Bitcoin 2026-07-29 daily brief — AI stance: bearish

Perpetual funding is neutral and must not be read as crowded. Open-interest-weighted across venues it annualizes to 9.23% (0.008432% per 8 hours), below the 0.01% per 8 hour exchange default that corresponds to roughly 11% per year; the single major venue tracked separately reads 9.53%, a divergence of only 0.29 percentage points, so there is no venue-level concentration story either. Over the last two weeks the weighted rate has roughly doubled from 0.0045% per 8 hours on 07-17 to 0.0084% now, meaning longs are paying more than they were — a mild rebuild of long bias off a very soft base, still at or under the neutral line. The operational consequence is important for how a decline would look: there is no leveraged-long overhang to detonate, so the bearish case does not run through a funding squeeze. It runs through spot supply that nobody is bidding for, which is a slower, grindier mechanism than a cascade — and one that a long-liquidation-watching framework will not see coming.

Market-wide futures open interest stands at $47.27B, down 7.4% from the 07-21 peak of $51.06B — the aggregate is de-risking as price chops. The one venue tracked separately at $1.47B is at a 14-day high, so venue-level positioning is rebuilding even as the market-wide figure falls, which is consistent with the offshore-led picture in spot. Options open interest is $34.74B, near its 14-day high, on $3.22B of daily volume; the surface is calm and balanced — 30-day at-the-money implied vol 35.0% at the 16.7th percentile of the last 90 days, 25-delta 30-day skew 0.0476 at the 30th percentile of the last 30 days, and a positive 30-to-90-day slope in contango. The 07-31 monthly expiry matters mainly for what it removes: the biggest clusters at $70,000 and $72,000, including a roughly $2.5B call spread, sit 10% to 13% out of the money with BTC at 63,874 and will roll off worthless. There is no upside pin and no dealer hedging flow to lift spot into or out of that expiry — after 07-31 the tape loses a standing bullish talking point and gains nothing.

Modest and one-sided against dip-buyers. On 07-29 long liquidations were $25.9M against $22.8M short, a 0.88 ratio, with total activity only at the 35.6th percentile of the last year — nothing resembling a flush. Across the last 14 sessions longs lost roughly $324M against roughly $260M for shorts, and the largest single event was $71.7M of long liquidations on 07-27's -2.5% day. Every bounce in this range has been paid for by leveraged longs getting carried out. Two readings follow: no capitulation event has occurred, so the compressed-spring argument for a violent bottom is not yet supported; and the pattern of longs bearing the cost is characteristic of distribution ranges rather than accumulation ones.

The US spot discount is at the practical extreme. Coinbase trades 10.63 basis points below offshore, and the two-week path is a widening one: -9.34 basis points on 07-17, tightening to -4.22 by 07-23, then blowing back out to -10.63 by 07-29 — negative in every one of the last 13 sessions and now beyond the roughly 10 basis point decile threshold on an empirical range of about plus or minus 24. The framework tags this offshore-led buying, but the flow data settles which side is driving it: with US spot ETFs at -707 BTC on the 7-day average and -1,287 on the 30-day, this is US-side distribution setting the discount rather than aggressive offshore accumulation lifting the offshore price. That distinction matters for the horizon — an offshore accumulation reading would be constructive, whereas persistent domestic selling into a thinner offshore bid is the same supply overhang showing up in the cross-venue spread.

ALIGNED — both the macro tape and the flow-and-positioning read push the same way, down. The real tension is inside macro rather than between macro and BTC: the rates-and-dollar leg turned supportive after the hold (10-year to 4.61% from 4.71%, DXY to 100.91 from 101.50, September hike odds trimmed) while the volatility-and-growth leg turned hostile (VIX 20.66 from 18.21, S&P -1.5%, Brent +6.8% to $89.79). Gold at $4,133.7 with M2 up 5.53% year over year settles which leg BTC is trading on: the liquidity and debasement bid demonstrably exists and is being expressed in metal, not in bitcoin. An asset that cannot bid on a soft dollar and easier front-end yields is trading as high-beta risk, and the risk leg is the one that just broke.

The marginal buyer has gone. US spot ETFs shed 903 BTC on 07-29 and average -1,287 BTC per day over 30 days — the 7th historical percentile, and 2.02 times the roughly 447 BTC of daily issuance, so ETFs alone are distributing double what miners create. Add issuance and the market must absorb about 1,734 BTC per day, roughly $111M at current price, from sources that are not the two structural bids. The second bid has also flipped: Strategy holds 843,775 BTC at an average near $75,476, has made no purchases for five consecutive weeks, sold $544.5M of its own stock between 07-20 and 07-26 and raised its dollar reserve to $3.75B — a corporate treasury accumulating dollars against a bitcoin position that is underwater at 63,874. Institutional flow is rotating rather than returning: ether ETFs took $103.9M in the week of 07-20 to 07-24 against $33.79M for bitcoin, roughly triple, repeating the prior week, and bitcoin ETFs have recovered only about 3.3% of the $8.2B that left the category through mid-July. Miner economics are deteriorating in parallel — Puell 0.732 at the 23rd percentile, fees at 511 sats per transaction in the 11th percentile despite 663,070 transactions at the 98.6th percentile (record throughput with no fee bid, which means no competition for blockspace and no urgency), hash ribbon at 0.9898 in the 10th percentile, and an estimated -13.9% difficulty adjustment on 08-11 as mining capital rotates to AI infrastructure. On the regulatory side the Clarity Act has no floor vote scheduled and Senate leadership indicates none before the recess beginning 08-08, removing the year's principal catalyst. The single genuine offset: that -13.9% difficulty cut mechanically restores miner margins, and historically that marks the end of forced miner selling rather than the beginning of it — a reason to expect the bottom to form later in Q3, not a reason to buy this week.

