Friday failed one marker and reversed Thursday’s leverage in one snapshot. This is the running scoreboard as of 06:10 UTC on Saturday 5 September 2026, every figure pulled at that clock unless a source and date say otherwise. The lead is in today’s news piece; this is the arithmetic. One thing settled: J1, as a fail. Nothing passed. Two grade tonight, at the retarget.

J1: set at 5bp, graded at 3bp, on the largest payrolls beat of the year

J1 was set on Thursday from Guide #40’s finding that 2026’s payrolls days had moved the 2-year Treasury 4.6bp on average across eight releases: the marker asked for at least 5bp on 4 September’s close against 4.34%. The Treasury’s file has 4.37%. +3bp; fail. With the ninth release the payrolls-day mean is 4.4bp (the nine absolute changes: 5, 3, 1, 5, 2, 12, 3, 6, 3), still above the ordinary session’s 3.7bp and still resting on 5 June’s 12bp. The intraday peak, per Reuters, was 4.4246%, which against the Treasury’s 4.34 would have been +8.5bp and a pass; the close is what the marker named, and today’s field guide is about why. The thesis J1 tested was Thursday’s: that the 5bp fall on Waller was the market “waiting for Friday.” Friday came, beat by 109,000, and the close moved three. A speech out-moved the print. Guide #40 said the payrolls half of its result rested on one day; it still does.

Payrolls release (2026)2-year change on the day
9 Jan+5bp
6 Feb+3bp
6 Mar−1bp
3 Apr+5bp
8 May−2bp
5 Jun+12bp
2 Jul−3bp
7 Aug−6bp
4 Sep+3bp
Mean absolute (9)4.4bp
Mean absolute, ex-5 June (8)3.5bp

US Treasury daily par yield curve, 2026 file, close-to-close changes in the 2-year CMT, computed here 5 September 2026. Release dates are the BLS first-Friday schedule as used in Guide #40. Of the 170 daily changes in the file, 64 (37.6%) are 5bp or larger.

The ETF column: Friday’s row was blank

At 06:10 UTC the 4 September row of Farside Investors’ table showed a dash in every fund column and 0.0 in the total, which is the unpopulated-cell state, not a zero-flow day; the row will fill during Saturday. We print nothing for it. Thursday’s row is unchanged from yesterday’s print: IBIT 454.0, FBTC 74.4, BITB 24.8, ARKB 137.7, BTCO 0, EZBC 0, BRRR 0, HODL −19.6, BTCW −5.2, MSBT 7.7, GBTC 8.2, mini BTC 48.8, total 730.8. Yahoo Finance called it the best day in nine months; our table shows fourteen rows and we can confirm only that it is the largest of those. H2 (September net above zero on 30 September) stands at +$595.4 million after three of the month’s 21 sessions, with Friday’s row pending and Monday a holiday.

Open interest: the 5,065 coins that arrived Thursday left on Friday

Binance USDT-margined perpetual open interest at the 00:00 UTC snapshot went from 112,717.545 BTC on 4 September to 107,912.535 BTC on 5 September: −4.263%, or 4,805.01 coins closed. It is the largest one-day decline in the thirty daily snapshots we hold (7 August to 5 September); the runner-up was −4,546.71 coins into 18 August, and the last comparable round trip was −3,152.85 into 29 August, the day the previous “leverage arrived” story reversed. The dollar figure fell from $9,156.55 million to $8,591.58 million, −6.170%; the implied mark fell 1.992%, from $81,234.44 to $79,616.16; and the identity closes: 0.95737 × 0.98008 = 0.93830, the dollar ratio to five places. Net of the two days, open interest is 260.47 coins (+0.242%) above Thursday morning’s 107,652.063. Thursday’s rally was, by Guide #37’s test, about half bought with new contracts; Friday’s fall closed almost all of them. Live open interest at 06:10 UTC was 107,572.078 BTC. G1 (below 100,000 BTC at any September 00:00 snapshot) is 7.91% away, from 12.72%.

