Bitcoin 2026-10-09 daily brief — AI stance: bearish

Market-wide, open-interest-weighted funding is about 4.1% annualized, below the roughly 11%/yr exchange-default neutral level, so perpetual positioning is not crowded in either direction. The single large venue runs warmer at about 5.3%/yr, about 1.1 points above the all-venue aggregate, so what residual long bias exists is concentrated there rather than market-wide. Over the past 14 days single-venue funding ranged from about -2.0% to +9.3%/yr and the OI-weighted figure from +0.7% to +9.1%/yr, never reaching the neutral baseline. Funding has stopped carrying a bullish lean since the 2026-10-04 top but has not flipped to the sustained negative prints that mark a washed-out bottom.

All-venue futures open interest is $51.8B, down from the $55.7B peak on 2026-10-06 (about -7%) and the lowest of the 14-day series, confirming de-leveraging rather than leveraged dip-buying. The single large venue shows the same shape at $6.10B, down from $6.37B on 10-04. Options tell the other half: options open interest rose to a 14-day high of $38.4B and volume to $7.4B, with the 30-day put-call skew at the 87th rolling percentile and 30-day ATM IV at 34.8% (49th 90-day percentile), up about 0.5 vol points in 5 days. Capital is rotating from directional futures into downside protection ahead of the 2026-10-14 CPI. Net: lightly positioned, defensively hedged, with no funding-driven squeeze fuel in either direction.

The 2026-10-06 to 10-08 stretch was a long flush: the pipeline feed shows $130M of long liquidations on 10-07 and $236M on 10-08 against $10M and $23M of shorts, and news reports put aggregate long liquidations near $1.17B, the largest since early June. On 2026-10-09 the pattern flipped to $22.7M of shorts versus $12.8M of longs, a 1.78x ratio, with total liquidations only at the 23rd percentile of the past year. Read: forced long selling is largely done and the bounce caught some fresh shorts. That removes the mechanical downside accelerant but does not supply a buyer.

The Coinbase-versus-offshore spot spread is -7.54 bps, inside the +/-10 bps extreme band but the most negative since -8.07 bps on 2026-10-01 and the second-widest of the 14-day series. The spread has been negative on all 14 days and widened from -1.19 bps on 10-07 to -5.61 and -7.54 bps on 10-08 and 10-09, meaning the 2026-10-09 bounce was led offshore while US spot was absent or a net seller. That matches the ETF outflows and the Coinbase Prime government deposit and is the main reason the bounce is not trusted.

CONFLICT. Equities and VIX say risk-on; the dollar, the 10-year and the CPI setup say tighten, and BTC's own flows (negative ETF 7-day average, Coinbase discount, government coins at Coinbase Prime) side with the tightening leg. The view follows rates, dollar and flow because that is what BTC has actually tracked for 14 days.

Flow is the story. US spot ETF net flow was +256 BTC on 2026-10-09 (about $21M) after -5,822 BTC on 10-07 and -2,988 BTC on 10-08; the 7-day average is -679 BTC (28th percentile) against +4,789 BTC on 2026-09-28, while the 30-day average is still +856 BTC. Reported US government wallets moved about 9,261 BTC to Coinbase Prime on 10-07/08, roughly 11 days of current 30-day average ETF demand and about 20 days of miner issuance at the implied 456 BTC/day, with no sale confirmed. The Coinbase premium is -7.54 bps, the most negative since 2026-10-01. Miners are stable: hash ribbons ratio 1.017 and rising for 14 straight days, Puell 1.07. Transaction count is at the 98th percentile but fees are at the 11th, so throughput is not demand-driven. In-window regulatory items (Thailand spot ETF framework effective 2026-10-16, FinCEN rule withdrawals) are marginal; the SEC comment deadline on 2026-10-20 and the FOMC on 2026-10-27/28 fall after the forecast window and are later context.

hawkish. The fed funds feed reads 3.75% after the September hike reported as a move to a 3.75%-4.00% range; the 10-year is 5.28%, up from 5.18% on 2026-09-26 and near its 14-day high of 5.31%; Cleveland Fed President Hammack, who has said more than one hike is needed, speaks on 2026-10-12. Market pricing for a hike at the 2026-10-28 FOMC has slipped to about 22% from 51% a week earlier, so this is hold-with-hiking-bias rather than active tightening. M2 is still growing 5.66% y/y, so broad liquidity is not contracting. Fear & Greed sits at 59 (Greed): sentiment has not capitulated despite the pullback.

US 10-year 5.28%, up 10 bps over 14 days with a 5.31% high on 2026-10-07; direction is up into a CPI print carrying a 3.6% y/y consensus and Brent above $100. No credit-spread feed is supplied, so no credit read is made.

DXY 102.21 is the high of the 14-day series, up about 1.1% from 101.10 on 2026-09-27, moving in step with the 10-year from 5.18% to 5.28%. A rising dollar plus rising yields is the combination this pullback has tracked: BTC fell 2.2% over the 14 days while the S&P rose 1.7%. At this stage a firm dollar caps rallies rather than forcing new lows on its own.

Risk-on. S&P 500 at 7,811.54, up from 7,683.69 on 2026-09-28 and within 0.1% of its 14-day high, with VIX at 14.84, the low of the series. BTC has decoupled to the downside from this equity tape, which says the drag is rates, dollar and BTC-specific supply rather than broad risk aversion.

First reference is the 2026-10-08 close at 81,686 (-1.1%); then the 50-day average at 80,715 (-2.2%) and the one-week straddle low at 80,456 (-2.6%). The straddle's own pricing implies roughly a one-in-six chance of settling below 80,456 on 2026-10-16 and about double that for touching it intraday during the window. Below, the two-week straddle low at 79,112 (-4.2%) and the 30-day low at 75,590 (-8.4%) are the next references; the 200-day average at 71,853 (-13%) is outside any realistic one-week range and marks the trend-regime boundary rather than a target. A disorderly path needs a new catalyst such as confirmed government sales or a CPI print well above +0.6% m/m.

moderate. Deribit DVOL is 36.39, only the 13th percentile of the past year, and 30-day ATM IV is 34.8%, while 30-day realized is 39.0% and 7-day realized 31.5%; implied is cheap relative to recent realized, a complacent setup into a scheduled event. At-the-money straddles price about +/-0.7% by 2026-10-11, +/-1.2% by 10-12, +/-2.2% by 10-14 (that expiry settles at 08:00 UTC, before the 12:30 UTC CPI release), +/-3.1% by 2026-10-16 (80,456 to 85,544, the first expiry covering CPI, PPI and the Warsh appearance) and +/-4.7% by 2026-10-23 (79,112 to 86,888). The 50-day average at 80,715 sits inside the one-week range, so the options market itself assigns a material chance of a gate test. These ranges say how far, not which way.