The Five Numbers
Before buying any ETF, check these in order:
| # | What to Check | Good Benchmark | Why It Matters |
|---|---|---|---|
| 1 | Expense ratio | Under 0.20% for index ETFs | The only cost that's guaranteed regardless of performance |
| 2 | AUM (fund size) | $500M+ for core holdings | Signals liquidity and reduces closure risk |
| 3 | What index it tracks | Published, rules-based, well-defined | "S&P 500 ETF" and "S&P 500 ESG ETF" don't hold the same stocks |
| 4 | Tracking difference | Within ±0.10% of expense ratio | Shows whether the fund actually delivers what it promises |
| 5 | Bid-ask spread | Under $0.05 for liquid ETFs | A hidden cost every time you buy or sell |
That's the whole framework. Everything else — fund provider name, star ratings, marketing materials, how recently it launched — is secondary to these five. Walk through each one below.
Expense Ratio: The Only Guaranteed Cost
The expense ratio is the annual fee charged as a percentage of your investment. It's deducted continuously from the fund's daily value — you never receive a bill, and you never notice it until you do the math.
On $10,000 invested, a 0.03% expense ratio costs $3 per year. A 0.50% expense ratio costs $50. That $47 gap sounds trivial. Compounded over 30 years at 8% annual returns, the difference between a 0.03% and a 0.50% fund on a $10,000 starting investment grows to over $6,000 in total return. Scale that to $100,000 and you're looking at $60,000 in lifetime cost from a fee choice you made once at account opening.
When a higher expense ratio is justified
Sometimes. If you want exposure to U.S. small-cap value stocks with a genuine factor tilt, AVUV at 0.25% has no direct 0.03% equivalent — the cheapest comparable option is around 0.15%–0.25%. You're paying for a specific, evidence-based strategy that isn't available cheaper. That's a defensible tradeoff.
What isn't defensible: paying 0.75% for ARKK on the theory that active management in disruptive technology will reliably outperform a 0.10% tech index ETF like QQQ. The evidence for consistent active outperformance after fees is thin. The higher fee is certain; the outperformance is not.
Start with expense ratio before anything else. It's the only variable in the fund's control that directly reduces your return with certainty. Two funds tracking the same index — one at 0.03%, one at 0.09% — will produce returns that differ by exactly that 0.06% annually, forever. There is no scenario where the 0.09% fund wins on cost. Check the full expense ratio guide to see how that gap compounds over decades.
AUM: Why Fund Size Matters More Than You'd Think
Assets Under Management (AUM) is the total dollar value of everything the fund holds. It's a proxy for two things that directly affect you: liquidity and closure risk.
Liquidity
Larger funds have more trading volume, which means tighter bid-ask spreads (covered in #5). A fund with $500M in AUM trades millions of shares per day. A fund with $15M in AUM might trade a few thousand. That difference shows up as cents on each share — but it adds up, especially if you trade frequently or hold a large position.
Closure risk
Fund companies close ETFs that are too small to be profitable. When a fund closes, you receive the cash value of your holdings — which sounds fine until you realize that closure usually triggers a taxable event, and often happens when the fund has underperformed (the worst time to be forced out of a position).
Vanguard has closed small ETFs. iShares has closed dozens. It's not a fringe event. A fund below $50M in AUM is meaningfully at risk of closure within 2–3 years if assets don't grow.
A freshly launched ETF with $10M in assets and strong marketing is a bet that the fund grows to viable size. Sometimes it does. Often it doesn't. If the underlying exposure is available in a larger, older fund, there's no reason to take on the closure risk just to be an early adopter.
What Index It Tracks: The Most Overlooked Variable
Two ETFs can share a name, a category, and even a Morningstar classification while tracking completely different indexes. This is the variable most investors skip — and it determines what you actually own.
The "S&P 500" problem
SPY, IVV, and VOO all track the S&P 500 Index. They are genuinely comparable — different providers, same index, different expense ratios (SPY: 0.09%, IVV: 0.03%, VOO: 0.03%). The performance difference between IVV and VOO over 20 years is rounding error.
But "SPYG" (S&P 500 Growth ETF) and "SPYV" (S&P 500 Value ETF) track sub-indexes that carve the S&P 500 into halves. "ESGU" tracks the MSCI USA ESG Leaders Index — an S&P 500-ish fund that excludes certain companies and applies ESG screens. Same general neighborhood, very different holdings.
This matters even more for ESG and impact investing ETFs, where two funds sharing the same label can use completely different methodologies. One might exclude all fossil fuels; another might include them if their ESG score is strong relative to sector peers. The five-number framework applies here more than anywhere. See how it applies to ESGV, ESGU, DSI, and VOTE →
The "total market" problem
VTI tracks the CRSP US Total Market Index — roughly 3,700 U.S. companies. ITOT tracks the S&P Total Market Index — roughly 1,600 companies. Both are called "total market" ETFs. VTI is more diversified by company count; ITOT concentrates more in larger companies. Neither is wrong, but they're not the same fund.
