Bitcoin 2026-09-18 daily brief — AI stance: bullish

Market-wide perp funding is 6.13%/yr on an open-interest-weighted basis across exchanges, comfortably below the roughly 11%/yr exchange-default neutral rate, so longs are not paying a premium for this rally. One major venue printed -0.75%/yr (shorts paying longs) on the breakout day, diverging 6.9 percentage points from the aggregate, which shows the leaning-short crowd on that venue was the one that got squeezed. Over the prior 13 sessions that venue's funding ran 2.6% to 9.8%/yr, also below baseline; there has been no crowded-long funding print in two weeks. Positioning has room to build before funding signals stretch; a move toward 15-20%/yr market-wide would be the first warning.

Market-wide futures open interest is $56.4B (Coinglass aggregate), up from $51.2B on 09-17 and above the 14-day peak of $53.4B on 09-05; the single-venue figure of $6.15B rose 8.7% on the same day, so the expansion was broad. Options OI is $41.3B with $3.67B traded on 09-18; put/call OI into the 09-25 quarterly is reported at 0.52, about two calls per put. Read together: leverage is rebuilding on the long side at un-crowded funding, the short side has been cleared, and the options crowd is positioned for upside. That is constructive for continuation but means the next sharp down day would liquidate longs rather than shorts. The quarterly expiry on 2026-09-25 with reported max pain at $72,000-75,000 is the positioning event at the edge of the window.

The 09-18 breakout forced $178.3M of short liquidations against $8.3M of longs (ratio 21.45), with total liquidations at the 89th percentile of the past year: a classic squeeze of shorts who sold the 09-10 to 09-17 dip. Three sessions earlier, on 09-15, the mirror image printed with $145.6M of longs liquidated on a -3.36% day, and 09-11 saw both sides hit ($72.8M longs, $74.8M shorts). Two-sided cascades within a week describe a regime where each 5% move is amplified by forced closes. With shorts largely cleared, the next cascade risk sits with the fresh longs that drove aggregate OI up 10% in one day.

Coinbase premium is -9.46 bps, meaning US spot on Coinbase trades nearly a tenth of a percent below Binance. The 14-day trend is the concern: it has fallen almost every session from +1.8 bps on 09-05 through -4.2 bps (09-11), -8.0 bps (09-16 and 09-17) to -9.46 bps on the breakout day, and now sits just above the -10 bps bottom-decile threshold. Offshore is leading this rally; US spot demand is not confirming it, which is in tension with the +4,014 BTC ETF print for 09-18 (ETF creations settle through authorized participants and need not show up as Coinbase spot buying). Historically a premium this negative is tagged with regional de-risking. A flip back toward zero would confirm US buyers joining; a break through -10 bps would strengthen the reversal case.

CONFLICT. The macro tape (first Fed hike in three years on 09-16, 10-year at 5.01%, DXY above 100, BoJ hike on 09-18) argues for de-risking; the local tape (new 90-day high, $178.3M of short liquidations, OI-weighted funding at 6.13%/yr, ETF inflows two days running, zero top triggers) argues for continuation. This brief sides with the local tape for the 5-7 day window: BTC rallied 5.73% two days after the hike, VIX fell to 14.81, the S&P held 7,650 and gold sits at 4,416, so risk assets are reading the backdrop as fiscal and debasement-driven (M2 +5.41% YoY) rather than as a liquidity drain. The macro headwind is real on a multi-week horizon and is the reason confidence is medium rather than high.

Flow reversal plus squeeze. The supplied ETF series shows net inflows of about 2,088 BTC on 09-17 (Farside: +$159.5M, IBIT +$183.7M, FBTC -$16.6M) and about 4,014 BTC on 09-18 (79th percentile, roughly 8.4x daily miner issuance, roughly $325M at the day's price) after outflows of 5,958 BTC (09-15, -$450.4M) and 3,886 BTC (09-16, -$295.9M); the 7-day average is still -794 BTC (25th percentile) while the 30-day average is +1,296 BTC (54th percentile). Directional interpretation of these flows is a hypothesis, not admitted research. Regulatory: the Clarity Act cloture vote failed 49-50 on 09-15, effectively ending 2026 Senate market-structure work, while the House Financial Services Committee advanced the strategic Bitcoin reserve bill H.R. 8957 by 28-21 on 09-16 with purchase provisions stripped. Miner economics: Puell 1.06 (47th percentile) neutral; hash ribbons ratio 0.9994 (15th percentile) still in the capitulation-signal state; fees 448 sats/tx (11th percentile) despite transaction count at the 99th percentile. Cycle: MVRV-Z 0.905 and NUPL 0.342 (both 39th percentile); LTH-SOPR 0.905 (22nd percentile) means long-term coins that move are on average sold at a loss, consistent with a recovery from the 90-day low of 58,566 rather than late-cycle distribution; STH-SOPR 1.011 (85th percentile) shows recent buyers taking profit into the high. Event: Deribit Q3 quarterly expiry 2026-09-25 08:00 UTC, about $14.63B of BTC options, put/call 0.52, reported max pain $72,000-75,000, roughly 8-12% below spot. Bitcoin Core 32.0 targets 2026-10-10, outside this window.

