The arithmetic, and the paperwork that does not exist

Strip away the commentary and the position is arithmetic. Overcoming a filibuster takes 60 votes. Republicans hold 53 seats. That leaves roughly seven Democratic votes to find, and the bill has already passed the House and cleared the Senate Banking Committee, so this is the only hurdle left.

Except there is a step before the votes, and it is the one worth watching. Ending debate requires a cloture motion to be filed, and cloture is deliberately slow: the petition sits for a mandatory intervening period before the chamber can vote on it, and then debate time follows. As of Monday no cloture motion had been filed for CLARITY. Not a failed one. None.

What the Senate actually did today is the clearest signal available. It convened at 3pm Eastern with a single scheduled roll-call at about 5:30pm, and that vote was a cloture motion on H.R. 6500, a continuing resolution to fund the government. CLARITY did not appear on the day's agenda at all.

Run the calendar forward and the window is narrow rather than theoretical. On the practical timetable, a petition filed by around August 5 allows a vote on ending debate around August 7, with the recess following the first week of August. That is the entire remaining path, and the first step in it has not been taken.

Thune quietly downgraded his own goal

Of everything published on this bill in the last week, one sentence moved our reading of it more than the rest. Majority Leader John Thune has expressed a desire to at least begin consideration of the bill before the recess.

Read that carefully. Beginning consideration is not passing a bill. It is putting it in the queue so that work resumes in September. A week ago the framing was whether the Senate would vote before the break. Now the person who decides what reaches the floor is describing success as merely starting. That single shift in language explains the market pricing better than any headline about support.

And the pricing has moved a long way. The Polymarket contract on CLARITY being signed into law in 2026 now sits in the high twenties to low thirties. It was around 34% when we wrote about it yesterday, 38% the day before that, 43% a week ago, and it peaked near 82% in February. That is a fall of roughly fifty points across the year and about ten points in the last week alone.

None of that decline came from bad drafting or industry opposition. It came from the calendar.

What the seven Democrats are actually asking for

This is the part that reframes last week's endorsement wave, and it is worth being precise about because it changes what would have to happen for the bill to move.

Ruben Gallego of Arizona and Angela Alsobrooks of Maryland have tied their floor support to strong conflict-of-interest rules. Separately, Thom Tillis of North Carolina and Gallego have developed a bipartisan compromise that is under White House review.

So the sticking point is conflict-of-interest provisions, not market structure, custody, or the treatment of tokens. Now put that next to what we reported yesterday: BlackRock, Fidelity, Franklin Templeton, Goldman Sachs and SoFi all publicly backed the bill in late July. Every one of those statements was about regulatory certainty and a level playing field. None of them addressed conflicts of interest, which is the thing the missing votes are conditioned on.

That is why five of the largest names in finance could endorse a bill in the same week its odds fell. They were not answering the question that is being asked. A compromise text sitting in White House review is a far more relevant signal than any of those press statements, and it is also the reason the situation is not hopeless: the negotiation is live, it just may not conclude inside four days.

Monday answered the question we asked on Sunday

On Sunday we set out one thing to find out: whether the reclaim of $62,500 would survive real weekday participation, since the weekend had done it on light volume. Monday answered clearly, and the answer arrived in the afternoon.

Hour (UTC)CloseChangeVolumeBuy share
07:00$62,621flat$79.2M48.1%
09:00$62,8370.47% higher$36.7M69.0%
11:00$62,5600.29% lower$17.7M45.6%
13:00$63,3511.04% higher$70.6M55.8%
14:00$63,9670.97% higher$151.9M63.7%
15:00$63,6950.43% lower$87.3M53.1%

The 14:00 hour is the one that matters: $151.9 million, the largest of the day, at a 63.7% buy share, and price rose 0.97% inside it. Inside that hour, the 14:25 candle carried $24.6 million at a 70.6% buy share and moved price 0.53% higher, and the 14:50 candle ran 76.0%. This is not drift. It is size crossing the spread to buy.

The instructive comparison sits in the same session. The morning's largest five minute candle was 07:05, at $24.5 million, which is almost exactly the same size as the 14:25 candle. Its buy share was 41.7%, and price moved 0.02% lower. Same money, opposite side, and the difference in outcome was the entire day.

Bitcoin trades near $63,880, up about 1.20% over 24 hours, on a Monday buy share of 53.9% across $0.75 billion. The high was $63,993, which stopped seven dollars short of $64,000, and the low of $62,300 held just above Saturday's $62,275.

A correction to yesterday's main caveat

The central hedge in yesterday's article was wrong, and since it was the caveat we leaned on hardest it needs correcting directly rather than in passing.

We wrote that Sunday's volume was about 45% below the previous Sunday's $0.51 billion and called it the thinnest tape in the sample. That was measured at 11:26 UTC on a day that was less than half over. Sunday finished at $0.53 billion, which is slightly above the previous Sunday and squarely in line with the four Sundays before it: $0.58B, $0.89B, $0.54B, $0.51B. There was nothing unusual about it.

