The Nine-Day $3.04B Streak Broke on One Speech — IBIT ETF Is Now the Entire Category

U.S. spot Bitcoin ETFs took in $924.5 million during the week of August 24–28, and the week ended with the largest single-session outflow in a month.

The sequence matters more than the total. Nine consecutive trading days of net inflows ran from August 17 through August 27, accumulating $3.04 billion. On August 28 the complex posted a $201.9 million net outflow, snapping the streak on the same session Federal Reserve Chair Kevin Warsh spoke at Jackson Hole and Bitcoin fell 3.34% to $77,413.77.

BlackRock's iShares Bitcoin Trust contributed $938.3 million of the weekly total, which is more than the category took in net. Every other fund combined was negative on the week by roughly $14 million. Grayscale's Bitcoin Mini Trust added $81.83 million, lifting its lifetime net inflow to $2.86 billion.

IBIT's cumulative historical net inflow now stands at $63.36 billion.

The prior week was larger still. The five sessions from August 17 to 21 delivered $1.92 billion, the biggest weekly total since early October 2025. Across the two weeks through August 28, the complex absorbed roughly $2.8 billion — the strongest concentrated institutional bid of 2026.

The daily record inside that stretch is worth reading. August 17 brought $297.6 million, August 18 $189.3 million, August 19 $517.2 million, and August 26 another $232.1 million. Then the reversal.

Bitcoin's price traced the same arc. It broke out of a mid-year range near $63,000, ran to $80,209.61 on August 27, fell to $77,838.65 on August 28, and trades at $78,173.52 into the final session of the month, up roughly 23% for August.

The thesis for this forecast: August was the best month U.S. spot Bitcoin ETFs have had in 2026 by a wide margin, and the calendar year is still net negative. The complex has stopped functioning as a broad institutional adoption channel and become a single-issuer allocation vehicle where BlackRock's subscriptions determine whether the category prints green or red. The August 28 reversal, arriving one day after net assets touched approximately $101 billion, is what that structure looks like when the macro turns.

August's $3.3 Billion Was the Strongest Month of 2026

The monthly number is genuinely impressive and it deserves the headline it received.

August month-to-date net inflows reached roughly $3.3 billion, making it the strongest month of 2026 and approximately double April's $1.97 billion, which had been the prior high-water mark.

The turnaround from July was violent. July finished with $172.43 million in net inflows across the entire month, the worst month of 2026, and closed on a $265.4 million outflow on July 31. The Crypto Fear and Greed Index sat at 25 — deep in extreme fear territory — as August began.

Then the first three trading days of August delivered $626 million. August 3 alone brought $170.1 million across seven funds with no ETF recording an outflow, led by IBIT at $111.4 million, Fidelity's FBTC at $33.4 million and Franklin Templeton's EZBC at $9.2 million. August 4 added $211.5 million, August 5 $244.4 million, August 6 $128.7 million and August 7 $98.8 million.

That first week totaled $853.54 million, the largest weekly figure since mid-April, with IBIT accounting for $693 million.

The middle of the month broke down entirely. August 10 posted a $144.7 million outflow, August 11 a marginal $4.9 million inflow, then August 12 at -$61.2 million, August 13 at -$131.1 million and August 14 at -$57.6 million. Four of five sessions negative.

Then the August 17 restart, which produced the nine-day, $3.04 billion run.

So August's $3.3 billion is not a smooth accumulation. It is two strong bursts separated by a week of redemptions, with the whole month bookended by outflows on July 31 and August 28.

That shape is the defining characteristic of this product category in 2026: violent, streak-driven, and entirely dependent on macro conditions that have nothing to do with Bitcoin.

The Year Is Still $2.8 Billion Negative

The number nobody leads with is the one that frames everything.

Despite the strongest month of 2026, U.S. spot Bitcoin ETFs remain net negative by approximately $2.8 billion for the calendar year.

The damage came in the first half. Spot Bitcoin ETFs recorded $5.4 billion in net outflows during H1 2026 — the first negative half-year since the products launched in January 2024. The longest outflow streak in the category's history ran 13 sessions from May 15 to June 3, 2026, shedding $4.37 billion.

August's $3.3 billion recovered roughly 61% of the first-half hole. It did not close it.

