Bitcoin 2026-08-08 daily brief — AI stance: neutral

Open-interest-weighted funding across venues annualizes to 3.56%, roughly a third of the ~11% per year implied by the exchange-default eight-hourly rate. Longs are paying, but well under the neutral baseline, so neither side is crowded. One detail deserves naming because it looks alarming and is not: the single-venue rate more than tripled day over day, from about 2.1% annualized on 08-07 to 6.7% now, and the all-venue weighted figure rose from about 2.8% to 3.56%. That shape — a multiple-fold one-day jump — is exactly what turned the 2026-06-07 read bearish. Three things make it a different signal this time. The absolute level is still under the neutral baseline, where June's spike carried it through; the all-venue figure barely moved, rising 28% against the venue's 224%, so this is one desk repositioning rather than a market-wide bid; and this is a Saturday book, where funding prints on thin open interest and mean-reverts into the Monday session. The all-venue weighted rate runs 3.14 percentage points below the single-venue read, which locates the modest long lean on one exchange rather than across the market. Over the fortnight the lean has faded rather than built — that same venue was running 8.4% to 9.5% annualized in late July.

All-venue futures open interest stands at $48.25B, up 1.9% from $47.34B on 07-27 and off the $49.50B peak set on 08-05 — essentially flat commitment across a fortnight in which price also went nowhere. Options open interest is $25.96B against $35.99B on 07-31, but that step down is the end-of-July expiry rolling off rather than positioning being abandoned; it has rebuilt steadily from $25.26B on 08-01, and options volume of $2.82B on 08-08 is mid-range for the window. The composite picture is light, non-directional positioning across both futures and options, which is the same thing the volatility compression is saying from another angle: 7-day realized volatility at 10.15% annualized against 90-day at 35.24% is not a market taking a side.

The Saturday print of $2.10M shorts against $367k longs — a 5.72 to 1 ratio and a total at the 0.27th percentile of the past year — is a weekend snapshot and should not be read as a positioning fact on its own. The multi-session decay behind it is the real evidence: long liquidations ran $71.7M on 07-27 and $71.3M on 07-31, then collapsed to $6.1M, $6.6M, $8.2M and $8.1M across 08-04 through 08-07. Whatever leverage this market was carrying into the month-end flush has been cleared over eight consecutive sessions. That caps the fuel available for a squeeze higher and equally caps the depth of any liquidation-driven flush lower — a move from here has to be spot-driven, and it will not cascade.

The Coinbase premium at -4.87 bps is a US discount, comfortably inside the ±10 bps decile threshold, so the bias is neutral. It is nominally the narrowest reading in the 14-day window, but it is also a thin Saturday book and I would not build a case on the single print — the two prior weekend sessions in the window came in at -9.73 and -9.46 bps, so it is not obviously an artifact either, just not load-bearing. The nine-session trend is the readable signal: the discount has narrowed steadily from -13.78 bps on 07-30 through the -8 to -11 bps band of early August. That eases the picture of US-side selling pressure. The caveat is material — the premium stayed negative through the largest ETF inflow week since April, which means the institutional bid is arriving through primary-market creations sourced offshore or over-the-counter rather than turning up as visible Coinbase spot demand. The flow is real; it is just not the kind that lifts the US order book.

CONFLICT — and the conflict is the point. The risk-appetite side of macro (S&P +4.65% in eight sessions, VIX 14.90, DXY through 100 at 99.60) and the on-chain read (3 of 8 bottom triggers firing, MVRV-Z at the 24th percentile, no leverage crowding) both point the same way, toward the long side. The rates side does not: Fed funds at 3.63%, a 10-year at 4.69% that has round-tripped its dovish move, and a September hike priced at 44%. More decisive than either, the alignment that does exist is not being expressed in price — BTC rose 1.81% while the S&P rose 4.65% and the dollar fell 1.87%. When the tailwinds are that clear and the beta asset does not respond, the correct inference is that something local is offsetting them.

ETF flow is the dominant BTC-specific narrative and it cuts both ways. Five consecutive positive sessions ran Monday 08-03 through Friday 08-07 at 2,677 / 3,293 / 3,784 / 2,140 / 1,568 BTC, totalling 13,463 BTC, of which 11,216 BTC was absorption above miner issuance. The 30-day flow average crossed from -425 BTC on 08-06 to +12.7 BTC on 08-07, its first positive reading in the window, with the 7-day average at +1,834 BTC. Weekly US spot BTC ETF inflows reached $853.5M, the largest since the week ending April 17, and Santiment reports wallets holding 10 to 10,000 BTC adding over 20,000 BTC since 07-29. The counterweight is that markup has lagged flow — price went 63,550 to 64,906 across those five sessions, or 2.13%, and remains capped 2.04% below the 30-day high — while daily flow decelerated three sessions running from 3,784 to 2,140 to 1,568 BTC on ETF trading volume of $8.19B, down 9% week over week and the second-lowest full trading week since October 2024. Miner economics are the structural soft spot: Puell at 0.756 (25th percentile) and the Hash Ribbons ratio at 0.9863 falling on every one of the last 14 days, meaning the 30-day hash average is grinding further under the 60-day. That stress is why Hash Ribbons registers as a bottom trigger — historically the capitulation is the signal — but the same stress is a live source of miner distribution in the interim. Network usage is anomalous in a way worth naming: 653,881 transactions puts throughput at the 98.5th percentile while fees sit at 396.8 sats per transaction, the 10th percentile — heavy blockspace demand that is not fee-competitive and therefore does nothing for the miner revenue line. On catalysts, BIP-110 mandatory signaling opens 08-09 at block 961,632 with miner support reported at 0-3% against a 55% threshold, so the fork is tracking toward failing lock-in rather than splitting the chain, which removes rather than adds tail risk. The BTCPay Server and LND exploit, following the Coldcard firmware flaw that drained roughly $89M across some 4,585 addresses, is an active drag on self-custody confidence.

