Blog · ETF fundamentals

How to Analyze an ETF: 7 Metrics That Actually Matter

Cognitor · EN

Most investors buy the story, not the structure — and that gap is where surprises live. Before you add any ETF to a watchlist, you need a repeatable checklist: expense ratio, liquidity, index-tracking quality, true portfolio concentration, income treatment, currency exposure, and — for synthetic structures — counterparty and collateral quality. Running all seven takes less than an hour and eliminates a category of mistakes that headlines never warn you about.

1) Expense ratio (TER): the visible but incomplete scorecard

TER is the annual headline fee expressed as a percentage of assets — deducted daily from net asset value, not billed separately. For two funds tracking the same index with similar structure, the lower TER almost always wins on cost alone. SPY (0.0945%) versus a comparable S&P 500 product at 0.20% looks marginal on day one, but on a $100,000 position held for 20 years that gap compounds to roughly $2,300 in additional drag, before any tracking difference.

The trap is comparing TER across different mandates. A 0.75% emerging-markets fund and a 0.03% U.S. large-cap fund serve completely different risk and geographic exposures — TER comparisons only carry weight inside the same index and structure category. Also note: TER does not capture trading spreads, brokerage commissions, or taxes. It is one number, not the whole cost story.

2) Liquidity and tradability: knowing your real execution cost

Average daily volume, bid–ask spread, and assets under management jointly determine how much slippage you absorb on entry and exit. A fund with $500M in AUM and a 1-cent spread on a $100 price represents a 0.01% round-trip friction — negligible for a long-term holder. The same position in a thinly traded sector ETF with a 30-cent spread doubles your visible cost in a single trade.

Illiquid ETFs are not automatically bad. If you are holding for three to five years and the mandate fits your objective, the spread may be dwarfed by the return differential. The problem appears when the underlying index becomes dislocated and the authorized participant mechanism struggles — you can see premiums or discounts to NAV widen sharply. Check the fund issuer's creation/redemption disclosures and historical NAV premium/discount data before sizing.

3) Tracking difference and index methodology: what you actually own

Tracking difference (TD) measures how much the fund's annual return deviates from its benchmark — positive TD means the fund outperformed (e.g., via securities lending income), negative means it lagged. Tracking error measures the volatility of that deviation over time. Both matter: a fund with tight tracking error is consistent; a fund with low average TD is cheap in practice. IEF, for instance, consistently posts TD near zero against the ICE 7–10 Year Treasury index because the underlying bonds are highly liquid and lending revenue offsets operating costs.

Index methodology is not boilerplate. Read it. Float-adjustment rules determine whether mega-cap dominance is amplified or capped. Rebalance frequency affects turnover and potential front-running. Capping rules in sector indices (like SMH's semiconductor weighting constraints) can make "the same theme" behave differently across providers. Understanding methodology is how you avoid the surprise of owning what the name implies but not what the structure delivers.

4) Portfolio concentration: "passive" is not "balanced"

Pull the top-10 holdings and their aggregate weight. In QQQ (Nasdaq-100), the top five names historically represent 40–45% of the fund — you are not buying equal exposure to 100 companies, you are buying a concentrated tech-and-growth portfolio with 100 names attached. The index is market-cap weighted, so price appreciation in the largest names continuously increases their dominance until a rebalance forces adjustment.

Single-country skew in regional or emerging-market ETFs deserves the same attention. An ETF labeled "Latin America" may be 60% Brazil and 20% Mexico; one labeled "Asia" may be overwhelmingly concentrated in three markets. Before you layer in thematic exposure, map what you already own — double-counting a sector via two different "diversified" ETFs is a real and common mistake.

5–7) Distributions, FX, and structure: three items most investors skip

Income treatment: accumulating share classes reinvest dividends inside the fund (compound automatically, no distribution tax event); distributing classes pay out. For tax-deferred accounts, accumulating may be preferable; for income-oriented strategies, distributing makes cash flows visible. Not all U.S.-listed ETFs offer both classes — confirm with the fund's prospectus.

Currency of exposure versus currency of listing: GLD trades in USD but the gold price is globally referenced in USD regardless — no FX layer. VWO, however, holds equities across dozens of emerging-market currencies; the USD is a translation layer on top of underlying FX risk. For non-USD investors, this distinction changes the risk profile materially. Map where FX exposure actually lives, not where the ETF quotes.

For synthetic replication (swap-based ETFs), read collateral and counterparty disclosures carefully. Physical replication holds the securities; synthetic holds a swap and a collateral basket — counterparty default, even if remote, represents a structural risk with no parallel in physical funds. Cognitor does not replace this checklist — it adds six independent specialist lenses and five SENIOR verdicts on the curated 40 US-listed ETFs so you can see where macro, rates, earnings, and flows create agreement or divergence across the universe. General information only.

FAQ

Which of the 7 metrics is most important?

It depends on your time horizon and position size. Long-term, buy-and-hold investors typically weight TER and tracking difference most heavily, since those compound over years. Active or tactical traders focus on spread, volume, and NAV premium/discount — execution friction dominates short holding periods. Run all seven and let the use case decide which to prioritize.

Does past performance predict future ETF results?

Past index behavior provides context but no guarantee of future returns. What past data does usefully reveal is tracking consistency, volatility patterns, and how the mandate behaved across different rate and risk environments — treat it as structural evidence, not a return forecast.

Is a large AUM always a positive signal?

Large AUM typically supports tighter spreads and more reliable creation/redemption mechanics, which reduces execution risk. But AUM alone does not determine tracking quality, mandate fit, or factor exposure. A newer fund with lower AUM and tighter tracking difference can be the better choice for your specific objective.

What exactly is tracking error and how is it different from tracking difference?

Tracking difference is the average annual gap between fund return and index return — it tells you the chronic cost or benefit of ownership. Tracking error is the standard deviation of that gap — it tells you how consistent the fund is in following the index. A fund can have low average tracking difference but high tracking error (inconsistent), or vice versa. Always read the issuer's factsheet definition because methodologies differ.

How does Cognitor help with this checklist?

Cognitor's structured weekly research on the curated universe runs Panel → SENIOR → PRIME across ~40 US-listed ETFs — covering macro, earnings, rates, flows, geopolitics, and market psychology. It surfaces where specialist views converge and where they split, adding context to the scenario implications of what you own. It is not a buy/sell signal and does not replace the checklist above.

Cognitor provides general financial information and educational research — not personal investment advice or a recommendation to buy or sell any security.

Alternate languages: EN · ES · PT