Cumulative net inflows since the January 11, 2024 launch stand at approximately $54.6 billion, having sat at $52.8 billion on August 19 and $51.8 billion through August 14. That trajectory shows how much of the lifetime total the past three weeks contributed — nearly $3 billion of a $54.6 billion cumulative figure arrived in fifteen sessions.

The read for anyone underwriting institutional adoption: the cumulative number is a stock, not a flow. It accumulates and rarely reverses because the assets sit in advisory and retirement accounts rather than trading books. But a rising cumulative total tells you nothing about current demand, and current demand in 2026 has been net negative.

The comparison across the complex sharpens it. Ethereum ETFs took $1.52 billion across ten sessions in August, making it their strongest month since the July 2024 launch. XRP ETFs took $110.49 million in the week ending August 28, their best of 2026 and the first week above $100 million since December 2025. Solana ETFs took $153.87 million across nine sessions.

All three of the smaller complexes posted their best periods on record or for the year in the same window. Bitcoin's complex — twenty times larger than Ethereum's — posted its best month of a losing year.

That is rotation, and it is visible in the flows.

IBIT Took $938 Million of a $924 Million Week

The concentration statistic is the single most important structural fact about this product category.

BlackRock's iShares Bitcoin Trust recorded $938.3 million of net inflows during the week of August 24–28. The entire category took $924.5 million. IBIT accounted for more than 100% of net flow, meaning the other eleven U.S. spot Bitcoin ETFs were collectively negative by roughly $14 million.

That is not diversified institutional demand. That is one fund.

The dominance runs through the whole month. During the eight-day version of the streak that pulled in $2.8 billion, IBIT drew $2.02 billion — 72% of everything that came in. During the three-day restart at the start of August, IBIT captured $479 million of $626 million, roughly 76%. In the first week of August it took $693 million of $853.54 million, about 81%.

Across every measurement window in August, IBIT ran between 72% and 100% of category flow.

The fund's scale explains why. IBIT holds $59.09 billion in assets at a 0.25% expense ratio against roughly $99 billion in category assets — approximately 60% of the complex in a single vehicle. Its cumulative net inflow of $63.36 billion exceeds the entire category's $54.6 billion, because other funds have net redeemed.

A structure this top-heavy is stable while flows are positive and fragile the moment the flagship reverses. That is not theoretical. IBIT's inflows come from allocation decisions at a limited number of institutional and advisory platforms rather than from dispersed retail demand, which makes the reversal function lumpy rather than smooth.

The same pattern now appears across every crypto ETF category. BlackRock's ETHA supplied $1.02 billion of Ethereum's $1.42 billion nine-day streak, 72%. Bitwise's BSOL took $25 million of a $33.5 million Solana session, 75%. Bitwise and Franklin hold 64% of XRP's cumulative flow.

One issuer decides each category. That is the state of the crypto ETF market in August 2026.

GBTC Has Bled 447,280 BTC Since Conversion

The structural drag on this complex has never stopped, and it is large enough to distort every category-level number.

Grayscale's GBTC has shed 447,280 BTC since its conversion from a closed-end trust, with cumulative net outflows reaching $23.94 billion. It continues to see redemptions.

At current prices, 447,280 BTC is roughly $35 billion of Bitcoin that left one fund. Against a category cumulative net inflow of $54.6 billion, that means the non-GBTC funds have gathered close to $79 billion gross while the category netted $54.6 billion.

That drag is structural and it is not finished. GBTC carries a materially higher expense ratio than IBIT's 0.25%, which gives every remaining holder a standing reason to switch. Each switch registers as a GBTC outflow and an IBIT inflow, inflating IBIT's dominance statistics while contributing nothing to net category demand.

How much of IBIT's $63.36 billion cumulative inflow is genuine new money versus GBTC rotation is impossible to determine from public flow data, and it is the most important unanswered question about this product category.

Grayscale has hedged the problem with its Bitcoin Mini Trust, which took $81.83 million last week and holds $2.86 billion of lifetime net inflows. That is the fee-competitive vehicle capturing some of what leaves GBTC.