How to actually check
Every ETF's factsheet (and the ETF BFF factsheet pages) lists the index tracked. Read it, not just the fund name. Specifically: what does the index include and exclude? Is it market-cap weighted or does it tilt toward specific factors? Is the index from S&P, MSCI, CRSP, or a proprietary index the fund company invented themselves?
Some ETFs track indexes created by the fund's own parent company. There's no independent oversight of what goes in or comes out. The index methodology can change, and you may not notice until it affects performance. Not automatically disqualifying — but worth understanding before you buy.
Tracking Difference: What the Fund Actually Delivers
Tracking difference is the gap between what the fund actually returned and what its benchmark index returned over a given period. It's different from tracking error, which is a measure of how volatile that gap is — and tracking difference is the one that puts money in or takes money out of your account.
You'd expect a 0.03% expense ratio fund to underperform its index by 0.03% annually. But real tracking differences can be better or worse than the stated fee, for two main reasons:
- Securities lending revenue: Large ETF providers lend their holdings to short sellers for a fee. That income flows back to the fund, partially or fully offsetting the expense ratio. VTI's tracking difference has at times been smaller than its 0.03% ER — meaning the fund barely underperformed its index at all, because lending income covered most of the fee.
- Cash drag and rebalancing costs: Index changes, dividend reinvestment, and cash held in the fund create small performance drags. In funds with more frequent turnover or illiquid underlying holdings, these can add up.
How to find tracking difference
ETF.com and Morningstar publish annual tracking difference figures. For broad U.S. index ETFs from major providers, tracking differences are typically within 0.05% of the expense ratio — genuinely negligible. For small niche ETFs with illiquid underlying holdings, the gap can be 0.20%–0.50% on top of the stated fee. That's a meaningful real-world cost that doesn't appear on the fund's marketing page.
For broad index ETFs from Vanguard, iShares, and Schwab, don't overthink tracking difference — it's essentially the expense ratio ± a few basis points. Where it matters: niche ETFs, commodity ETFs, and leveraged ETFs, where the cost structure is more complex and tracking difference can meaningfully diverge from the stated ER.
Bid-Ask Spread: The Hidden Transaction Cost
Every ETF trade has a bid-ask spread — the difference between what buyers will pay (bid) and what sellers will accept (ask). When you buy an ETF at the ask price and sell at the bid price, you pay the spread twice: once going in, once going out.
What the numbers look like in practice
| ETF | Approx. Share Price | Typical Spread | Round-Trip Cost on $10,000 |
|---|---|---|---|
| SPY (S&P 500) | ~$600 | $0.01 | ~$0.33 |
| VTI (Total Market) | ~$280 | $0.01 | ~$0.71 |
| IEFA (Intl Developed) | ~$55 | $0.01 – $0.02 | ~$1.80 |
| Small niche ETF | ~$30 | $0.10 – $0.30 | ~$33 – $100 |
For major ETFs from large providers, the spread is irrelevant — a few cents on a trade is not a real cost. For thinly traded niche ETFs, the spread can be 0.3%–1.0% of the position value, which is a significant hidden cost that compounds every time you trade.
Practical rules for managing spread
- Trade during regular market hours (9:30 AM – 4:00 PM ET). Spreads widen at the open and close.
- Use limit orders for any ETF with an AUM under $100M. A market order on a thinly traded fund can execute at a meaningfully worse price than you expected.
- For major broad-market ETFs (VTI, SPY, BND, etc.), market orders during regular hours are fine — spreads are negligible.
What Doesn't Matter as Much as People Think
The fund provider's brand
SPY (State Street/SPDR), IVV (iShares/BlackRock), and VOO (Vanguard) all track the S&P 500. Over 20 years, their returns differ by the expense ratio gap and essentially nothing else. Pick the cheapest version of the exposure you want. If two funds have the same index and the same fee, pick the one with more AUM. The logo on the fund doesn't compound.
Star ratings
Morningstar and similar ratings are backward-looking. A 5-star rating means the fund performed well over the past 3–5 years. It tells you almost nothing about future performance. A fund that gained 5 stars by riding a sector boom is not a better fund than one that returned 5% in the same period by following a boring total-market index. Check the five numbers above instead.
How recently the fund launched
Older funds aren't better funds by virtue of age. What age does tell you: a fund that launched in 2018 has less AUM history and less fee competition pressure behind it. A fund launched in 1993 (SPY) has survived multiple market cycles and has essentially no closure risk. Longevity is a proxy for AUM stability, not fund quality.
