Bitcoin 2026-08-23 daily brief — AI stance: neutral

Open-interest-weighted funding across venues annualizes to 10.49%, and the single major venue carried in this feed prints 10.95%. The exchange-default eight-hour rate annualizes to roughly 11%, so both readings sit at or just under the baseline — this is a market paying essentially nothing to be long. The gap between that single venue and the all-venue aggregate is -0.46 percentage points, inside the half-point threshold at which one venue's positioning would be worth separating out, so there is no meaningful crowding concentration to point at. The significance is what this says about the past week rather than the present level: BTC gained 23.47% between 08-16 and 08-23 and perpetual funding never left baseline for it. That places the buying in spot and ETF hands rather than in leveraged perps, and rallies financed at baseline funding do not carry the forced-unwind risk that crowded ones do. It is the single strongest argument against a violent retrace here.

Market-wide futures open interest is $55.66bn, up 18.9% from $46.79bn on 08-11 against a 21.97% price gain over the same window. OI grew slower than price, so notional leverage per dollar of market value did not expand, and it has since eased from the 08-21 peak of $56.89bn while price held near 77.5k — a mild deleveraging into strength. The single venue carried separately in this feed shows $1.45bn, roughly thirty times smaller and essentially flat across the fortnight from $1.426bn to $1.452bn, so effectively all of the aggregate growth happened elsewhere; that venue's book tells you nothing about this move. Where positioning genuinely changed is options: open interest is up 47.8% since 08-11 to $38.3bn, volume ran $8.4bn, $9.4bn and $12.0bn on 08-20 through 08-22 against a $2.0-2.5bn baseline, and with 25-delta 30-day skew at -0.0389 in the 3.3rd percentile of its trailing month, the marginal new position is a hedge rather than a chase. Net: an unlevered spot advance, hedged rather than pressed in options, with futures leverage neutral and no crowding on either side of the perpetual book.

The three sessions that made the move were a short-side massacre: $811m of shorts against $45m of longs on 08-19, an 18-to-1 ratio; $262m against $27m on 08-20; $470m against $114m on 08-21. Roughly $1.54bn of shorts versus $186m of longs across the three days. Then it stopped abruptly. On 08-22 and 08-23 the ratio inverted, with $62.4m and $43.1m of longs liquidated against $26.2m and $31.7m of shorts, and total activity fell back to the 58.6th percentile of the trailing year. The reading is unambiguous: the short base that fueled the advance has been cleared out, so continuation-by-squeeze has no fuel left, and the marginal liquidation pain has already rotated to the long side. That is a stall condition, not a reversal condition — but it does mean the next leg has to be bought outright rather than squeezed.

The Coinbase premium is -2.66 bps — offshore still fractionally in front of US spot, but well inside the ±10 bps decile that marks an extreme, so the snapshot itself is neutral. The trend carries the information. The series printed -9.09 bps on 08-11 and bottomed at -10.85 bps on 08-15, a genuine bottom-decile US discount, then compressed through -6.81, -5.01, -2.21 and -1.59 to today's -2.66. That is 8.2 bps of recovery in eight sessions, and it independently corroborates the ETF tape: US demand did not lead this rally, but it went from clearly absent to roughly parity over exactly the days the ETF creations landed. It has not crossed zero, which matters — a sustained positive premium would be the confirmation that the institutional bid is driving rather than following, and until then the US side is a participant, not the leader.

CONFLICT

The rally has an identifiable buyer. US spot BTC ETFs took roughly $1.9bn in the week ended Aug 21, the largest 2026 week, reversing a combined $392m outflow the prior week, with combined BTC and ETH net AUM up 25.4% to $96.1bn. In coin terms the daily series ran +4,631 BTC on 08-17, +2,931, +7,472, +8,344 on 08-20 and +3,920 on 08-21, with flow-to-miner-issuance hitting 19.4x and 21.0x on 08-19 and 08-20 and excess absorption of 7,087 and 7,947 BTC on those two days; 30-day flow z-scores of 2.41 and 2.47 are two-sigma events. The policy stack behind it is specific and dated: the SEC proposed Regulation Crypto Assets on Aug 18 — a $5m-over-four-years startup exemption, a $75m-per-twelve-months fundraising exemption, and a conditional safe harbor from investment-contract status — and the White House says it remains committed to CLARITY Act floor action when the Senate returns in September, the bill having cleared Senate Banking 15-9 in May. The discordant note is miner economics: hash ribbons at 0.9865 sit in the 9th percentile and have declined every single day of the fourteen-day series, while Puell is 0.881 in the 35th percentile. Miners are capitulating straight through a 22% price rally, which is a cost-structure problem rather than a demand signal, and it is why a bottom-cycle trigger is firing at a 90-day high. Note the flow data is stale by design: 08-22 and 08-23 are null (weekend) and the next Farside tabulation is the 08-24 event, so everything above is as of Friday's close.

