How to analyze an ETF the way professionals do
Updated October 9, 2026. Competitor details last checked August 24, 2026.
On ETF Copilot's daily calculation, as of October 9, 2026, the US equity market trades at a P/E of 18.9 — more expensive than 60% of the last 10 years, against a median of 18.1, while Frontier Markets sits 21.6% below its 10-year median. That spread is why analysing a fund has an order: the same multiple on two funds means opposite things depending on what they hold, so what a fund owns has to be established before any number about it can be read.
Professional research runs in that order: what the fund owns, then how good those businesses are, then what you pay for them, then what could go wrong, then what it costs to run. For an ETF that is six steps — holdings, growth, profitability, valuation, concentration, and cost.
Most guides open with the expense ratio. It is the most written-about and least differentiating number in the whole exercise, which is why it comes last here.
The six steps, in order
- 0. Holdings — what the fund actually owns, and how much sits in the top ten.
- 1. Growth — how fast the companies inside it are growing.
- 2. Profitability — whether that growth is being earned or bought.
- 3. Valuation — what you pay for those earnings.
- 4. Concentration and overlap — how much of this you already own elsewhere.
- 5. Cost, size and liquidity — the practical checks.
Each step only makes sense after the one before it. A price-to-earnings ratio tells you nothing until you know what produces it, and a fee comparison tells you nothing until you know the two funds do the same job. Run the order backwards and you get the familiar mistake: two funds picked because they were cheap, which turn out to hold the same forty companies.
Read every number two ways
A single number about a fund is almost never useful. A price-to-earnings ratio of 27 only means something next to something else — so compare it twice:
- Against its own past — where does this sit in the fund's own range over its history?
- Against the market — where does it sit against every other fund?
The two often disagree, and the disagreement is the useful part.
Dividend Global ex-US trades at a P/E of 14.2: more expensive than 73% of its own last 10 years, and cheaper than 83% of the other 29 segments. Of every segment we track, it is the one whose two readings are furthest apart — the same number, read two ways, gives two different answers.
The same reading taken sector by sector is where the two comparisons come apart: Semiconductors is 33.3% above its 10-year median, while Frontier Markets is 21.6% below its 10-year median. A fund is expensive or cheap only relative to something, and which something you pick changes the answer.
Find out what it actually holds
Fund names are marketing. A fund called "innovation", "quality" or "future tech" tells you what the provider wants you to think it holds, not what is in it. A fund that is 37% technology will look expensive and fast-growing however well it is run, because technology is.
Write down two things before anything else:
- The top ten holdings and what percentage each one is.
- The sector split.
Every number after this is read in that light.
Where to find it: free and without an account on Yahoo Finance, Stock Analysis and ETFdb. All three show top-10 weights; ETFdb also publishes top-15. ETF Copilot looks through every holding rather than the top ten — what the ETFs you own actually hold.
Growth
A fund has no revenue of its own. Its growth is the growth of the companies inside it, weighted by how much of each it holds.
Growth comes first because it changes fastest. A fund whose companies were growing at 15% and are now growing at 6% has changed in a way its price may not show yet.
Where to find it: Morningstar computes it at fund level, and ETF Copilot charts it against the market and the fund's own past.
Profitability
Growth without profit is expansion somebody else is funding. Profitability — return on equity, margins — tells you whether the growth in step 1 is being earned or bought.
It moves slowly, so a real change against the fund's own history usually means something structural about what it holds rather than a change in market mood.
Where to find it: ETF Research Center publishes return on equity at fund level, and ETF Copilot charts it on every fund page.
Valuation
Only now does the price matter. A fund's price-to-earnings ratio is the weighted average of its holdings', and a high one is not a problem if steps 1 and 2 explain it. Expensive-and-growing is a different fact from expensive-and-slowing.
Against peers is easy to find free. Against the fund's own history is the one almost nothing publishes — and usually the more useful, because a whole peer group can get expensive together, which a peer comparison cannot show you.
Where to find it: peers free on ETFdb and ETF Research Center. History: Koyfin charts it on a paid plan. ETF Copilot charts a fund's own valuation history against its category and the whole US equity ETF universe on every fund page.
