Bitcoin 2026-09-30 daily brief — AI stance: bearish

Open-interest-weighted funding across venues annualizes to 9.11% per year, with the single major venue at 9.31% and a divergence of only 0.19 percentage points. Both sit under the roughly 11% per year exchange-default baseline, so longs are paying nothing extra to be long after a 33% two-month rally. Over the past 14 days the single-venue 8-hour rate dipped negative twice (Sept 18 and Sept 25) and never exceeded 0.0095%; the OI-weighted rate peaked at 0.0094% on Sept 19. This is a market that has not re-levered since the Sept 21 squeeze.

Aggregate futures open interest is 52.7B, down 14.4% from the Sept 21 peak of 61.6B and lower in every session since except a small uptick today from 52.3B, consistent with the deleveraging analysts described as profit-taking rather than forced. Options open interest fell from 52.9B on Sept 23 to 36.0B on Sept 26, consistent with the late-September quarterly expiry rolling off, and has since rebuilt to 36.5B with a put-leaning skew (30-day 25-delta at +2.6 vol points, the highest in 30 days) and a positive 7-to-30-day term slope of 3.4 vol points. The single-venue perpetual book at 6.0B is down from 6.7B on Sept 22, tracking the aggregate. Net read: positioning is light and hedged, so the macro catalyst decides direction rather than a squeeze.

Sept 30 saw 48.8M of shorts liquidated against 26.2M of longs, a 1.86 ratio, with total liquidations at the 57th percentile of the past year. The Sept 21 session liquidated 392M of shorts on the 6.47% up day. From Sept 22 to Sept 29 longs were the larger casualty on 6 of 8 sessions (88M on Sept 23, 65M on Sept 28) while the 83,500 range floor held, so weak longs have been flushed inside the range without breaking it; today's flip back to shorts being hit shows bears pressing the floor are still punished. The pattern argues against a liquidation-driven cascade and for a macro-driven move instead.

Coinbase premium is -3.0 bps today and has been negative in 12 of the past 14 sessions, ranging from -9.5 bps on Sept 18 to +1.2 bps on Sept 22, with a -5.2 bps print on Sept 29. None of these cross the plus or minus 10 bps extreme threshold, so the bias is officially neutral, but the persistence of an offshore lead while the digest reports a nine-session US ETF inflow streak is notable: US spot is not leading the tape, which fits a mild risk-off tilt rather than institutional accumulation.

CONFLICT. Trend, cycle and derivatives positioning are neutral to constructive, while yields, dollar, equities and the options skew lean negative. I side with the macro tape for this horizon because the two scheduled catalysts are macro, bitcoin's September resilience coincided with a flow bid that is now rolling over, and a deleveraged book means positioning cannot generate the move on its own.

Flow momentum is the live narrative and it is decelerating. The 30-day ETF flow average is 2,155 BTC per day (76th percentile) and the 7-day average 2,275 BTC per day (70th), down 55% from 5,035 on Sept 25; the Sept 30 print is -1,503 BTC (26th percentile, 30-day z-score -0.92) and the flow-to-issuance ratio fell from 24.8x on Sept 21 to -3.4x. That daily print conflicts with the digest's nine-session $3.1B inflow streak, so the single day is unconfirmed; the deceleration in the averages is not. Whether fading inflows lead price is a hypothesis, not demonstrated alpha. Miner economics are stable: Puell 1.03, hash ribbon ratio 1.010 rising for 12 straight sessions and back above 1.0 since Sept 20, fees at the 13th percentile while transaction count sits at the 97th. Short-term-holder MVRV of 1.138 implies a cost basis near 73,400, so a pullback to 81,000 would not put recent buyers underwater. The Block on Sept 29 described price testing a long-term-holder supply cluster after leverage cleared; Reserve Risk at the 14th percentile says long-term conviction is intact. Bitcoin Core 32.0 targets Oct 10, outside the window and not a price catalyst. The SEC staking FAQ of Sept 25 has no bitcoin-specific read-through.

hawkish. The policy rate is 3.63% against a 10-year at 5.24% and M2 growing 5.66% year on year. The September dot plot median for year-end 2026 was revised up to 4.1%, which implies roughly two further hikes; soft August PCE on Sept 30 cooled October 27-28 hike odds but Forbes on Sept 29 reported markets still bracing for two hikes with December the favoured window. Fear & Greed sits at 71 (Greed), so sentiment has not yet flinched at the tightening. Inside the 5-7 day window the high-impact inputs are September payrolls on Oct 2 and the FOMC minutes on Oct 7; CPI on Oct 14 and the Oct 27-28 FOMC are outside the horizon.

US 10-year 5.24%, up 30 bp from 4.94% on Sept 19-21 and 13 bp since the Sept 25 brief, at levels last seen in 2007 per the Sept 29 coverage. The 161 bp gap over the 3.63% policy rate is term premium and fiscal pricing rather than Fed-driven. There is no credit-spread feed in this context, so no spread read is offered.

DXY 101.58 is up 1.36% from 100.22 on Sept 17, higher in 9 of the last 10 observed sessions and at the top of its 14-day range. A dollar bid on yield differentials is a headwind through the risk-asset channel. The offset is gold at 4,190 rising alongside the dollar, which signals a hard-asset bid that partly explains why bitcoin held its September gains; for this horizon the risk channel is the one that transmits through payrolls day.

Fading risk-on, not yet risk-off. S&P 500 at 7,651.5 is down 1.46% from the Sept 21 high of 7,764.7 with three consecutive down sessions; VIX 16.34 is the high of the 14-day window, up from 14.21 on Sept 22 but below the 20 threshold. Bloomberg on Sept 30 described speculative positioning winding down across risk assets, and bitcoin's 2.4% intraday gain to 85,594 that day was fully erased, matching that tape.

Base case for the bearish lean is a drift to the 81,170 to 81,240 shelf (Sept 19-20 closes before the Sept 21 spike), about 3% below spot and inside the one-week straddle range, most plausibly on a hot payrolls print on Oct 2 or hawkish minutes on Oct 7. A break of that shelf exposes the 50-day average at 77,271 (7.5% below), which coincides with the 30-day straddle-implied low of 77,659 and sits above the 30-day low of 75,590; my subjective odds of reaching it within the window are roughly one in four, conditional on the shelf breaking first. The short-term-holder cost basis near 73,400 and the 200-day at 71,261 are the structural floors below, and reaching them in 5-7 days would need a 12% move that exceeds the 30-day straddle range. The upside tail is live too: a daily close above 86,600 would make the 30-day straddle-implied high of 90,341 the next reference, and that is the scenario a long-only trend book wants.

low implied, compressed realized, expansion due. Deribit DVOL 35.23 is at the 5.5th percentile of its 1-year range; 30-day ATM IV is 35.5% (52nd percentile of 90 days) and has risen 1.6 vol points in five days. Realized vol is 9.67% over 7 days versus 41.3% over 30 days and 38.2% over 90, so the options market is pricing roughly 2.7 times the daily move the last week delivered (implied daily sigma 1.66% versus about 0.6% realized). The at-the-money straddle prices a move of plus or minus 1.45% (82,290 to 84,710) by the Oct 2 08:00 UTC expiry, which settles before the 12:30 UTC payrolls release, plus or minus 2.05% (81,790 to 85,210) by Oct 3, which captures it, and plus or minus 3.75% (80,871 to 87,129) by Oct 9. Payrolls day carries only about 0.6 points of extra premium, so the event is not richly priced. These ranges say how far, not which way.