Spot Bitcoin ETFs recorded a net outflow of 942 BTC on May 21, worth about $72.66 million; over the past 7 days, the cumulative net outflow was 15,915 BTC, about $1.23 billion. On the same day, Ethereum ETFs saw a net outflow of 15,222 ETH, worth about $32.44 million; over the past 7 days, the cumulative net outflow was 114,871 ETH, about $244.79 million.
If the time window is extended to an earlier week—up to the week of May 15—spot Bitcoin ETFs saw total net redemptions of about $1 billion, ending the prior trend of net inflows totaling $3.4 billion over six consecutive weeks. The pace of this outflow has shown a clear acceleration: on Monday, it was still a small net inflow of $27.29 million; on Tuesday it quickly flipped to a net outflow of $233 million; on Wednesday, the single-day net outflow reached $635 million, the highest of the week; on Thursday it briefly returned with a net inflow of $131 million; and on Friday pressure returned again, with 11 spot Bitcoin ETFs totaling a net outflow of $290 million. Ethereum ETFs also faced net outflows for five consecutive trading days, withdrawing about $254.46 million in total, bringing total assets under management down to about $12.93 billion.
From the long-cycle perspective of asset size, since Bitcoin ETFs launched in January 2024, they have累计 net inflows of about $58.34 billion, still leaving a gap of about $2.47 billion versus the historical peak of about $61.19 billion set in October 2025. Despite recent outflows, the ETF total asset size remains above $100 billion, and there has been no systematic loosening of institutional “core” holdings.
Which macro factors drove this round of ETF outflows
The macro backdrop for the large-scale outflows from ETFs has been broad-based inflation data coming in significantly above expectations. April CPI rose 3.8% year over year, the highest level since the fall of 2023; PPI surged to 6.0% year over year, close to the 2022 highs. Core PPI rose 5.2% year over year, the largest increase in more than three years. Energy costs rose sharply as oil prices climbed amid Middle East geopolitical tensions, with the energy price component in CPI rising 17.9% month over month.
These inflation readings caused markets’ expectations for the Federal Reserve’s monetary policy path to be substantially reconfigured. The implied probability of the Fed raising rates in December 2026 embedded in CME FedWatch jumped from about 2% to about 28%, and the yield on the 30-year U.S. Treasury returned to the 5% threshold. In other words, mainstream market expectations shifted from “rate cuts this year” to “the possibility of further rate hikes.” The more hawkish signals released in the Fed meeting minutes for April further reinforced this expectation, pressuring valuations of risk assets.
At the same time, risk appetite is rotating to other asset classes. In the U.S. stock market, AI-related names such as NVIDIA, Google, and Apple have continued pushing to historical highs, with significant gains on their first day after new listings. Together, these structural factors have driven the timing of crypto ETF outflows; at its core, it is a systematic repricing of risk assets under an inflation shock.
Why did Solana ETFs see net inflows against the backdrop of large BTC and ETH outflows
While both Bitcoin and Ethereum ETFs experienced large net outflows, Solana ETFs showed a clear pattern of inflows. As of May 21, Solana ETFs recorded a net inflow of 8,312 SOL in a single day, worth about $723,000; over the past 7 days, the cumulative net inflow was 203,326 SOL, worth about $17.69 million. Extending the window to the entire month of May, Solana ETFs cumulatively attracted net inflows exceeding $103 million. In an earlier round of 11 trading days, Solana ETFs pulled in as much as $1.12 billion in net inflows, with net outflow on zero of the 11 days in a row.
This divergence in fund flows reflects a deeper shift in institutional allocation logic. Large institutional investors are changing their crypto allocation strategy from “buying across the board” to a more granular “coin-by-coin” selection process. Solana’s differentiated competitive advantages in network performance, throughput capacity, and the developer ecosystem—especially its unique appeal in consumer-oriented crypto products and DeFi applications—help explain this behavior.
It is also worth noting the concentration of Solana ETF inflows. In a given week, Bitwise’s BSOL ETF accounted for about 92% of total inflows into the Solana ETF category, meaning structural risk from funds being highly concentrated in a single product cannot be ignored either.
How do Q1 institutional 13F holdings reveal tactical adjustments in crypto ETF positions
The Q1 2026 13F holding reports reveal different operational paths taken by major U.S. institutions in crypto ETFs. By institution type, the position adjustments show three typical directions:
The first is actively reducing exposure for risk aversion. Harvard Management’s IBIT holdings fell from about $266 million at the end of 2025 Q4 to about $117 million at the end of 2026 Q1, a cut of roughly 43%; meanwhile, all ETHA it still held in the prior quarter was fully liquidated. Goldman Sachs also sharply reduced its crypto ETF exposure, cutting its iShares Ethereum Trust spot position by about 74% and exiting all XRP- and Solana-related ETFs. Millennium Management’s IBIT holdings declined by about 43.8%, with ETHA holdings dropping in parallel.
