Three metrics. Three different truths.
Every mutual fund advertisement quotes a return number. But which return? Calculated how? From which date? Most retail investors — and even many advisors — use CAGR, XIRR and rolling returns interchangeably, not realising that each metric answers a fundamentally different question.
Using the wrong metric can mean the difference between an apparently stellar 24% return and a sobering 6% — for the exact same fund, the exact same exit date, just a different entry point.
This report explains each metric clearly, demonstrates the pitfalls with real data, and gives you a decision framework for when to use which number.
CAGR
The smoothed annualised growth rate from a single start date to a single end date. Ideal for lump sum investments. Hides all volatility and path dependency.
XIRR
The internal rate of return adjusted for multiple cash flows at irregular dates. The only correct metric for SIP investors. SEBI-mandated for SIP return disclosures.
Rolling Returns
Not a single number — a distribution of outcomes across all possible entry points for a given holding period. Reveals consistency, not just average performance.
Compound Annual Growth Rate
CAGR represents the smoothed rate at which an investment would have grown if it had compounded at a constant rate every year. It assumes a lump sum invested on Day 1 and withdrawn on the last day, ignoring everything in between.
Example
You invest ₹1,00,000 in a fund on 1 January 2019. On 1 January 2024, the corpus is ₹1,76,234. That is a 5-year CAGR of exactly 12% — regardless of whether the fund went up steadily or crashed 40% in 2020 and recovered sharply.
"CAGR is like a flight's average speed. If the journey took 5 hours at an 'average' 600 km/h, it doesn't tell you that the plane was grounded for two hours in Mumbai due to fog."
The start-date trap
CAGR's biggest vulnerability is that it can be manipulated simply by choosing a favourable start date. Two investors in the same fund will report wildly different CAGRs depending solely on their entry point. This is the mechanism behind most misleading mutual fund advertisements.
Extended Internal Rate of Return
XIRR is CAGR's smarter sibling. It handles multiple cash flows at different dates — exactly what happens in a Systematic Investment Plan (SIP), or when you make partial redemptions or top-up investments. XIRR solves for the single annualised discount rate that makes the Net Present Value of all dated cash flows equal to zero.
Technically, XIRR is the solution to this equation, computed iteratively. Excel, Google Sheets, and every online SIP calculator solve it for you. But understanding what it represents is critical for correct interpretation.
Why XIRR > CAGR for SIP investors
Consider a ₹10,000/month SIP for 5 years. The first instalment compounds for 60 months; the last instalment barely compounds at all. CAGR on a "total invested" basis completely misrepresents this asymmetry. XIRR accounts for it precisely by tagging each cash flow to its exact date.
Key insight: rupee cost averaging lifts XIRR above lump sum CAGR
In a fund with significant volatility, the SIP investor's XIRR often exceeds the lump sum investor's CAGR. The reason: SIP investors automatically buy more units when the NAV is depressed (market crashes) and fewer when it is elevated. This rupee cost averaging effect is invisible in CAGR but fully captured in XIRR.
Rolling Returns
Rolling returns do not give you one number — they give you a distribution of outcomes. You choose a holding period (say, 3 years) and compute the CAGR for every possible 3-year window in the fund's history: Jan 2010 to Jan 2013, Feb 2010 to Feb 2013, and so on daily until the last possible window.
The result is hundreds or thousands of data points that collectively answer the question: "Across all market cycles, how has this fund performed for investors who stayed invested for 3 years?"
"A fund with an average 3-year rolling return of 12% and 96% positive windows is a fundamentally different proposition from a fund showing 15% trailing CAGR from a cherry-picked start date."
The key outputs from rolling return analysis are: the minimum return (worst case for a patient investor), the median return (typical experience), the percentage of windows with positive returns, and the beat rate against the benchmark. These four numbers tell you far more than any single CAGR.
Nippon India Large Cap Fund — A Three-Lens Analysis
Launched on 8 August 2007, Nippon India Large Cap Fund is one of India's most tracked large-cap equity funds, benchmarked against the Nifty 100 TRI. It has grown from ₹8,676 crore AUM in March 2020 to over ₹51,690 crore by 2026, making it an ideal real-world specimen for this analysis.
Fund snapshot as of May 2026
Managed by Sailesh Raj Bhan (since August 2007) and Bhavik Dave (since August 2024). Invests ≥80% in Nifty 100 large-cap stocks. Expense ratio: 0.71% (direct plan).
Lens 1: CAGR — The marketing number
The table below shows how dramatically CAGR shifts depending on the entry date, despite the same May 2026 exit date at ₹~88.9 NAV (Regular plan). This is the start-date trap in its purest form.
| Entry point | Approx. NAV | Holding period | CAGR to May 2026 | Context |
|---|---|---|---|---|
| Aug 2007 — inception | ₹10.00 | ~19 years | 13.3% | Inception CAGR |
| Jan 2014 | ₹28.00 | ~12 years | 10.5% | Decent entry |
| Mar 2020 — COVID low | ₹29.00 | ~6 years | ~24% | Cherry-picked bottom |
| Jan 2022 — near peak | ₹74.00 | ~4 years | ~6% | Peak entry |
Lens 2: XIRR — What the SIP investor actually earned
A monthly SIP of ₹10,000 over 10 years (₹12 lakh total invested) in this fund grew to ₹29.78 lakh — an XIRR of 17.35%. The same SIP in the benchmark (Nifty 100 TRI) grew to only ₹27.01 lakh, or 15.54% XIRR.