Series (Binance)3 Sep 00:004 Sep 00:005 Sep 00:00Fri → Sat
Perp OI, coins107,652.063112,717.545107,912.535−4.263%
Perp OI, US$m8,321.789,156.558,591.58−6.170%
Implied mark77,302.6181,234.4479,616.16−1.992%
Funding, per 8h (00:00 settlement)0.007287%0.008558%0.001014%9.371% → 1.110% ann.
Global accounts long/short1.21480.78161.0165back above 1
Top-trader positions long/short1.97741.92612.1587+0.2326

Own pull, 5 September 2026, 06:10 UTC. Funding annualised as rate × 3 × 365. Account and position ratios are Binance’s daily-period series at the 00:00 UTC row.

Funding collapsed through the day. Friday’s three settlements were 0.008558%, 0.006498% and 0.004455% per eight hours — 9.4%, 7.1% and 4.9% annualised — and the 00:00 Saturday print was 0.001014%, 1.110% annualised, the lowest settlement since 16 August 00:00 (0.000966%) in the 110-row window we hold from 30 July 16:00. The window mean is 0.006701% (7.337% annualised); 26 settlements printed at the 0.01% cap and none was negative. The crowd measures crossed back: Binance’s global long/short account ratio went from 0.7816, Friday’s low, to 1.0165 — retail accounts that had sold into Thursday’s rally bought Friday’s dip — while the top-trader position ratio rose from 1.9261 to 2.1587, the highest of the ten daily rows we hold (27 August onward; the previous high was 2.1462 on 27 August). Ten rows is not a series; it is printed because it is the extreme we have recorded, with the date attached.

Basis: the front contract held at 2.9%, and the perpetual now pays less than it

With the COIN-M index at $79,614.81, the 25 September quarterly at $79,740.7 carried a 0.1581% premium over 20.08 days, 2.875% annualised, against 2.849% on Friday morning; I3 settled on Friday and stays settled. The December contract at $80,665.2 is 1.3193% over 111.08 days, 4.335% annualised, from 4.119%. September sits 104bp below the 3-month bill (3.91%) for a second morning; December is 34bp above the 4-month bill (4.00%). The relationship that flipped is between the perpetual and the future: on Friday the perpetual’s 9.371% funding was 652bp above the September basis; on Saturday, at 1.110%, it is 177bp below it. A perpetual that pays less than the dated contract is a perpetual whose longs have left, which is what the open-interest row says in coins. G2 (September basis above 8.00% on 25 September) is 513bp short with 20 days left.

Difficulty: 90 blocks to the retarget, projected +1.20%, and what would have to happen for D2 to pass

The projection at 06:10 UTC was +1.200% at 95.54% of the period, 90 blocks from the retarget at height 965,664, average block time 593.19 seconds so far, estimated arrival 21:00 UTC tonight. The morning path has been +1.415% (1 September), +0.796%, +0.234%, +1.468%, +1.200%. D2 asked for the adjustment to land in [−1.50%, 0.00%]; F1 asked for at or above zero. For D2 to pass from here the remaining 90 blocks would have to average roughly 12.4 minutes — a straight extrapolation of the period’s time so far puts the adjustment at about +1.15%, near mempool’s projection, and it would take the last 90 blocks arriving about a quarter slower than the period’s average pace, for all 90, to pull the total below zero. That is not impossible for 90 blocks; it is unlikely, and we say so in advance. Today’s difficulty update has the hashrate, pool and fee detail. Sunday’s markers piece grades both on the realised figure.

Price, sentiment, Strategy

Bitstamp’s 4 September candle: open $81,265.00, high $81,426.37, low $78,645.01, close $79,676.60 (−1.955%), 2,226.88 BTC of volume. The fall is the 39th-largest of 2026’s 247 daily changes. The week, Friday close to Friday close, is +2.366% from $77,835.18; September to date is +1.407% from $78,571.17; the year is −8.937% from $87,496. Six of the eight next-days after a 5%-plus gain in 2026 have been red, median −1.176%; the lead has the list. Fear & Greed printed 73 on Saturday, one below Friday’s 74, which had equalled the 2026 high set on 25 August; it is the 17th consecutive Greed day since 20 August. Strategy’s $80,318 tranche is −$2.95 million at Friday’s close and −$3.40 million at the snapshot, back on the side of the bar it started on; H1 is settled and stays settled. MSTR closed $142.80 (−1.39%, S&P Global via StockAnalysis), +12.17% on the week from $127.31. M1 (an 8-K by Tuesday disclosing at least one coin bought 31 August to 6 September) grades Tuesday.

