Key takeaways
- Time-weighted measures the investment. Money-weighted measures your experience of it.
- Annualize with a geometric root, never by dividing total return by years.
- The arithmetic average of volatile returns always exceeds the compound growth rate.
- IRR handles irregular timing but ignores scale and can have multiple solutions.
- Maximum drawdown predicts whether a plan gets followed better than volatility does.
Total return ignores time
A dollar gain or total percentage change does not say how long capital was invested. Annualization converts a beginning and ending value over a known period into an equivalent compound yearly rate.
IRR is money weighted
Internal rate of return is sensitive to the size and timing of deposits and withdrawals. It is useful for an investor’s experienced return or a project’s cash flows, but unusual cash-flow patterns can produce multiple or misleading solutions.
Time-weighting isolates portfolio performance
Time-weighted return links subperiod returns around external cash flows so large deposits and withdrawals do not dominate the result. It is often better for evaluating an investment manager who does not control cash-flow timing.
Compare like with like
Use the same gross-or-net convention, fee treatment, tax treatment, date precision and compounding basis. A precise number built from inconsistent cash flows is not a valid comparison.
Time-weighted and money-weighted returns answer different questions
A time-weighted return removes the effect of contributions and withdrawals, which makes it the right measure of how an investment or a manager performed. A money-weighted return, equivalent to an internal rate of return, incorporates the size and timing of cash flows, which makes it the right measure of what the investor actually experienced. The two can differ substantially for the same portfolio over the same period, and neither is wrong. They are answers to different questions.
Why your return differs from the fund's published return
Published fund returns are time-weighted. If you added money after a rise and before a fall, your money-weighted return will be lower than the fund's, and if you added before a rise it will be higher. This gap is a measure of contribution timing instead of the fund, and it is why comparing a personal account statement to a published performance figure often produces confusion.
Annualizing correctly means compounding, not dividing
Converting a multi-period return to an annual figure requires taking the geometric root, not dividing by the number of years. A total return of 60% over five years is not 12% a year. The fifth root of 1.60 minus one, which is meaningfully lower. Dividing overstates the annual figure, and the error grows with both the return and the number of periods.
Averages of returns are misleading without the geometric mean
The arithmetic average of a sequence of returns exceeds the compound growth rate whenever the returns vary, and the gap widens with volatility. A portfolio that falls 50% and then rises 50% has an arithmetic average of zero and an actual result of a 25% loss. Any performance claim quoting an average return without specifying whether it is arithmetic or geometric should be treated as ambiguous.
Internal rate of return and its limitations
IRR solves for the discount rate that makes a series of cash flows net to zero. It handles irregular timing well, which is what makes it useful for private investments and property. Its limitations are real: a series with multiple sign changes can have several mathematically valid solutions, it implicitly assumes reinvestment at the IRR itself, and it says nothing about the size of the gain. A high IRR on a small, short investment can be worth less than a lower one on a larger, longer one.
Risk-adjusted measures put return in context
Comparing returns without reference to the risk taken to earn them rewards leverage and concentration. Ratios that divide excess return by a volatility measure make two strategies comparable on a per-unit-of-risk basis. They depend on choices, which risk-free rate, which period, which volatility definition, so they are best used to compare options measured identically , not as absolute scores.
Drawdown measures the experience, not the average
Maximum drawdown is the largest peak-to-trough decline over a period. It captures something averages and standard deviations do not: the worst moment an investor had to sit through. Two strategies with the same return and volatility can have very different drawdowns, and drawdown is often the better predictor of whether a plan will actually be followed to completion.
After-fee, after-tax is the only return you keep
Gross returns are a property of the investment. Net returns are a property of your situation. Expense ratios, advisory fees, trading costs, and the tax treatment of distributions all reduce what is actually retained. Comparisons run on gross figures systematically favour higher-cost and less tax-efficient options, because the thing that separates them has been removed from the comparison.
