$681M Bitcoin ETF Outflow Shows Crypto Demand Is Thin

The crypto story that matters from the last few days is not that bitcoin bounced back above roughly $83,000 after briefly trading near $80,400. Price did stabilize, but the cleaner signal came from the regulated fund channel. U.S. spot bitcoin ETFs lost $681.1 million over the trading week, calculated from the SoSoValue series cited by The Block as -$89.9 million + $118.9 million - $487.1 million - $244.1 million + $21.1 million = -$681.1 million. Farside Investors, using a slightly different table, gives nearly the same answer: -$89.8 million + $118.8 million - $484.9 million - $244.1 million + $21.1 million = -$678.9 million. The difference is $2.2 million, calculated as $681.1 million - $678.9 million = $2.2 million, or about 0.32%, calculated as $2.2 million / $681.1 million * 100 = 0.32%. That is close enough for the investment conclusion: bitcoin demand through ETFs turned negative just as macro policy became less forgiving.

Context

The timing matters. The Federal Reserve minutes released on October 7 covered the September 15-16 meeting and showed a unanimous decision to raise the federal funds target range by a quarter percentage point. In plain arithmetic, a quarter point is 0.25 percentage point, and the new target range of 3.75% to 4.00% has a midpoint of 3.875%, calculated as (3.75% + 4.00%) / 2 = 3.875%. The minutes also said most participants viewed another increase by year-end as likely appropriate, while inflation was still above the Fed's 2% goal.

That is an awkward backdrop for an asset whose public-market adoption story has leaned heavily on ETF demand. Spot bitcoin funds are not the entire market. They do, however, translate a fuzzy institutional-demand narrative into daily dollars. When the ETF complex is taking in cash, a part of the market is mechanically buying bitcoin exposure. When the complex is losing cash, that flow support is absent or reversed.

CoinDesk's live market coverage captured the immediate stress: bitcoin ETFs lost about $244 million on Thursday after about $485 million left on Wednesday, giving a two-day withdrawal of about $729 million, calculated as $244 million + $485 million = $729 million. The same report said bitcoin fell to about $80,400 and then recovered near $82,000. The price rebound was visible, but the fund-flow arithmetic says the rebound happened without a broad ETF bid.

The analysis

Start with the weekly ETF number. The SoSoValue-linked daily path in the public reports was: Monday -$89.9 million, Tuesday +$118.9 million, Wednesday -$487.1 million, Thursday -$244.1 million, and Friday +$21.1 million. The week is therefore -$89.9 million + $118.9 million - $487.1 million - $244.1 million + $21.1 million = -$681.1 million. The sign matters. A positive number would mean net creations across the funds. A negative number means net redemptions.

Now compare that with Farside Investors. Its published rows show Monday -$89.8 million, Tuesday +$118.8 million, Wednesday -$484.9 million, Thursday -$244.1 million, and Friday +$21.1 million. The Farside sum is -$89.8 million + $118.8 million - $484.9 million - $244.1 million + $21.1 million = -$678.9 million. The two independent readings do not match to the decimal, but they match in direction and scale. The small gap of $2.2 million equals $681.1 million - $678.9 million = $2.2 million, which is tiny next to the weekly flow.

The day that carried the week was Wednesday. In the SoSoValue-linked table, Wednesday was -$487.1 million. Its share of the weekly outflow was 71.5%, calculated as $487.1 million / $681.1 million * 100 = 71.5%. In the Farside table, Wednesday was -$484.9 million, and its share of Farside's weekly -$678.9 million total was 71.4%, calculated as $484.9 million / $678.9 million * 100 = 71.4%. So this was not a slow, evenly distributed bleed. It was one heavy redemption day, followed by another negative day, then a modest Friday inflow.

Thursday deepened the drawdown. Using the SoSoValue-linked figures, the two-session outflow was -$487.1 million - $244.1 million = -$731.2 million. Friday's +$21.1 million inflow replaced only 2.9% of that two-session loss, calculated as $21.1 million / $731.2 million * 100 = 2.9%. That is the arithmetic behind the phrase 'thin rebound.' A green Friday flow was better than another outflow, but it did not repair the damage from the prior two sessions.

The Block reported that bitcoin ETFs hold $105.84 billion in net assets and have drawn $57.11 billion in cumulative net inflows since launch. Against that asset base, the weekly outflow was small but noticeable: $681.1 million / $105.84 billion * 100 = 0.64%. Against cumulative inflows, it was larger: $681.1 million / $57.11 billion * 100 = 1.19%. Those percentages do not describe a structural exit from bitcoin. They do show that the ETF channel can remove real marginal demand quickly when rates and risk appetite turn.

That is why the Fed context belongs in a crypto article. The minutes noted a 3.75% to 4.00% target range, whose midpoint is 3.875% as calculated above. A higher policy-rate path raises the hurdle for long-duration and non-yielding assets. Bitcoin does not have cash flows, coupons, or dividends. Its case in public markets often rests on scarcity, liquidity, and adoption. ETF outflows pressure the adoption leg of that case at the same moment that the discount-rate leg becomes less helpful.

Risks and counterpoints

The first counterpoint is scale. A $681.1 million outflow is only 0.64% of $105.84 billion in ETF net assets, calculated as $681.1 million / $105.84 billion * 100 = 0.64%. That is not a mass withdrawal. It is a weekly flow reversal. If ETF buyers return quickly, the conclusion that demand is thinning would be weakened.

The second counterpoint is measurement. Farside's -$678.9 million and the SoSoValue-linked -$681.1 million differ by $2.2 million, calculated as $681.1 million - $678.9 million = $2.2 million. Flow data can revise as issuers finalize share counts. The broad conclusion survives that gap, but precision beyond the first decimal place is not the story.

The risk that would make this analysis wrong is straightforward: the ETF channel could flip back to sustained inflows while bitcoin holds its price range. The key assumption is that ETF flows are a useful read on marginal institutional demand. If that assumption breaks because direct buying, offshore liquidity, or derivatives positioning absorbs the ETF redemptions, then the conclusion that demand is thin breaks with it.

What to do with it

Treat the $681.1 million weekly outflow as a stress gauge, not a trading signal. It says the bitcoin market has less visible ETF support than it had before the Fed minutes and the midweek selloff. It does not say what bitcoin must do next, and it does not justify buying or selling a specific security.

The practical read is to watch whether Friday's +$21.1 million inflow becomes a pattern. One day repaired only 2.9% of the prior two-session outflow, calculated as $21.1 million / $731.2 million * 100 = 2.9%. A string of inflows would change the evidence. Another large outflow would confirm that the ETF bid is still retreating. Until that arithmetic improves, the crypto rebound is better described as price stabilization than renewed demand.

Sources

• Farside Investors Bitcoin ETF Flow
• The Block: Ether ETFs extend outflow streak; Bitcoin funds snap three-week inflow run
• TFTC Bitcoin ETF Flows Today
• CoinDesk: Bitcoin slips from day's highs, set to close week with a loss
• Federal Reserve: Minutes of the FOMC, September 15-16, 2026