Funding rates flip positive across majors — but the gamma still cuts both ways
BTC perpetual funding flipped positive on Binance, OKX, and Bybit for the first time in 23 days, but dealer gamma above $100k strike means leverage is being unwound, not built.
The 08:00 UTC funding tick on Wednesday put Binance BTC-USDT perpetuals at +0.0094% (10.3% annualised), OKX at +0.0081% (8.9%), and Bybit at +0.0076% (8.3%). It is the first day in twenty-three that all three venues printed positive funding simultaneously, per Coinglass’s aggregated tape. ETH perpetual funding flipped on the same tick — Binance +0.0067%, OKX +0.0052%, Bybit +0.0058%. SOL and BNB pairs are still mixed, with Deribit’s perp staying in negative territory at -0.0021%. The pattern matters: this is the cleanest cross-venue alignment in BTC funding since the late-March squeeze that took spot to $108,000 before unwinding sixteen days later. Read alongside Tuesday’s $487m of net ETF inflows, the tape is signalling a coordinated re-engagement of leveraged longs at the margin.
The trap, and there is always one in funding-rate analysis, is that positive funding is not the same as fresh long exposure. Perp open interest on the three majors rose only 4,100 BTC notional week-on-week against a $487m spot ETF print. That dispersion — positive funding without a meaningful OI build — tells you the move is being driven by short covering and basis rebalancing, not aggressive directional longs. The gamma profile in the options book, which I covered separately on the ETF side, is the proof: dealers are long gamma above $100,000 strike, which mechanically dampens funding-driven momentum. If you are trading this tape, the funding flip is a signal to size down, not up.
Why funding flipped now, mechanically
Perpetual funding is an arbitrage tax: when perp price trades above index price, longs pay shorts to keep the contract anchored. The mechanism cleared in 2017 and has run essentially unchanged on Binance and OKX. Funding goes positive when one of three things happens — fresh leveraged longs lift the perp, short covering compresses the basis, or the index spot itself ticks up faster than perp market-makers can respond. This week looks like the second and third. The 24h spot volume on Coinbase Prime printed $2.1bn against a trailing 30-day average of $1.6bn — that is ETF-creation-related buying, not retail. Perp longs followed, not led.
| Venue | BTC-USDT funding (8h) | BTC perp OI (USD) | ETH funding (8h) | ETH perp OI (USD) |
|---|---|---|---|---|
| Binance | +0.0094% | $6.81bn | +0.0067% | $3.94bn |
| OKX | +0.0081% | $2.34bn | +0.0052% | $1.27bn |
| Bybit | +0.0076% | $3.07bn | +0.0058% | $1.61bn |
| Deribit | +0.0042% | $1.12bn | -0.0021% | $0.78bn |
| Bitget | +0.0061% | $1.84bn | +0.0033% | $0.92bn |
The gamma overlay says size down, not up
Deribit’s BTC options book carried $14.2bn of open interest into the funding flip, with dealer gamma exposure now positive in the $100,000-$112,000 corridor. That long-gamma position is what makes positive funding less informative than it would otherwise be. Mechanically: when dealers are long gamma, they sell into rallies and buy into dips to maintain delta-neutral books. That dampens realised volatility but also dampens the persistence of any funding-driven momentum. The same +0.0094% funding tick in February — when dealers were short gamma below $90,000 — produced a 4.2% rally in twenty-four hours. Wednesday’s same print has so far produced 0.8% of spot drift. Same funding signal, opposite gamma context, completely different price response.
The 25-delta risk reversal — the cleanest single measure of options skew — sits at -0.7 vol points on Deribit’s 30-day surface, against a 90-day average of -2.8. That tightening of the skew means the market is paying less for downside protection, which is consistent with a dampened-vol regime. The implied/realised gap on 30-day BTC vol now sits at +6 vol points (implied 47%, realised 41%). That is roughly the historical median, which means the variance premium is fairly priced — there is no obvious systematic short-vol trade to put on, and there is no obvious systematic long-vol trade either. If you are positioning around this funding flip, your edge has to come from spot direction, not from vol structure.
Where the basis trade fits
CME front-month BTC futures settled Tuesday at an 8.8% annualised premium to spot, calculated against the 16:00 ET reference fix. That basis is the cleanest carry trade available in the asset class: short CME front-month against long IBIT or FBTC creation baskets locks in roughly 8.5% net of fees, against three-month SOFR at 4.31% per the New York Fed. That spread — roughly 420bp over funding — is the magnet pulling cash into the asset. When the basis blows out above 10%, hedge funds and prop shops front-run the next ETF allocation. When it compresses below 6%, they unwind. Wednesday’s 8.8% sits in the constructive middle of that band.
