A friend bought a spot bitcoin ETF in his brokerage account and asked me afterwards whether he had picked the right one. My first reaction was that it barely matters, since they all hold the same coin and the fees are all a fraction of a percent. Then I worked through what those fractions cost over a ten year hold and had to walk some of that back. The differences are real, they just hide in places the comparison tables never show you.
So this is the note I wish I had sent him before he bought. How the fee actually gets charged, why waivers muddy the comparison, and why the fund with the lowest headline number is sometimes the most expensive one to own.
You never see the bill
An expense ratio is an annual percentage of the fund's assets, but nobody invoices you for it. The fund accrues the fee daily, roughly the annual rate divided by 365, and pays the sponsor out of what the fund holds. For a spot bitcoin ETF that means the trust covers the fee out of its bitcoin, either transferring coins to the sponsor or selling a small amount. The bitcoin backing each share shrinks a little every day, which is why the fund's value per share drifts below the raw bitcoin price over time even when tracking is otherwise perfect.
On your statement this shows up as nothing at all. No line item, no deduction, no cash leaving your account. You just underperform bitcoin itself by the fee, every year, for as long as you hold. That invisibility is a big part of why people ignore it. If your broker debited your account every January to cover it, a lot more people would shop around.
The compounding is what gets you on long holds. Take a fee of a quarter of a percent, which is typical for the cheaper spot bitcoin products. One year costs you roughly a quarter of a percent of the position, which is genuinely trivial. Twenty years costs you roughly five percent of the final stack, because the drag compounds the same way returns do. In a product charging around one percent, the twenty year drag runs closer to a fifth of what the position would have grown into, all from a number that looked like rounding error on day one.
There is a historical footnote that shows how few people act on this. When spot bitcoin ETFs launched in the US, the largest product by far was a converted legacy trust charging several times what the new entrants charged. Billions eventually migrated out, but billions also stayed, partly from inertia and partly because selling in a taxable account meant realizing years of capital gains. That second part is worth sitting with, because in a taxable account your first pick is semi-permanent. Switching funds later means selling, and selling means a tax bill.
Fee waivers expire, and the expiry is the fine print that matters
Many issuers waive some or all of the fee for an introductory window, either for a fixed number of months or until the fund crosses an asset threshold, whichever comes first. The number on the marketing page might be zero while the prospectus fee is a quarter of a percent, and the prospectus number is the one you will pay for the other nineteen years of a twenty year hold.
My rule is that if your intended holding period is longer than the waiver window, you compare funds on the post-waiver fee and treat the waiver as a small bonus rather than the price. Waiver terms live in the prospectus and in the fee table footnotes on the issuer's site, so read the footnotes before you compare anything. Also note whether the waiver dies on a date or at an asset level, because a popular fund can blow through an asset threshold much faster than the calendar suggests, and nobody emails you when it happens.
When the cheap fund is the expensive one
An expense ratio is a per-year cost. A bid-ask spread is a per-trade cost. Mixing up those units is the most common mistake I see in fee comparisons, and it is the reason a cheaper fund can end up costing you more.
Every time you buy and later sell, you cross the spread. For the biggest, most liquid bitcoin ETFs the spread is typically a few basis points, and for the small ones it can be many times that. Suppose fund A charges five basis points less per year than fund B, but trades ten basis points wider. If you buy once and hold for many years, fund A wins on fees. If you move in and out a few times a year, fund A is decisively more expensive despite the lower headline number, and no fee table will ever tell you that.
Liquidity matters in a second way here. Bitcoin trades around the clock and the ETF only trades during exchange hours, so the ETF price has to catch up to whatever bitcoin did overnight and over the weekend. Big funds with active market makers handle that gap tightly. Small funds can trade at noticeable premiums or discounts to the value of the bitcoin inside them, especially in the first and last minutes of the session, which means you can quietly overpay on the way in and get underpaid on the way out. Use limit orders, avoid the open and the close, and glance at the fund's premium and discount history, which issuers are required to publish.
There is one more failure mode I would weight heavily. Funds that stay small tend to get closed, and a closure is a forced sale at a time you did not choose. In a taxable account that is a realized gain whether you wanted one or not. A slightly higher fee at a fund that will obviously still exist in a decade is often cheaper than the bargain fund once you price that in.
The five-minute check
Here is the whole comparison, in the order I would run it, for the two or three funds on your shortlist.
- Open each issuer's own product page and note the expense ratio. Then find the fee table footnote or search the prospectus for the word waiver. Write down the post-waiver fee and what ends the waiver, a date or an asset level.
- During regular market hours, pull a live quote and look at the bid and the ask. Divide the difference by the price to get the spread in percentage terms. Do this at a calm mid-session moment rather than at the open.
- Check assets under management and average daily volume. There is no magic threshold here, you are flagging the outlier that is dramatically smaller than its peers.
- Estimate your own behavior honestly, meaning the years you expect to hold and the round trips you expect per year. Total annual cost is roughly the post-waiver expense ratio plus the spread times your round trips.
- If this is a taxable account, pick as if you cannot change your mind later, because changing your mind comes with a tax cost attached.
That is the whole exercise, five minutes and no spreadsheet, and it covers the three ways these products actually differ in cost, the ongoing fee, the trading friction, and the waiver fine print.
I will admit the stakes are modest. Between two large, liquid, low-fee funds the difference over a decade might be a fraction of a percent, and bitcoin's own volatility will dwarf it a hundred times over. But the comparison costs five minutes once, the savings accrue for as long as you hold, and it builds the habit of pricing the boring parts of a trade before the exciting parts. The market scorecards we built at Blockcircle come from that same instinct, surfacing costs and mechanics that never make it onto a marketing page, and an ETF fee table is the gentlest possible place to start practicing. Run the check before the buy order goes in, then stop thinking about it.