May Sees Record Bitcoin ETF Outflows as Hidden Block Trade Shows Interest Still Live

Record outflows, off-exchange transfer

Last month's sequence of events cannot be reduced to the Bitcoin selloff. It marks one of the most violent waves of outflows ever observed on US spot Bitcoin ETFs since their launch in January 2024. The complex recorded ten consecutive sessions of net outflows from 15 May to 29 May totaling $2.96bn. Selling pressure peaked on 27 May, with $733.4mn of net outflows across the complex in a single session. IBIT, the largest ETF in the market, accounted for $527.8mn of that figure on its own – its second-worst session on record.

This sequence cannot be read as a uniform institutional exit from Bitcoin. On 26 May, at the heart of the redemption wave, a block trade of approximately $1.29bn on IBIT was reported off the visible order book, comprising 29.2mn shares exchanged at $43.16. According to Alex Thorn, head of research at Galaxy Digital, this transaction stands as the largest off-exchange block trade ever observed on a crypto ETF since the launch of US spot Bitcoin ETFs in January 2024. The execution took place on the secondary market off-exchange, most likely through a dark pool or broker-dealer internalization, and not as a primary creation or redemption of the trust with BlackRock.

The distinction is essential. ETF outflows measure pressure on the primary market, namely creations and redemptions processed by authorized participants that mobilize physical Bitcoin at the custodian. The block trade generates no creation or destruction of shares. It does not add to outflows, does not cancel them, and does not appear in any public flow statistic. What it does reveal is different. At the precise moment the primary market was recording massive redemptions, an IBIT exposure of systemic size was circulating on the secondary market between two institutional players, without any mobilization of the underlying Bitcoin.

Why outflows matter for the Bitcoin price

Spot Bitcoin ETFs are not mere quotation instruments. Their primary mechanism, based on creations and redemptions of share blocks (baskets), can have a direct impact on the underlying asset. In the case of IBIT, the trust holds primarily Bitcoin and adjusts its assets when primary flows occur, in coordination with its custodian. Persistent redemptions can therefore generate pressure on the Bitcoin market, either directly or through the intermediation mechanisms of the authorized participants.

It would nevertheless be excessive to mechanically equate each dollar of outflow with an equivalent spot sale. Transmission depends on operational factors such as authorized participants' inventories, settlement timelines and hedging strategies. Moreover, the relationship is bidirectional. A Bitcoin decline can trigger ETF outflows through de-risking, rebalancing or loss of momentum, while these outflows can in turn amplify the pressure on the price. In May, the most robust dynamic is not a single causality but a negative feedback loop between price, ETF sentiment and primary flows.

Price action confirms this pressure. Based on daily data, Bitcoin reached its monthly intraday peak on 6 May at $82,823, then gradually pulled back to $75,849 on 26 May, then accelerated lower, breaking below $73,000 between 27 May and 29 May (with lows around $72,400). This drop occurred even as the S&P 500 was reaching new highs, weakening the case for a broad risk-off move.

Shift from April to May

In April 2026, BTC gained 12.1%, the S&P 500 9.8%, while gold declined by 2.9%. The two risk assets rose together; gold underperformed alone. In May, the picture inverts for BTC, with a gap that widens starting 15 May. BTC closed the month at −6.2%, the S&P 500 continued its advance at 5.0%, and gold declined only 1.6%. The decoupling of BTC from equities, an 11-point gap in a single month, cannot be attributed to a risk-off reversal, since the S&P is making new highs.

Net flows follow a dynamic of equivalent amplitude and opposite sign. In April, IBIT recorded $2.01bn of cumulative net subscriptions on an aggregate complex that collected $2.02bn. IBIT captured nearly the entire positive flow, with the rest of the complex essentially flat ($8mn cumulative). In May, the complex lost $2.41bn cumulatively across 20 sessions, of which $1.41bn through IBIT alone (59% of the total) and $1bn across the rest of the complex. The gap between the $2.96bn lost over 15-29 May and the $2.41bn over the full month corresponds to the positive net subscriptions accumulated between 1 and 14 May ($555mn), before the shift.

ETF flows, the dominant driver of BTC price in May

Between August 2025 and the end of May 2026, the correlation between daily IBIT flow and daily BTC return stands at +0.28, a highly significant coefficient. Transmission between ETF flows and BTC price therefore exists structurally, not as a one-off artifact. Over the 63 sessions of the past three months (March through May 2026), this correlation intensifies to +0.59, suggesting that BTC price sensitivity to ETF flows has progressively increased as the institutional base of IBIT has expanded. Over the narrow window of May 2026, the correlation reaches +0.76 on IBIT and +0.81 on the aggregate complex.

The shift has a defensible mechanical explanation. When net primary flows are modest relative to BTC spot liquidity, their price impact is diluted by market noise. When they become large and persistent, as in May with ten consecutive sessions of outflows totaling $2.96bn, transmission to the spot market becomes mechanical. The custodian mobilizes physical BTC to honour redemptions, and this volume becomes a significant fraction of the daily order book. ETF microstructure then takes precedence over the usual macro drivers of price.

Ultimately, the May sequence should not be reduced to a simple institutional exit from Bitcoin. ETF outflows were both real and historically significant, but their interpretation requires a multi-factor lens: momentum-driven de-risking, sensitivity to macro conditions, particularly real yields and financial tightening, and institutional portfolio constraints. At the same time, the IBIT block trade highlights a key limitation of flow data, which captures primary-market activity but not secondary ownership transfers. Taken together, May reflects a phase of institutional dispersion, where visible reductions in exposure and invisible redistribution coexist within a market structure now deep enough to obscure part of the true positioning dynamics.