Hawkish, with a deteriorating-growth counterweight that has not yet changed the committee. The effective fed funds rate is 3.63% inside a 3.50-3.75% target range, and the July 28-29 minutes released Aug 19 show a 9-3 hold in which all three dissents — Hammack, Kashkari, Logan — voted for a quarter-point hike, with many participants saying tightening would likely be necessary if inflation does not decline. The 10-year is 4.69% and M2 is growing 5.53% year over year. Two dated tests follow immediately: PCE on Aug 26, then Warsh's first Jackson Hole keynote as Chair on Aug 28, three weeks ahead of the Sept 15-16 FOMC. Cutting the other way, the growth data is soft — July payrolls fell 23,000 against roughly +85,000 expected, Q2 GDP came in at 1.5% below expectations, and preliminary August UMich sentiment dropped to 51.0 from 55.2 against a 54.5 consensus with one-year inflation expectations at 4.3%. That is a stagflationary mix, and it cut September hike odds without making the committee dovish. One instrument reading has to be discarded: the Fear & Greed gauge shows 25, Extreme Fear, but it is dated 2026-07-18 — 36 days stale and entirely pre-rally. It is not usable as a live sentiment read and should not be cited as one.

The 10-year is 4.69%, effectively unchanged over fourteen days from 4.72% on 08-11. The one real move came Aug 19, when Treasury said liquidity-support buybacks in the 10-to-20-year and 20-to-30-year nominal coupon sectors would at least double from $2bn to at least $4bn per operation, effective Sept 9 through the Nov 4 refunding quarter — that took 5.7bp off the 10-year to 4.647% and 9bp off the 30-year to 5.196%, after the long end had reached its highest level since 2007. Duration supply being absorbed by the issuer is liquidity-positive at the margin and is a September-dated tailwind. There is no credit-spread feed in this stack, so I have no credit read at all — that is an absent read, not a benign one, and a widening in credit is a blind spot for this brief.

DXY is 98.821, down from 100.020 on 08-12 — a 1.2% slide over eleven sessions and the lowest print in the fourteen-day series. A dollar softening into a Fed with three live hike dissents is the market discounting the rhetoric, and it removes the most reliable macro headwind for BTC over exactly the window in which the asset gained 21.97% (63,614 on 08-11 to 77,589 on 08-23). Worth pairing with gold at 4,707 and M2 growth of 5.53%: if the market believed this Fed would successfully compress inflation, neither of those would look the way it does. The hard-asset bid is being priced independently of the policy rate, which is the specific channel that lets BTC rally against a hawkish committee.

Drifting, not risk-off. The last S&P print is 7,674.37 on 08-21, down 1.6% from 7,798.99 on 08-13, with VIX at 15.13 on 08-21 — a soft tape with no stress bid at all. Both series are null for 08-22 and 08-23, so these are last-available rather than current. The point is the divergence: BTC gained roughly 22% over a week in which equities fell. This move has no equity beta underneath it, which means it cannot lean on one — it has to be carried by its own flow, and if that flow pauses there is no broader risk-on tide to hold it up.

The downside is a ladder of thin shelves, because the 69,221-to-78,447 zone was covered in three sessions and has almost no traded volume behind it. The first genuine air pocket is 72,661, the 08-20 close and the last print before the 08-21 vertical day, 6.4% below spot. Beneath that, the 08-19 close at 69,221 sits directly on top of the 200-day average at 69,026 — a confluence at roughly -11% that is the strongest level on the chart and where I would expect a first flush to be defended. Below there the 50-day at 65,149 is -16.0% and the 30-day low of 62,780 is -19.1%; reaching either implies the ETF-flow thesis failed outright, which the 08-24 Farside tabulation and the following week's prints would have to confirm before price got there. These probabilities are event-conditional rather than stationary, which is the whole reason 30-day implied is bid over 90-day: the tail is concentrated on 08-26 (PCE) and 08-28 (Warsh), with the 09-04 payrolls report and the 08-28 BLS benchmark revision behind them. A hot PCE print into a committee that already carries three hike dissents is the specific mechanism that takes this to 72,661 and through it. Absent that, the unlevered composition of the advance — baseline funding, no long crowding, shorts already cleared — argues the 69k confluence holds and the move digests sideways instead.

moderate

DVOL at 43.18 sits at the 53.2nd percentile of its own trailing year, squarely mid-distribution. On a like-for-like horizon there is no mispricing to report: 30-day realized is 44.17%, DVOL is 43.18 and 30-day ATM implied is 42.27% — all three within two points of each other. The 87.8th-percentile reading on ATM IV is measured against a 90-day lookback only, short enough to reflect an event pin rather than a regime shift. The 67.89% 7-day realized figure is a trailing artifact of three sessions (+6.92% on 08-19, +4.85% on 08-20, +7.66% on 08-21) and is already decaying, with 08-22 at -1.85% and 08-23 at +0.77%. Moderate base, a near-term event bid on top of it, and a burst that is rolling off.