hawkish. The FOMC raised the target range 25bp to 3.75-4.00% on 2026-09-16, the first hike in three years, stating inflation remains elevated; the supplied effective fed funds print of 3.63% predates that move's pass-through. The 2027 dot plot split eight officials for another hike, six on hold, four for cuts. The 10-year Treasury is at 5.01%, up 24bp from 4.77% on 09-05. M2 is still growing 5.41% YoY, so tightening is via the price of money rather than an outright money-supply contraction. The Bank of Japan added a 25bp hike to 1.25% on 2026-09-18, a 31-year high, in a 7-2 split. Crypto sentiment has not been dented: Fear & Greed 56 (Greed), up from the range chop, but well short of euphoric.

10-year Treasury yield at 5.01%, up 24bp over 13 sessions and rising or flat every day of the window (4.77% on 09-05, 4.95% by 09-12, 5.01% on 09-18), pushed by the Hormuz pipeline attack oil spike and the hike. No credit-spread feed is supplied here, so no read on credit conditions is offered.

DXY 100.21, up about 1.5% from 98.78 on 09-08 and above 100 since 09-16. A rising dollar alongside 5% Treasury yields is the classic headwind for a non-yielding dollar-denominated asset, yet the 81,265 high was printed with DXY at its 14-day high, so the correlation has not been binding over the past week. A push through about 101 with a stalling S&P is where the dollar would start to matter inside this window.

Risk-on and complacent. S&P 500 at 7,650.5, within 0.3% of its 14-day high of 7,673.5 (09-08), with VIX at 14.81, the lowest of the 14-day window and down from 17.71 on 09-16 immediately after the FOMC. Equities absorbed the first hike in three years without a drawdown. No NDX feed is supplied here.

Downside map from 80,874: first support 79,700-80,300 (early-September closes of 79,822 on 09-05 and 80,329 on 09-06, two-day straddle low 79,701); a repeat of the 09-15 day (-3.36%) lands near 78,150, into the 78,300-78,500 cluster of the 09-08, 09-09 and 09-14 closes and the 09-25 straddle low of 78,320; a full retrace of the breakout day returns to the 09-17 close of 76,371 (-5.6%); the 30-day low is 73,021 (-9.7%) and the 50-day MA 72,471 (-10.4%), which coincides with the reported $72,000-75,000 options max-pain zone. Probability framing: the straddle pricing implies roughly a one-in-five chance of trading below 78,320 at the 09-25 expiry; because realized vol is running 1.5 to 2.5 times implied, that is a floor estimate rather than a fair one. Below 76,400 there is no in-range support until 73,000.

moderate. Implied vol is low: DVOL 35.46 at the 6th percentile of the past year and 30d ATM IV 33.4% at the 20th percentile of the last 90 days, down 2.8 vol points in five days. Realized vol is not: 52.3% over 7 days, 47.3% over 30 days, 38.0% over 90 days, with four days of +5% or more (+5.73% on 09-18, +5.01% on 09-03, +7.0% on 08-21, +5.24% on 08-20) and one of -3.36% (09-15) in the past month. Options are underpricing the movement actually occurring. Straddle-implied ranges from the Deribit book (index 81,212 at 03:17 UTC 09-19): about +/-0.9% (80,249 to 81,751) by the 09-20 expiry, +/-1.6% (79,701 to 82,299) by 09-21, +/-3.3% (78,320 to 83,680) by the 09-25 quarterly, +/-5.0% (76,939 to 85,061) by 10-02. These state how far the market expects price to travel, not which way.