Why would anyone buy an ETF that yielded BTC minus 12% per year?

[nibor]

I ask this as everyone is very excited about a futures backed ETF. But due to the roll cost is not going to track BTC price - but is going to track its price less 12% (+fees) per year!

You can see this from CME quotes:
https://www.cmegroup.com/markets/cryptocurrencies/bitcoin/bitcoin.quotes.html#

Looking at right now:
Oct 21: 60285
Nov21: 61035

So if you are a futures back ETF what you do is you buy the nearby future - so in this case the Oct21. Then over a period of about a week at the end of the month you gradually sell your Oct21 futures and buy the Nov21 futures. But here you see the problem - you are selling low and buying high - you just lost $750 vs your benchmark of BTC price - over 1%. And you have to do this every month!

Obviously you could instead buy the Dec22 future so you do not have to roll for 14 months. But that has zero liquidity (open interest is 1!), and was $3510 more that the Oct 21 future at the last close - but due to lack of liquidity would have to pay a lot more for it if bought in volume. So this would probably be even worse.


So WHY is the roll so expensive when you could do a risk free arbitrage I hear you ask!

Problem is the arbitrage is very capital intensive. Take now arb would be:
Buy 5BTC @ 60k = $300k
Sell 1 Future @ $60,285 (you just made a $1,425 paper profit by the end of the month!).
But to buy the future you need to lodge $101,842 of margin at CME (initial might be 10% more than this):
https://www.cmegroup.com/markets/cryptocurrencies/bitcoin/bitcoin.margins.html
And you need cash at hand - as you are short and if the futures price increases you need to lodge more margin fast or the CME will close your position.

So this arb you need $450k (50k for some liquidity to cover margin calls - I assume if you need more could close out position) in CASH.
And you are making $750 * 5 * 12 = 45k a year. So a 10% return - but needs to be carefully managed to be sure if BTC price spike CME does not close you out - exposing you to the crash just after the spike - as you are now naked long BTC - and this could easily wipe out years of profits.

Now you might be able to borrow some money from someone and lodge the 5 BTC as collateral which would reduce your capital need and increase return - but not sure who would take on this loan!

[1715144271]

1715144271

[1715144271]

1715144271

[1715144271]

1715144271

[1715144271]

1715144271

[Quickseller]

The CME futures are cash settled, so it is not possible to know for sure that any arb trade will be profitable. If you buy one month and short another, your ability to exit either position will depend on the market for the contract you wish to exit. If for some reason the market to crash in December 2021, after the November 21 contract expires, it is possible the December 21 contract will go way down while the November 21 contract doesn’t move.

I’d rather be trading a futures contract that allows for the delivery of the underlying, which is much easier to arb.

The futures curve usually is influenced by the internet rates associated with borrowing the underlying and with borrowing dollars, along with costs associated with holding the underlying, if any.

[mk4]

I'm pretty sure none of the people excited about the Bitcoin futures ETF was excited to use it in the first place, but more of excited because it sort of gives more "legitimacy" to bitcoin now, and it definitely gives a lot of press coverage.

But yea, an actual bitcoin non-futures ETF is where it's at.

[Quickseller]

But yea, an actual bitcoin non-futures ETF is where it's at.

An ETF that tracks a deliverable futures contract would be much better than what we have. I would compare the current futures-based ETF to the grayscale investment trust, there isn't much to guarantee it will actually track the price of bitcoin.

[mk4]

But yea, an actual bitcoin non-futures ETF is where it's at.

An ETF that tracks a deliverable futures contract would be much better than what we have. I would compare the current futures-based ETF to the grayscale investment trust, there isn't much to guarantee it will actually track the price of bitcoin.

It's actually worse than GBTC if we're looking at it in the investing perspective. Futures ETFs has expiration dates, so you can't literally just hold it in your portfolio passively whereas you can with GBTC.

[Poker Player]

Yes, the truth is that it doesn't make much sense. Well, maybe the only one is that there are institutional investors who can't buy Bitcoin directly, and have less trouble buying the ETF, but I understand that they can buy GBTC, as mk4 points out.

In the stock market, ETFs make sense because it is a single product in which you have many assets, for example, in an ETF of the S&P 500 you have the 500 companies in a single product, and there it makes sense to pay fees. But paying for an ETF that only has one component and paying large fees is more difficult to justify.

[Quickseller]

But yea, an actual bitcoin non-futures ETF is where it's at.

An ETF that tracks a deliverable futures contract would be much better than what we have. I would compare the current futures-based ETF to the grayscale investment trust, there isn't much to guarantee it will actually track the price of bitcoin.

It's actually worse than GBTC if we're looking at it in the investing perspective. Futures ETFs has expiration dates, so you can't literally just hold it in your portfolio passively whereas you can with GBTC.

There isn't any expiration date on the ETF that was recently approved. The ETF invests in front-month CME BTC futures contracts. As was explained in the OP, the price on future dated futures contracts tends to be higher than the current month futures contract. So when the futures contracts the ETF is holding expires, it will need to sell the contracts it holds, and purchase the next months contract at a higher price. Over time, this will eat into the ETF's returns.