Anatomy of a Short Squeeze: Bitcoin's 45-Hour Run to $79.6K, Molecule by Molecule
Bitcoin ran from $64.5K to $79.6K in 45 hours. We replay it in consolidated order flow: $2.2B of taker buys in 30 minutes, $1.09B of shorts liquidated.
Bitcoin added roughly a quarter of its value in under a week, and most of it in 45 hours. The news explains the direction. The perpetual-futures plumbing explains the violence.
From the August 14 low of $62,484 to Friday's high of $79,603, the move measured +27.4%. From the moment it actually ignited, Wednesday, August 19 at 12:00 UTC, to the top, it was +23.5%.
The headlines have their version. The Treasury announced it was at least doubling its liquidity-support buybacks of longer-dated debt, from $2B to $4B per operation, and Secretary Bessent followed up by saying the ceiling could go higher still. Trump hosted crypto and prediction-market executives at the White House the same day to push the Clarity Act. US spot ETFs took in $517M, their largest inflow day since early May.
That's the why on the macro layer. This post is about the other layer: what the perpetual futures plumbing was doing before, during, and after. We replayed the whole episode through our consolidated feed: trades, liquidations, open interest, funding, and order-book state across Binance, Bybit, OKX and Hyperliquid, archived every 15 minutes. The short version: this was a textbook short squeeze that matured into a real position build, and the fuel for it was loaded two days before the news hit.
Hover or use ←/→ to scrub. Click a moment to pin it. Price, open interest and hourly liquidations share one timeline.
BTC price, consolidated · USD
Open interest, consolidated · USD
Liquidations per hour · USD (Binance + Bybit + OKX)
Data table (hourly)
| Hour ending (UTC) | Price | Open interest | Liquidated | % shorts | Funding bps | Funding pct |
|---|---|---|---|---|---|---|
| Aug 19 · 00:45 UTC | $64,530 | $15.65B | $0.1M | 26% | 0.45 | 39th |
| Aug 19 · 01:45 UTC | $64,389 | $15.57B | $0.7M | 3% | 0.31 | 28th |
| Aug 19 · 02:45 UTC | $64,360 | $15.54B | $0.6M | 1% | 0.22 | 22th |
| Aug 19 · 03:45 UTC | $64,312 | $15.44B | $0.2M | 14% | 0.20 | 21th |
| Aug 19 · 04:45 UTC | $64,315 | $15.43B | $0.0M | 96% | 0.21 | 22th |
| Aug 19 · 05:45 UTC | $64,288 | $15.40B | $0.0M | 1% | 0.23 | 23th |
| Aug 19 · 06:45 UTC | $64,241 | $15.33B | $1.5M | 0% | 0.24 | 24th |
| Aug 19 · 07:45 UTC | $64,332 | $15.33B | $0.3M | 98% | 0.23 | 23th |
| Aug 19 · 08:45 UTC | $64,422 | $15.30B | $0.6M | 69% | 0.31 | 28th |
| Aug 19 · 09:45 UTC | $64,339 | $15.28B | $0.1M | 86% | 0.31 | 28th |
| Aug 19 · 10:45 UTC | $64,373 | $15.30B | $0.0M | 18% | 0.30 | 27th |
| Aug 19 · 11:45 UTC | $64,436 | $15.29B | $0.4M | 100% | 0.32 | 28th |
| Aug 19 · 12:45 UTC | $64,920 | $15.31B | $2.9M | 99% | 0.31 | 28th |
| Aug 19 · 13:45 UTC | $65,016 | $15.40B | $9.0M | 93% | 0.29 | 27th |
| Aug 19 · 14:45 UTC | $65,430 | $16.02B | $16.4M | 100% | 0.28 | 26th |
| Aug 19 · 15:45 UTC | $68,672 | $16.11B | $450.5M | 99% | 0.26 | 26th |
| Aug 19 · 16:45 UTC | $68,563 | $16.47B | $0.7M | 51% | 0.97 | 81th |
