Bitcoin 2026-10-03 daily brief — AI stance: bullish

Market-wide, open-interest-weighted perpetual funding is running at about +0.93% annualized, so longs are paying shorts almost nothing. One major venue is at -2.03% annualized, meaning shorts there are paying longs. Both figures are far below the roughly 11%/yr neutral baseline, so this is not a crowded, stretched or extreme long market; it is a market where leverage has stepped aside. The 2.96pp gap between the single venue and the aggregate shows that venue's long interest retreated hardest after the Oct 2 rejection. Over the last 14 days the single venue mostly ran between 4% and 9% annualized, and news described aggregate funding climbing from about 3% to 10% annualized during the Sept 30 to Oct 2 OI build; that build has now been reset to baseline.

Neutral and deleveraged. Market-wide futures open interest is 53.38bn on 2026-10-03, down 13% from the 61.56bn peak on 2026-09-21 and just above the 52.34bn trough of 2026-09-29, which news describes as near a 12-month low. The rebuild to 54.50bn on Oct 2 was partly unwound on the rejection. On a single major venue OI is 6.13bn, down from 6.70bn on Sept 22. Options open interest is 36.21bn after the September quarterly expiry removed about 16bn on 2026-09-26 (52.23bn to 36.00bn); options volume of 6.74bn on Oct 3 is elevated for a weekend session. Term structure is in contango with 30-day IV 4.4 vol points above 7-day IV, so no event premium is loaded into the coming week. Net: no crowded side, two-way liquidations, and a book that is positioned for continued range rather than a break.

Two-sided range chop. On 2026-09-21 the +6.5% day liquidated 391.6m of shorts; 2026-09-23 (-2.1%) took 88.3m of longs; 2026-09-28 took 65.2m of longs; 2026-10-02 took 100.7m of shorts and 55.0m of longs on the push to ~87,187. News then reports $433.6m of crypto liquidations, 74% longs, on the reversal to ~84,600, but the dataset's Oct 3 row shows only 1.6m long and 1.9m short at the 2nd percentile of 1y, which looks like a partial-day capture. Over the 14-day window roughly 733m of shorts and 480m of longs were liquidated. Both sides are being punished at the range edges, which is characteristic of a consolidation rather than trend-following leverage.

The Coinbase premium is -1.55 bps on 2026-10-03, well inside the +/-10 bps extreme threshold. The 14-day series has been negative on 13 of 14 days, ranging roughly -2 to -8 bps, with the only positive print at +1.24 bps on 2026-09-22 and the most negative at -8.07 bps on 2026-10-01, close to the extreme band. Offshore spot has led US spot for the entire fortnight even while US ETFs reported net inflows, which is a mild caution: the US institutional bid is not what is holding the range, and the Oct 1 reading coincided with the ETF 7-day average collapsing toward zero.

CONFLICT. The macro tape (dollar near a 17-month high, 10y near a 2002 high, CPI consensus reaccelerating to 3.7%, a hike debate for Oct 27-28) points down. The local tape (price above both rising averages, futures OI down 13% from peak with funding at baseline, 0/8 top indicators, Fear & Greed 67) points up or sideways. The brief sides with the local tape for the 5-7 day window because BTC has already absorbed this exact macro configuration for three weeks while gold broke, the weak payrolls just softened the hike risk, and the in-window calendar (ISM services Oct 5, FOMC minutes Oct 7, jobless claims Oct 8, Michigan sentiment Oct 9) contains no first-tier release. The first-tier macro tests, CPI on 2026-10-14 and the FOMC on 2026-10-27/28, fall after the forecast horizon.

Three threads. First, ETF flows: the impulse that accompanied the Sept 17-21 rally has faded, with daily net flow falling from 11,536 BTC (30d z-score 2.72) on 2026-09-21 to 375 BTC on 2026-10-02, the 7-day average dropping from 5,035 BTC (Sept 25) to 690 BTC (Oct 2), and the flow-to-issuance ratio at 0.85 on Oct 2, meaning ETF demand no longer exceeds new supply. News reports YTD flows flipped positive at +$886.8m after $6.6bn of inflows since Sept 17, and Farside logged +$102.7m on Oct 1. Whether fading flow with a stable price is bullish resilience or a missing marginal buyer is a hypothesis, not validated research. Second, supply: news reports whales sold more than 30,000 BTC into the Oct 2 push to ~87,187, and LTH SOPR at 1.169 (spiking to 1.458 on Sept 22-23) shows long-term holders realizing profit into strength. Miner economics are stable: Puell 1.00 (43rd percentile) and the hash ribbons ratio rising for 14 straight days to 1.0128. Third, regulation: the SEC proposed an adviser self-custody framework on Oct 1 and approved six 3x leveraged ETPs including the first referencing BTC and ETH on Oct 2, with more rulemakings signalled; Citi raised its 12-month BTC forecast to $113k. These are structural tailwinds, not five-day catalysts. Cycle position is mid-range with MVRV-Z 1.01 (42nd percentile) and NUPL 0.367 (43rd percentile).

