Bitcoin futures ETFs and the cost of rolling contracts

The US futures ETFs still trading

This is the whole surviving universe. Two of the funds below reference ether rather than bitcoin — they sit on the same issuer shelves, use the same wrapper and carry the same roll mechanics, so excluding them would give a misleading picture of the category's health. Assets are shown as reported by the issuer or an aggregator on the date noted in the caption.

Notice the shape of that list. One fund holds ninety-odd per cent of the money and the other six are rounding errors — three of them under $20m, which is below the level at which a sponsor can comfortably pay for an audit, a listing and a Cayman subsidiary. BITS deserves a specific warning: it is a hybrid that holds CME bitcoin futures alongside a position in Global X's own blockchain equity ETF, so it charges 0.65% on top of the 0.50% that fund already charges on that sleeve. If you hold both, you are paying twice for the same equities. That is a different animal from the coin funds on the mining ETF page.

The forward curve

What a futures ETF actually owns

Start with the thing most summaries skip. Open the holdings page of a bitcoin futures ETF and you will not find bitcoin. You will find Treasury bills, a money market position, and a small number of futures contracts with expiry months printed next to them. The Treasuries are collateral. The contracts are the exposure.

A CME bitcoin futures contract is an agreement to settle the difference between the contract price and the exchange's reference rate on a specific date, in cash. Nobody delivers coins. Each contract has a fixed expiry, which creates the structural problem at the centre of this page: a fund that wants permanent exposure has to hold a temporary instrument. Before the contract it owns expires, the manager must sell it and buy a contract expiring later. That transaction is the roll, and it happens forever, every cycle, for as long as the fund exists.

This is not a flaw in any particular fund's execution. It is the definitional property of holding derivatives instead of an asset. A spot bitcoin ETF buys coins, puts them in cold storage with a custodian and does nothing further. There is no expiry to manage, because bitcoin does not expire. The whole difference between the two wrappers falls out of that one sentence, and we set it out at length on spot versus futures bitcoin ETFs.

Contango, backwardation and what the roll actually costs

The price of a futures contract is not a forecast. It is roughly today's price plus the cost of carrying the position to the expiry date — financing, opportunity cost of the cash, and whatever premium the market demands for leveraged exposure without funding it directly. For bitcoin, that carry has almost always been positive. Traders have been willing to pay for deferred exposure, which means the contract expiring in three months usually trades above the one expiring next month, which usually trades above spot.

An upward-sloping curve is called contango, and it is precisely the wrong shape if your job is to roll. Work through it. The fund owns the September contract, currently worth $77,000. It expires, so the fund sells it and buys the October contract, which costs $77,600. Bitcoin itself has not moved. The fund now controls the same economic exposure but has spent $600 per contract to keep it. Do that every cycle and the fund's return diverges downwards from bitcoin's return even if the manager makes no mistakes at all. That divergence is the roll cost, sometimes called roll yield when it is positive and roll drag when it is not.

Two features of roll cost make it particularly awkward for a long-term holder. The first is that it is invisible in the expense ratio. A fund that charges 0.95% and bleeds another couple of percentage points a year on the curve has a true cost of ownership nowhere printed on the fact sheet — it shows up only as tracking difference, after the fact, in the total return chart. The second is that it compounds against you in exactly the periods you would most want exposure, because sustained bull markets are when the futures premium tends to be widest. Anyone comparing headline fees across wrappers should read our breakdown of what these funds really cost before deciding.

Why spot funds destroyed the category commercially

BITO's launch on 18 October 2021 was a genuine event. It was the first US exchange-traded fund of any kind to give ordinary brokerage accounts bitcoin exposure, it arrived during a bull market, and it took in more money on day one than almost any ETF in history. For twenty-seven months it had no spot competitor, because the SEC would not approve one.

Then on 10 January 2024 the Commission approved eleven spot products in a single order, Release No. 34-99306, and the economics inverted overnight. The new funds charged 0.20% to 0.25%, held bitcoin directly, and had no curve to roll. Against that, a 0.95% fee plus roll drag was not a competing product; it was the same exposure with a permanent handicap. The market worked out the implication quickly.

VanEck did the arithmetic fastest and most publicly. Its futures fund XBTF last traded on 30 January 2024, barely three weeks after the approval order and immediately after VanEck's own spot fund HODL began trading. The sponsor simply closed the inferior wrapper rather than ask investors to keep paying for it. Its ether equivalent EFUT followed, last trading on 16 September 2024. Hashdex took the third available route and converted its futures fund DEFI into a spot fund in March 2024 — which bought the product two more years before it was liquidated in August 2026, becoming the first US spot bitcoin ETF ever to close.

That last number matters. It is fashionable to write the futures category off entirely, and that overstates the case. BITO still holds more money than nine of the twelve US spot bitcoin funds. It has an established options market, a monthly distribution that some holders genuinely want, and a structure that certain accounts can buy when they cannot buy a commodity trust. The category is diminished. It is not extinct.