Notice that the 10-year SIP XIRR of 17.35% is significantly higher than the lump sum CAGR since inception of 13.3%. The reason is rupee cost averaging: SIP investors who continued their ₹10,000/month through the March 2020 crash accumulated units at drastically depressed prices, which supercharged their eventual returns as markets recovered. CAGR cannot show this — XIRR captures it perfectly.
Lens 3: Rolling Returns — The analyst's tool
The table below shows 3-year rolling return windows for the fund versus the Nifty 100 TRI across different market periods. This is the data that should drive fund selection decisions, not a single trailing CAGR.
| 3-Year window | Fund CAGR | Nifty 100 TRI | Alpha | Outcome |
|---|---|---|---|---|
| Jan 2013 → Jan 2016 | 18.2% | 16.1% | +2.1% | Beat benchmark |
| Jan 2015 → Jan 2018 | 9.6% | 9.8% | −0.2% | Marginal miss |
| Jan 2017 → Jan 2020 | 11.4% | 10.2% | +1.2% | Beat benchmark |
| Mar 2020 → Mar 2023 | 28.4% | 24.6% | +3.8% | Strong alpha |
| Jan 2021 → Jan 2024 | 14.9% | 13.1% | +1.8% | Beat benchmark |
| Jan 2022 → Jan 2025 | 9.2% | 10.4% | −1.2% | Underperformed |
Source: Based on publicly available NAV data and CRISIL research. Rolling returns are illustrative approximations for educational purposes.
The fund beat its benchmark in 5 of 6 rolling windows shown — a solid consistency record. The one underperformance (2022–2025) coincides with a period when mid and small-cap stocks dramatically outperformed large-caps as a category — making this a category headwind rather than a fund management failure. This nuance is invisible in trailing CAGR but fully visible through rolling return analysis.
Which number should you trust?
The answer depends entirely on what question you are trying to answer. All three metrics are correct — they just answer different questions.
Evaluating a lump sum investment → use CAGR
CAGR is the right tool when you have a single start date and a single end date. Use it to compare how a fund has performed versus a benchmark or a peer fund over the same defined period. Just ensure both comparisons use the same start and end dates.
Evaluating a SIP or any multiple-cash-flow investment → use XIRR
XIRR is the only mathematically correct metric when cash flows occur at irregular dates. For every SIP portfolio statement, XIRR is the number that represents your actual annualised return. Comparing XIRR to the benchmark's XIRR tells you whether your SIP in this fund beat the index fund option.
Selecting which fund to invest in → use Rolling Returns
Rolling returns are the decision-making tool. When comparing Nippon India Large Cap vs ICICI Prudential Bluechip vs Mirae Asset Large Cap, don't just compare 5-year trailing CAGR. Compare the median 3-year rolling return, the percentage of positive 3-year windows, and the consistency of benchmark-beating across different market regimes.
Interpreting fund advertisements → apply healthy scepticism
Fund advertisements use CAGR almost exclusively because it is the most manipulable metric. The single most important habit a retail investor can develop: whenever you see a CAGR, ask "from which date?" Then check the rolling returns from a data platform like Value Research, AdvisorKhoj, or Moneycontrol to see whether that performance is durable or coincidental.
"The three questions every investor should ask before trusting a return number: What was the start date? What happens to this return if I change the start date by 6 months? And how often has this fund delivered positive returns over my intended holding period?"
The exam question
For finance students: "A fund manager presents a 5-year CAGR of 22%. List three follow-up questions." The answers are: (1) What is the exact start date — does it coincide with a market bottom? (2) What is the 3-year rolling return distribution — median, minimum, and percentage of positive windows? (3) For a systematic investor, what would the XIRR have been over the same period?
These three questions cannot be answered with a single trailing CAGR. They require all three lenses working together.
Summary comparison
| Attribute | CAGR | XIRR | Rolling Returns |
|---|---|---|---|
| What it measures | Lump sum annualised return | Return on any set of cash flows | Distribution of returns across all entry points |
| Cash flows handled | Single invest + single exit only | Multiple, irregular dates | Each window is a separate CAGR calculation |
| Output | Single percentage | Single percentage | Range: min, median, max, % positive |
| Best used for | Lump sum evaluation, benchmark comparison | SIP return measurement, personal portfolio return | Fund selection, consistency analysis, advisor due diligence |
| Main weakness | Start-date sensitive; ignores path | Does not show consistency; depends on cash flow timing | Requires longer data history; more complex to compute |
| SEBI requirement | Point-to-point performance | Mandatory for SIP return claims | Not mandated; used by analysts and researchers |
| Nippon Large Cap (real data) | 13.3% since inception | 17.35% (10-yr SIP) | Beat benchmark in 5/6 rolling 3-yr windows |