Bid-Ask Spreads and Trading Costs

What you actually pay

Quote 82.39 × 82.43. The spread is €0.04, or 0.049% of the €82.41 midpoint. Convention: the round-trip cost is the full spread (you buy at the ask, sell at the bid), so half of it applies to each transaction. On a €10,000 purchase that's about €2.40 one way: forgettable once, meaningful if repeated weekly.

Two refinements that separate careful investors from casual ones:

• The effective spread can be tighter than the quoted one. Orders sometimes fill between the displayed bid and ask (price improvement), especially where multiple venues compete or a broker's routing finds a better price. Comparing your fill against the midpoint at the moment of execution is the honest scorecard.
• Displayed size matters for larger orders. If your order exceeds the size shown at the best price, the remainder fills at worse levels ("walking the book"). For retail sizes in mainstream funds this essentially never binds; for €100k+ in niche funds it can.

The four ingredients, and what each implies

1. Basket trading cost. A market maker hedging your trade must transact the underlying. An S&P 500 basket is nearly free to trade; a frontier-market or off-the-run credit basket is not. This sets the floor: a fund's spread can never be much tighter than its holdings' spreads, which is why "why is this EM small-cap fund's spread 0.60%?" answers itself.

2. Hedging risk and time. The longer the maker must hold unhedged exposure, and the more volatile the market, the wider the quote. Hence spreads widen in fast markets and at the open, before prices settle.

3. Competition on the book. More makers quoting the same fund on the same venue means tighter prices. This is why the same ISIN can show 0.06% on Xetra and 0.25% on a thinner listing (Chapter 12's fragmentation, Europe's structural quirk), and why checking a second venue is sometimes worth more than checking a second fund.

4. Fair-value uncertainty. The dominant term when the underlying market is closed: the maker prices from futures and correlations and charges for the guesswork. This is the entire mechanical basis of the awake-underlying rule.

What normal looks like

Orientation ranges for typical retail-size orders in mid-session, with the underlying open (as of 2026, consistent with LSEG/Dimensional 2024–25 European tick-data averages; flagship funds sit at the tight end):

Use the table as a smell test: a flagship world ETF quoting 0.40% is telling you something about the venue or the moment, not about the fund.

Checking your own execution

The three-step audit, worth doing once a year rather than every trade:

1. Record the midpoint at the moment you submit (most platforms show bid/ask; screenshot if needed).
2. Compare your fill. Filled at or inside the ask when buying is normal and fine; filled meaningfully worse suggests routing or timing problems.
3. Cross-check the venue. For European multi-listing funds, compare the same ISIN's quote on a major venue at the same second. Persistent gaps mean your broker's default routing is costing you, and the fix is either a venue selection setting or a different broker (Chapter 14).

The same audit catches the two most common self-inflicted errors: trading before the underlying opens, and trading a foreign-currency line while paying FX both ways.

When size becomes an institutional question

Retail orders in mainstream funds are absorbed by the standing quote. The threshold where that changes is roughly when your order approaches or exceeds displayed size, or represents a noticeable share of the fund's typical daily volume in an illiquid segment. At that point the professional tools apply: splitting the order across the day, using the closing auction (Chapter 71), or, for genuinely large amounts, requesting a risk price from an ETF desk via RFQ (Chapter 88): the mechanism where the ETF's real liquidity, the basket's, is accessed directly. Chapter 69's central point applies throughout: volume is not liquidity; the basket is. A fund trading 5,000 shares a day on a mega-cap basket can absorb far more than a busy-looking fund on an illiquid one.

A worked example

Same investor, same €20,000 purchase of a Europe-listed world ETF, four ways:

• Monday 09:02 CET, market order, USD line: spread 0.28% (opening + closed US underlying), plus 0.25% FX each way. Total cost ≈ €96.
• Monday 09:02, marketable limit, EUR line: spread cost ≈ 0.28% one-way ≈ €28, no FX. €28.
• Monday 15:45, marketable limit, EUR line, default venue: spread 0.08% → €8.
• Monday 15:45, marketable limit, EUR line, primary venue: spread 0.05% → €5.

The gap between the worst and best version is roughly €90 on a single trade, for decisions that cost nothing but attention; and repeated monthly, the difference compounds into a fee-level annual drag that no TER comparison would ever reveal.