How should I compare two mutual funds, and which measures actually decide it?

Three ways a comparison goes wrong before it starts

Most fund comparisons are invalid for one of three boring reasons, and no amount of analysis afterwards can rescue them:

• Different windows. Fund A's 5-year number against Fund B's 3-year number is not a comparison. Different windows contain different markets.
• Different plans. A direct plan against a regular plan builds in a handicap of roughly 0.75 points a year that has nothing to do with the manager. Compare direct with direct.
• Different benchmarks. A mid-cap fund beating the NIFTY 100 has not shown you anything. Each fund must be judged against the index that matches its mandate — and against a TRI, which counts dividends.

Get those three right and the comparison is already better than most.

What to compare, in order of how much it tells you

1. Rolling returns, not point-to-point. The worst, median and best over every window — plus how often a window ended negative. This answers "what does holding this normally feel like", which one CAGR cannot.
2. Risk-adjusted return. Sharpe and Sortino, over the same window with the same risk-free rate. Two funds with the same return and different Sharpe ratios are not equally good.
3. Maximum drawdown and recovery time. The deepest fall, the dates it happened between, and how long it took to get back. This is the number at which people actually sell.
4. Up and down capture. How much of the benchmark's rises and falls each fund took. Two funds with identical returns can have completely different shapes.
5. Portfolio overlap. If you are choosing between them, this matters less; if you are considering holding both, it decides it. Two funds at 70% overlap are one fund with two fee structures.
6. Sector and market-cap exposure. Where each fund actually is, from its disclosed holdings, rather than what its name implies.

Expense ratio belongs on this list too — but it is not in our data. Take it from the AMC factsheet, and remember the cheaper fund has to be beaten by the dearer one every year, not just once.

Comparing debt and hybrid funds

Debt funds need a different list, and applying the equity one to them produces nonsense. Credit quality, average maturity and duration decide a debt fund's behaviour; its beta against an equity index is noise. That is why debt, gilt, liquid and arbitrage funds here carry no equity benchmark, no beta and no alpha at all. Printing those numbers would make the page look more complete and the analysis less true.

Hybrid funds sit between, and the honest comparison starts with the equity share. A fund holding 70% equity and one holding 25% are not competitors whatever the category label says — compare them on the blend they actually hold, which the disclosed portfolio shows.

Comparing fund managers

A manager's record is attached to the funds they ran, over the periods they ran them — not to the fund's whole history. When a manager changes, the track record above the change belongs to someone else. We publish the measures for the scheme and state the window; attributing them to a person is a judgement we leave to you, because the data does not carry a clean manager timeline.

What you can do is check whether the fund's behaviour changed after the handover — drawdown, capture ratios and sector concentration before and after will tell you more than any profile.

Comparing three or more at once

Yes — and with three or more, overlap becomes the most interesting output rather than a footnote. Compared pairwise, a shortlist of four "diversifiers" frequently turns out to be two ideas. On the compare page every pair's overlap is reported alongside the stocks they share.

What a comparison will never produce here is a winner. You get the measures side by side with their windows stated; the ranking is yours to make. MF Analyser is not a SEBI-registered Investment Adviser or Research Analyst, and a "winner" badge is a recommendation however it is drawn.