A mutual-fund return table becomes readable only after you identify what each number measures. Absolute return, compound annual growth rate (CAGR), a point-to-point annualised return and a cash-flow-based result do not share one denominator. Record the exact scheme, plan, option, dates, NAV or cash-flow series, benchmark label and calculation method before comparing anything.
Direct answer (52 words): Treat every return figure as the output of a defined data recipe. First freeze the investment identity and observation dates. Then identify whether the inputs are two NAVs, a sequence of periodic NAVs, or dated investor cash flows. Reproduce only the stated formula, preserve rounding, and leave missing definitions unresolved rather than guessing.
Reviewed by GFS Research Desk.
Why two correct percentages can disagree
A reader may see one percentage in a factsheet, another on an account screen and a third in a household spreadsheet. The instinct is to ask which one is wrong. Sometimes none is wrong: the figures may answer different questions.
A point-to-point scheme return usually describes the movement between two values in a specified series. CAGR expresses that movement as a compounded annual rate over a stated span. An absolute return leaves the whole-period change unannualised. A cash-flow-based calculation uses the dates and amounts of several contributions or withdrawals rather than only a starting and ending NAV. Rolling returns repeat a chosen holding-period calculation across many start dates; they are another dataset again.
The label matters as much as the percentage. A table that does not reveal its series, dates, method and assumptions cannot be reproduced from the displayed number alone. It should be marked “definition missing,” not silently translated into a familiar metric.
This article was checked on 29 July 2026. It uses no live scheme return, current ranking or tax rate. SEBI’s mutual-fund master circular is the official regulatory document reviewed for the disclosure framework. AMFI’s NAV History and Other Data pages are official industry routes for dated published data. A current AMC scheme document, factsheet and disclosure remain necessary for any named scheme.
Start with the return identity card
Before touching a calculator, copy these fields from the source:
| Field | What to record | Error it prevents |
|---|---|---|
| Scheme identity | Exact scheme, plan and option | Mixing direct and regular plans or different options |
| Series identity | NAV series, benchmark series or investor cash flows | Comparing a product series with a personal experience |
| Start and end | Exact dates and values used | Replacing stated observations with nearby dates |
| Period length | Days, months or years, as defined | Annualising with an unstated convention |
| Method label | Absolute, CAGR, annualised, rolling or cash-flow based | Treating unlike outputs as interchangeable |
| Benchmark label | Exact benchmark name and return-series label shown | Comparing to a different index construction |
| Source and as-of date | Document, page/table and publication date | Treating an old table as current |
| Notes | Reinvestment, rounding and unavailable fields | Hiding assumptions inside arithmetic |
Do not infer a plan or option from a shortened scheme name. Do not replace the table’s dates because another source has newer data. Newer data creates a new calculation; it does not reproduce the old one.
Four calculations that answer different questions
1. Absolute return: whole-period change
For a simple two-value illustration, absolute return is:
`(ending value / starting value - 1) × 100`
If a hypothetical series moves from 10 to 12, the whole-period change is 20%. The illustration says nothing about how long the move took until dates are added. Calling it an annual figure without the period would add information that the inputs do not contain.
2. CAGR: a constant annual-equivalent path
For a span expressed as `n` years, the mechanical form is:
`[(ending value / starting value)^(1/n) - 1] × 100`
CAGR compresses a start value, end value and time span into one compounded annual-equivalent rate. It does not say that the series actually earned that rate each year. It hides the path between the endpoints, including volatility and drawdowns. The time convention and precise dates therefore belong beside the output.
3. Point-to-point annualised return: method still matters
A table may label a result “annualised” for a particular period. That label is not permission to invent the day-count, observation rule or rounding method. Use the methodology supplied with the table. If the methodology is absent, record the output as published and state that independent reproduction is limited.
4. Cash-flow-based return: the investor ledger is the input
When money enters or leaves on several dates, a start-NAV/end-NAV calculation does not represent those dated cash flows. A cash-flow-based result depends on every amount, direction and date included. Missing one instalment, switch, withdrawal or valuation date can change the output. It is a household-record calculation, not automatically the scheme’s published point-to-point return.
That difference is not a defect. Scheme performance and an investor’s experience are different questions. Neither should be relabelled to force agreement.
