XIRR vs CAGR vs absolute return
Three measures, three jobs. Absolute return says how much you gained. CAGR annualises a single investment. XIRR annualises many payments made on many dates. Using CAGR on a SIP — the common error — understates your return substantially.
The three measures, side by side
One scenario: ₹5,000 a month for 36 months, so ₹1,80,000 paid in, and the folio is now worth ₹2,20,000.
| Measure | Result | What it means | Right here? |
|---|---|---|---|
| Absolute | 22.2% | Total gain as a share of total paid in. No time dimension at all. | Only as a headline |
| CAGR | 6.92% | Treats the whole ₹1,80,000 as invested on day one, which it was not. | No — understates it |
| XIRR | 13.44% | Weights every instalment by how long it was actually invested. | Yes |
Look at the gap between the CAGR and XIRR rows. Same money, same period, same outcome — and CAGR reports roughly half the rate. Not because one is wrong arithmetically, but because CAGR is answering a question that does not apply.
Why CAGR breaks on a SIP
CAGR assumes one amount, invested at one moment, held for one period. Its formula has room for
exactly two values and a duration: (end ÷ start)1÷years − 1.
A SIP has none of that structure. Your 36th instalment was invested for one month, not three years. Feeding ₹1,80,000 in as a day-one "start value" pretends money you paid last month has been compounding since the beginning — so to explain the same end value, the implied rate has to be far lower.
Average holding time in a 36-month SIP is about 18.5 months, a little over half the period. That is roughly the factor by which CAGR misstates it.
The convention trap that makes XIRR look inflated
Here is the subtlety almost nobody explains. XIRR is an effective annual rate. A SIP calculator's "expected return" input is usually a nominal rate compounded monthly.
12% nominal compounded monthly is 1% a month, and 1.0112 − 1 = 12.68% effective. So a SIP you projected at 12% will report an XIRR near 12.68%, and nothing is broken.
This is why people conclude a fund "beat the projection" when it did exactly what was modelled. The XIRR calculator shows both figures side by side for precisely this reason.
Which to use, decided
- One investment, one exit — CAGR. A lump sum, an FD, a property. Use the CAGR calculator.
- Multiple payments on multiple dates — XIRR. Any SIP, any portfolio you have added to, any case with a partial withdrawal. Use the XIRR calculator.
- Comparing two funds — whichever is stated consistently for both. AMCs quote XIRR for SIP returns and CAGR for lump-sum returns; comparing one fund's XIRR against another's CAGR is meaningless.
- Absolute return — only as a headline, and only with the period attached. "Up 22%" without "over three years" is not information.
Two ways to fool yourself with XIRR
Short periods and cherry-picked windows
Annualising a short period produces nonsense. A 6% gain over four months annualises to about 19%. That is arithmetically correct and practically meaningless. Treat any XIRR computed over less than a year with deep suspicion.
The start date decides the answer. An XIRR measured from a market bottom looks spectacular; the same fund measured from the preceding peak looks poor. When comparing funds, use the same window for both, and prefer longer windows.
Questions people actually ask
Which return does my mutual fund statement show?
For SIP investments, XIRR — because that is the only correct measure for staggered purchases. For a lump-sum holding, CAGR.
If a statement shows both, they are describing different things and should not be compared with each other.
Can XIRR and CAGR ever be the same?
Yes, when there is only one investment and one exit. With a single cash flow in and a single value out, XIRR reduces to CAGR — the general case collapsing to the specific one.
How do I calculate XIRR in Excel or Google Sheets?
Put dates in one column and amounts in the next — investments negative, redemptions
and the current value positive — then =XIRR(amounts, dates).
The most common error is sign: if every number is positive there is no solution and the function returns an error.
Is a higher XIRR always better?
Not on its own. A higher XIRR earned by taking far more risk, or measured over a shorter and luckier window, is not a better outcome — it is a different bet.
Compare XIRR only across similar categories, over identical periods, and alongside how much volatility you had to sit through.