How it unfolds, step by step
The setup
An account starts at ₹10,000. Exactly two years later it ends at ₹12,100, with nothing added or withdrawn in between. Absolute return is the change over the start. That is ₹2,100 on ₹10,000, or 21%. It covers the whole period, not a year.
What happens next
The money sat undisturbed, so the 21% can be spread evenly. The steady yearly rate that turns ₹10,000 into ₹12,100 over two years is 10%. Note what happened to the number. The total is 21% and the yearly equivalent is 10%. Neither is wrong. They answer different questions. Quoting the 21% as a yearly figure, or the 10% as the total, is where most comparisons go wrong.
What it means — and does not
CAGR earns that 10% only because nothing moved. Change the account so money arrives on dates, and it stops describing anything. Take a second account. On 1 January ₹1,000 is paid in, and another ₹1,000 follows on 1 December. On 31 December the account is worth ₹2,150. Total paid in is ₹2,000, so the simple percentage is 7.5%. But the second ₹1,000 was invested for only a month. A figure that credits it with the whole year understates the rate the money actually earned. XIRR discounts every flow from its own date on a 365-day basis. It reports 13.96%, nearly double, from the same two payments.
- Capital value (₹)
- Smooth 10% path at one year: ₹11,000 · 11,000
₹10,000 grows to ₹12,100 over two years, which is 21% in total and 10% a year. The dip to ₹8,500 in between changes nothing about the reported CAGR. That is exactly the point. One steady rate describes every path that starts and ends in the same place.
Show the underlying data
| Year | Capital value (₹) |
|---|---|
| 0 | 10,000 |
| 1 | 8,500 |
| 2 | 12,100 |
What this does not establish
Each measure stops holding at a different point. Absolute return stops being comparable the moment two periods differ in length. A 21% gain over two years and a 21% gain over seven are not the same result. CAGR stops holding as soon as money moves in or out. It also hides the path completely. The account above reports 10% whether it climbed steadily or fell to ₹8,500 in between and recovered. So it describes neither the volatility nor the worst point along the way. XIRR needs cash flows of both signs to have a solution at all. Unusual patterns can allow more than one mathematical answer, or fail to converge. A solver that returns an error has not found a return of zero. It has found no return. And every one of the three describes what already happened. None of them is a forecast.
Things to check
5Is the percentage being quoted the whole period, or one year of it?
Did any money go in or out during the period? If it did, CAGR is the wrong tool.
Are the two figures being compared over the same length of time?
For a dated calculation: is every contribution negative, the closing valuation positive, and is every date the real one?
Does the figure say anything about the worst point along the way? None of these three does.
Common mistakes
5Reading a total return as an annual one. The account above is 21% in total and 10% a year. Both describe it, and neither stands in for the other.
Applying CAGR to a schedule of contributions, by dividing the closing value by the total paid in. That treats late money as though it had been invested from the start.
Reading a smooth annualised rate as a description of the journey. It is a single equivalent rate, and every path that ends in the same place reports it.
Treating an XIRR solver error as a zero return, when it means the calculation did not resolve.
Comparing a money-weighted figure such as XIRR against a time-weighted one, which answers a different question about the same account.
On 1 January ₹1,000 goes in, and another ₹1,000 on 1 December. The account is worth ₹2,150 on 31 December, so the simple percentage on the ₹2,000 paid in is 7.5%. What does the dated calculation report?
- 7.5% as well — an annualised method cannot change a figure measured over exactly one year
- About 13.96%
- Less than 7.5%, because spreading the contributions out dilutes the return
Show the answer
B. About 13.96%
The second ₹1,000 was invested for a month, not a year. The simple percentage divides the gain by everything paid in and credits all of it with the whole period. That understates the rate the money actually earned. XIRR discounts each flow from its own date and reports 13.96%. The first answer is the tempting one, because the period really is one year. But it is the money, not the calendar, that was in for less time.
Frequently asked
3Which one should I use for a monthly contribution?
The dated calculation. A schedule of payments is exactly the case CAGR cannot describe. It has no way to know that later money was invested for less time. In the example above, the simple percentage reads 7.5% and XIRR reads 13.96%.
Why is my XIRR higher than my absolute return?
Usually because much of the money arrived late. The simple percentage divides the gain by everything paid in and credits all of it with the full period. XIRR credits each amount only for the days it was actually invested.
Does a good CAGR mean the investment was not risky?
No. CAGR is computed from two values, the first and the last. So it cannot see anything that happened in between. The page /learn/drawdown-recovery-arithmetic works through why the worst point along the way is a separate question.