| Aug 19 · 17:45 UTC | $68,019 | $16.19B | $1.8M | 16% | 0.97 | 81th |
| Aug 19 · 18:45 UTC | $68,400 | $16.23B | $1.8M | 98% | 0.97 | 81th |
| Aug 19 · 19:45 UTC | $68,369 | $16.23B | $1.3M | 83% | 0.95 | 81th |
| Aug 19 · 20:45 UTC | $68,638 | $16.32B | $3.8M | 98% | 0.93 | 79th |
| Aug 19 · 21:45 UTC | $69,653 | $15.99B | $108.5M | 98% | 0.94 | 80th |
| Aug 19 · 22:45 UTC | $69,323 | $15.94B | $1.5M | 28% | 0.94 | 80th |
| Aug 19 · 23:45 UTC | $69,301 | $15.93B | $0.1M | 44% | 0.94 | 80th |
| Aug 20 · 00:45 UTC | $69,547 | $16.02B | $1.1M | 91% | 0.88 | 77th |
| Aug 20 · 01:45 UTC | $69,614 | $16.07B | $0.6M | 30% | 0.88 | 77th |
| Aug 20 · 02:45 UTC | $69,300 | $16.07B | $1.4M | 53% | 0.88 | 77th |
| Aug 20 · 03:45 UTC | $69,183 | $16.13B | $1.5M | 10% | 0.88 | 77th |
| Aug 20 · 04:45 UTC | $69,327 | $16.21B | $0.1M | 100% | 0.88 | 77th |
| Aug 20 · 05:45 UTC | $69,546 | $16.30B | $0.5M | 99% | 0.88 | 77th |
| Aug 20 · 06:45 UTC | $69,761 | $16.46B | $0.9M | 92% | 0.88 | 77th |
| Aug 20 · 07:45 UTC | $69,602 | $16.47B | $2.0M | 84% | 0.88 | 77th |
| Aug 20 · 08:45 UTC | $71,480 | $16.59B | $103.6M | 100% | 1.00 | 85th |
| Aug 20 · 09:45 UTC | $71,978 | $16.71B | $22.7M | 97% | 1.00 | 85th |
| Aug 20 · 10:45 UTC | $71,786 | $16.74B | $12.6M | 85% | 1.00 | 85th |
| Aug 20 · 11:45 UTC | $72,123 | $16.84B | $2.3M | 84% | 1.00 | 85th |
| Aug 20 · 12:45 UTC | $71,648 | $16.58B | $3.1M | 23% | 1.00 | 85th |
| Aug 20 · 13:45 UTC | $71,641 | $16.48B | $4.7M | 16% | 1.00 | 85th |
| Aug 20 · 14:45 UTC | $71,513 | $16.47B | $2.3M | 73% | 1.00 | 85th |
| Aug 20 · 15:45 UTC | $72,212 | $16.64B | $13.8M | 95% | 1.00 | 85th |
| Aug 20 · 16:45 UTC | $72,622 | $16.73B | $4.8M | 90% | 0.68 | 61th |
| Aug 20 · 17:45 UTC | $72,729 | $16.68B | $0.9M | 65% | 0.68 | 61th |
| Aug 20 · 18:45 UTC | $72,467 | $16.65B | $2.8M | 63% | 0.68 | 61th |
| Aug 20 · 19:45 UTC | $72,681 | $16.65B | $0.4M | 58% | 0.68 | 61th |
| Aug 20 · 20:45 UTC | $72,730 | $16.64B | $2.7M | 94% | 0.68 | 61th |
| Aug 20 · 21:45 UTC | $72,705 | $16.57B | $0.5M | 78% | 0.68 | 61th |
| Aug 20 · 22:45 UTC | $72,654 | $16.60B | $0.3M | 10% | 0.68 | 61th |
| Aug 20 · 23:45 UTC | $73,074 | $16.64B | $12.1M | 100% | 0.67 | 60th |
| Aug 21 · 00:45 UTC | $73,658 | $16.92B | $50.3M | 99% | 0.52 | 46th |
| Aug 21 · 01:45 UTC | $74,710 | $17.16B | $55.4M | 99% | 0.53 | 47th |
| Aug 21 · 02:45 UTC | $74,692 | $16.77B | $41.7M | 91% | 0.54 | 48th |
| Aug 21 · 03:45 UTC | $74,651 | $16.80B | $3.0M | 30% | 0.54 | 48th |
| Aug 21 · 04:45 UTC | $74,932 | $16.99B | $0.9M | 70% | 0.54 | 48th |
| Aug 21 · 05:45 UTC | $75,249 | $17.13B | $3.9M | 85% | 0.54 | 48th |
| Aug 21 · 06:45 UTC | $75,364 | $17.16B | $3.8M | 61% | 0.54 | 48th |
| Aug 21 · 07:45 UTC | $76,580 | $17.34B | $59.4M | 98% | 0.54 | 48th |
| Aug 21 · 08:45 UTC | $77,643 | $17.71B | $31.1M | 97% | 0.86 | 76th |
| Aug 21 · 09:45 UTC | $78,201 | $17.45B | $63.2M | 66% | 0.86 | 76th |
| Aug 21 · 10:45 UTC | $77,841 | $17.46B | $0.1M | 13% | 0.86 | 76th |
| Aug 21 · 11:45 UTC | $76,862 | $16.94B | $10.3M | 10% | 0.86 | 76th |
