The cut continued a volatile trading history. In Q4 2025, Brevan Howard sold about 31.2 million shares, reducing its position by approximately 86%, from 36.7 million shares to about 5.5 million and cutting the reported value from roughly $2.4 billion to $275 million. Later filings and reporting show that its exposure subsequently changed again, underscoring why a single quarter-end balance should not be treated as a permanent investment view.

Graham Capital also made a large reduction—but the exact figures are unclear

Available reporting says Graham Capital reduced its IBIT position by approximately 72% in Q2. Another report describes a cut closer to 75% and a remaining position worth about $9 million.

The supplied evidence does not consistently establish Graham’s exact Q1 and Q2 share counts. One portfolio data source instead lists approximately 2.0 million shares worth $67 million and a 38% reduction, while another earlier-period record shows 925,862 shares. Because these figures conflict with the reported Q2 cut of roughly 72%, the safest conclusion is limited to the direction and scale of the move: Graham materially reduced its disclosed IBIT exposure, but its precise before-and-after holdings and remaining value cannot be verified confidently from the available material.

Other institutional filings show a split market

The Q2 filings do not show a uniform institutional response to Bitcoin’s decline.

• JPMorgan increased its reported IBIT position to about 10.4 million shares, worth approximately $355.7 million, from roughly 8.3 million shares and about $162 million in Q1.
• Morgan Stanley also reported more IBIT shares, rising from approximately 13.4 million to 16.5 million, a 23% increase. Its reported position was worth about $549 million at quarter-end.
• UBS’s filing reflected a more complicated change in exposure. Reporting describes a shift in its disclosed options mix, while another source puts its June 30 IBIT holding at approximately 2.5 million shares worth $90 million.
• Bridgewater’s activity is not established by the supplied evidence, so it should not be included in a claim that all major institutions accumulated IBIT or Bitcoin ETFs.

Morgan Stanley’s filing also illustrates why instrument-level comparisons matter. Reports describe a reduction in external ETF units alongside the appearance of a new Morgan Stanley-branded Bitcoin trust; the filings do not prove that this represented a direct transfer of assets between the two products.

The aggregate numbers were more resilient than the headline cuts

Across the institutional universe covered by the available Q2 analysis, reported Bitcoin ETF holdings rose 7.5%, even though Bitcoin declined by roughly 14% during Q2. That means additions elsewhere outweighed the reductions by Brevan Howard and Graham Capital in the aggregate data.

The result is a market with several simultaneous signals: some macro funds cut spot ETF exposure, large banks reported larger IBIT positions, some institutions held steady, and others changed instruments or options exposure. The evidence supports a description of heterogeneous positioning, not a single institutional consensus.

A difficult market backdrop

The reductions came against a weak crypto-market backdrop. Bitcoin was reported down approximately 32% to 33% year to date by the end of the first half of 2026, while Q2 alone saw a decline of roughly 13.4% to 14.2% depending on the dataset.

ETF flows also became less supportive. U.S.-listed spot Bitcoin ETFs recorded more than $465 million of outflows across July 23–24, as concerns about higher interest rates outweighed improving policy momentum. In the week ended August 10, the 13 funds saw approximately $389.7 million in net withdrawals, reversing the prior week’s $853.5 million inflow.

These conditions may help explain why some managers reduced direct spot exposure. But the available evidence does not establish a single cause for any firm’s trade, and it does not independently verify every element of the broader policy narrative around stalled U.S. crypto legislation.

Why a 13F reduction is not a definitive Bitcoin forecast

Form 13F data is useful for tracking disclosed U.S.-listed holdings at a reporting date, but it is not a real-time portfolio record. The SEC’s 13F data infrastructure organizes periodic filings and datasets rather than continuous trading information.

A reported decline in IBIT therefore does not necessarily mean that a manager became structurally bearish on Bitcoin. It may reflect portfolio rebalancing, hedging, an arbitrage or basis-trade adjustment, liquidity needs, or a move from spot ETF shares into options or another instrument. Brevan Howard’s repeated changes in both spot and options exposure make that limitation particularly important.

The reverse is also true: a larger reported ETF position does not prove a durable bullish thesis. A 13F generally cannot reveal short positions, many derivatives, non-U.S. holdings, activity after the quarter ended, or whether exposure is held for clients or for the manager’s own balance sheet.

The most defensible reading of Q2 is therefore narrow: Brevan Howard and Graham Capital made very large disclosed IBIT reductions, while other institutions added or restructured Bitcoin ETF exposure. The filings show a divided market—not definitive proof that institutional investors as a whole abandoned Bitcoin.

