What are rolling returns?
A single trailing return - “this fund returned X% over 5 years” - is one data point from one start date.
A rolling return recalculates that same annualized return again and again, sliding the start date forward across the fund’s whole history. The result is a range: the worst, median, and best annualized return an investor could actually have gotten for that holding period.
CAGR = (NAVend / NAVstart)1/N − 1
Repeat this for every possible start date in the fund’s history. That full set of CAGR values, not any single one, is the rolling return.
Real example: Parag Parikh Flexi Cap Fund
Rolling CAGR for Parag Parikh Flexi Cap Fund - Direct Plan - Growth, as of 2026-08-28.
CHART
Rolling CAGR by holding period - min, median, max
Each group is one holding-period length. The gap between the bottom and top bar is the range of outcomes an investor could have gotten depending purely on start date.
| Holding period | Min CAGR | Median CAGR | Max CAGR |
|---|---|---|---|
| 1Y | -21.3% | 17.4% | 100.1% |
| 3Y | 0.7% | 19.3% | 37.6% |
| 5Y | 4.2% | 19.1% | 33.6% |
| 7Y | 14.2% | 19.3% | 22.3% |
| 10Y | 17.1% | 19.0% | 21.5% |
The 1-year window swings from -21.3% to 100.1%, almost entirely luck of timing. By 10 years it’s 17.1% to 21.5%, a much tighter range.
How to use it in practice
Rolling returns calculate the return over a fixed period, such as three years, starting on every possible date, and then summarise the results. They show how a fund has done across many start dates, instead of one. Look at the median, the worst and the best results, and at the share of windows that beat the benchmark. A fund whose median is good and whose worst case is bearable is more dependable than one with one great year.
Common mistakes
See Rolling Returns vs Point-to-Point Returns and Calendar-Year Return Bias for worked examples on a real fund.
Rolling windows overlap, so adjacent results are not independent and a long run of good windows may reflect only one good market period. The window length also changes the picture: three-year and five-year rolling returns can tell different stories. Make sure you compare funds over the same window and the same dates.
Frequently asked questions
Why are rolling returns better than point-to-point returns?
Point-to-point returns depend on the start and end dates and can be flattered by a lucky window. Rolling returns show the range of outcomes across many start dates.
What rolling window should I look at?
One that matches your own holding period. For equity funds, three and five years are common.
What does the median rolling return mean?
It is the middle result across all windows, so half of the windows did better and half did worse. It describes the typical experience.
WHAT THIS MEANS FOR YOU A single trailing return tells you what happened to one investor on one timeline. The rolling range tells you what a real range of investors, entering on different dates, actually experienced. Check the rolling spread before judging any fund on one headline number - especially over shorter holding periods, where that spread is widest.
METHODOLOGY Rolling CAGR computed across all valid overlapping windows using the fund’s full NAV history through 2026-08-28, Direct Plan - Growth option. Data: AMFI-disclosed NAV history via Punji. Not investment advice.