hawkish, on hold with hike risk. Fed funds is 3.75% while the 10y yields 5.24%, so the long end sits roughly 150bp above the policy rate and is doing the tightening. The news calendar describes a live debate over whether the Fed hikes or cuts at the Oct 27-28 meeting, with September CPI consensus at 3.7% y/y versus 3.4% prior for the Oct 14 release. The one dovish element is M2 growing 5.66% y/y, so broad liquidity is still expanding. September payrolls of +29k against 84k consensus, with unemployment at 4.2% and 60k of downward revisions, cut hike odds and pulled the 10y from 5.29% on Oct 2 to 5.24% on Oct 3. Fear & Greed is 67 (Greed): constructive sentiment, short of extreme.

The 10y at 5.24% on 2026-10-03 is 30bp above its 4.94% level of 2026-09-20 and 2026-09-21, and news cites 5.34% on Oct 1 as the highest since 2002. The direction over the last 24 hours is down (5.29% to 5.24%) on the weak payrolls. There is no credit-spread feed in this context, so no statement about IG or HY spreads can be made.

DXY was 101.92 on 2026-10-02 (Oct 3 not supplied, weekend), up from 100.30 on 2026-09-20, and news places it near a 17-month high. A rising dollar at this level is a conventional headwind for BTC, and it is the force that knocked gold down more than 3% on the week. BTC, by contrast, rose 4.4% from 81,169 (Sept 20) to 84,743 over the same dollar advance, so the usual inverse relationship has not been binding over the last two weeks.

Risk-on but stalled near highs. S&P 500 closed 7,722.72 on 2026-10-02, 0.5% below the 7,764.70 level of 2026-09-21 and inside a 14-day range of 7,651 to 7,765. VIX printed 15.31 on Oct 2, down from 16.39 on Oct 1, and stock futures rose on the payrolls miss. NDX is not supplied here. Equities are not signalling stress; the stress is showing in gold and duration, not in stocks.

First support is the options-implied one-week floor at 82,621 (-2.5%), which sits just under the two-week range floor of 83,479 set on 2026-09-28 (-1.5%); a close through 82,600 is roughly a one-in-four outcome by Oct 9 in straddle terms. Below that there is little structure until the 50-day average at 78,556 (-7.3%), which is about a two-sigma weekly move in options terms, and the 30d low at 75,590 (-10.8%). The short-term-holder cost basis implied by STH-MVRV of 1.149 is about 73,800 (-12.9%); a test there would push a large share of the 82% of UTXOs currently in profit underwater and is where recent buyers would be most likely to capitulate. The long-term-holder cost basis implied by LTH-MVRV 1.718 is near 49,300 and is not relevant to this horizon. The realistic downside case for the week is a failed range with a close in the 82,000 to 83,500 zone; the tail case is a macro-driven break that reaches the 50-day average.

low. DVOL is 35.36, the 6th percentile of its 1-year range, and 30-day ATM IV is 32.7% at the 13th percentile of 90 days, having fallen 2.1 vol points over five days. Realized vol is 15.3% over 7 days, 37.5% over 30 days and 38.2% over 90 days, so the last week has been unusually quiet against the month. The Deribit at-the-money straddles price a move of about +/-0.95% by Oct 5 (84,197 to 85,803), +/-1.65% by Oct 6 (83,598 to 86,402), +/-2.8% by Oct 9 (82,621 to 87,379) and +/-4.6% by Oct 16 (81,092 to 88,908). Those are the options market's own estimate of how far price travels, not which way. The one-week implied one-sigma move is about 3.5% (implied daily sigma 1.56% over five days). With implied vol at the 6th percentile, these ranges will widen quickly if vol re-expands.