What we do
We take each fund's holdings and the weight of each one, then combine the underlying companies' fundamentals in those proportions. That produces a portfolio-level set of figures for the fund: trading multiples, profitability, growth expectations, and how the weight is spread. Those figures are what we score, compare and rank, and the same holdings drive overlap, concentration and exposure.
The look-through
Each holding is matched to a company, and that company's figures enter the fund's totals at the weight the fund holds it. Lines we cannot match — cash, derivatives, unlisted or unmapped positions — are left out rather than estimated, and we track how much of a fund's weight is actually covered. A fund with too little covered weight is not scored.
The four scores
Each fund carries four scores from 0 to 10, ranked within its category and mapped to that scale. Higher is better in all four, Risk included. They are independent: a fund can print 9 on Quality and 2 on Valuation.
Value
EV/EBITDA and P/E, each on trailing (LTM) and forward (NTM) figures, weighted across the holdings.
Growth
Annualized analyst consensus earnings growth of the holdings, plus dividend yield.
Quality
Return on equity, return on invested capital, and EBITDA margin.
Risk
3-year volatility and risk-adjusted return, plus the largest and average 5-year declines.
Two reference frames
Each valuation metric is measured twice: against the fund's category, and against the fund's own history. A semiconductor fund at 25× forward earnings is expensive against the broad market and ordinary against its peers. The historical comparison is reported as a percentile — the share of the fund's past that traded cheaper than it does now — so a fund can read inexpensive against peers and expensive against itself. Both numbers appear on the page.
Key outputs
Comparison and overlap
Symmetric overlap between two funds, plus the directional share: how much of A sits inside B, by weight and by count, with peer pairs ranked. We publish percentages, counts and ranks, never a constituent list. How we measure overlap.
Concentration
Top-10 weight, and how many holdings are needed to reach half the fund's weight. A fund can hold 500 names with most of its weight in ten, which a holdings count alone will not show. How we measure concentration.
Exposure
Theme, size and single-stock weight computed from the holdings: how much of a fund is semiconductors, or one megacap, rather than what its name implies. How we measure exposure.
Daily Ideas
A daily screen across the covered universe, led by the fundamental scores with flows and price action as context. Each selection is published with the figures that drove it.
Ranked lists
A ranked list sorts qualifying funds by one of the four scores, or by a combination of them where the list expresses a broader idea such as a core holding or income. Lists rebuild daily with the board, and no placement is for sale. See the live lists.
Data and freshness
Holdings, prices and company fundamentals come from our market-data provider. The portfolio-level figures, scores, percentiles and rankings are our own work on top of that data. The board rebuilds daily, and every page carries the date of the data behind it.
Coverage and limits
Roughly a thousand US-listed equity ETFs. Bond, commodity and currency funds are out of scope, since a company-level look-through does not describe them. A fund needs at least a year of trading history and enough of its holdings matched before it is scored. Below that, the page shows no score rather than one built from part of the book.
What we don't do
No buy, sell or hold calls. No price targets, no return forecasts. No issuer pays for coverage, placement or a ranking position. We state what each figure is, what it was measured against, and when it was computed.
Scores are research support, not investment advice. Past performance does not guarantee future results.