The perp-CME basis spread is the other half of the trade. Perpetual funding annualised at 10% versus CME term basis at 8.8% gives you 120bp of intra-venue spread. That is the arb that keeps perp funding from running away from CME — when perp funding gets too hot, the carry trade is to short perp and long CME, collecting both legs. The fact that the cross-venue spread is only 120bp tells you the funding flip is not yet extreme. Anything above 200bp would attract enough basis-trade capital to compress perp funding back toward CME equivalent within hours. We track the live spread in the market hub.
ETH funding is more interesting than BTC’s
The headlines will focus on BTC because the dollar tape ran first. But the cleaner signal is in ETH. Ether perpetual funding has been more persistently negative than BTC’s through Q2 — twenty-eight of the last forty trading days printed negative on Binance — reflecting a stickier short-bias driven by the staking-yield discount and continued unwind of the spot-ETF disappointment trade. Wednesday’s flip to +0.0067% on Binance ETH-USDT is therefore a stronger directional signal than the BTC flip. ETH OI rose 31,000 ETH notional on the same tick, against a flat-to-down BTC perp OI build. Read together: this is the cleanest evidence in two months that the ETH-as-laggard trade is unwinding.
The ETH/BTC ratio printed 0.0341 at the time of the funding tick, up from a 22-month low of 0.0298 on 14 May. A 14% bounce off the floor is the kind of move that triggers systematic rotation flows from ETH-underweight portfolios. If the funding flip persists for three more sessions and ETH/BTC clears 0.0360, expect the dispersion trade — long ETH, short BTC perp — to attract real capital. That is the trade to watch, not the headline BTC funding rate.
What the historical record says about funding flips
I pulled the last fourteen instances of a coordinated three-venue BTC funding flip from negative to positive over 2024-26, defined as Binance, OKX, and Bybit all printing positive on the same eight-hour tick after at least ten sessions of negative or mixed funding. The forward returns are not as one-directional as the bull case implies.
The signal is positive on the short horizon and negative on the medium. That is consistent with positive funding marking the end of a short-covering rally rather than the beginning of a sustainable trend. The trade structure that has worked best historically is short-dated long, exit on day five to seven, then reassess. The trade that has lost money most consistently is buying the funding flip and holding through month-end. We do not curve-fit; the sample is fourteen and noisy. But the pattern is consistent enough that I would not size aggressively here.
The leverage on perp books is not what it was
Aggregate BTC perp OI sits at $14.2bn across all venues, against a 2024 cycle peak of $34bn pre-FTX-equivalent vintage. That is a meaningful deleveraging. The MiFID-style exchange surveillance regimes that came online in 2025 — including ESMA’s MiCA derivatives rulebook — have compressed maximum leverage on EU-registered venues to 20x retail, 50x professional. Binance and OKX, both operating outside that regime, still offer 125x and 100x respectively, but their institutional clients are increasingly migrating to CME and Deribit for futures exposure where the basis trade is cleaner and the counterparty risk lower. The net effect: BTC perp funding is a less levered, more institutional signal in 2026 than it was in 2021-22. Read the funding tick accordingly.
What to watch from here
Three things. First, whether funding holds positive on the next 16:00 UTC tick — single-tick flips reverse routinely; a third consecutive positive tick is meaningful. Second, whether OI builds on the next funding cycle. Positive funding with flat OI is short covering and exhausts. Positive funding with $200m+ of fresh OI is real leveraged engagement. Third, whether the basis on CME widens above 10% annualised, which would pull the cash-and-carry trade back into the asset and accelerate the spot bid. Track the live funding tape and basis chart in our tools section; the next ECB and FOMC dates are on the events calendar.
The honest read: a coordinated three-venue funding flip is information, but it is information about what just happened, not what is about to happen. The dealer gamma profile, the muted OI build, and the historical post-flip distribution all point to a tape that is constructive for spot but does not yet justify scaling perp leverage. The cleanest trade here is the basis trade — short CME front, long ETF creation baskets — which earns 8.5% net regardless of which direction spot goes. The directional trade requires patience the funding rate is not asking for.