How “cheaper than 80%” is actually calculated (advanced)
"Cheaper than 80% of similar ETFs" sounds precise. The honest version has two important qualifiers hiding behind that one number: similar to *what*, exactly — and how many funds is '80%' actually counted out of?
What the chip is actually claiming
A How this fund ranks against other funds that follow a similar strategy — “cheaper than 80%” means 4 out of 5 similar funds cost more than this one. More → chip like "cheaper than 80% of similar ETFs" is a plain relative-rank claim: line this fund up against its peer group on one metric — cost, in this example — and 4 out of every 5 of those peers cost more. It says nothing about whether the fund is good value in absolute terms, only where it sits among funds doing a similar job. That distinction matters more than it sounds.
How the peer group is actually built
The peer group is the closest set of funds we can identify sharing the same broad strategy and asset class — not every ETF on the platform, and not an official index-provider category. Two consequences worth knowing: it's an honest, computed comparison, not a licensed benchmark segment (the same distinction our How closely the fund’s actual return has matched the index it targets — a small gap is normal; a bigger one means it strayed further from what it is meant to copy. More → figures make); and "similar" is a strategy-and-asset-class match, not a guarantee the peers are equally good substitutes for each other in every other way (currency, domicile, replication can still differ within a peer group).
The math, in plain terms
Once the peer group is set, the percentile itself is simple: for a cost metric, count what share of the peer group this fund undercuts; for a return metric, what share it beats. "Cheaper than 80%" on a peer group of 50 funds means roughly 40 of those 50 cost more than this one. There's no smoothing or weighting hidden in that number beyond the plain rank — which is also exactly why the size of the peer group matters so much to how much weight the percentile can bear.
Where it gets shakier — small peer groups
A percentile computed against 300 broad world-equity funds is a fairly stable read. The same style of percentile computed against 6 funds in a niche strategy swings hugely if just one or two of those funds are unusual — one outlier can move a fund from the 40th to the 80th percentile without the fund itself changing at all. Treat a percentile on a thin peer group as a much softer signal than the same-looking number on a deep one; we don't currently surface the peer-group size alongside the chip, which is worth knowing when you're weighing how much to lean on it.
Using it well
The practical rule: a percentile is a second opinion, not a first one. Look at the fund's own absolute number first — its actual The yearly running cost of the fund, shown as a % of your money. €0.20 per €100 a year at 0.20%. Lower is cheaper. More → , its actual return — then use the percentile to sanity-check whether that number is ordinary or unusual for what the fund is trying to do. A fund that's cheap in absolute terms but only mid-pack on its percentile just means its whole peer group is cheap; that's still useful context, not a contradiction.