Renting Conviction: What Bitcoin-ETF Flows Actually Measure
Author: @DarkoB1995 @IOSGVC
ETF inflows are read as a thermometer for how much big money believes in Bitcoin. Week to week, they mostly measure something else: a hidden interest-rate trade being switched on and off. Here's how to tell, how big the trade really is, and why it's been quietly leaving.
The number everyone watches
Every week the Bitcoin ETFs publish how much money flowed in or out, and the number gets read as a verdict. A big inflow means institutions are piling in; an outflow means they're losing faith. The flow figure has quietly become the market's headline gauge of conviction.
The trouble is that not everyone buying an ETF is betting on Bitcoin. Some of the largest buyers don't care which way the price goes at all — and once you account for them, the weekly flow number turns out to measure their activity far more than anyone's belief. To see why, you have to meet a different kind of buyer.
A buyer who doesn't care about the price
There's a classic, boring trade called cash-and-carry. A Bitcoin "future" is just a contract to buy or sell Bitcoin at a set price on a future date, and most of the time that future trades a little above today's price — say Bitcoin is $100 now, but a contract expiring in three months sells for $103.
A trader can pocket that $3 gap with no view on the price at all:
- Buy 1 Bitcoin today for $100 (very often by buying the ETF).
- Sell the future at $103, promising to deliver in three months.
Watch what happens at expiry. If Bitcoin rips to $120, the trader makes $20 on the coin but loses $17 on the contract — net $3. If it craters to $80, they lose $20 on the coin but gain $23 on the contract — net $3. If it goes nowhere, still $3. The profit is the same in every case. The direction cancels out; traders call this "delta-neutral." That $3 gap, expressed as a yearly rate, is the basis — effectively an interest rate the trader earns for parking money in the trade, and it's worth doing as long as it beats what they'd get risk-free in Treasury bills.
Why this pollutes the headline
Here's the catch. The first leg — buy 1 Bitcoin — is very often done by buying the ETF. So a delta-neutral trader with zero opinion on Bitcoin shows up in the data as an ETF inflow, indistinguishable on the surface from a true believer.
When a lot of cash-and-carry is being put on, inflows look strong and the "institutions are accumulating" story writes itself — even though that money is hedged and would reverse the moment the trade stopped paying. The flow number, in other words, doesn't just measure conviction. It measures how busy the arbitrage desks are. The question is how to separate the two — and how much of each there really is.
How you can tell them apart
The cash-and-carry trader leaves a second footprint. Every dollar of Bitcoin they buy is matched by a dollar of futures they sell short on the CME, the regulated U.S. exchange where institutions trade Bitcoin futures. A true believer leaves only the first footprint; the arbitrageur leaves both.
And the second footprint is public. Every week the U.S. derivatives regulator publishes a report showing how much each type of trader is long or short on the CME. One category — leveraged funds, essentially hedge funds — is where the cash-and-carry crowd lives. So you can line up, week by week, the money flowing into the ETFs against the new short positions these funds are putting on. If "demand" were really conviction, the two shouldn't have much to do with each other. If a lot of it is the hidden trade, they should move together.
What the data shows: week to week, flows follow the futures, not the price
They move together, tightly. Across every week since the ETFs launched, weeks with more new futures-shorting are weeks with more ETF inflows — almost one-for-one. About half of the entire week-to-week swing in flows can be explained by that single thing: how much new shorting the funds did. The correlation is 0.70, the kind of link you see between two things that are clearly connected rather than coincidental.
The part that should give a believer pause: the price itself explains essentially nothing. Test whether that week's Bitcoin return predicts ETF flows and the answer is statistically indistinguishable from zero. The weekly money wasn't chasing performance; it was moving in lockstep with a hedged interest-rate trade.
So as a weekly signal, ETF "demand" is mostly arbitrage. The flow number is a poor thermometer for conviction because its swings are the basis trade being switched on and off, not anyone changing their mind about Bitcoin.
But how much of the money is the trade?
This is where the easy version of the argument — "it's all fake" — falls apart, and the real story is more interesting. The basis trade dominates the week-to-week swings, but it was never most of the money.