The other funds occupy a distant tier. Fidelity's FBTC took $33.4 million on the strongest single day of the August restart. Franklin's EZBC took $9.2 million. ARK 21Shares' ARKB, Bitwise's BITB, VanEck's HODL, Invesco's BTCO, Valkyrie's BRRR and WisdomTree's BTCW round out the twelve-fund complex with amounts that rarely move the aggregate.

For the forecast, the practical implication is that the category's headline number understates the demand for Bitcoin exposure and overstates the health of the ecosystem. New money is arriving. Old money is leaving a legacy wrapper at a similar pace.

Fifty-Four Percent of 2026 Sessions Have Been Negative

The most damning statistic about this year sits in the session-level data.

Across the 666 trading sessions from January 11, 2024 through August 14, 2026, net flows were negative on 266 of them — about 40% of all sessions. But the share has deteriorated every year: 31% of sessions in 2024, 40% in 2025, and 54% so far in 2026.

More than half of 2026's trading days have seen more money leave U.S. spot Bitcoin ETFs than enter them.

That is a regime change, not noise. In the launch year, seven of ten sessions brought net creations. This year, fewer than five in ten do. The product category has moved from accumulation to two-way flow.

The extremes confirm it. The longest outflow streak in the category's history ran 13 sessions from May 15 to June 3, 2026, shedding $4.37 billion. The longest inflow streak — 19 sessions — ran from May 13 to June 7, 2024, in the launch year.

Set against that history, the nine-day August run looks less impressive. It was the longest positive stretch since April and it was less than half the length of the 2024 record.

The pattern of how streaks end is equally instructive and it has repeated three times this year. A seven-session July streak collected nearly $1 billion before snapping abruptly on July 23 with a $225.18 million outflow. The August restart ran three days, then broke. The nine-day run ended with $201.9 million out.

None of the three tapered. All three reversed in a single session on a macro trigger.

That is the behaviour of a flow driven by allocators watching the same rate screens rather than by conviction holders accumulating a long-term position. When the discount rate moves, the whole bid disappears at once.

Expect the September pattern to look the same.

The Nine-Day Streak: $3.04 Billion and How It Broke

The August 17–27 run is worth reconstructing because its ending explains the next month.

It started with $297.6 million on August 17, added $189.3 million on August 18, then $517.2 million on August 19 — the largest single-day figure of the entire streak. Total net assets across the funds stood at $84.3 billion at that point, with cumulative net inflows since launch at $52.8 billion.

August 19 is also the day the U.S. Treasury's expanded bond buyback policy hit the market, compressing long-end yields and triggering the liquidation cascade that took Bitcoin from $63,000 toward $80,000.

The streak ran another six sessions, adding $232.1 million on August 26 among them, and reached $3.04 billion cumulative by August 27. Net assets touched approximately $101 billion that day with Bitcoin at $80,209.61.

Then Warsh spoke. September Fed hike odds jumped from 35.4% to 59.7%. The two-year Treasury yield ripped 11.97 basis points to 4.352%. The dollar index rose 0.4% to 99.57. Bitcoin fell 3.34%. The ETF complex posted a $201.9 million outflow, and net assets settled near $97.6 billion.

One speech ended a nine-session, $3.04 billion accumulation.

The composition of the day matters. The same session saw Ethereum ETFs take in $102.18 million and XRP ETFs $26.20 million. Capital did not leave crypto — it left Bitcoin specifically and rotated down the risk curve into the smaller wrappers.

That rotation is the second signal, and it is the more concerning one for IBIT. Allocators reaching for beta after the leader has already run 23% is late-cycle positioning, and it typically precedes broad redemptions rather than continued rotation.

Watch the first three sessions of September. Consecutive outflows would confirm distribution.

Net Assets Touched $101 Billion and Settled at $97.6 Billion

The asset base is the number that determines whether flows move price, and it has been moving fast.

Combined net assets across U.S. spot Bitcoin ETFs briefly reached approximately $101 billion on August 27 before settling near $97.6 billion following Friday's outflow and the drop in Bitcoin. On August 19 the same measure stood at $84.3 billion.

That is a $17 billion increase in eight sessions, driven far more by price than by flow — $3.04 billion of inflows against a Bitcoin move from roughly $68,400 to $80,209.61.