It gets worse for the original framing. Sunday's full-day taker buy share came in at 54.5%, and that is the highest of the last five Sundays, against 49.8%, 48.6%, 51.1% and 51.6%. Sunday also closed 1.19% higher, not the 0.56% we reported mid-session. So the weekend signal was stronger than we credited, not weaker, and the caveat we attached to it was an artefact of the clock rather than a fact about the market.

The lesson is narrow and worth keeping: do not characterise a session's volume before the session closes. A ratio like the buy share is readable at any point because it is normalised. A total is not, and comparing a half day against a full day produces exactly the kind of confident wrong sentence we published.

Where the bid is coming from, and it is not CLARITY

If the legislative odds fell roughly ten points in a week and Bitcoin still rose, the buying is clearly not a bet on the bill. The board says where it is coming from.

PairTaker buy share, 24h24h change
Bitcoin54.3%1.20% higher
Ethereum51.5%0.58% higher
Paxos Gold51.0%0.61% lower
BNB50.3%1.19% higher
Solana47.9%0.95% higher
XRP47.7%0.56% higher
Tether Gold34.8%0.50% lower

Look at Tether Gold. Its taker buy share has now printed 67.9%, 47.2%, 40.3%, 38.2% and 34.8% across five readings since July 28. That is a straight line down over six sessions, and today's is the lowest we have recorded. Both tokenised golds are lower on the day while Bitcoin, Ethereum and BNB are higher.

This is the same hedge unwind we identified on Friday, still running, and it has now stopped being a story about everything falling together. On Friday every asset on the board was below a 50% buy share, gold included, and we called it a de-risking. Today four of the seven are above 50% and the weakest by a wide margin is gold. That is what a rotation looks like when it finally shows up: money leaving the inflation hedge and landing in crypto, on a day when crypto's own biggest legislative catalyst got worse.

Sentiment is following slowly. The Fear and Greed index reads 28, up from 27, 27 and Friday's 25, which remains the only Extreme Fear print of the stretch. Total crypto market capitalisation is $2.267 trillion, up 0.66%, with Bitcoin dominance at 56.5% and Ethereum at 9.9%.

What is actually scheduled this week

A note on how this section has changed. Yesterday we retired the practice of attaching if-then predictions to price levels, after two consecutive instances of our own conditions firing in both directions within an hour. So what follows is a list of scheduled events and things we will measure, without thresholds that claim to confirm a direction.

  • Whether a cloture petition is filed, and when. This is the single checkable fact of the week. On the practical timetable a filing by around August 5 permits a vote around August 7. No filing by midweek and the bill waits for September.
  • The Tillis and Gallego compromise text. It is under White House review. Movement there is more informative than any further corporate endorsement, because conflict-of-interest language is what the missing votes are conditioned on.
  • The Polymarket contract. High twenties to low thirties now, from 43% a week ago and about 82% in February. It has been the fastest reacting instrument on this story all week.
  • Tether Gold's buy share. Five readings, all lower, now at 34.8%. We will keep reporting it because it is currently the clearest available read on where the crypto bid is coming from.
  • $64,000 overhead and $62,275 below. Today's high stopped seven dollars short of the round number, and Saturday's low is the floor that has held twice. We are noting them as reference points, not as triggers.
  • August ETF flows. July closed at $172.4 million for Bitcoin against $365.2 million for Ether. The first full week of August shows whether that gap was a July artefact.

How to read this as a trader

Three things from today, in order of how much money they are worth.

First, the procedural detail beats the narrative. All week the headlines have been about endorsements, and all week the probability has fallen. The fact that actually mattered was mechanical and dull: no cloture petition on file, and the only roll-call on Monday's calendar was a funding motion. If you want to know whether a bill is moving, look at the cloture ledger and the daily floor schedule, not at who tweeted support. The same discipline applies to the majority leader's language, where the shift from passing a bill to beginning consideration of one is worth more than a dozen supportive statements.

Second, size on the correct side is the whole signal. Today produced two five minute candles of almost identical value, $24.5 million at 07:05 and $24.6 million at 14:25. The first was 41.7% buy and moved price nowhere. The second was 70.6% buy and moved it 0.53% higher, and the hour containing it took $151.9 million at 63.7%. Volume alone told you nothing. Volume plus the side told you everything, which is the entire argument for watching the split rather than the bar height.

Third, be suspicious of your own caveats. Yesterday we hedged a genuine signal with a volume comparison that turned out to be an artefact of measuring a day before it ended. Sunday was a normal Sunday with the strongest buy share of the last five, and we talked ourselves partly out of it. Publishing the correction is not throat clearing: the same error would have had you doubting a real bid for a full session. Our live dashboard separates aggressive buying from aggressive selling minute by minute across the most liquid pairs, which is how the 14:00 hour was visible while it was happening rather than afterwards. For the arc behind this, see yesterday's endorsements and the 34% odds, Friday's break and the 49.5% July, the two pre-release bids, and the Fed's 9 to 3 hold. If the buy and sell split is new to you, the volume guide covers the mechanics.