The cut continued a volatile trading history. In Q4 2025, Brevan Howard sold about 31.2 million shares, reducing its position by approximately 86%, from 36.7 million shares to about 5.5 million and cutting the reported value from roughly $2.4 billion to $275 million. Later filings and reporting show that its exposure subsequently changed again, underscoring why a single quarter-end balance should not be treated as a permanent investment view.

Graham Capital also made a large reduction—but the exact figures are unclear

Available reporting says Graham Capital reduced its IBIT position by approximately 72% in Q2. Another report describes a cut closer to 75% and a remaining position worth about $9 million.

The supplied evidence does not consistently establish Graham’s exact Q1 and Q2 share counts. One portfolio data source instead lists approximately 2.0 million shares worth $67 million and a 38% reduction, while another earlier-period record shows 925,862 shares. Because these figures conflict with the reported Q2 cut of roughly 72%, the safest conclusion is limited to the direction and scale of the move: Graham materially reduced its disclosed IBIT exposure, but its precise before-and-after holdings and remaining value cannot be verified confidently from the available material.

Other institutional filings show a split market

The Q2 filings do not show a uniform institutional response to Bitcoin’s decline.

• JPMorgan increased its reported IBIT position to about 10.4 million shares, worth approximately $355.7 million, from roughly 8.3 million shares and about $162 million in Q1.
• Morgan Stanley also reported more IBIT shares, rising from approximately 13.4 million to 16.5 million, a 23% increase. Its reported position was worth about $549 million at quarter-end.
• UBS’s filing reflected a more complicated change in exposure. Reporting describes a shift in its disclosed options mix, while another source puts its June 30 IBIT holding at approximately 2.5 million shares worth $90 million.
• Bridgewater’s activity is not established by the supplied evidence, so it should not be included in a claim that all major institutions accumulated IBIT or Bitcoin ETFs.

Morgan Stanley’s filing also illustrates why instrument-level comparisons matter. Reports describe a reduction in external ETF units alongside the appearance of a new Morgan Stanley-branded Bitcoin trust; the filings do not prove that this represented a direct transfer of assets between the two products.

The aggregate numbers were more resilient than the headline cuts

Across the institutional universe covered by the available Q2 analysis, reported Bitcoin ETF holdings rose 7.5%, even though Bitcoin declined by roughly 14% during Q2. That means additions elsewhere outweighed the reductions by Brevan Howard and Graham Capital in the aggregate data.

The result is a market with several simultaneous signals: some macro funds cut spot ETF exposure, large banks reported larger IBIT positions, some institutions held steady, and others changed instruments or options exposure. The evidence supports a description of heterogeneous positioning, not a single institutional consensus.

A difficult market backdrop

The reductions came against a weak crypto-market backdrop. Bitcoin was reported down approximately 32% to 33% year to date by the end of the first half of 2026, while Q2 alone saw a decline of roughly 13.4% to 14.2% depending on the dataset.

ETF flows also became less supportive. U.S.-listed spot Bitcoin ETFs recorded more than $465 million of outflows across July 23–24, as concerns about higher interest rates outweighed improving policy momentum. In the week ended August 10, the 13 funds saw approximately $389.7 million in net withdrawals, reversing the prior week’s $853.5 million inflow.

These conditions may help explain why some managers reduced direct spot exposure. But the available evidence does not establish a single cause for any firm’s trade, and it does not independently verify every element of the broader policy narrative around stalled U.S. crypto legislation.

Why a 13F reduction is not a definitive Bitcoin forecast

Form 13F data is useful for tracking disclosed U.S.-listed holdings at a reporting date, but it is not a real-time portfolio record. The SEC’s 13F data infrastructure organizes periodic filings and datasets rather than continuous trading information.

A reported decline in IBIT therefore does not necessarily mean that a manager became structurally bearish on Bitcoin. It may reflect portfolio rebalancing, hedging, an arbitrage or basis-trade adjustment, liquidity needs, or a move from spot ETF shares into options or another instrument. Brevan Howard’s repeated changes in both spot and options exposure make that limitation particularly important.

The reverse is also true: a larger reported ETF position does not prove a durable bullish thesis. A 13F generally cannot reveal short positions, many derivatives, non-U.S. holdings, activity after the quarter ended, or whether exposure is held for clients or for the manager’s own balance sheet.

The most defensible reading of Q2 is therefore narrow: Brevan Howard and Graham Capital made very large disclosed IBIT reductions, while other institutions added or restructured Bitcoin ETF exposure. The filings show a divided market—not definitive proof that institutional investors as a whole abandoned Bitcoin.