The distinction is important because it identifies which variable actually drives the asset base. At 6.52% of Bitcoin's total market capitalisation, the ETF complex is large enough that its flows move spot, but the complex's own reported assets are dominated by mark-to-market rather than subscription activity.

The 6.52% ratio is the number to track over time. It represents how much of Bitcoin's float has migrated into vehicles whose flows are governed by macro allocation committees rather than by conviction holders. In March 2025 the comparable figure was 5.65% on $94.35 billion of net assets and $36.05 billion of cumulative inflows.

So over roughly seventeen months, cumulative inflows rose from $36.05 billion to $54.6 billion — $18.5 billion of new money — while the ownership share moved from 5.65% to 6.52%.

That is genuine, steady penetration. It is also slow.

For comparison, Ethereum's complex holds $12 billion to $13 billion in assets, XRP's $1.44 billion, and Solana's roughly $1.22 billion of cumulative inflows. Bitcoin's wrapper is between seven and eighty times larger than the alternatives, which is why its flows have a measurable spot impact and theirs largely do not.

The threshold that matters for the forecast: once ETF assets pass roughly 10% of market capitalisation, the complex becomes the marginal price setter on most sessions rather than merely an influence. At the current pace of penetration, that is a 2028 event.

The Treasury Buyback Is the Actual Driver

The cleanest explanation for August's flows has nothing to do with Bitcoin adoption.

On August 19, the U.S. Treasury announced it would double the maximum size of liquidity-support buyback operations for longer-dated bonds, with operations of at least $4 billion apiece beginning September 9. The mechanism injects liquidity into the Treasury market and compresses yields at the long end.

Long-dated yields fell. Non-coupon assets became relatively more attractive. Capital moved into spot Bitcoin ETFs as a route to expressing that. Short liquidations totaled $3.5 billion between August 19 and 22, with $1.29 billion closing inside a single hour on August 19 — the fastest concentrated squeeze of 2026.

The flow data lines up exactly. August 17 delivered $297.6 million, August 18 $189.3 million, and August 19 — buyback announcement day — $517.2 million, the largest single session of the streak.

That sequencing tells you what the marginal ETF buyer is actually trading: the rate path, not Bitcoin.

The corollary is that the same buyer leaves when the rate path reverses, which is precisely what August 28 demonstrated.

Two forward implications follow. First, the September 9 buyback operation is the single most important scheduled event for Bitcoin ETF flows this month — larger than any crypto-specific catalyst. If a $4 billion-plus operation compresses long yields the way the announcement did, the flow impulse gets a second wind.

Second, the effect of a policy already announced and priced is far smaller than the effect of its announcement. The first execution rarely repeats the surprise.

The counterweight is the Fed. Warsh's remarks pushed September hike odds to 58%, with the December reading at 80%. A hike raises the front-end opportunity cost of holding a non-yielding asset even as Treasury suppresses the long end — the two arms of U.S. policy working against each other, with Bitcoin ETF flows caught between them.

Warsh, 58%, and the Allocator's Discount Rate

The macro variable that ends streaks is the one to model.

Warsh told Jackson Hole that inflation data are more concerning than labour-market trends, that inflation is unlikely to return to target on its own, and that the Fed will have work to do if policymakers are not confident underlying inflation is heading to 2%. He cited PCE at 3.7% against the 2% objective and described financial conditions as not restrictive.

Hike odds moved from 35.4% to 59.7% within ninety minutes. The two-year closed at 4.36%, the ten-year at 4.72%, the thirty-year at 5.21%.

For an ETF allocator, the calculation is arithmetic rather than ideological. A zero-yield asset held in a taxable or advisory account competes directly against a two-year Treasury paying 4.36% with no volatility. Every basis point of expected tightening widens that gap and reduces the size of the allocation a committee is willing to approve.

That is why the flows track the rate screen rather than the Bitcoin chart.

The counter-framing is credible and comes from serious sources. CME pricing at 58% is a lean rather than a done deal — the Fed tends to validate market expectations once pricing crosses 60% to 70%, and certainty reads above 90%. Jim Bianco called it a lean hike, not a done deal. Robin Brooks of the Brookings Institution argued a September hike would be aimed at anchoring the ten-year and avoiding a repeat of the post-July 29 bond selloff, making it performative tightening whose principal aim is keeping financial conditions loose — which would leave the debasement trade intact.