NAV is an input, not the complete interpretation
AMFI’s NAV History page provides an official route to dated NAV records. A downloaded NAV does not by itself establish the complete return methodology. The reader must still match the scheme, plan, option and dates, and must check whether a published table applies any stated convention.
A NAV series also does not answer why a return differed from a benchmark. Expenses, portfolio implementation, cash, tracking, timing and other factors may be relevant in particular contexts, but the return table alone does not establish causation. A difference is an observation first. Causal claims need separate evidence.
Do not mix a scheme NAV series with a benchmark level merely because both have dates. Confirm the benchmark identity and whether the published comparison uses the stated benchmark return series. If the label is abbreviated, consult the current factsheet, Scheme Information Document (SID), Key Information Memorandum (KIM) and relevant addenda.
A six-step audit for any published table
- Save the source. Record the document title, publisher, page or table, access date and as-of date. A cropped image without notes is incomplete evidence.
- Freeze identity. Match the exact scheme, plan and option. Keep benchmark and additional-benchmark labels separate where the source does so.
- Classify the input. Is the figure based on two NAVs, repeated observation windows, a benchmark series or dated household cash flows?
- Copy the method. Preserve the source’s metric label, period convention, reinvestment assumption and rounding note. Write “not stated” for absent fields.
- Reproduce cautiously. Use the formula only when all required inputs and conventions are available. Show intermediate values and do not “fix” a small difference by changing dates.
- Explain the limit. State what the result cannot determine: future return, future risk, liquidity at a required date, tax outcome or personal suitability.
A small reproduction gap may come from rounded displayed values, a date convention or a different underlying precision. That is a prompt to inspect the source notes, not proof of error or misconduct.
What can be compared—and what should remain separate
Two figures can be placed side by side educationally when they use the same scheme identity, comparable series definitions, the same observation period and disclosed methods. Even then, the comparison is descriptive. It does not become a quality score.
Keep figures separate when one is absolute and the other annualised; one is point-to-point and the other rolling; one is a scheme series and the other uses personal cash flows; plans or options differ; benchmark labels differ; or the as-of dates do not align. A blank comparison cell is more accurate than a fabricated bridge.
Return is also only one dimension of a document review. It does not replace the scheme objective, asset-allocation framework, Riskometer, expense disclosure, portfolio snapshot, liquidity terms or tax review. A higher historical figure does not resolve those other questions.
Honest limitations
This method can organise evidence and reveal denominator mismatches. It cannot validate an unpublished calculation, recover missing cash flows, identify a person’s tax treatment, predict a scheme’s path or decide whether any scheme fits a household. Official pages, document formats and methodologies may change. Recheck the current scheme material and applicable official disclosure before acting on a live question.
Frequently asked questions
Is absolute return always wrong for periods longer than one year?
No. It can accurately describe total change over the stated period. The problem is presenting it as though it were an annual rate or comparing it with CAGR without preserving the labels and period.
Does CAGR show the return earned in every year?
No. CAGR is a compounded annual-equivalent summary between endpoints. The actual path can vary substantially between those dates.
Why can my account result differ from a factsheet return?
Your result may include dated contributions, withdrawals or other transactions, while the factsheet may show a defined scheme return for a standard observation period. Check both input recipes before interpreting the gap.
Can I calculate a scheme return from two NAVs?
You can calculate a simple point-to-point change when the exact series and dates are known. Do not assume that it reproduces a published figure whose methodology or conventions are different or unstated.
Should I compare direct-plan and regular-plan return rows?
First preserve each row’s exact plan, option, dates and disclosure context. A numerical comparison is incomplete without those identities and does not produce a personal action conclusion.
What should I do when the benchmark label is unclear?
Record the label exactly and consult the current factsheet, SID, KIM and addenda. Do not substitute a similarly named index or series.
Does the highest historical return identify the strongest scheme?
No. A historical return number is not a complete assessment of mandate, risk, expenses, portfolio, liquidity, taxes or future outcomes, and this article does not rank schemes.
Useful GFS routes
> Mutual fund investments are subject to market risks. Read all scheme-related documents carefully.
> This content is educational and is not investment advice or a recommendation. Verify independently before acting.
> Past performance is not indicative of future returns.