ETF Bid-Ask Spread vs Expense Ratio: Two Costs on Different Clocks

The short answer

The spread is paid per trade. The expense ratio is paid per day. Neither number means anything on its own, because the total cost depends on how long you hold and how often you trade.

Put them on the same clock and a simple rule appears: for a fund with a 0.02% quoted spread and a 0.25% annual fee, one month of holding costs about as much in fees as one round trip costs in spread. Below that, the trade is the bigger cost; above it, the fee is.

That arithmetic is the whole article. Everything below is how to do it with real published numbers.

What each cost actually is

The bid-ask spread is the difference between the highest price a buyer will pay and the lowest price a seller will accept. If the quote is 45.21 bid and 45.23 ask, the spread is 0.02 on a mid of 45.22, or about 0.044%. A market order pays roughly half of that on entry and half on exit, so a full round trip costs approximately the quoted spread. The exact amount depends on order size, timing and how the order is routed.

The expense ratio is the fund’s annual operating cost as a percentage of assets. It is deducted continuously from the portfolio, which is why a fund’s NAV grows slightly more slowly than its holdings. It appears on a product page as a small annual number and on your statement not at all — it is not billed, it is absorbed.

The two costs differ in three ways that matter:

That last row is the one people skip. A spread is avoidable but not free: the limit order that avoids it may not fill, and while it rests the market may move away. The expense ratio is unavoidable for as long as you hold the fund.

The published numbers for one fund

iShares publishes both figures on the IBIT product page. Read on 2026-10-28, the page showed:

• Net assets: $65,369,061,493 as of Oct 27, 2026.
• Shares outstanding: 1,432,920,000 as of Oct 27, 2026.
• NAV: $45.62 as of Oct 27, 2026.
• Closing price: $45.22 as of Oct 26, 2026.
• Premium/discount: −0.01% as of Oct 26, 2026.
• 30-day median bid/ask spread: 0.02% as of Oct 26, 2026.
• Sponsor fee: 0.25%, as stated in the prospectus.

Two cautions about reading that list, because the page mixes dates. The closing price is dated Oct 26 while the NAV is dated Oct 27, so subtracting one from the other produces a meaningless number. The premium/discount figure the page publishes is the correct paired comparison, and it was −0.01%. The 30-day median spread is also dated Oct 26, and it is a median over a month rather than a current quote.

With those caveats, the two costs are 0.02% and 0.25%. The question is how to compare them.

It is worth knowing exactly what that 0.02% measures, because Rule 6c-11 defines the calculation rather than leaving it to the fund. The rule says the median bid-ask spread is computed by taking the national best bid and offer at the end of each 10-second interval on each trading day of the last 30 calendar days, dividing each bid-ask difference by the midpoint, and taking the median of those values — and it must be expressed as a percentage rounded to the nearest hundredth.

Two things follow. First, the figure samples the quiet parts of the day as well as the busy ones, which is why it is a better planning number than a single quote. Second, the rounding is coarse: 0.02% is the nearest hundredth, so the underlying median could be anywhere from about 0.015% to 0.025%. Treat the published spread as a two-decimal number with a band around it, not as an exact cost.

Putting both on the same clock

Annualise the spread, or pro-rate the fee. Either direction works.

One round trip at 0.02%. If you buy and sell once, the spread costs about 0.02% of the position.

The fee over an arbitrary holding period is 0.25% × days ÷ 365.

The crossover is visible at one month: 0.0205% of fee against 0.020% of spread. Hold for a year and the fee is more than twelve times the spread.

The same arithmetic in dollars, on a $100,000 position:

• One round trip at 0.02%: $20.
• One year at 0.25%: $250.
• Break-even: 250 ÷ 20 = 12.5 round trips per year.

So a buy-and-hold investor pays the fee and barely notices the spread. A strategy that trades in and out monthly pays the spread twelve times, which is roughly equal to a full year of fees — and it pays both.

What moves the spread, and what does not move the fee

The spread is a market variable. It widens when volatility rises, when volume falls, around the open and close, and during stress. The 30-day median published on a product page deliberately smooths those swings, which makes it a reasonable planning number and a poor estimate of what you will pay at 9:31 a.m. on a volatile morning. The spread, depth and slippage guide separates those three costs.

The mechanism that keeps an ETF’s spread tight is creation and redemption: authorized participants can exchange baskets of the underlying for shares, and the arbitrage between the ETF price and the basket value pulls the two together. That mechanism is why a large, liquid ETF can quote 0.02% where a small one quotes several times that.

The article on this site covering creation and redemption explains the mechanics. This one does not repeat them; it takes the mechanism’s output — a tight spread — as an input.

The expense ratio behaves the opposite way. It is set in the prospectus, disclosed on the product page, and does not respond to volume or volatility. Some funds waive part of it for a period, and the page will say so; a waived fee is a temporary number, and the prospectus figure is the one that persists.

Premium/discount is a third number

A common error is to fold the premium or discount into either cost. It is neither.

• The spread is the cost of crossing the market.
• The expense ratio is the cost of holding the fund.
• The premium or discount is the difference between the market price and the NAV at a point in time.

If you buy at a 0.3% premium and later sell at a 0.3% premium, the premium roughly cancels and you have paid only the spread and the fee. If you buy at a premium and the premium collapses before you sell, you have lost the difference in addition to both. That is a risk of timing, not a fee, and it is why the premium/discount is reported separately.

IBIT’s −0.01% on Oct 26 is a tiny discount: the market price was a hair below NAV. Reading it as “the fund is cheap” would be reading a one-day snapshot as a signal.

A note on what this is not

Two related topics have their own articles and are not re-explained here.

• Tracking difference and tracking error describe how closely a fund follows its index over time. That is a different question from what a trade costs, and the tracking difference versus tracking error guide keeps the two separate.
• Creation and redemption describe how shares are made and destroyed. That is the mechanism behind the tight spread, not the cost of it; the authorized participant guide covers the mechanics.

This article answers one question: given a spread and an expense ratio, which one costs more for the way you actually use the fund?

A checklist

1. State the holding period. Without it neither number can be ranked.
2. Count your round trips per year. Twelve round trips at 0.02% equals a year of a 0.25% fee.
3. Use the median spread for planning and the live quote for execution. They are different numbers with different purposes.
4. Check whether the fee is waived. A waiver is temporary; the prospectus number is the durable one.
5. Keep the premium/discount separate. It is a price relative to NAV, not a fee, and it can work for or against you.
6. Compare like with like across funds. A cheaper fee with a wider spread can cost more for a short holding; a tighter spread with a higher fee can cost more for a long one.

Risk note. IBIT’s net assets, shares outstanding, NAV, closing price, premium/discount, 30-day median bid/ask spread and sponsor fee were read from the iShares product page on 2026-10-28, with the as-of dates shown above. Spreads and premiums move continuously and the figures here are a snapshot, not a quote. The break-even arithmetic assumes one round trip at the quoted spread and a constant fee; it is an illustration, not a projection of any fund’s returns.