Under that reading, a September hike is survivable for ETF flows and possibly bullish.

Monday's tape offers partial evidence. The two-year eased two basis points to 4.32%, the dollar backed off, and Bitcoin recovered from $77,413.77 to $78,173.52 — absorbing a live U.S. strike on Iranian territory and a 3.5% Brent move to $91.20 without reacting.

Bitcoin trades the Fed. Nothing else on the tape moves it.

The Week That Prices September

The flow forecast for September routes entirely through four U.S. macro releases and one Treasury operation.

ISM Manufacturing PMI and July JOLTS land Tuesday, September 1. ADP private payrolls print Wednesday, September 2. Challenger layoffs, jobless claims and ISM Services arrive Thursday, September 3, with Fed Governor Waller speaking. The August employment report closes the week Friday, September 4.

The labour data has been deteriorating and it is the ETF complex's route back to inflows. July nonfarm payrolls fell 23,000 against an +83,000 consensus, with government down 53,000 and private up 30,000. May and June were revised down a combined 103,000. Participation slid to 61.4%, the lowest outside the COVID era since the mid-1970s. Average hourly earnings grew 3.2% year over year, the slowest since May 2021. The preliminary benchmark revision came in at -79,000.

The Chicago Business Barometer collapsed 10.5 points to 47.1 in August, the first contraction in four months, with prices paid accelerating. Consumer sentiment fell to 51.7. Capital Economics forecasts +90,000 for August with unemployment unchanged at 4.2%.

A payrolls print below 40,000 pushes hike odds under 45% and restarts the inflow streak. A print above 130,000 with firm wages pushes odds past 70% and produces a second consecutive negative week.

Then Treasury's first doubled long-dated buyback operation on September 9, U.S. CPI on September 11, and the FOMC on September 16.

For the flow data itself, the sessions to watch are the first three of September. Bitcoin's three prior 2026 streaks all ended with a single-day reversal followed by continued redemptions rather than an immediate restart. Three consecutive outflow days would confirm the August accumulation was tactical positioning that is now unwinding.

Watch the split against Ethereum and XRP too. Continued rotation out of IBIT into the smaller wrappers is a warning that allocators are reaching for beta rather than adding exposure.

Bitcoin ETF Inflows Forecast: Outflow Risk Through Payrolls, $1.5B Upside Week

The complex enters September with the best month of a losing year behind it and a broken streak in front of it.

The bull case has five legs. August delivered roughly $3.3 billion, the strongest month of 2026 and approximately double April's $1.97 billion. The nine-day run from August 17 to 27 pulled in $3.04 billion, the longest positive stretch since April, and the two weeks through August 28 absorbed $2.8 billion. Cumulative net inflows since the January 2024 launch reached approximately $54.6 billion, with net assets touching $101 billion on August 27 before settling near $97.6 billion. IBIT's $63.36 billion of lifetime net inflows and $59.09 billion of assets at a 0.25% expense ratio give the category a flagship with genuine distribution reach. And Treasury's first doubled long-dated buyback executes September 9 — the same mechanism that produced August's entire impulse.

The bear case has five. The calendar year remains net negative by approximately $2.8 billion after $5.4 billion of first-half outflows, the first negative half-year on record. Fifty-four percent of 2026 sessions have been negative, against 40% in 2025 and 31% in 2024. IBIT accounted for more than 100% of last week's net flow, meaning the other eleven funds were collectively negative — a single-issuer dependency. GBTC has bled 447,280 BTC and $23.94 billion since conversion and continues redeeming. And all three 2026 streaks ended with an abrupt single-session reversal on a macro trigger rather than a taper.

The verdict is neutral with outflow risk through September 4. Base case for the first full week of September: net flows between -$400 million and +$700 million, with a midpoint near flat. A weak payrolls print below 40,000 restarts the streak and puts a $1.5 billion week in play, taking Bitcoin back toward $81,455. A print above 130,000 produces a second consecutive negative week and pushes Bitcoin toward $72,400. Watch $77,000 in spot and three consecutive outflow sessions as the confirmation signal.

Trade the payrolls print. This complex has stopped trading Bitcoin.

That's TradingNEWS