The scoreboard

MarkerBarSettlesReading, 5 Sep 06:10 UTCStatus
H1Bitstamp close ≥ $80,31830 Sep$81,265.00 on 3 SepSettled — pass (3 Sep)
H2Sept Farside net > $030 Sep+$595.4m after 3 sessions; 4 Sep row blankOpen, positive
H330y CMT ≥ 5.25% on 17 Sep17 Sep5.24%Failing by 1bp
D2Retarget in [−1.50%, 0.00%]~21:00 UTC 5 Sep+1.200% projectedFailing by 1.20pp, grades tonight
F1Retarget ≥ 0.00%~21:00 UTC 5 Sep+1.200% projectedPassing by 1.20pp, grades tonight
G1Perp OI < 100,000 BTC, any Sept 00:0030 Sep107,912.535Open, 7.91% away
G2Sept basis > 8.00% on 25 Sep25 Sep2.875%Failing, 513bp short
I110y real ≥ 2.50%, any close to 16 Sep16 Sep2.43%Open, 7bp away
I2BOJ hikes 18 Sep18 SepOpen
I3Sept basis < 4.00% at a 06:00 snapshot before 25 Sep25 Sep2.849% on 4 SepSettled — pass (4 Sep)
J12y CMT 4 Sep close ≥ 5bp from 4.344 Sep4.37, +3bpSettled — FAIL
K1CLARITY cloture ≥ 60 votes by 18 Sep18 SepOpen
L1Bitstamp close 16 Sep ≥ $81,265.0016 Sep$79,676.60 (4 Sep)Open, below
M1Strategy 8-K by 8 Sep, ≥ 1 BTC bought 31 Aug–6 Sep8 SepOpen
N1 (new, P1)FedWatch hike probability ≥ 60% on the evening of 11 Sep, per two major outlets11 Sep52.6–60% on 4 SepOpen

Three settled (H1, I3 pass; J1 fail), two grading tonight, ten open. Grading rules as written when each marker was set; a marker is graded on the source named in it and nothing else.

What we got wrong, and what to watch. Thursday’s reading that the crowd had “sold into the move” on a 0.7816 account ratio was one row; it reverted in one row, which is the reason we print row counts. The claim that the front-month basis had “un-inverted” is holding, but the more useful number today is the perpetual sitting below it. On the calendar: Farside’s Friday row; the retarget around 21:00 UTC; Labor Day Monday; Strategy’s 8-K window Tuesday; the CPI on Friday 11 September at 08:30 ET; the CLARITY cloture vote on 15 September; the FOMC decision on the 16th; the BOJ on the 18th.

Method: prices, funding, open interest, basis and on-chain figures in this article are pulled directly by Bitcoin Mastery at the timestamp stated — Bitstamp BTC/USD daily candles for closes, Binance BTCUSDT spot and USDT-margined perpetual for intraday, open interest, funding and account ratios, Binance COIN-M quarterly contracts for basis, mempool.space for difficulty, hashrate, pool shares and fees, alternative.me for the Fear & Greed series, Farside Investors’ table for ETF flows (every named fund column is read from a print of that column with its header) and US Treasury CMT par yields, nominal and real, for rates. Where a third-party figure is cited we name the source and its date; where two sources disagree we print both. Every streak or extreme figure is published with the first date of its series in the same sentence.

Disclaimer: This article is for informational purposes only and does not constitute investment advice. Cryptocurrencies are volatile and you can lose money. Nothing here is a recommendation to buy or sell any security, digital asset or exchange-traded fund, including MSTR. Do your own research and consult a licensed financial advisor before making investment decisions.