Worked with real numbers
What this looks like in the IRR calculator
The guidance above is easier to judge against figures. Using the IRR calculator worked example, moving initial cash flow from $8,000 to $12,000 changes the estimated IRR from 12.6% to 29.4%.
| Initial cash flow | Estimated IRR | Total net cash flow |
|---|---|---|
| $8,000 | 29.4% | $9,500 |
| $9,000 | 24.1% | $8,500 |
| $10,000 | 19.7% | $7,500 |
| $11,000 | 15.9% | $6,500 |
| $12,000 | 12.6% | $5,500 |
Open the IRR Calculator to use your own numbers →
The inputs behind those figures
| Input | Value | Definition |
|---|---|---|
| Initial cash flow | $10,000 | Enter the initial cash flow for this scenario. |
| Year 1 inflow | $2,500 | Enter the year 1 inflow for this scenario. |
| Year 2 inflow | $3,000 | Enter the year 2 inflow for this scenario. |
| Year 3 inflow | $3,500 | Enter the year 3 inflow for this scenario. |
| Year 4 inflow | $4,000 | Enter the year 4 inflow for this scenario. |
| Year 5 inflow | $4,500 | Enter the year 5 inflow for this scenario. |
What goes wrong
Common mistakes
Each of these produces an answer that looks reasonable, which is why they survive review. To catch them in IRR, time-weighted and annualized return, rerun the IRR calculator with a different assumption and check whether the result moves in the direction the guidance predicts, since an error that survives that test is usually in one of the inputs and not in the arithmetic.
| The mistake | Why it misleads | Do this instead |
|---|---|---|
| Dividing a multi-year return by the number of years | Returns compound, so dividing overstates the annual figure and the error grows with the horizon. | Take the geometric root: the nth root of the growth factor, minus one. |
| Comparing your account's return to a fund's published return | Yours is money-weighted and includes your contribution timing. Theirs is time-weighted and excludes it. | Compute a time-weighted return if you want to judge the investment, money-weighted if you want to judge your outcome. |
| Quoting an average return without saying which average | The arithmetic mean of volatile returns is always higher than the compound rate actually achieved. | Report the geometric mean, or state clearly which is being used. |
| Ranking investments on IRR alone | IRR ignores the size of the investment and assumes reinvestment at its own rate. | Look at IRR alongside the absolute gain and the capital committed. |
People also ask
Frequently asked questions
Is IRR the same as CAGR?
No. CAGR uses only beginning value, ending value and time; IRR incorporates intermediate cash flows.
Which return should I use for my portfolio?
Money-weighted return reflects your cash timing. Time-weighted return better isolates the investments’ performance.
Why is my return lower than the fund's stated return?
Almost always because of contribution timing. The fund reports a time-weighted return that ignores when money arrived. Your account experienced a money-weighted return that reflects it. If you added money before a decline, your return will trail the fund's even though you held exactly the same investment.
Which return should I report to compare two managers?
Time-weighted, because it removes the effect of cash flows that the manager did not control. Money-weighted is the right measure for judging your own decisions, including the timing of contributions, which is a separate skill from selecting the investment.
Can IRR give more than one answer?
Yes, when the cash flow series changes sign more than once. Mathematically several rates can solve the equation, and the calculator will return one of them. In that situation IRR is not a reliable single measure and the cash flows should be examined directly or a modified rate used.
What is a good Sharpe ratio?
It depends entirely on the asset class, the period and the risk-free rate used, which is why absolute thresholds are unreliable. The ratio is most useful for comparing options measured over the identical period with the identical assumptions, and close to meaningless when comparing figures computed differently.
How do fees change the comparison?
They reduce the return every year and therefore compound against the balance. Two funds with identical gross returns and a one-percentage-point fee difference diverge steadily, and over a long horizon the gap is a large share of the terminal value. Always compare net of fees.
Which input moves the estimated IRR most in the IRR calculator?
Ranked by how far each moves the estimated IRR across the range tested: initial cash flow (8.2%, 42%), year 2 inflow (1.7%, 8.7%), year 1 inflow (1.7%, 8.6%) and year 3 inflow (1.7%, 8.4%).
How much does initial cash flow matter in the IRR calculator?
The worked example uses $10,000. With the other inputs left at the worked example, moving initial cash flow from $9,000 to $11,000 takes the estimated IRR from 24.1% to 15.9%, a swing of 42% of the worked-example figure.