| Aug 21 · 12:45 UTC | $77,436 | $17.18B | $2.4M | 100% | 0.86 | 76th |
| Aug 21 · 13:45 UTC | $77,306 | $17.13B | $4.1M | 23% | 0.86 | 76th |
| Aug 21 · 14:45 UTC | $77,155 | $17.06B | $0.2M | 31% | 0.86 | 76th |
| Aug 21 · 15:45 UTC | $77,505 | $17.14B | $0.0M | 98% | 0.86 | 76th |
| Aug 21 · 16:45 UTC | $77,354 | $17.03B | $0.3M | 88% | 1.00 | 85th |
| Aug 21 · 17:45 UTC | $77,365 | $17.02B | $0.0M | 37% | 1.00 | 85th |
| Aug 21 · 18:45 UTC | $76,928 | $16.89B | $2.6M | 13% | 1.00 | 85th |
| Aug 21 · 19:45 UTC | $77,146 | $16.85B | $0.4M | 75% | 1.00 | 85th |
| Aug 21 · 20:45 UTC | $77,441 | $16.92B | $2.6M | 81% | 1.00 | 85th |
| Aug 21 · 21:45 UTC | $78,232 | $17.09B | $14.9M | 99% | 1.00 | 85th |
| Aug 21 · 22:45 UTC | $78,511 | $17.17B | $7.3M | 88% | 1.00 | 85th |
| Aug 21 · 23:45 UTC | $78,373 | $17.14B | $0.8M | 57% | 1.00 | 85th |
| Aug 22 · 00:45 UTC | $77,825 | $16.96B | $1.5M | 6% | 1.00 | 85th |
| Aug 22 · 01:45 UTC | $77,903 | $16.96B | $0.7M | 66% | 1.00 | 85th |
| Aug 22 · 02:45 UTC | $78,500 | $17.09B | $2.7M | 81% | 1.00 | 85th |
| Aug 22 · 03:45 UTC | $78,648 | $17.19B | $2.7M | 91% | 1.00 | 85th |
| Aug 22 · 04:45 UTC | $78,671 | $17.16B | $3.2M | 79% | 1.00 | 85th |
| Aug 22 · 05:45 UTC | $77,180 | $16.66B | $39.9M | 16% | 1.03 | 87th |
| Aug 22 · 06:45 UTC | $77,402 | $16.85B | $1.2M | 42% | 1.03 | 87th |
The setup: shorts got comfortable
Rewind to the week before ignition. BTC had spent two weeks grinding between $62.5K and $65.5K. On the surface, nothing. In the positioning data, three things were quietly lining up.
Open interest was leaking out. Consolidated BTC open interest fell from $16.05B on August 16 to a trough of $15.26B at 12:30 on August 19. About $790M of positions closed in three and a half days while price went nowhere. Conviction, on both sides, was low.
Funding compressed to a two-year floor. By the morning of August 19, the cross-venue 8h funding rate sat at 0.26 bps, the 26th percentile of the past 730 days. More telling: through the night of August 18 (01:00–08:30 UTC), the rate briefly printed negative, at the 8th percentile of two years. For a few hours, shorts were the ones paying to hold, which is exactly the condition that makes a short position expensive to keep and cheap to abandon. Annualized, being long BTC perps cost about 2.8%: essentially free carry. Positioning had leaned short at the exact moment shorting stopped earning anything.
Perps traded at a discount. Basis, measured as the Binance perpetual against the spot index, averaged −4.6 bps through August 19's morning. Perpetuals persistently below spot. No leverage-driven froth anywhere in the term structure.
None of this predicts a rally. What it does is define the fuel: a market leaning short, at near-zero cost, into thinning open interest. If a catalyst shows up, the exit is narrow.
Ignition: 30 minutes, $2.2B of market buys
The catalyst showed up in the US morning session on Wednesday, August 19. As the Treasury buyback headlines circulated, BTC ground from $64.5K to $65.9K between 12:00 and 15:00. Then the narrow exit got tested.
