The $47 Fee Buried in the Tax Slip of My Bitcoin ETF

The number that began this was $47.14, sitting on a 1099 composite form from my brokerage in February, under a category I had to look up: fee accruals and adjustments related to a bitcoin ETF I'd held for eleven months. I'd snagged $10,000 of the thing as a boring, mainstream way to hold bitcoin exposure inside a regular brokerage account, and I'd assumed the story ended at the expense ratio — 0.25%, about $25 a year, visible and honest. The $47 was something else. It was my share, apparently, of the fund's own transactional plumbing, passed thru in ways the headline fee never shows.

That number sent me into the fund's annual report, which is the least-read document in finance and possibly the most honest. What I found reframed how I think about every ETF I own, not just the crypto ones.

Here's the mechanism nobody explains at the point of sale. When demand for a bitcoin ETF rises, the fund doesn't go buy coins on an exchange like a retail trader. Authorized participants — giant market-making firms — deliver actual bitcoin to the fund, and in exchange receive newly created fund shares, which they can sell into the hot market. When demand falls, the process runs backward: shares get redeemed, coins leave the fund. This creation-redemption loop is why the share price tracks the coin price at all. It's an arbitrage machine, and the $47 on my tax slip was a crumb of its operating cost.

the middlemen who do the actual creating

The names are worth knowing cuz they explain the fees: authorized participants are a handful of giant trading firms, and they only bother delivering coins and minting shares cuz the spread pays them to. Nothing about this is charity. When you buy an ETF on a busy day, you are one customer of a machine whose staff gets paid from the gap between your price and the coin's price. Efficient? Remarkably. Free? Never.

the tracking gap that shows up in the boring numbers

Compare the fund's yearly return to bitcoin's own price move and you'll see a gap of a few tenths of a percent. That gap is the whole story in one number: the expense ratio takes its slice, the coins that back the shares get sold or moved to meet creations and redemptions, those movements are taxable events inside the fund in some structures, and the trust holds alil cash for operational costs instead of being 100.00% coin. None of this is a scandal. All of it is real, and it compounds. Over a decade, a 0.4% annual gap on a $50,000 position is approximately $2,000 you never see leave.

why the premium over spot can never quite die

Some days the ETF trades a hair above the value of its underlying coins; some days below. The arbitrage machine is supposed to crush those gaps, and mostly does — but crushing em costs money, and the cost lives in that same plumbing. During January of this year's inflow surge, some funds traded at a persistent modest premium simply cuz creations couldn't be manufactured faster than demand arrived. I'd spent years thinking of an ETF price as the coin price with a fee attached. It's actually the coin price plus or minus the current cost of the machine that keeps em welded together.

what this meant for my own account in April

I rebalanced in April, selling about a third of the position, and the sale produced a modest short-term capital gain that I owed ordinary income rates on — the ETF's convenience inside a taxable brokerage account cuts both ways, cuz there's no way to sell just the gains the way you can slice fractions of a coin. I ran the numbers against just holding coin directly in self-custody: the ETF costs me approximately 0.25-0.4% a year in visible and buried fees; direct holding costs me operational hassle and a hardware device, but zero annual drag. Neither answer is wrong. But now I'm choosing with the real price list instead of the advertised one.

the wrapper question, answered with numbers

Over any ten-year stretch, the choice is approximately this: the ETF keeps about 99.6% of the coin's return and costs me no weekends; direct custody keeps about 100% of the return and costs me a device, a backup ritual, and the obligation to never lose a piece of paper. On $10,000 the difference is a few hundred dollars a decade. On $500,000 it's a new car. The wrapper matters in proportion to the size of the position, which is the one variable my original purchase never weighed.

the argument I had with my brother-in-law at Easter

hed just snagged his first bitcoin ETF and was delighted there was "nothing to learn." I made the mistake of explaining creations and redemptions over ham, and his response was fair: "Does any of this change whether I should own it?" Honest answer: not much. The ETF remains the right tool for most people who want exposure inside a standard account, with normal statement reporting and a clean 1099. What the $47 changed was my expectations, not my allocation. The fund will never return exactly what the coin returns. The spread is not a bug being fixed. It is the toll for the bridge, charged every day, in slight enough amounts that you only meet it when you go looking.

the line item I now read every February

I still own the fund, sized a little modest, alongside a direct position. Every February I pull the 1099 and the annual report and read the fee and adjustment lines like a landlord reading a water bill — not because I can change em, but because unexamined numbers drift. The price of convenience in this asset class is invariably quoted in basis points and invariably paid in dollars, and the dollars are never printed on the purchase confirmation. Mine printed eleven months later, in a box I had to research, labeled with a number I'd never have connected to anything: $47.14, the yearly rent on a machine I didn't know I was using.