Concentration and overlap
"Diversified" is a word, not a measurement. A fund holding 500 companies can be more concentrated than one holding 50, if its top ten are 37% of it. Read the top-ten weight, not the number of holdings.
Then check it against what you already own. Two funds with different names and different providers routinely share more than half their holdings. Hold both and you are not diversified — you are doubled up, paying two fees. It is the most common mistake in a retail ETF portfolio, and it is invisible unless you look for it.
Where to find it: top-10 weight free on Yahoo Finance, Stock Analysis and ETFdb. For overlap, ETF Research Center's free tool is the deepest available. ETF Copilot publishes concentration and overlap across the market.
Cost, size and liquidity
These come last on purpose. Among funds that pass the first four steps, the fee gap is usually tiny next to the differences you have already found.
Three checks, then stop thinking about them:
- Is the fee far out of line with near-identical funds?
- Is the fund large enough not to be at risk of closing?
- Does it trade enough that the spread will not cost more than the fee saves?
Where to find it: everywhere. Expense ratio, assets and average volume are on every free site, and on the fund provider's own page.
ETF analysis tools
5 complete, with context4 complete, context partial or gated3 present but bare2 partial or indirect1 absent— blocked, not verifiable
| Tool | Price /mo | Fund profile cost, AUM, volume | Holdings what it owns | Concentration step 4 | Metrics vs market steps 1–3 | Metrics vs own history steps 1–3 |
|---|---|---|---|---|---|---|
| ETFdb / VettaFi | Not published | 5 | 4 | 5 | 4 | 1 |
| Morningstar Investor | $34.95 | 5 | 4 | 4 | 4 | — |
| ETF Copilotthis site | $20.00 | 4 | 4 | 5 | 5 | 5 |
| Koyfin | $39.00 | 4 | 3 | 2 | 4 | 4 |
| ETF Research Center | $29.00 | 3 | 4 | 4 | 4 | 1 |
| Stock Analysis | $9.99 | 5 | 4 | 4 | 1 | 1 |
| Yahoo Finance | $7.95 | 5 | 4 | 3 | 2 | 1 |
| Google Finance | Free | 2 | 1 | 1 | 1 | 1 |
Capability by step, checked on each provider's own site, monthly price where published.
Cost, size, holdings and concentration — steps 0, 4 and 5 — are well served almost everywhere, and mostly free. If those are all you need, several tools here will do.
Steps 1 to 3 are where the table separates: whether a tool can put a fund's growth, profitability and valuation next to the market, and next to the fund's own past.
What order do professional investors analyze a fund in?
Exposure first, then business quality, then price, then risk, then cost. For an ETF that means holdings, then growth and profitability, then valuation, then concentration and overlap, then fees and liquidity. Holdings come first because every later number is read in the light of what the fund owns; cost comes last because among funds that pass the first four checks it rarely decides anything.
Is a low expense ratio the most important thing?
It is the most written-about and the least differentiating. Between two funds tracking similar exposures the fee gap is usually a few hundredths of a percent, while the differences in what they hold and what they overlap with can be very large. Treat cost as a bar to clear, not a way to choose.
How many holdings should an ETF have?
The count barely matters; the weights do. A 500-holding fund with 37% in its top ten behaves more like a concentrated fund than a 50-holding fund with 12% in its top ten. Read the top-ten percentage.
Can I just look at past performance?
Past returns tell you what happened, not what a fund holds now. Holdings change, index rules change, and the companies inside change. The six steps above describe the fund as it stands today; a return series describes a fund that no longer exactly exists.
What does ETF Copilot score a fund on?
Four factors, each on a 0–10 scale where higher is better: Value, Growth, Quality and Risk. Each is built from the aggregated fundamentals of what the fund holds, measured against the fund's own history and against the rest of the US equity ETF universe — which is steps 1 to 3 of this method, updated daily. How the scores are built.
Run these six steps on any fund →
Nothing on this page is investment advice, a recommendation, or a forecast of returns.