15:00–15:30 UTC. Price jumped from $65,896 to $68,284: +3.6% in 30 minutes. Net taker delta in that single 30-minute window was +$2.22B, the largest print of the entire week, with adjacent windows adding another ~$870M. Someone, or rather many someones, crossed the spread at size, simultaneously, on four venues.
The forced sellers were buyers. Trailing one-hour liquidations peaked at $450.5M by 15:45 UTC, 98.9% of it shorts. That's the signature of a squeeze: positions that must buy to close, buying into a rising tape, making the next short's liquidation price closer.
Volume went vertical. The 12:00–16:00 consolidated candle traded 406,180 BTC, roughly $27B of notional in four hours. Our 15-minute volume gauge hit 40.8× its trailing weekly median at the peak of the burst.
Funding repriced the same day. At the next settlement the cross-venue rate stepped from 0.26 to 0.97 bps, from the 26th to the 81st percentile in one funding cycle. The market went from "paying nobody" to "longs pay about 11% annualized" within hours.
One more detail worth pausing on: this was not only covering. Open interest, which bottomed at $15.26B just after noon, two and a half hours before ignition, rose $670M by midnight while price gained 7.5%. If the move were purely shorts closing, open interest would have fallen. New longs were entering on top of the covering flow. The squeeze provided the spark; fresh positioning provided the follow-through.
Two more waves, same fingerprint
Squeezes rarely resolve in one candle. The tape printed two more forced-buying waves with an identical fingerprint:
| Wave (UTC) | 1h liquidations | Share shorts | Price milestone |
|---|---|---|---|
| Aug 19, ~15:45 | $450.5M | 98.9% | $64.5K → $68.5K |
| Aug 19, ~21:45 | $108.5M | 97.5% | through $69.6K |
| Aug 20, ~08:45 | $103.6M | 99.5% | through $71.3K |
Between the waves, the buying never really stopped. Daily net taker delta on BTC: +$4.12B on August 19, +$2.45B on August 20, +$2.44B on August 21. Order-book imbalance spent most of the climb negative, asks stacked above price and sellers offering into strength, and the market ate through them anyway. Passive sellers kept underestimating the bid.
Across the three days, our feed counted $1.21B of BTC liquidations on Binance, Bybit and OKX, $1.09B of it shorts (90%). That's a floor, not a total: Hyperliquid publishes no comparable liquidation feed, and press tallies across all assets and venues ran into the billions. For scale, the largest cascades in crypto history clear this in a single hour. This was a sustained bleed, not one detonation.
And here's the molecule most commentary missed: basis stayed negative through almost the entire rally. Out of 319 archived 15-minute snapshots from August 19 to now, perps printed above spot index only 19 times. A 23% rally in which the perpetual trades at a discount the whole way up is not a leverage mania. It's consistent with spot-led demand (that $517M ETF day) dragging derivatives along behind it. ETH, for reference, outran BTC on the same window: +32.7% from its $1,922 breakout to $2,550, with about $17B of notional in its own ignition candle.
The top: half a billion of late open interest, gone in 15 minutes
The character of the tape changed at the top, and the change is measurable.
Into Friday 09:00 UTC, price printed its high (our 15-minute archive caught $79,323; the consolidated candle high was $79,603) and open interest peaked at $18.01B, up $2.76B or +18.1% from Wednesday's trough. More than $535M of that open interest arrived in the final 45 minutes before the print. The marginal participant was no longer a short covering; it was a late long chasing.
They were tested immediately. Within 15 minutes of the high, open interest dropped $613M ($695M within 30). The 30-minute net taker delta flipped to −$635M, the most negative of the week. The liquidation mix inverted: the hour into 09:45 forced $63.2M with the long share jumping to 34%, and the following hours ran 87–90% longs. Fittingly, basis chose exactly this window to print its only meaningful positive reading of the whole episode: +3.5 bps at 12:15, the one moment perps briefly led. That was the top, at least so far.
Saturday morning brought the second test, and it carries the most instructive number of the aftermath. At 05:15 UTC price slipped from $78.5K to $77.3K and open interest fell $447M in 15 minutes, but only about $41M of the hour was forced liquidation. A 10:1 ratio of voluntary closing to forced closing. Friday's top flush was longs being carried out; Saturday's dip was longs walking out. Different mechanism, and the second one is healthier for the structure that remains.
Where the tape is, right now
As of 07:40 UTC on August 22, with BTC at $77,389 (−2.8% from the high):
Open interest is holding. $16.83B. The market kept about $1.6B (+10.4%) of the positions built during the run. This wasn't a round-trip; the squeeze converted into a standing position base.
