# Bitcoin ETFs: $169 million out and the week still in the green

> Bitcoin ETFs lose 2,123 BTC in the latest session, $169.47 million, but the week closes with 3,344 BTC net and ether adds its fourth week in the red.
> By José Maldonado · Bitcoin · 2026-10-27

Spot bitcoin ETFs closed the latest session with 2,123 BTC of net outflow, $169.47 million in the [flow tally](https://blockchain.news/flashnews/bitcoin-etfs-daily-outflows-reach-169m), and even so the seven-day balance remains positive: 3,344 BTC of net inflow and $266.96 million. The daily figure ate around two thirds of what the week had accumulated. That tug of war between the session and the weekly balance is the information that matters, far more than the headline of the day.

The week of October 19 to 23 left two sessions of heavy selling. On Thursday the 22nd, [$635.49 million](https://news.bitcoin.com/bitcoin-etf/blackrock-drives-305m-bitcoin-etf-comeback-with-676m-inflow/), the largest withdrawal since June 25, and the weekly balance held in the green after two weeks in negative territory ($229 million and $681 million). Bitcoin closed Sunday the 25th around $80,048, its third consecutive weekly close in the red and the weakest since September, after falling 7.4% from the October 5 open.

The price is 36% below the [all-time high of $126,080](https://decrypt.co/381802/bitcoin-dips-below-80000-ahead-of-fed-decision-midterms), the Fed decides on Wednesday the 28th and the midterm elections arrive on November 3. Against that backdrop, a week in the green does not mean institutional money is comfortable.

## The day that wiped out two thirds of the week

The math is worth doing before drawing conclusions: $169.47 million of outflow against $266.96 million of inflow over seven days leaves the daily withdrawal at 63.5% of the accumulated total. In a single session they gave back almost two thirds of what was captured the previous week. The pattern repeats with a regularity that no longer surprises anyone who follows these products: money comes in during rallies and leaves as soon as the price loses a round level.

BlackRock concentrated the move. Its iShares Bitcoin Trust (IBIT) lost 1,358 BTC in the session, 64% of the group's entire outflow, and Fidelity shed 391 BTC. Grayscale gave up 368 BTC across its two vehicles, while Bitwise and ARK 21Shares closed flat on the day.

The table of the five largest funds shows where the money came from and why the aggregate moved so much with so few participants. A single session does not pass sentence on the market: it shows which fund makes the call when the price moves.

Fund

Latest session (BTC)

Last seven days (BTC)

BlackRock IBIT

\-1.358

+6.792

Fidelity FBTC

\-391

\-1.522

Grayscale GBTC and BTC

\-368

\-589

ARK 21Shares ARKB

0

\-1.229

Bitwise BITB

0

\-233

Group total

\-2.123

+3.344

The month's precedent helps size up Thursday. On October 7, $484.9 million left in a single session, the largest withdrawal since June 25, as we reported when covering [bitcoin ETFs and their worst day since June](https://observatorioblockchain.com/en/bitcoin/etfs-bitcoin-485-millones-salida-octubre/). Fifteen days later, that mark had already been beaten by $150 million. So far in October, the accumulated balance of US bitcoin funds remains negative, with net outflows of around $314 million.

## The week in the green is held up by a single fund

IBIT absorbed $675.96 million in the week of October 19 to 23, more than double the $266.96 million that entered the category as a whole. Fidelity lost $84.80 million, ARK 21Shares withdrew $99.64 million and Grayscale another $51.86 million. The positive balance does not describe a market coming in as a bloc, but one manager buying while three others take the opportunity to trim.

That detail changes how the weekly flow should be read. A fund that captures more than its entire group combined is not measuring institutional appetite, but commercial distribution power. IBIT is the vehicle with the most volume and the best access on advisor platforms, and when a portfolio model rebalances, the money shows up there first. The useful question is not how much came in, but who decided to come in and why they chose that ticker.

There is a second layer of concentration, less visible on the charts. Coinbase is the sole or primary custodian of 84.5% of these funds' bitcoin, and the form count is in the analysis on [custody concentration in crypto ETFs](https://observatorioblockchain.com/en/bitcoin/concentracion-custodia-etf-cripto/).

. Buying four ETFs from four different managers can end up being a single operational bet. The diversification that appears on the investor's statement does not match the one that exists on the custodian's balance sheet, and that asymmetry only shows up on the day there is a problem with the keeper of the keys, not before.

## Ether is left without a last-minute buyer

Ethereum funds lost 6,960 ETH in the session ($16.07 million) and 85,801 ETH over seven days ($198.15 million). Last week was the fourth consecutive in negative territory, a streak bitcoin has already broken and ether has not. The price follows the move: the coin is around $2,320 this Monday, with the crypto market correcting on the weakness of tech stocks and the Fed meeting.

The asymmetry has an explanation that does not depend on the chart. When risk appetite shuts down, marginal money keeps what can count as a store of value and lets go of what needs a growth thesis. Bitcoin has had that story since 2020 and has sold it to advisors for six years; ether has it only halfway, tied to an economy of applications that has yet to prove stable flows. Four weeks of redemptions look less like a bump and more like an allocation being trimmed first.

It is worth noting that the comparison is made on different bases. Ether funds manage a much smaller asset base, so the same outflow in dollars weighs considerably more on their portfolio.

That amplifies the percentage and the sense of a stampede, without any large institutional seller being needed behind every red session. The difference in size explains the tone of the move, not its direction.

## What the ETF buyer does when the price falls

The week left a round trip worth looking at slowly. Between Monday and Wednesday, $1.074 billion entered bitcoin funds and the price touched highs of $84,777. On Thursday and Friday, $773.5 million left, with the price back toward $80,000.

The [research DWF Labs published on Thursday](https://cointelegraph.com/markets/bitcoin-etfs-rebound-weekly-inflow-301-million) on that behavior points to an uncomfortable pattern. Since January 2024, when these products began trading, the funds recorded net outflows in seven of the eight bitcoin drops of at least 15%, and the firm sums it up bluntly: ETF investors do not buy the dips as many people believe. With the price where it is, the finding stops being a statistical curiosity.

This is where the weekly flow gains value as a thermometer. A fund that comes in on the rally and leaves on the drop is not a long-term partner, but a tenant on a short lease with bags by the door.

## The week that looks to the Fed

The Fed publishes its rate decision on Wednesday the 28th and the crypto market arrives at that appointment with bitcoin stuck at $80,000, a level it has closed above for almost six weeks. Losing it on a sustained basis would open the door to the 50-week moving average, near $78,600, and to a different conversation about the cycle. Reclaiming it comfortably would give breathing room back to the funds that bought in last week's rally.

For the coming days there are two metrics worth more than any headline. The first is whether the seven-day balance of bitcoin ETFs holds in positive territory when the price tests $79,000 again, because that will show whether last week's buyer was tactical or structural. The second is whether the pace of ether redemptions moderates, a sign that the rotation is over. Everything else is noise from a single session.

The $169 million figure that opens this piece is accurate and also misleading on its own. What portrays the ETF investor in October 2026 is the sum of the table's two columns: a group that buys as a bloc when the price rises, sells as soon as it corrects and, four weeks later, still wants nothing to do with ether.