hawkish. The FOMC voted 9-3 on 2026-07-29 to hold the target range at 3.50%-3.75%, effective funds 3.63%, with all three dissents in favour of a hike and no Summary of Economic Projections or dot plot at this meeting; the June dot plot had already penciled one quarter-point increase by end-2026. A hold whose only dissents are hawkish is not a pivot. It lands the day before June PCE (07-30) and two weeks before July CPI (08-12), with Brent having jumped 6.8% to $89.79 on renewed Middle East airstrikes and Hormuz shipping threats, and OPEC+ reportedly set to pause output increases for three months from October — an inflation impulse arriving into a committee that already has three members voting to tighten, and with Governor Cook noting 2026 headline inflation is tracking about a percentage point above year-ago expectations. M2 is growing 5.53% year over year and jobless claims printed 187,000 the week ending 07-18, the lowest since 1969: there is no labour-market crack to force easing. Live sentiment reads alternative.me Fear & Greed at 25, Extreme Fear — but that value is dated 07-18 and is 11 days stale, so treat it as directional, not current.

The 10-year sits at 4.61%, down 10 basis points from 4.71% on 07-24 and 07-25, as market-implied September hike odds were trimmed following the hold. Yields easing while three governors vote to hike is an unstable configuration — it resolves on the 07-30 PCE and the 08-12 CPI, not before. There is no credit-spread feed in this dataset, so no credit read is available and none should be inferred from the equity move. The scheduled rates risk inside the horizon is the 08-05 Treasury quarterly refunding, against a previously estimated $671B of privately-held net marketable borrowing for the July-September quarter; a heavier-than-expected coupon schedule pushes yields back up independent of anything the Fed does, which is a direct discount-rate headwind for a zero-cashflow asset.

DXY 100.91, down 0.59% from 101.50 on 07-27 and effectively flat across the last two weeks (100.75 on 07-17 to 100.91 now) — range-bound just above 100 with a modest post-Fed dip. At this level the dollar is not the binding constraint on BTC. What matters is the non-response: on the same dollar dip, gold trades at $4,133.7 while BTC sits 22.3% below its 90-day high. A soft dollar in an early-bear phase removes a headwind without supplying a buyer, and this tape is showing exactly that.

Risk-off turn, and it is fresh. The S&P closed 7,316.15, down 1.5% on the day and 2.6% from the 07-21 high of 7,509.20, with VIX at 20.66 versus 18.21 the prior session — a 13% jump back above 20. This dataset carries no separate Nasdaq feed, so this is an S&P-only read. The asymmetry matters: equities are only 2.6% off their high while BTC is 10.97% below its 200-day, meaning BTC is not being carried by equity strength on the way up but retains higher beta on the way down. If the 07-29 equity break extends through PCE and the 08-07 payrolls print, BTC has less cushion than the index does.

30-day realized vol of 33.34% annualized is about 1.75% per day, so a one-sigma move over the next five to seven sessions is roughly 3.9% to 4.6% — around 61,400 to 61,900 on the downside from 63,874. The structural map is more informative than the sigma. The 50-day at 63,320 is the first and only shelf; below it the entire 14-day range offers no traded structure until the 30-day low of 58,519, 8.4% lower, which is precisely why a break there is likely to be fast rather than orderly. The market is not positioned for that: put-to-call open interest at 0.52 versus 0.76 in late June, 30-day implied vol at the 16.7th percentile of 90 days, and the Deribit 30-day index at 37.45 in the 21st percentile of the year all say downside protection is cheap and unowned, which is the condition under which a modest catalyst produces an outsized move. The mechanism is an inflation surprise on 07-30 PCE or 08-12 CPI with Brent at $89.79, or a heavier-than-expected 08-05 refunding schedule against $671B of estimated quarterly borrowing. Upside is bounded and near: 66,257 is 3.7% away, the 200-day at 71,743 is 12.3% away, and the $70,000 and $72,000 option clusters roll off worthless on 07-31. Rough odds for the horizon: about 60/40 that price closes below 63,874 in five to seven sessions, roughly one in four that the 50-day gives way and 58,519 is tested, and low single-digit odds of a true capitulation low — that last one is deliberately low because liquidation activity sits at only the 35.6th percentile of the year and funding is not crowded, so the leverage that produces capitulation lows is not present to be flushed. The realistic bear case here is a grind, not a crash, with a fat tail attached to two inflation prints.

low