The stillest session in three weeks

Start with how little happened, measured properly.

Saturday August 15 traded between $62,920 and $63,188. That is a range of $268, or 0.42% of the close, and it is the narrowest full session in our last twenty. The runner up is August 8 at 0.63%. Today is running a $190 range, roughly 0.30%, and that figure is partial because the session has not closed, a distinction we now label after getting it wrong twice.

SessionRangeAs % of closeVolumeBuy share
Fri Jul 31$2,9444.68%$1.30B47.2%
Sat Aug 1$8751.39%$0.48B49.9%
Sat Aug 8$4080.63%$0.37B48.8%
Fri Aug 14$1,0821.72%$0.78B54.1%
Sat Aug 15$2680.42%$0.34B57.5%
Sun Aug 16, partial$1900.30%$0.12B52.6%

Three consecutive Saturdays have tightened in order: 1.39%, then 0.63%, then 0.42%. Against July 31's 4.68% and July 27's 3.36%, the market has given up roughly nine tenths of its daily range in a fortnight. Bitcoin sits at $63,009, down 2.92% from the August 9 close, inside a twenty session band of $62,275 to $65,745.

And the buying was the strongest of any weekend in the sample

Now the part that makes the stillness strange. That narrowest session was not quiet on both sides. It had the highest weekend taker buy share we have recorded in three weeks.

Saturday closed at 57.5%. The four weekend sessions before it read 49.9%, 54.5%, 48.8% and 45.6%. Across the last twenty hours the aggregate is 60.1% on about $231 million, and inside that stretch the hourly readings include 71.7%, 70.6%, 71.4%, 79.7%, 66.5% and 70.2%. A 79.7% hour means roughly four out of every five aggressive dollars were lifting offers.

Price response to all of that: none. Saturday closed 0.07% higher. Today is 0.12% lower.

Taken alone, a 60% buy share reads as clear demand and would have been the headline of this article a week ago. Taken alone it would also have been wrong, which is the whole point of what follows.

The other side, published alongside as promised

On Friday we wrote that a buy share reading without the ETF flow beside it is an incomplete picture, and that we would stop presenting it as a complete one. Here is the pair.

WeekBitcoin ETF net flowIBITBTC price move
Ended Aug 7$853.5M in$693M inRose to $65,474
Aug 10 to 14$389.7M outLed the outflows2.92% lower

The week of August 10 to 14 saw about $389.7 million leave US spot Bitcoin ETFs, reported as the largest weekly withdrawal in six weeks. Friday alone accounted for $56.2 million, the third consecutive outflow day, taking that streak to roughly $180 million. IBIT led it with $55.5 million out, and Fidelity's FBTC lost $6.8 million. Bitwise's BITB went the other way with about $6.1 million in.

Put the two weeks side by side and the swing is roughly $1.24 billion between consecutive weeks, with the same fund at the centre of both. IBIT took in $693 million in the first week of August and led the selling in the second. Cumulative net inflows across all the funds since launch stand at about $51.79 billion.

So the weekend resolves cleanly. Aggressive buyers on Binance spot are lifting offers at 60%, redeeming institutional holders are taking the other side through the ETF wrapper, and the price sits inside two hundred dollars because those two forces are close to equal. Neither series alone describes this. The buy share says demand. The flow says distribution. Together they say stalemate, which is exactly what the chart shows.

One honest limit on that: this is a single weekend, and both numbers are small in absolute terms. It is a clean illustration of the mechanism, not proof of a law.

Solana shows the same thing with the signs reversed

The mechanism is not a Bitcoin quirk. The same week produced a second, independent instance in the opposite direction, which is the more persuasive evidence.

Solana ETFs took in about $10.26 million across August 10 to 14, reported as their highest weekly inflow since May. Over the same days Solana's taker buy share on Binance spot read 44.4%, 43.9% and 43.7%, and the price fell 1.15% on the 14th and drifted lower after.

So: institutional money entering through the fund, aggressive sellers dominating on the exchange, and a price that went down. That is Bitcoin's situation with both signs flipped, in the same week, in a different asset. If you were reading only the Binance flow you would have concluded nobody wanted Solana, and you would have missed its best ETF week in three months.

For completeness, Ethereum ETFs were roughly flat at about $2.25 million out for the week, after BlackRock withdrew $23.8 million on August 10 and midweek inflows partly offset it.

Worth noting without building a story on it: the direction of the rotation inside crypto has now changed three times in a month. July finished with Ether funds ahead of Bitcoin's, $365.2 million against $172.4 million. The first week of August was overwhelmingly Bitcoin. This week Solana had the best of it. Three readings, three answers, and we are explicitly not calling that a trend, for reasons Friday's correction section covered at length.

The CPI that should have helped, and did not

The week's macro event runs against the price action, which is worth recording plainly.