Split every week's inflow into the part the futures-shorting explains (hedged) and everything else (directional), then add it up since launch. Of the roughly $55 billion that has flowed into the ETFs, the basis trade accounts for only about $1 billion on a net basis today — the rest is the steady, directional bid. That bid runs at about $0.4 billion a week, every week, regardless of the basis or the price, and compounded over two years it is essentially the entire mountain.
Looked at as a share of assets rather than flows, the same picture: the hedged sleeve peaked near 14% of ETF assets in 2024 and sits around 4–5% today. A meaningful minority at its height, a small slice now.
The right way to say it, then, is precise: ETF flows overstate the volatility of conviction, not its level. The weekly ups and downs are mostly rented — arbitrage capital arriving and leaving — but the assets that have actually piled up are mostly owned. The trade churns through the flow data; it was never the bulk of the balance.
And the trade is leaving
The hedged share didn't just stay small — it has been shrinking for two years. The leveraged-fund short pile built from about $3 billion at launch to roughly $14 billion by late 2024, then bled steadily back down to around $4.5 billion. The carry trade has been unwinding across the whole period, not just lately.
That matters for how you read the present. Into June the hedged book roughly halved again — the funds' short shrank from about $6.4 billion to $4.3 billion — while the ETFs bled $300–500 million a day. Off the headline, that looks like capitulation. With the futures data, it's a routine cleanup of an interest-rate trade that stopped paying. Same outflow number, two completely different stories.
When the basis compresses, the demand fades with it
The cleanest confirmation is what happens when the trade stops paying. When the $3 gap narrows toward what a trader could earn risk-free, the trade isn't worth it. If a big chunk of the weekly demand is really this trade, then weekly demand should weaken exactly when the gap compresses — and it does. Strip each series of its trend and look at the moments around a compression: ETF inflows drop below their normal pace and the funds buy back their shorts, both at once. The demand and the trade breathe together.
A true believer wouldn't care about the futures basis. This weekly "demand" clearly does.
Which comes first, and who's really doing it
First, the link is contemporaneous — it's tightest in the same week, with no clean lead or lag — and what directional evidence there is actually runs the other way: ETF flows tend to nudge the shorts, not the reverse. That fits a paired trade in which the ETF is bought first and the futures hedge follows, rather than the shorting somehow conjuring inflows. Second, the arbitrage crowd isn't the only mover. The flow tracks leveraged-fund shorts most tightly, but it also co-moves with directional institutions' positioning — both kinds of buyer are active. The claim isn't that every inflow is a hedge; it's that the hedged trade is the tightest, most reliable driver of the weekly swings.
Ether: the same trade, but the math barely works
Run the identical test on the Ether ETFs and the signature is there but weaker — the link to futures-shorting is looser, and the steady directional bid underneath is close to nonexistent. There's a clean reason. Holding spot Ether instead of the future means giving up the staking yield Ether pays, perhaps 3–4% a year. Net that out and Ether's basis is frequently negative — the carry trade often doesn't clear its hurdle at all. So the Ether ETFs have neither a strong conviction bid nor a robust arbitrage book behind them; they're simply smaller and noisier than their Bitcoin cousins.
How to read ETF flows from here
The takeaway isn't a price call. It's a way to read the flows. When the basis is fat, expect "institutional demand" to look strong and to be substantially hedged — and don't mistake the strength for belief. When the basis compresses, expect both the inflows and the shorts to fade together, and don't mistake the resulting outflows for a verdict on Bitcoin. The two numbers worth watching are the annualized basis relative to the T-bill rate, and the leveraged-fund net short in the weekly CME report. They will tell you how much of the next "demand" headline is real.
How we measured this
A few honest limits. The basis is built from the nearest CME futures contract against spot, with the last few days before each expiry dropped (their tiny time-to-expiry turns rounding into false spikes); a contract-by-contract series would sharpen the exact number but doesn't change the story. The link between flows and shorts is a strong co-movement rather than proof that one causes the other — the point is that they are two halves of the same trade. The futures-short figure is an upper bound on how much ETF buying is hedged, since some shorting hedges coins held elsewhere.
None of it changes the spine. Week to week, Bitcoin-ETF "demand" is mostly a hidden interest-rate trade, not conviction — the flows measure how busy the arbitrage is far better than they measure belief. The genuine bid is real, it is patient, and it is now most of what's left, because the rented part has spent two years going home.