Rolling returns: a fairer way to judge a fund

Rolling returns measure a fund's return over every possible window of a fixed length across its history — for example every 3-year period, rolled forward day by day. Instead of one number that depends on today's date, you see the full distribution of outcomes an investor could have experienced.

Why trailing returns can mislead

A fund's "3-year return = 18%" depends entirely on the start and end dates. Pick a different month and it could be 9% or 25%. If the window happens to start at a market bottom, the fund looks brilliant; start it at a peak and the same fund looks mediocre. Fund marketing naturally gravitates to whichever date range flatters. Rolling returns remove this luck by averaging across hundreds of overlapping windows — every possible entry date gets counted, including the unlucky ones.

A worked example

Consider two flexi-cap funds, each showing a trailing 3-year return of 16%. Roll the 3-year window across a decade of NAV history and the picture splits: Fund A's rolling 3-year returns average 14.5%, with a worst window of +6% and a best of +22%. Fund B also averages 14.5% — but its worst window was −4% and its best +34%. An investor who entered Fund B at the wrong time sat on losses after three full years; no Fund A investor ever did. Same trailing number, completely different reliability. That is the information a single date-stamped return hides.

What to look for

• Average rolling return — the typical outcome over that horizon.
• Minimum & maximum — the worst and best you'd have seen. A small gap means consistency.
• % of windows above a target — e.g. "beat 12% in 85% of 5-year periods."

How to use rolling returns when picking funds

Match the window to your goal: judge a fund you'll hold for five years on 5-year rolling windows, not 1-year ones. Within a category, prefer the fund whose minimum rolling return is highest — it protected its unluckiest investor best. Be careful with young funds: a fund with only four years of history has very few independent 3-year windows, so its rolling statistics are thin evidence. And remember that a fund with slightly lower average returns but far higher consistency is often the better long-term hold, because consistency is what keeps you invested through rough patches. As always, history is evidence, not destiny — mutual fund investments are subject to market risks.

→ Dhanik shows rolling-return distributions on every fund page, and you can backtest a fund's real SIP across different start dates in the Backtest tab.