Funding is elevated, not euphoric. 1.04 bps per 8h, the 87th percentile of two years. Longs now pay about 11.4% annualized for a position that cost 2.8% four days ago. Spicy, but this is not the 99th-percentile froth that marks blow-off tops.
Funding, applied to your position size — before the rally vs now.
| Pre-rally (0.26 bps) | Now (1.04 bps) | |
|---|---|---|
| Per 8h settlement | $0.26 | $1.04 |
| Per day | $0.78 | $3.12 |
| Per 30 days | $23.40 | $93.60 |
| Annualized | $285 | $1,139 |
Rates: 0.26 bps/8h (pre-rally, 26th percentile) vs 1.04 bps/8h (Aug 22, 87th percentile of 730 days). Longs pay when funding is positive. Live funding · More calculators
The book is thinner and leaning ask. Displayed depth within 10 bps of mid averages about $67M combined against roughly $145M before the breakout (post-volatility quoting, cut roughly in half), with order-book imbalance at −0.15. Realized 24h volatility is running at 86% annualized. Thin books cut both ways: they amplified the way up, and they'll amplify whatever comes next.
Shorts haven't capitulated as a class. Even in the last 24 hours, $228M was liquidated on BTC and 57% of it was still shorts. And our state feed keeps flagging prints of "open interest rising while price falls", which is new shorts re-engaging against the new range.
What history says about the 87th percentile
The reflex read is "funding this hot means a squeeze or a collapse is imminent." Our own base rates say: mostly, neither.
730 days of BTC history. Overlapping windows collapsed; n = effective samples.
- Hit rate up
- 51%
- Effective n
- 87
- Median max drawdown
- −1.33%
Across 730 days of history, hours where BTC funding sat at or above the 85th percentile were followed by a median +0.10% over the next 24h (87 effective samples after collapsing overlapping windows), −0.60% over 72h (43 samples, 44% hit rate up), and −0.02% over 7 days, with a median worst drawdown of −4.05% along the way (26 samples). In plain terms: after funding gets this crowded, the historical edge in either direction is roughly zero, but the ride gets bumpier. Small samples, honestly disclosed; treat them as context, not a forecast. We reached the same conclusion the last time we measured what extreme funding actually precedes, on a different window and a wider sample.
That's the general shape of this whole episode, actually. The news explained the direction. The positioning data explained the violence: two weeks of shorts building at zero cost, a $2.22B half-hour that forced them out, $1.09B of forced short buying, an 18% open-interest build, and a funding market that repriced the cost of the crowd's new consensus by 4× in three days.
Price charts tell you what happened. Microstructure tells you why.
FAQ: the August 2026 bitcoin short squeeze
What caused the bitcoin short squeeze in August 2026?
Two layers. Positioning loaded the fuel: by August 18 the cross-venue funding rate had compressed to the 8th percentile of the past 730 days, briefly negative, while open interest bled $790M in three and a half days. Shorts were crowded and holding at near-zero cost. The macro catalysts lit it: the Treasury at least doubling its long-dated liquidity-support buybacks to $4B per operation, a White House meeting with crypto executives, and a $517M spot-ETF inflow day, all on August 19. In 30 minutes that afternoon (15:00–15:30 UTC), price jumped +3.6% and hourly liquidations hit $450.5M, 98.9% of them shorts.
How does a short squeeze work in crypto?
A liquidated short is closed by a forced market buy. That buying pushes price higher, which moves the next short's liquidation price closer: a feedback loop. In this episode the loop is visible in the numbers: $2.22B of net taker buying crossed the spread in a single 30-minute window, and three distinct liquidation waves ($450.5M, $108.5M, $103.6M per hour) each ran 97.5–99.5% shorts. Result: +23.5% in 45 hours. The liquidation heatmap is the tool for seeing where the next rung of that ladder sits before it gets hit.
How long did the squeeze last?
The forced-buying phase is measurable in the liquidation mix: from ignition at 15:00 UTC on August 19 until roughly 09:45 UTC on August 21, about 43 hours in all, liquidations ran overwhelmingly short-side. After the $79.6K top, the mix flipped: the following hours ran 87–90% longs. There's no universal duration for a squeeze; this is what this one measured.
How much bitcoin was liquidated during the rally?
