Bitcoin Futures Basis Falls Below Treasury Yields for Second Time on Record
Bitcoin's 3-month futures basis has paid less than the 2-year U.S. Treasury every month since February 2026, only the second such stretch on record. The first one ended at the 2022 cycle low.
The carry trade that once paid 20%-plus is now losing to a 2-year government note, and the only prior comparable stretch ended at the 2022 cycle low.
Key takeaways
- Bitcoin's 3-month annualized futures basis has paid less than the 2-year U.S. Treasury yield continuously since February 2026, only the second time on record this has happened, per Glassnode's Week 30, 2026 "Paid to Wait" report.
- The one prior stretch ran from August 2022 into January 2023. It ended at the cycle low, after which Bitcoin began its recovery.
- Carry traders and arb desks are exiting. The marginal buyer of Bitcoin is no longer a hedge fund running a delta-neutral book. That shift in who owns the bid matters more than the basis number itself.
Bitcoin's once-lucrative futures carry trade is now underwater relative to the risk-free rate. According to Glassnode's Week 30, 2026 report, the 3-month futures annualized basis has been paying less than the 2-year U.S. Treasury, a spread that has been negative every month since February. Spot volume has fallen to its lowest level since 2019, and BTC is trading in the low-to-mid $60,000 range at time of writing.
The carry trade mechanics are straightforward: long spot Bitcoin (or a spot ETF), short the futures contract, and pocket the premium between the two. During the 2021 bull market, that premium ran above 20% annualized at peak across major venues. Today the math no longer clears. A dollar deployed in the carry trade earns less than a dollar sitting in government paper, and that is before pricing in execution risk, basis blowout risk, and counterparty exposure on the short leg. In risk-adjusted terms, the real effective cost to carry is deeply negative for any allocator with a mandate.
The Record Shows One Prior Comparison
Glassnode's report is direct on the historical context:
"Three-month futures basis, the yield on the cash-and-carry trade that anchors institutional participation in crypto, has been paying less than a 2-year Treasury since February. Only one other stretch on record has run this long, from August 2022 into January 2023, and it ended at the cycle low."
That prior period ran roughly five months and concluded at the bottom of the most severe drawdown of the last cycle. What followed was the recovery. That is one data point, not a law, but it is the only comparable data point that exists.
Bitcoin spot volume sitting at its lowest since 2019 fits the same picture. Futures volume has also declined from February's peak, consistent with arb desks pulling capital when the risk-adjusted math turns negative.
Who Owns the Marginal Bid Now
The more consequential shift is structural, not numeric. Carry traders are yield-seeking by definition. When the spread goes negative in risk-adjusted terms, they leave. What remains are holders who are in Bitcoin because they believe in the asset, not because of a funding rate differential. The "weak hand" arb money is exiting. Historically, that is the market cleaning itself up ahead of a sustained move.
The Bitcoin ETF era introduced institutional carry desks into the asset class at scale. Many of those desks used spot ETFs as the long leg of the trade. If carry money is now exiting, ETF flows become the diagnostic instrument. Two paths: ETF outflows track the basis compression, which would suggest that institutional Bitcoin demand was primarily arb-driven rather than conviction-driven. Or ETF inflows decouple and recover while the basis stays low, confirming that the ETF holder base is accumulating on fundamentals, not harvesting a premium. The on-chain work from Glassnode on long-term holder positioning is worth watching alongside ETF flows for exactly this reason.
The falsifiable thesis: the collapse in the basis is not a failure signal but a market efficiency signal. Carry tourists are out, inefficiencies are compressing, and a cleaner market structure is forming where the marginal buyer is a long-term holder. That thesis breaks if the basis stays sub-Treasury through another full cycle leg without a price recovery, and if HODLer-driven demand measured through ETF inflows and on-chain accumulation fails to replace carry-driven volume.
What to Watch
The trigger to confirm or deny the maturation thesis is straightforward: ETF inflows recovering while basis stays suppressed would validate it. ETF outflows tracking the basis lower would not. The August 2022 analog resolved in roughly five months. The current stretch began in February. The on-chain accumulation data from long-term holders, tracked against ETF flow direction, is the signal. The bond market backdrop remains relevant too: if Treasury yields move materially, the basis comparison resets.
Sources
Frequently Asked Questions
The carry trade goes long spot Bitcoin (or a spot ETF) and short a Bitcoin futures contract, capturing the premium between futures price and spot price as yield. When that premium collapses below Treasury yields, the incentive for institutional arb desks to participate disappears. Their exit changes who sets the marginal price. Arb desks are indifferent to Bitcoin's long-term value. Long-term holders are not.
The historical record has exactly one prior data point: the August 2022 to January 2023 stretch, which ended at the cycle low and preceded Bitcoin's recovery. That makes it a genuinely ambiguous signal, not a clear directional call. The answer depends on whether conviction-driven demand steps in to replace carry-driven volume. Watch ETF inflows and on-chain accumulation from long-term holders.
Yes, indirectly. Many institutional basis trades used spot ETFs as the long leg. If carry money exits, ETF volume and inflows may fall independent of Bitcoin's fundamentals. The diagnostic question is whether ETF outflows track the basis compression or decouple from it. Decoupling would confirm that ETF holders are genuine accumulators, not arb tourists.