Whether it's popular or trending
Popularity correlates with liquidity, which is good. But an ETF that's trending because of a market story — AI, clean energy, crypto, biotech — is usually popular right when the theme is fully priced in. The five-number framework cuts through the trend narrative: is this fund cheap, large, well-tracked, and based on a clearly defined index? If yes, it may be worth owning. If the five numbers look bad but the marketing story is good, skip it.
Thematic ETFs (clean energy, AI, metaverse, cannabis) are launched when themes are popular, which is often after the relevant stocks have already run up. They typically carry expense ratios of 0.40%–0.75%. Their indexes are proprietary. Their holdings overlap unpredictably with broad market ETFs you probably already own. If you're attracted to a thematic ETF, ask: what specific exposure am I buying that VTI doesn't already give me, and is the extra fee worth it?
The One-Sentence Gut Check
Before buying any ETF, try to fill in this sentence:
"This fund holds [what], tracking [which index], at [what cost], and I want it because [specific reason]."
If you can complete it cleanly, you understand what you're buying. If it requires more than one sentence and you're still not sure what it holds — or why you specifically want this exposure versus just owning more VTI — that's a signal to keep researching before buying.
VTI: "This fund holds every publicly traded U.S. company, tracking the CRSP US Total Market Index, at 0.03%, because I want broad U.S. equity exposure at the lowest cost." ✓
AVUV: "This fund holds U.S. small-cap value stocks with a factor tilt, tracking a rules-based index, at 0.25%, because I want tilted small-cap value exposure that isn't available at lower cost." ✓
A trending thematic ETF with 0.65% ER: "This fund holds... some tech companies... tracking a proprietary index... because AI is going to be big?" ✗ — This is a marketing story, not an investment thesis. Do more work or skip it.
Use the ETF comparison pages to run two candidate ETFs side by side on these five numbers, and the individual factsheet pages to pull up the numbers for any ticker you're considering.
One ETF concept a week. Free, forever.
Plain-English ETF education in your inbox. New guides like this one when they drop. No jargon, no stock tips. Reviewed by a CFA® Charterholder.
Free forever · No spam · Unsubscribe anytime
Frequently Asked Questions
What should I look for when choosing an ETF?
Five things, in order of importance: expense ratio (the annual fee — under 0.20% for broad index ETFs), AUM (fund size — at least $500M for core holdings), what index it tracks (read the actual index name, not just the fund name), tracking difference (how closely the fund matches its benchmark in practice), and bid-ask spread (the cost of entering and exiting a trade, which matters more for small illiquid funds than for major ones).
Does it matter which provider makes the ETF?
Less than most people think. SPY (SPDR), IVV (iShares), and VOO (Vanguard) all track the S&P 500 — their 20-year returns differ only by their expense ratio gap. SPY charges 0.09%, the other two charge 0.03%. The provider's name isn't the variable that determines your return. The index, the fee, and the fund size are. Among providers with equally low fees and large AUM on the same index, it genuinely doesn't matter much.
How many ETFs do I actually need?
For most investors, 1–3. VTI alone — every publicly traded U.S. company, market-cap weighted, at 0.03% — covers more diversification than most portfolios need. Adding VXUS brings international stocks. Adding BND brings bonds. That's the complete 3-fund portfolio. More ETFs add complexity and overlap; they don't automatically add diversification or improve returns.
Should I buy an ETF that just launched?
Be careful. New ETFs have minimal AUM, wider bid-ask spreads, no performance track record, and meaningful closure risk if they don't attract assets. The bar should be: does this fund offer exposure I specifically want that isn't available cheaper and more reliably in an established fund? If the answer is no, wait until the fund proves itself or stick with the established alternative.
Is a higher star rating a reason to buy an ETF?
No. Morningstar ratings look backward — they reflect past performance over 3–5 years. Past performance does not reliably predict future returns, especially for equity ETFs where returns are driven by market conditions the fund doesn't control. A 5-star rating for a sector ETF probably means the sector performed well recently. That tells you something about where the market has been, not where it's going. Use the five numbers in this guide instead.
The short version
- Check expense ratio first — it's the only cost that's guaranteed regardless of performance. Under 0.20% for broad index ETFs. Under 0.10% is better.
- Fund size (AUM) matters for liquidity and closure risk. Core holdings should be $500M+. Avoid anything under $50M for long-term positions.
- Read the actual index name, not just the fund name. "S&P 500 ETF" and "S&P 500 Growth ETF" don't hold the same stocks.
- Tracking difference tells you what the fund actually delivered vs. its benchmark. For major broad-market ETFs it's essentially the ER. For niche funds it can diverge meaningfully.
- Bid-ask spread is negligible for liquid ETFs like VTI or SPY. It can be 0.3%–1.0% for small thinly traded funds — a real, hidden cost.
- The gut check: if you can't state in one sentence what the fund holds, which index it tracks, what it costs, and why you want that specific exposure, keep researching.
ETFs Mentioned in This Guide
Hover any ticker for a live data preview — click to open the full factsheet.