The notice went out on 3 August. The Hashdex Bitcoin ETF, which trades under the ticker DEFI, will change hands for the last time on 17 August. On or about 28 August it pays out cash and stops existing. The notice gives the reasons in one sentence, and not one of them is about bitcoin.
This article covers three things. What the notice says, and what it takes care not to say. How big the fund was when measured against bitcoin rather than against other funds. And what the industry's own tally shows about why a fund this small was still standing.
The reason to bother is that a first death invites a large reading. When a fund that holds one thing shuts down, the easy story is that the thing is in trouble. That story would be wrong here, and the fund's own paperwork is the first witness against it. The size of the fund is the second.
What the Notice Says
An exchange-traded fund is a fund whose shares trade on a stock exchange, the same way a share does. A spot bitcoin fund holds real coins and nothing else. DEFI was the only fund in its trust, so closing it closes the whole structure.
Hashdex reports $14.7 million in the fund as of 30 July. Anyone still holding shares at the close on 17 August is paid in cash. The amount will be what the fund's holdings are worth on the day it winds up, after the costs of closing it and after whatever bitcoin does in the meantime.
Then comes the part worth reading twice. The sponsor says it weighed "assets under management, trading liquidity, operating costs, investor interest, and how each fund fits within the firm's broader index-based product range." That is five reasons and they are all the same kind of reason. How much money is in it. How easily it trades. What it costs to run. Who wants it. Where it sits on the shelf.
Nowhere in the notice is there a word about bitcoin's prospects. A firm closing a fund because it has soured on the asset has every reason to say so, and this one does not say so. Bitcoin is not the reason.
The "first ever" framing needs care too. CoinDesk hedges it twice, calling this "what may be the first liquidation of a U.S. spot bitcoin offering" and "what appears to be the first closure" of such a fund. Hashdex's own announcement makes no claim to a first. So the honest version is that this appears to be the first, and no source here states it flatly.
One more detail changes the shape of the story. DEFI was not one of the funds that arrived when spot bitcoin ETFs were cleared in January 2024. CoinDesk reports that Hashdex launched it as a bitcoin futures fund in September 2022, and that it only became a spot product in late March 2024, "nearly three months after IBIT's launch". It did not lose a fair race from the same starting line. It arrived late, in a different shape, and never made up the ground.
How Big Is $14.7 Million
This site holds no data on fund flows. What it holds is fifteen years of daily coin data, so the way to size a fund here is to convert it into bitcoin terms and see what it buys.
Take a normal day of 2026, meaning the middle day of the year so far, with half the days busier and half quieter. On that day, exchanges reported $10.14bn of spot bitcoin trading. The fund's entire assets equal 0.145% of that, which is about two minutes of trading.
Reported volume can be inflated, so put the same question to a count that cannot be faked. Coins moving on-chain into and out of exchange wallets are settled transfers, written to the ledger, and on that same middle day they came to $3.39bn. The fund is 0.43% of one day of it: about six minutes.
The third comparison is the one that sticks. Miners bring new bitcoin into the world at a fixed pace, and on the middle day of 2026 that new supply was worth $33.94m. The whole fund, every dollar of it, is 43.3% of a single day's mining. The network replaces it in about ten hours.
The sector it leaves is not small in the same way. CoinDesk puts total net assets across the US spot funds at $77.6bn. On 23 May, the last day this site has coin data for, every bitcoin in existence was worth $1.535 trillion, so the funds hold the equivalent of 5.06% of the market. Convert those dollars at bitcoin's close on 30 July and they come to 1.21 million coins, 6.06% of the 20.03 million that exist. Measured in trading days, the entire American ETF industry is worth 7.7 days of spot volume.
Two limits on all of this, stated rather than buried. The coin figures stop on 23 May 2026, which is where this site's on-chain data ends, while the fund's own number is dated 30 July. And neither Hashdex nor CoinDesk gives DEFI's coin count, so none is estimated here.
Where the Money Went
The interesting question is not why a $14.7m fund closed. It is why so little of the money ever reached it.
CoinDesk's tally of the era runs like this, and the figures are CoinDesk's rather than this site's. BlackRock's IBIT has taken in $60.5bn. Fidelity's FBTC has taken in $9.95bn. The sector as a whole has netted $51.5bn in its lifetime and holds $77.6bn today. Grayscale has bled $27.47bn out the other side.
Add the first two and the arithmetic stops being background. $60.5bn plus $9.95bn is $70.45bn, which is $18.95bn more than the entire industry has netted since it started. Two funds took in more than everybody put in, and that is only possible because money walked out of Grayscale while it walked into them. The two measures are not identical, since CoinDesk describes the first pair as inflows and the sector total as net inflows, so treat the gap as a description of the shape rather than a balanced ledger.
It still says the thing plainly. "The industry raised $51.5bn" and "two funds raised $70.45bn while one lost $27.47bn" describe the same period. The second is the honest one.
Against that, DEFI's $14.7m is 0.019% of the sector's assets. One dollar in every 5,279. A firm looking at a fund that size, with its own costs and its own auditors and its own board, does not need a view on bitcoin to decide it is not worth the paperwork.
What to Take Away
Three findings and two dates.
The first is that the sponsor's stated reasons are all fund economics. A reader can check that in one page of primary text, without taking anybody's word for it.
The second is that the fund was worth about two minutes of daily trading and about ten hours of mining. That is the measured version of "too small to mean anything".
The third is what a winner-take-most market looks like from underneath. Two funds hold the field. One fund has lost $27.47bn. The rest compete for what is left.
The dates are 17 August, when the shares stop trading, and 28 August, when the cash goes out.
What this site cannot yet do is measure the concentration itself. The figures above are CoinDesk's, and they are labelled that way on purpose, because a number this page did not compute is a number this page does not own. When this site can count the flows itself, and the count survives a check against the published totals, the concentration of the ETF era becomes a dated series here rather than an adjective borrowed from a news story.
Until then the useful habit is the one this piece ran on: read the notice before the headline. The notice weighed money, costs, trading and shelf space. Bitcoin is not the reason, and it never claimed it was.