Bitcoin cleared $80,000 on Tuesday for the first time in 102 days, printing $81,272.62 on Binance BTCUSDT and sitting at $80,759.99 at 06:09 UTC. Two positioning series flipped on the same candle, in opposite directions, and both flips argue against the story most of the coverage is telling.

The Crowd Went Net Short Into a Three-Month High

Binance publishes two long/short ratios that are routinely quoted as if they were the same statistic. They are not, and this desk has now watched them disagree three times in a week. The global long/short account ratio counts accounts: how many holders are net long versus net short, one vote each. The top-trader long/short position ratio weighs the largest accounts’ positions by size.

On Tuesday’s daily readings the account ratio printed 0.9448. Below 1.00 means more accounts on Binance futures are net short bitcoin than net long — at the highest price in three months.

Read the two right-hand columns against the left one. The account ratio has spent the entire rally falling — from 1.4845 on 18 August to 0.9448 on Tuesday, a 36% decline — while price rose 24.8%. The retail-shaped side of this market did not buy the move. It sold into it, and by Tuesday it was net short.

Meanwhile the top-trader position ratio climbed for five consecutive readings to 2.0859 on Monday, a series high, then fell for the first time in six readings to 2.0203 on Tuesday. The top-trader account ratio fell harder, 1.1561 to 1.0210. So the size accounts, which bought the entire rally including both losing sessions of the weekend, took something off on the breakout.

What That Kills, and What It Does Not

It kills the tidiest available bear story. If this were a retail squeeze now unwinding as retail exits, you would need retail to be long. Retail is short, and has been getting shorter for a week. There is no crowd left to shake out; the crowd already left, and the exit was gradual rather than violent.

It also complicates the tidiest bull story. The accounts that carried the move trimmed into the high. One reading is not a trend — the position ratio is still at 2.02, an extreme level by its own recent history, and a 3.1% reduction from a series high is a trim, not a reversal. But it is the first non-confirming print in six days, and we would rather flag it on the day it happens than after it becomes obvious.

The uncomfortable version, and the one we hold: concentrated size sitting two-to-one long into a Wednesday inflation print and a Friday keynote from a Fed chair whose committee has three members voting for a rate hike is not a comfortable configuration, regardless of which way it resolves.

The Funding Pin Broke — Downwards

For ten consecutive eight-hourly settlements, from 21 August 16:00 UTC through 24 August 16:00 UTC, Binance BTCUSDT funding printed exactly 0.0100% — the baseline rate, to four decimal places, three and a third days without deviation. That run ended at the 25 August 00:00 UTC settlement, which printed 0.0058%.

It broke in the direction nobody watching a breakout expects. Across the full twenty-settlement window we pulled — 18 August 16:00 UTC through 25 August 00:00 UTC, covering a range of roughly $18,700 — the maximum funding rate is exactly 0.0100%. Not one settlement exceeded the baseline. On the candle that took bitcoin to a three-month high, the price of being long fell.

Open Interest Is Building — But Read It in Coins

Coin-denominated open interest on the Binance perpetual bottomed at 105,531.28 BTC on Monday’s daily reading, a series low for the move. Tuesday’s reading is 107,215.99 BTC, up 1.60%; the live endpoint returned 107,883.381 at 06:10 UTC. That is the second consecutive build after seven sessions of decline, and it is the first evidence in this rally that new positions are being opened rather than old ones closed.

It is also, still, 4.26% below the 15 August peak of 111,988.29. Dollar-denominated open interest over the same window went from $7.057bn to $8.465bn, up 19.95%. Same book. One number says the market is 4% smaller; the other says it is 20% bigger. The difference is entirely price. If you read only one, read coins.

The Marker Board

Every marker on this desk is written with a date, a deadline type, a named instrument and a numeric bar, published before it settles, and graded in public whether it flatters us or not.

Grading A1 — Pass, and a Good One

A1 was written on Monday and settled Monday evening when Strategy Inc filed its Form 8-K. The BTC Update table reads: no purchase, no sale, 840,447 bitcoin held, $63.36 billion aggregate cost, $75,385 average. That is the tenth consecutive reporting period without a purchase.

We grade design as well as outcome, because a marker that could not fail is worth nothing. A1 was genuinely open. It was written on the specific observation that the position had just crossed back above its cost basis for the first time in a quarter — which is simultaneously the condition under which a buyer returns and the condition under which a seller monetises at better prices. The filing resolved it, and it resolved with a third option we had flagged as possible but not favoured: neither. Fair-to-good marker; we would write it again.

The filing also disclosed something we did not anticipate, which is covered in full in today’s update: a brand-new $1.59 billion liquidity bucket called USD Cash, funded almost entirely by a single $2.01 billion week of stock sales, none of which bought bitcoin.

Grading X2 and Y1 — Both Failing, Both Against Us

X2 and Y1 were written on the premise that the Iran escalation would push crude and the long end higher. Both are moving the wrong way, and Monday made it worse rather than better. Brent settled 2.35% lower on the day of the sanctions announcement itself; the 30-year is 12.1 basis points below its bar with three sessions to run and would need a new 2026 high to clear it.

We are stating that plainly rather than quietly letting the markers expire on Friday, because the premise underneath them — that sanctions escalation is an inflation and energy shock — is the same premise underneath a great deal of the commentary currently attached to bitcoin’s rally. Our own instruments are voting against it.

Three New Markers

• B1 — at least one Binance BTCUSDT funding settlement prints above 0.0100% between now and 28 August 16:00 UTC. Window deadline; instrument named. Genuinely open, and it is the cleanest single test of whether a leveraged bid ever shows up in this rally. Twenty consecutive settlements have failed to do it.
• B2 — total U.S. spot bitcoin ETF net flow for August 2026, on completed Farside rows, reaches ≥ $3,424.9 million by month end, clearing October 2025 and making August the best month since. Requires $689.4m across four sessions, versus a trailing six-session average of $375.9m. Roughly a coin flip.
• B3 — Strategy’s next 8-K, expected Monday 31 August and covering 24–30 August, discloses a bitcoin purchase greater than zero. Scheduled deadline. Written specifically because Monday’s filing named “acquiring bitcoin” as the first permitted use of the new USD Cash pool, and this marker tests whether that language means anything within a week.

What Settles Next

• Wednesday 26 August, 08:30 ET — A3 settles on July core PCE from the BEA.
• Wednesday 26 August — the 25 August ETF row publishes, adding to Z3 and to B2.
• Friday 28 August — X2, Y1, Z2 and B1 all expire on the same day Chair Warsh delivers his Jackson Hole keynote at roughly 10:00 ET. This desk will publish a single multi-grade board that afternoon.
• Sunday 31 August, 23:59:59 UTC — A2 and B2 settle on the monthly close and the final August flow row.
• Monday 31 August — B3 settles on Strategy’s 8-K.

Open interest, funding and long/short ratio data pulled directly from the Binance futures API at 06:10 UTC on 25 August 2026. Price data from the Binance spot API. ETF flows from Farside Investors. Strategy figures from the Form 8-K filed with the SEC on 24 August 2026.