% that requires nothing. ROI measures money in and money out; the rest of the decision lives outside the formula.
Finally, it is nominal. A 6% return during 4% inflation is a 2% real return, and that is what actually changed your buying power.
Frequently asked questions
- What is a good ROI?
- It depends entirely on risk and timeframe. As reference points, US stocks have averaged roughly 10% nominally over the long run, and high-yield savings has recently paid around 4%. Anything promising far more than the market average deserves scrutiny of the risk being taken.
- What's the difference between ROI and CAGR?
- ROI is the total percentage gain over the whole period. CAGR converts that into an equivalent annual rate, which is what lets you compare investments held for different lengths of time.
- Should I include my own labour as a cost?
- For any project where you work on it — a rental property, a small business, a renovation — yes. Value your time at what you could otherwise earn. Excluding it makes labour-intensive investments look artificially strong.
- How do I calculate ROI with ongoing income?
- Add cumulative income to the final value, and put ongoing costs in the additional costs field. For irregular contributions or withdrawals, IRR is the more appropriate measure.