Three numbers, three different questions
Absolute return, CAGR and XIRR are not competing estimates of the same thing. They answer different questions, and using one where another belongs is how a perfectly ordinary fund comes to look excellent.
| Measure | The question it answers | When it is correct |
|---|---|---|
| Absolute return | How much did this grow in total? | Any time, but it is meaningless without also stating the period |
| CAGR | What steady annual rate turns the start value into the end value? | A single investment, made once |
| XIRR | What annual rate makes all these cashflows, on their actual dates, add up? | Money going in (or out) on more than one date |
One rule covers almost every case: one cashflow, use CAGR; many cashflows, use XIRR. A lumpsum has one. A SIP has as many as it has instalments.
The trap that catches most people
Here is the calculation people do without noticing they have done it.
A ₹10,000 monthly SIP has run for five years. Total invested: ₹6 lakh. Current value: ₹8.4 lakh. Gain: ₹2.4 lakh. That is 40%, and it gets quoted as the return.
Forty percent of what, over what? The first instalment has been invested for five years. The most recent one has been invested for a month. Treating them as a single ₹6 lakh block that grew 40% is not a rounding shortcut — it silently assumes every rupee was present from the beginning, which is false for all but one of the sixty instalments.
Worse, dividing that 40% by five years to get “8% a year” compounds the error with a second one, because returns compound rather than adding.
What CAGR does, and where it stops working
CAGR is the constant annual rate that would take a starting value to an ending value over a period. Put ₹1 lakh in, take ₹1.61 lakh out five years later, and the CAGR is 10% — the single rate that, compounded five times, produces that result.
It is exactly right for a lumpsum and it is a genuinely useful smoothing: the actual path was not 10% a year, but 10% is the rate that describes the outcome.
It breaks the moment there is more than one cashflow, because it has nowhere to put the dates. CAGR takes a start value, an end value and a duration. A SIP does not have a start value in that sense — it has sixty of them, on sixty dates.
What XIRR does
XIRR is the internal rate of return for cashflows on irregular dates. It takes the full list — every instalment with its date, and the final value with its date — and solves for the single annual rate that makes them balance.
The intuition: it asks what rate a bank account would have had to pay, given exactly those deposits on exactly those days, to end up with exactly this much. Money in for five years is credited with five years of growth; money in for one month gets one month's. Nothing is assumed about when it arrived, because the dates are inputs.
That is the whole of it. XIRR is not a more sophisticated CAGR — it is the same idea with the dates put back in. For a single cashflow, XIRR and CAGR return the same number, which is a useful check that you have understood both.
The same SIP, three ways
Take the ₹10,000 monthly SIP above: ₹6 lakh invested over five years, worth ₹8.4 lakh today.
| Stated as | Figure | Is it right? |
|---|---|---|
| Absolute return | 40% | Arithmetically true, and useless on its own — it does not say over how long, or that most of the money was invested for far less than the full period |
| “40% over 5 years, so 8% a year” | 8% | Wrong twice: it ignores the dates, and it divides rather than compounding |
| CAGR on total invested | ~6.9% | Wrong. It assumes the whole ₹6 lakh was present from day one, which understates the return |
| XIRR | The correct figure | Right. It uses each instalment's actual date |
Note the direction. Treating a SIP as a lumpsum understates the return, because it credits the money with more time than it had. This is why fund marketing rarely makes that particular mistake, and why comparison articles frequently do — a mis-computed SIP figure set against a correctly computed lumpsum CAGR produces a conclusion that is an artefact of the arithmetic.
Which one to use, in practice
- Lumpsum, held throughout. CAGR.
- SIP, or any schedule of contributions. XIRR.
- SIP that you later added to, paused, or partly redeemed. XIRR — this is precisely the case it exists for. Redemptions go in as negative cashflows on their dates.
- Comparing two funds. Whichever applies, but the same one for both, over the same period. Most published comparisons fail this test rather than the formula.
- Comparing a SIP against a lumpsum. XIRR for the SIP and CAGR for the lumpsum are both correct and both annualised, so they are comparable — which is the point of annualising at all.
One caution that applies to all three. An annualised figure describes the outcome, not the experience. A 12% XIRR is entirely compatible with having been 30% down at some point along the way, and it is the drawdown, not the XIRR, that determines whether somebody stays invested — the mechanism behind the behaviour gap. Read them together.
And because a SIP accumulates a new entry date every month, it is exposed to the whole distribution of outcomes rather than one draw from it — which makes rolling returns the natural companion to XIRR rather than a single trailing figure.
Computing it on your own transactions
XIRR is in every spreadsheet — list the dates in one column, the amounts in another with contributions negative and the current value positive on today's date, and apply the XIRR function. Doing it once by hand is worth the ten minutes, because it makes the dates-matter point concrete in a way reading about it does not.
FNOTrader's Mutual Funds app computes it from the full AMFI NAV history — around 34 million NAV rows — for any scheme and any contribution schedule, reporting XIRR alongside invested-versus-value and maximum drawdown, so the return and the ride are on the same screen. The formulas are published in the user guide.
Common questions
What is the difference between XIRR and CAGR?
CAGR is the constant annual rate that turns one starting value into one ending value, so it is correct for a single investment. XIRR is the annual rate that reconciles cashflows arriving on many different dates, so it is correct for a SIP. For a single cashflow the two produce the same number.
Why is XIRR used for SIP returns?
Because each SIP instalment has been invested for a different length of time. Any method that treats total invested as a single starting amount credits recent instalments with years of growth they never had. XIRR takes each instalment's actual date as an input, so no such assumption is made.
Is absolute return a useful measure?
Only alongside the period, and even then it says nothing about when the money went in. A 40% absolute return on a five-year SIP is arithmetically true and tells you almost nothing, because most of that money was invested for far less than five years.
Can I calculate CAGR on a SIP?
You can compute it, but it will be wrong — it assumes the entire invested amount was present from the first day. The error runs in one direction: it understates the SIP's true return, because it credits the money with more time than it actually had.
Can XIRR be compared with CAGR?
Yes, provided each is applied to the right thing. Both are annualised rates, so a SIP's XIRR and a lumpsum's CAGR over the same period are directly comparable — which is exactly why annualising is worth doing.
What if I paused my SIP or withdrew some money?
That is the case XIRR is built for. Every contribution is a cashflow on its date and every withdrawal is a negative cashflow on its date; the function solves for the single annual rate that reconciles all of them with the current value.
Does a high XIRR mean the fund was easy to hold?
No. An annualised return describes the outcome, not the experience — a 12% XIRR is entirely consistent with having been 30% down at some point. Read it alongside the maximum drawdown, which is what actually determines whether an investor stays invested.
Continue reading
More in Mutual Funds · App: Mutual Funds · Definitions: glossary · Free tools: calculators · All: every article