Rolling Returns vs Point-to-Point Returns — Why Fund Comparisons Can Mislead
"This fund delivered 18% annualised returns over the last 5 years." It's the kind of headline number that decides investment decisions every day — and it's also one of the easiest numbers to unintentionally (or deliberately) cherry-pick, precisely because of how it's calculated. Understanding the difference between point-to-point and rolling returns is what separates a genuinely informed comparison from one that's quietly been shaped by which exact dates were chosen.
What Point-to-Point Returns Actually Are
A point-to-point return measures a fund's performance between exactly two fixed dates — typically "today" and "X years ago today." It's the standard figure shown on nearly every factsheet, app, and rating platform: 1-year, 3-year, 5-year, and 10-year returns, each calculated from a single starting date to a single ending date.
The problem is that this single-window measurement is extremely sensitive to when exactly those two dates happen to fall. A fund's 5-year point-to-point return calculated as of today could look dramatically different from the same fund's 5-year return calculated just three months ago or three months from now, purely because of what the market was doing on those specific start and end dates — not because anything about the fund's actual investment quality changed.
A Concrete Illustration of the Distortion
Imagine a fund whose 5-year window happens to start right at a market bottom (say, just after a sharp correction) and end near a market peak. Its point-to-point 5-year return will look spectacular — not necessarily because the fund manager did anything unusually skillful, but simply because the starting point captured a market low and the ending point captured a market high. Shift that same 5-year window back by even six months, and the return figure can look meaningfully more ordinary, because the "lucky" starting point is no longer included.
This is exactly why fund houses sometimes time their marketing material around whichever specific trailing period happens to show the most flattering number — it isn't necessarily dishonest, but it does mean a single point-to-point figure can genuinely misrepresent how consistently a fund has actually performed.
What Rolling Returns Actually Measure
Rolling returns solve this by calculating the return for every possible window of a given length across the fund's entire history, rather than just one fixed window. For example, a 5-year rolling return analysis would calculate:
- The 5-year return from Jan 2015 to Jan 2020
- The 5-year return from Feb 2015 to Feb 2020
- The 5-year return from Mar 2015 to Mar 2020
- ...and so on, moving forward by a small increment (commonly one day or one month) each time, generating hundreds of overlapping 5-year return observations across the fund's full history. Instead of one single number, this produces a distribution of returns — showing the average, the best case, the worst case, and how consistently the fund performed across every conceivable entry and exit point, rather than just the one specific pair of dates a factsheet happened to highlight.
Why This Distinction Genuinely Matters for Decision-Making
- Consistency over luck — a fund with a strong point-to-point 5-year return but a wide, inconsistent rolling return distribution (great in some windows, mediocre or negative in others) is a fundamentally different, riskier proposition than a fund with a similar average return but a tight, consistently positive rolling return distribution.
- Comparing funds fairly — when comparing two funds' point-to-point returns, you're implicitly assuming both investors happened to enter and exit at the same two dates — an assumption that rarely holds true for real investors, who enter and exit on whatever date suits their own circumstances, not the dates that happen to flatter a particular fund's factsheet.
- Realistic expectation-setting — rolling returns give a far more honest sense of "what range of outcomes should I actually expect if I invest for this duration," since it's built from many real historical entry-exit combinations rather than one cherry-picked pair.
How to Actually Use Rolling Returns
When evaluating a fund seriously, look for (or calculate) its rolling return statistics over relevant holding periods — commonly 3-year, 5-year, and 7-year rolling windows for equity funds. Key things to check:
- Average rolling return — the central tendency across all windows, generally more reliable than a single point-to-point figure.
- Minimum rolling return — the worst-case outcome an investor would have experienced entering at the worst possible time within that history; this tells you the realistic downside even for a "long-term" holding period.
- Consistency of outperformance versus benchmark — what percentage of rolling windows the fund actually beat its benchmark, rather than just whether it beat the benchmark once, over the one specific window shown on the factsheet.
Why This Matters
A single point-to-point return is a snapshot; a rolling return distribution is closer to a full picture. For advisors and investors comparing funds, or evaluating whether a fund's headline "5-year return" genuinely reflects consistent skill rather than a fortunately timed measurement window, rolling returns are the more rigorous — and considerably harder to game — tool for the job.