July CPI, released August 12, rose 0.1% on the month after falling 0.4% in June, putting the annual rate at 3.4%. Core CPI rose 0.2% for an annual 2.5%. Both annual figures came down a tenth from June, and reporting noted the print lowered September rate hike odds.

Notice the gap between those two numbers. Headline at 3.4% against core at 2.5% is an energy story, and the components say so: gasoline is up 24.6% year over year, down from 26.7%, and fuel oil 39.1%, down from 42.9%. This is precisely the composition the FOMC described on July 29 when it attributed elevated inflation partly to supply shocks in sectors including energy. Core at 2.5% is close to the Committee's 2% target.

That is a friendly print for anyone hoping the three dissenters lose the September argument. It is the third such data point after core PCE easing to 3.3% and second quarter GDP at 1.5% against 2.1% expected, and after the weak July jobs report on August 7.

And crypto fell anyway. Bitcoin is down 2.92% since the August 9 close, the ETF outflows landed in the sessions immediately after the CPI release, and the post-print rally never arrived. Sentiment, oddly, went the other way: the Fear and Greed index reads 34, up from 29, 29, 27 and 29, its first clear move after eight readings pinned in a four point band. Total crypto market capitalisation is $2.250 trillion, Bitcoin dominance 56.1%.

Where the board sits today

Today's readings, offered as a description of today.

PairTaker buy share, 24h24h change
Bitcoin57.7%0.04% lower
Tether Gold57.2%flat
Paxos Gold48.0%flat
Ethereum47.8%0.02% higher
BNB45.3%0.85% lower
Solana43.7%0.04% lower
XRP40.5%0.27% lower

Bitcoin has the strongest aggressive bid on the board and is flat. Tether Gold is second at 57.2%, which is a large jump from the 34.8% we quoted on August 3 and the 45.3% of two days ago. We are reporting that number and saying nothing further about it, because a five reading run in this exact series is what we extrapolated into a trend a fortnight ago and it reversed inside a day.

Beneath the flat majors there has been real dispersion this week. Chainlink is reported up roughly 9% over seven days while Cardano is down more than 11%, so the calm at the index level is not calm underneath it.

What is scheduled, and it is close

Scheduled events and things we will measure, with no thresholds attached that claim to predict a direction.

  • FOMC minutes, Wednesday August 19. The written record of the 9 to 3 hold. What the three dissenters put on paper is the nearest read on whether September is still live for them after CPI, PCE, GDP and the jobs print all came in soft.
  • Jackson Hole, August 21 and 22. Chair Warsh's first symposium in the job, and he has made a point of saying less about the policy path rather than more. A chair who dislikes forward guidance at the one event built for it is worth watching on its own terms.
  • September 15, still carrying two things. The reported CLARITY procedural vote and day one of the Fed meeting that brings back the dot plot.
  • Daily ETF flows next to the buy share. Now a standing part of what we publish. Three straight outflow days into Friday is the current state; whether that streak extends or breaks is the single most informative number for next week.
  • Whether the compression resolves. A 0.42% session inside a five week range, into that calendar, is a market with its ranges collapsing ahead of known dates. We are noting the compression as a measured fact. We are not claiming to know which way it breaks, and anyone telling you they do is selling something.

How to read this as a trader

Two things, and the first is the reason this weekend was worth an article at all.

A single number can be perfectly accurate and completely misleading. The 60.1% buy share is correct. It is measured from closed candles, it is not an estimate, and it says something true: on Binance spot, aggressive buyers dominated. What it does not say, and cannot, is that demand exceeded supply, because a large part of the supply was arriving through ETF redemptions that never touch that order book. A week ago we would have published the 60% and called it a bid. This week we published the $389.7 million next to it and called it a stalemate, and the stalemate is what the price actually did.

The habit that follows is cheap to adopt: for any flow reading, ask which venue it covers, and then go find the flows it does not. Both series are free and daily. Solana this week is the tidiest example available, because there the exchange flow was bearish, the fund flow was its best since May, and reading either one alone would have produced a confident wrong answer.

Second, compression is information even when direction is not. Daily ranges have fallen from 4.68% on July 31 to 0.42% yesterday, roughly a nine tenths collapse in a fortnight, into a calendar with FOMC minutes, Jackson Hole and a doubled up September 15. That tells you position sizes built for the last two weeks are the wrong size for the next two. It tells you nothing whatsoever about which way it goes, and we have spent enough of the last fortnight correcting our own predictions to be careful about that distinction. Our live dashboard separates aggressive buying from aggressive selling minute by minute across the most liquid Binance pairs, which is one venue, precisely measured, and now published with its blind spot named. For the run-up to this, see Friday's full accounting of the venue limitation, the CLARITY arithmetic, July's 7.24% on a 49.5% buy share, and the Fed's 9 to 3 hold. If the buy and sell split is new to you, the volume guide covers the mechanics.