You read that a bitcoin ETF saw $400 million in inflows. It sounds like a fund manager made a decision. No such decision happened. The fund has no discretion at all. Understanding what actual
You read that a bitcoin ETF saw $400 million in inflows. It sounds like a fund manager made a decision.
No such decision happened. The fund has no discretion at all.
Understanding what actually occurred explains why two reputable sites can publish contradictory flow numbers for the same trading week.
Spot crypto ETFs work through authorized participants, usually large broker-dealers and market makers.
When demand for shares pushes the ETF price above the value of the bitcoin behind it, an authorized participant creates new shares. It delivers cash or bitcoin to the fund, and the fund issues shares in return.
When the ETF trades below the value of its holdings, the process reverses. The participant hands back shares and receives cash or bitcoin.
That is an inflow and an outflow. It is an arbitrage response to a price gap, not an investment opinion.
Creations and redemptions happen in large blocks, often 10,000 or 25,000 shares at a time. That is why flow data arrives in lumps rather than smoothly.
US spot bitcoin ETFs launched in January 2024 under a cash create model. The authorized participant delivers cash, and the fund’s own trading desk buys the bitcoin.
In-kind creation, where the participant delivers actual bitcoin, was permitted later. It is more tax-efficient and reduces the fund’s trading costs.
The distinction matters for one reason. Under cash create, there is a lag between the flow being recorded and the bitcoin actually being bought. Under in-kind, the coins move with the shares.
So a reported inflow does not always mean bitcoin was purchased that same day.
TermWhat it meansAuthorized participantFirm permitted to create and redeem shares directly with the fundCreationNew shares issued, recorded as an inflowRedemptionShares destroyed, recorded as an outflowCreation unitThe block size, often 10,000 or 25,000 sharesNAVNet asset value, the per-share value of what the fund holdsPremium or discountGap between market price and NAV that triggers the arbitrage
There are three independent reasons two sources can report different figures, and almost no article tells you which one is in play.
The first is the provider. SosoValue, Farside Investors and the issuers themselves use different settlement conventions. Same-day figures can differ by single-digit millions and both be correct.
The second is the window. Daily, weekly, month-to-date and cumulative-since-launch are all reported as flow numbers, often without a label.
The third is the basket. Bitcoin-only, bitcoin plus Ethereum, or every crypto ETF together produce very different totals from identical underlying data.
A figure that does not state provider, window and basket cannot be compared to one that does. That is the whole discipline.
Flows measure share creation. They are a reasonable proxy for net demand through the regulated wrapper, and nothing more.
They do not capture direct spot buying, over-the-counter desks, corporate treasury purchases, or anything happening outside US-listed products.
They also say nothing about who is behind the trade. A $400 million redemption could be one hedge fund unwinding a basis trade, or thousands of retail investors selling. The data does not distinguish.
Basis trades are worth knowing about specifically. A fund can buy ETF shares while shorting futures to capture the spread, which registers as an inflow but carries no directional view on price at all.
The same flow figure means different things depending on how much trading surrounded it.
A $385 million weekly outflow during heavy volume signals active repositioning. The identical number in one of the quietest weeks in two years is closer to drift.
Always pair a flow number with trading volume for the same period. If the article does not give you both, it has not given you enough.
Start with the direction and size, then immediately ask three questions. Which provider, which window, which basket.
Then check whether the figure is being compared to something measured the same way. Cumulative-since-launch numbers are frequently set against weekly numbers, which is meaningless.
Optimisus applied this in practice when covering bitcoin ETF flows that read stronger than they were, where an unbroken streak of positive days masked very small daily amounts.
For the product basics underneath all this, our explainer on bitcoin ETFs covers what the funds hold and how they are structured.
Supply-side mechanics work the same way in other corners of the market, as our piece on token unlocks and vesting lays out.
An ETF inflow is a market maker responding to a price gap. It is evidence that demand for shares outran supply on that day, in that product.
It is not a fund manager’s forecast, not a guarantee that coins were bought, and not a measure of total market demand.
Treat it as one useful signal among several, and always check what it was measured against.
This is not financial advice.
Optimisus covers crypto and technology news for readers who want the detail behind the headline.