The second is adding exposure against the trend. During a Bitcoin price pullback, JPMorgan consistently increased its holdings; its IBIT holdings grew by about 174%, adding about $162 million in market value. This difference indicates significant strategic divergence among institutions, and not all are trimming at the same time.
The third is cross-asset rotation. While the above institutions compressed their crypto ETF exposure, they generally reallocated positions into assets related to the AI compute power chain, forming a structural rebalancing of “less crypto, more AI.” On the crypto equity side, Goldman Sachs significantly increased its allocation to Circle (up 249%) and Galaxy Digital (up 205%), overall reflecting a strategy profile of “compress ETF exposure and pivot to selected individual stocks.”
Does large-scale ETF outflow signal that institutions are withdrawing from crypto assets long-term
From the data perspective, the current nature of ETF outflows looks more like a defensive tactical adjustment rather than a long-term withdrawal trend. Four dimensions support this view:
First, long-term cumulative holdings remain at high levels. As of mid-May, Bitcoin ETFs have累计 net inflows of about $58.34 billion, leaving a gap of about $2.47 billion versus the historical peak, but the absolute scale is still near historical highs. Ethereum ETFs累计 net inflows total $11.6 billion.
Second, the structural contradiction between whale selling pressure and ETF demand is worth watching. Bitcoin addresses with holding periods exceeding five years have sold about 38,400 BTC since the beginning of this year, equivalent to roughly three months of demand from a typical exchange-traded fund. Wallets holding 3 to 5 years have seen their share of holdings fall from 13% since the end of 2025 to below 10%. This behavior of cashing out by “old money” absorbs much of the buy-side demand that ETF inflows would otherwise represent, creating an overlapping effect with the timing of ETF outflows.
Third, long-term catalysts still exist. In March 2026, the SEC and CFTC jointly issued guidance on five categories of digital assets, clearly defining 16 crypto assets as digital commodities and moving them out of SEC jurisdiction. SpaceX disclosed in its IPO filing that it holds 18,712 BTC, further reinforcing the long-term narrative of “corporate allocation to Bitcoin.” These structural factors do not disappear due to short-term capital disturbances.
Fourth, capital rotation is happening within crypto asset categories, not through a systemic exit. Solana ETFs’ large net inflow in May and the performance of XRP ETFs both show that institutions are reallocating positions among different crypto assets. This “switching between assets” characteristic suggests this round of outflows is closer to defensive reconfiguration rather than a withdrawal of crypto asset categories themselves.
Taken together, as of May 21, U.S. Bitcoin ETFs had a weekly net outflow of $1.23 billion, while Ethereum ETFs had a weekly net outflow of $245 million. This round of outflows has been driven jointly by macro factors such as inflation data exceeding expectations, rising expectations of rate hikes, and higher U.S. Treasury yields. Institutions are responding defensively by compressing their exposure to crypto ETFs and adjusting the direction of asset allocation. However, ETF long-term cumulative holdings remain near historical highs, long-term structural catalysts have not disappeared, and capital is rotating within crypto asset categories rather than exiting systemically—these features together indicate that current ETF outflows are more likely a tactical adjustment rather than a trend reversal.
Frequently Asked Questions
Q: How large were the weekly net outflows for Bitcoin and Ethereum ETFs?
As of May 21, Bitcoin ETFs had a weekly net outflow of 15,915 BTC, about $1.23 billion; Ethereum ETFs had a weekly net outflow of 114,871 ETH, about $245 million.
Q: What mainly drove the ETF outflows in this round?
They were mainly driven by April CPI and PPI coming in far above expectations, rising expectations of rate hikes, higher U.S. Treasury yields, and geopolitical tensions in the Middle East, among other macro factors.
Q: How does the Solana ETF fund flow differ from BTC and ETH?
Solana ETFs maintained net inflows overall in May; on May 21 alone, they recorded a net inflow of 8,312 SOL, and over the past 7 days they accumulated net inflows of 203,326 SOL, showing a clear divergence from BTC and ETH ETFs.
Q: Does large-scale ETF outflow mean institutions are withdrawing from crypto assets?
Based on the current data, the outflows are more akin to a defensive tactical adjustment. Long-term cumulative ETF holdings are still near historical highs, and the evidence of capital rotating within crypto assets is clear; institutions have not systematically exited crypto asset categories.
Q: What changes in crypto ETF holdings occurred for institutions in Q1?
The Q1 13F filings show significant divergence in institutional actions: Harvard University and Goldman Sachs sharply reduced their exposure to crypto ETFs; JPMorgan increased its Bitcoin ETF holdings against the trend by about 174%; and some institutions reallocated positions toward AI compute-power related assets.