The stale claim about optimum roll, and what BITC does now

If you search for bitcoin futures ETFs you will find a large volume of writing describing Bitwise's BITC as an "optimum roll" fund — a product that scanned the CME curve and positioned itself in whichever expiry offered the least costly roll, rather than mechanically holding the front month. That was true when the fund launched on 20 March 2023. It has not been true since 3 December 2024.

On that date Bitwise abandoned curve optimisation entirely, renamed the fund to the Bitwise Trendwise Bitcoin and Treasuries Rotation Strategy ETF, and replaced the strategy with something structurally different: a binary long-flat momentum rotation. The fund is either fully in front-month bitcoin futures or fully in Treasuries, and the switch is governed by a 10-day versus 20-day exponential moving-average signal on the bitcoin price. There is no optimisation of anything. There is a trend rule.

Whether you like the new strategy is a separate question from whether you were told about it. A trend rotation is a legitimate design with a long history in managed futures, and moving to Treasuries in downtrends is a real risk-management choice rather than a marketing one. But it is emphatically not a cheaper way to hold bitcoin. It is an active timing strategy that will lag badly in choppy markets, whipsawing in and out on moving-average crossovers, and it costs 0.88% net against a 1.12% gross ratio. Anyone who bought it for roll efficiency owns something else now.

Futures funds that have closed

Three products in this lineage have gone. Their fates are worth reading together, because they show the three ways a sponsor can respond to a wrapper that has been made obsolete: close it, convert it, or keep paying for it.

A liquidation is orderly — the fund sells its positions and pays shareholders in cash — but the payout lands on a date you did not choose, and in a taxable account that turns an unrealised gain into a realised one. Our tax guide covers what that means in practice. The practical lesson for anyone holding one of the smaller funds in the table above is simply to know the risk exists. At $4m to $17m, a fund is not covering much beyond its own overheads.

The 1940 Act advantage that futures funds actually retain

Here is the one genuine structural point in favour of these products, and it is a real one rather than a consolation prize. US spot bitcoin ETFs are not ETFs in the strict legal sense. They are grantor trusts registered under the Securities Act of 1933, holding a commodity. They are not registered investment companies, they have no board of directors owing fiduciary duties to shareholders, and the protections of the Investment Company Act of 1940 do not apply to them. We unpack this at length in what a bitcoin ETF actually is.

BITO is different. It is a registered investment company under the 1940 Act, which brings a specific and enforceable package: an independent board majority, statutory limits on leverage and on transactions with affiliates, custody rules for fund assets, and mandated diversification and reporting. It cannot hold bitcoin directly under that Act, which is exactly why it holds futures — and why it routes part of the exposure through a wholly owned Cayman Islands subsidiary, a standard device that lets a 1940 Act fund take commodity exposure while staying inside its own qualifying-income rules.

The disclosed holdings go further than pure CME exposure. BITO holds CME bitcoin futures and Coinbase-listed futures alongside swap agreements, which is worth knowing if you assumed the fund was a clean CME tracker. It pays monthly distributions, largely funded by the Treasury collateral and the mechanics of the futures position rather than by anything bitcoin produces — bitcoin generates no income, so a distribution from a bitcoin-linked fund is being sourced from somewhere else. ProShares also quietly dropped "Strategy" from the fund's name, so the ticker now sits under the plain "ProShares Bitcoin ETF" label. Same fund, same 0.95%.

Wrapper, not asset

Who might still reasonably hold one

We would not recommend a futures fund to someone who simply wants long bitcoin exposure and has a normal brokerage account. The spot funds are better at that job on every measure that compounds. But three groups have defensible reasons, and it is worth naming them precisely rather than pretending the category has no use.

1. You specifically want the 1940 Act wrapper

   An independent board, statutory leverage constraints and Investment Company Act custody rules are not marketing. If your investment policy statement, trustee or compliance function distinguishes between a registered investment company and a commodity trust, that distinction is the whole argument, and it is the one thing a spot fund cannot offer you.

2. Your plan menu only admits registered investment companies

   Some employer plans, model portfolios and advisory platforms are built to hold 1940 Act funds and nothing else. In those accounts a spot bitcoin trust is not an expensive option; it is not an option at all. Paying 0.95% for access you would otherwise not have is a different calculation from paying 0.95% by choice.

3. You want the monthly distribution

   BITO pays monthly, which some holders value for cash-flow reasons even understanding that the money is not being generated by bitcoin. If income is the actual goal, though, compare it against the dedicated covered-call and income funds, where the distribution mechanics and the return-of-capital question are the entire point.

What none of these funds is, is a cheaper or a safer route to bitcoin. And they should not be confused with the other derivative-based products on this site: the 2x leveraged funds reset daily and carry volatility decay that dwarfs roll cost, while the inverse funds do the same thing in reverse. If you want a map of every category at once, the full crypto ETF list sets them side by side, and the ether fund list covers EETH's spot competitors.