Hawkish. Fed funds sit at 3.63% against a 10-year at 4.69%, and the live debate is over a September hike, not a cut. July payrolls fell 23,000 against a +80,000 forecast with 103,000 of downward revisions and average hourly earnings decelerating to 3.2% year over year, which cut the implied probability of a September hike from roughly 57% to 44% within hours. That is a softening at the margin of a tightening path, not a pivot to easing. The bond market has not ratified even that much: the 10-year dipped toward 4.64% on the print and is back to 4.69% as of 08-08, exactly where it sat on 07-27. M2 growth of 5.53% year over year is the one liquidity offset pointing the other way. Resolution points are July CPI on 08-12 at 12:30 UTC and the July 28-29 FOMC minutes on 08-19. On sentiment, the Fear & Greed reading of 25 (Extreme Fear) carries an as-of date of 2026-07-18 — 21 days stale and pre-dating both the payrolls miss and the ETF inflow week, so it should be read as a historical marker rather than a live gauge.

The 10-year yield is 4.69% on 08-08, unchanged from 4.69% on 07-27, with a 14-day range of 4.61% to 4.75%. The post-payrolls move toward 4.64% has fully retraced, so rates are offering no directional impulse into the CPI print. Against Fed funds at 3.63% that leaves roughly 106 basis points of positive slope. There is no credit-spread feed in this dataset — high-yield and investment-grade spreads are simply not observed here, so I am making no credit read rather than inferring one.

DXY closed at 99.60 on 08-07, down from 101.50 on 07-27 — a 1.87% decline over eleven sessions that took it through the 100 handle without a single meaningful bounce. This is the cleanest macro tailwind available to BTC right now, and BTC has captured almost none of it: price moved from 63,750 on 07-27 to 64,906, or 1.81%, over the identical window. A dollar decline of that size producing that little bid points to a supply-side constraint on BTC rather than a demand deficit.

Unambiguously risk-on and accelerating. The S&P 500 closed 7,757.64 on 08-07 versus 7,413.18 on 07-27, up 4.65%, with VIX at 14.90 against 20.66 as recently as 07-29. Equities are making highs while BTC sits 20.49% below its own 90-day high of 81,628. That relative weakness is the single most informative cross-asset fact in this brief, and the coverage of the Hashdex Bitcoin ETF closure names the mechanism directly — investors rotating toward AI returns. BTC is not currently being bought as the high-beta expression of risk appetite.

The first defense is the 50-day MA at 63,317, 2.51% below spot, then the 30-day low of 61,849 at -4.71%, then the 60- and 90-day low of 58,519 at -9.84%. Against a 7-day one-sigma of roughly 4.45% at current implied volatility, the range low sits just outside one sigma and 58,519 sits near two — so a break of the range low is a plausible CPI-day outcome while a break of 58,519 inside a week would require a shock rather than a data surprise. The structural mitigant is that there is nothing left to liquidate: eight sessions of decaying forced closes and open-interest-weighted funding at 3.56% annualized mean any decline has to be spot-driven and will not compound through forced selling. The structural aggravant is the mirror image — five sessions absorbing 11,216 BTC of ETF creations above miner issuance bought only 2.13% of markup, so if that flow stops there is no demonstrated bid underneath. Upside is the tighter constraint: 66,257 is 2.04% away and the 200-day MA at 70,265 is 8.26% away and declining, which means the realistic 5-7 day distribution is meaningfully skewed toward more room below than above.

low

DVOL at 34.23 is at the 0.8th percentile of the past year, 30-day ATM implied at 32.1% is at the 3.3rd percentile of its rolling 90-day range, and 7-day realized volatility is 10.15% annualized against 30-day at 29.54% and 90-day at 35.24%. The term structure is in contango with 90-day implied 5.68 volatility points above 30-day, and 25-delta skew is a mild put premium of 0.96 volatility points, itself at the 3.3rd percentile of its range. Every volatility measure available sits at or near a one-year floor. Compression at these percentiles has a short half-life.