What happened

The rebound did not survive the following week. After U.S. spot Bitcoin ETFs pulled in roughly $853.5 million over August 3–7 — their strongest week since mid-April, per CoinDesk — the next three sessions handed most of the enthusiasm straight back.

The tape from August 10–12:

Net across the three days: roughly −$198 million redeemed — nearly a quarter of the prior week's haul reversed in three sessions. The detail that matters is *who* did the redeeming. On the way up, BlackRock's IBIT took about 81% of the inflows. On the way down, the redemptions clustered in Fidelity's FBTC: it bled on all three days and led the tape on August 12 with a −$46.82M print, according to Odaily's SoSoValue-sourced flash. IBIT, by contrast, flipped back to a +$50.2M inflow on August 11 even as the aggregate stayed weak.

Ethereum offered no offset. Spot Ether ETFs posted a second straight day of small net outflows on August 11 (−$1.76M, after −$14.59M on August 10), cooling the five-week inflow streak we flagged on Aug 11.

A sponsor divergence, not just a red week

A choppy week after a strong one is unremarkable. The *shape* of this one is the story.

Spot Bitcoin ETF flows are creation and redemption activity, not screen trading. When an authorized participant redeems FBTC shares, Fidelity hands back Bitcoin and the fund's share count shrinks; the aggregate 'flow' number is the sum of those AP decisions across sponsors. So a week where FBTC leads the exits while IBIT holds or adds is telling you the selling is concentrated in one issuer's book — not a broad, category-wide retreat.

That asymmetry echoes the funds' different investor bases. IBIT's roughly $61.2 billion in cumulative net inflows dwarfs FBTC's ~$9.9 billion, and IBIT is the default liquidity venue with the deepest options market — the vehicle tactical desks re-enter first and exit last. FBTC's book skews more toward Fidelity-platform and advisor allocations, which can rebalance in lumpier, less liquidity-sensitive blocks. Same underlying asset, different order flow — the exact trade-offs we walk through in IBIT vs FBTC.

The upshot: the aggregate 'Bitcoin ETF flow' headline is only as representative as its lead fund on any given day. Last week that fund was IBIT and the number was green; this week the swing fund was FBTC and the number went red. Neither week is the whole market.

What it means for investors

Read the daily number, then read the fund behind it.

• Concentration cuts both ways. The same single-sponsor dominance that made the $853M week look powerful makes the aggregate fragile: when one fund's AP desks flip from creating to redeeming, the headline can swing $200M in three sessions without the underlying thesis changing. Watch the per-fund split, not just the total.
• FBTC-led redemptions ≠ IBIT-led redemptions. Selling concentrated in the second-largest fund, while the largest keeps adding, is a softer signal than a broad exit across sponsors. It has not (yet) become the latter — but a day when *both* IBIT and FBTC redeem together would be the one to worry about.
• Lopsided books widen the pre-market gap. When creations and redemptions are one-sided, the arbitrage that pins price to NAV works harder, and small premium/discount gaps are likelier to open at the pre-market open — the window Pre-Tick's estimation engine is built for. A fund seeing steady redemptions can gap slightly differently at 9:30 than one seeing inflows, even on the same overnight Bitcoin move.
• Context caps it. Spot Bitcoin ETFs remain roughly −$4.5B in net flows for 2026. A −$198M three-day stretch neither breaks nor rescues that picture; it simply confirms the rebound was episodic, as the early-August tape suggested.

None of this is investment advice. It is a reminder that with crypto ETFs, the sponsor doing the buying — or the selling — often matters more than the sign on the daily total.