At least $1.21B in BTC perpetual positions between August 19 and the morning of August 22 on Binance, Bybit and OKX: $1.09B of it shorts (90%), $119M longs. The peak single hour forced $450.5M. Treat these as floors: Hyperliquid publishes no comparable liquidation feed, and cross-asset press tallies ran into the billions.
What does a high funding rate mean for bitcoin right now?
BTC funding sits at 1.04 bps per 8h, the 87th percentile of two years, or roughly 11.4% annualized paid by longs (it was about 2.8% before the rally). Historically it's a crowding gauge, not a crash call: across 730 days, hours with funding at or above the 85th percentile were followed by a median +0.10% over 24h (87 effective samples), −0.60% over 72h (43), and a median worst drawdown of −4.05% within a week (26). Bumpier, not directional.
What happened to bitcoin open interest during the rally?
Consolidated open interest ran from a $15.26B trough (August 19, 12:30 UTC) to an $18.01B peak at the top (+18.1%), then shed $695M in 30 minutes during the top flush. As of August 22 it holds $16.83B, keeping +10.4% of the build. That's the difference between a pure squeeze (open interest round-trips) and a squeeze that converted into a standing position base (open interest holds).
Is the bitcoin rally over?
We don't forecast. Here's what the tape shows as of 07:40 UTC, August 22: price −2.8% off the $79.6K high, open interest holding +10.4%, funding at the 87th percentile, displayed depth about half of pre-breakout, and fresh shorts re-engaging (repeated "open interest up, price down" prints). The honest base rate after funding this crowded is a flat median with −4% median drawdowns along the way. Watch the funding percentile and the open-interest-versus-price divergence rather than anyone's prediction, ours included.
What this is and isn't
- Liquidation totals are floors. Hyperliquid publishes no comparable liquidation feed, so every liquidation number here covers Binance, Bybit and OKX only. Exchange liquidation feeds are also throttled and aggregated by the venues themselves, which biases published totals down.
- Basis is a single-venue proxy. "Perps at a discount" is measured as the Binance perpetual against its spot index. It is the cleanest continuous series we archive, not a cross-venue consensus.
- The base rates are small samples. 87, 43 and 26 effective observations. That is enough to say "the median is near zero and the drawdowns are real," and not enough to say anything sharper.
- Attribution is not proof. We can show that funding was at a two-year floor before the move and that forced short covering dominated the first 43 hours. We cannot show that the Treasury headline caused it. Positioning describes the mechanism; it does not identify the trigger.
- This is a snapshot. Every "right now" number is stamped 2026-08-22 07:40 UTC. The live modules have moved on since.
Methodology
Prices, volume, net taker delta, open interest and order-book metrics are consolidated across Binance, Bybit, OKX and Hyperliquid. Liquidation totals cover Binance, Bybit and OKX only. Basis is measured as Binance perp versus spot index. Funding is cross-venue, normalized to 8 hours; percentiles rank against 730 days of history. Net taker delta figures are over 30-minute windows; state metrics are archived every 15 minutes.
The interactive replay above reads one frozen slice of that archive: 319 rows from 2026-08-19T00:00:00Z to 2026-08-22T07:30:00Z. The liquidation field is a trailing one-hour window, so consecutive rows overlap; the hourly bars sample it at every fourth row, giving 79 non-overlapping hours that sum to $1,211.1M.
Forward-return base rates use Binance 1h closes, with overlapping windows collapsed so the sample counts shown are the effective ones, not the raw hit counts. Nothing here is trading advice; it's a description of what the tape did.
Sources
- Treasury doubles debt buybacks as Bessent moves to steady bond market — CNBC
- Treasury Announces Increased Sizes of Nominal Long-End Liquidity Support Buybacks Beginning September 9 — U.S. Department of the Treasury
- Bessent says Treasury buyback operation could be more than $4 billion — CNBC
- Spot bitcoin ETFs report $517 million in net inflows, largest in 3.5 months — The Block
- Live updates: Bitcoin extends gains as Bessent suggests more Treasury intervention — CoinDesk
- Crypto Executives Join Trump in Push for US Digital Asset Legislation — Bloomberg
- White House Convenes Meeting With Crypto and Prediction Market CEOs — PYMNTS
Related
→ What extreme funding rates actually precede, measured
→ What a negative funding rate means
→ Cumulative Volume Delta: the perpetual futures trader's guide to order flow
→ How to read a liquidation heatmap
→ Shorts paid for the rally: 25 days of negative funding
Watch the same data we used for this post, live and free, refreshed every 30 seconds: Liquidations, Funding, Positioning, Book depth and the Footprint chart.