US employers added 162,000 jobs in August, the Bureau of Labor Statistics reported at 08:30 ET on Friday, against a Dow Jones consensus of 53,000 and a Reuters consensus of 56,000. It was the largest monthly gain since March; June was revised up by 11,000 to +31,000 and July by 44,000, from −23,000 to +21,000, so the two prior months together are 55,000 higher than first reported. The unemployment rate was unchanged at 4.1%, average hourly earnings rose 0.3% on the month and 3.1% on the year, and the average workweek edged up to 34.4 hours. The BLS’s own framing puts the number beside a 31,000 average monthly gain over the prior twelve months. Food services and drinking places added 59,000 and local government education 42,000; the information industry lost 23,000.
Bitcoin, which had closed at $81,265.00 on Thursday after Governor Waller’s speech, closed at $79,676.60 on Bitstamp on Friday, down 1.955% — a session that ran from $81,426.37 to $78,645.01 and gave back $1,588.40, or 40.08%, of Thursday’s $3,963.17 gain. Bloomberg had the intraday fall at as much as 3.5% and the low at $78,649; our Bitstamp candle has the low $4 lower. As of 06:10 UTC on Saturday the price was $79,578.48 on Binance, 0.123% below Friday’s close. The S&P 500 fell 0.38% to 7,718.60, the Dow 271.86 points to 53,414.25 and the Nasdaq 0.29% to 26,506.99, per CNBC’s market report, all three giving back part of Thursday’s best day in a month ahead of the Labor Day weekend.
The number Waller said would not matter, and the number that moved anyway
Thursday’s lead quoted Governor Christopher Waller’s framing of this report a day before it printed: “I don’t expect that the employment data will deviate much from what we have been seeing. So my decision on the appropriate stance of policy will be heavily influenced by what we learn about August inflation.” The employment data deviated: 162,000 against a 12-month average of 31,000 is, per the BLS, the largest monthly gain since March, and the +55,000 revision removes the negative July that had anchored the “weakening labor market” case. Whether it changes his vote is, on his own words, not the question; the CPI on Friday 11 September is. But the rates market does not wait for a governor’s trigger variable, and by 09:05 ET the CME FedWatch probability of a hike at the 15–16 September meeting was 52.6% against 49.4% on Thursday, per IBTimes’ reading of the tool; CNBC had it at 58% later in the session and other reports put it near 60% by the close. Three snapshots, three numbers, one direction: the market gave back roughly half of what Waller had taken out on Thursday, when it moved from about 67% to about 50%.
The 2-year: about 8–9bp at the peak, 3bp at the close, and a marker that failed
On Thursday this desk set a marker, J1, on the simplest version of the question “does payrolls move rates”: the 2-year Treasury’s constant-maturity close on 4 September would be at least five basis points from Thursday’s 4.34%, in either direction. The bar came from Field Guide #40, which had scored 2026’s eight payrolls days at a 4.6bp average move and set the marker slightly above the mean. The Treasury’s file, published Friday evening, has the 2-year at 4.37%, +3bp. J1 failed. It failed on a day when the print beat the consensus by 109,000, which is the useful part of the result: a big surprise in the data produced a small move in the close, and Thursday’s speech, with no data attached, had produced a bigger one (−5bp).
The intraday story is different, and both belong in the same paragraph. Reuters reported the 2-year “briefly peaked at 4.4246%, its highest since January 2025,” and was “last up 4 basis points at 4.37%” in the afternoon; Bloomberg’s headline had an eight-basis-point climb; the 10-year reached 4.812% before settling near 4.78%. Against the Treasury’s 4.34 the peak was +8.5bp, so the peak-to-close give-back on the 2-year was about 5–6 basis points, and a claim graded on the peak would have passed. Our markers grade on the close, for reasons today’s field guide sets out: the Treasury’s figure is a bid-side quotation taken at or near 3:30 PM Eastern, it is the only version of the number that is published once and never revised, and it is the one every other row in the 2026 file was built from. Reuters’ “+4bp” against the Treasury’s “+3bp” is a difference of reference, not of fact: Reuters had Thursday’s level at 4.33% in its morning report, the Treasury file has 4.34%.
| Tenor | 3 Sep close | 4 Sep close | Change |
|---|---|---|---|
| 3-month bill | 3.89 | 3.91 | +2bp |
| 2-year | 4.34 | 4.37 | +3bp |
| 5-year | 4.52 | 4.54 | +2bp |
| 10-year | 4.77 | 4.78 | +1bp |
| 30-year | 5.25 | 5.24 | −1bp |
| 10-year real (TIPS) | 2.42 | 2.43 | +1bp |
| 10-year breakeven | 2.35 | 2.35 | 0bp |
US Treasury daily par yield curve and daily real yield curve, 2026 files, pulled 5 September 2026, 06:10 UTC. Breakeven is nominal minus real, computed here. Reuters’ intraday figures: 2-year peak 4.4246%, 10-year peak 4.812%.
Two things in the table matter beyond J1. The 30-year closed at 5.24%, one basis point below the 5.25% bar of marker H3 (30-year at or above 5.25% on 17 September), which had been passing by exactly zero on Thursday and is now failing by one; a hot payrolls print that flattened the curve rather than steepening it is the shape of a market repricing the next meeting and not the long-run path, as Thursday’s move was in the other direction. And the 10-year’s 1bp rise was entirely real rate, with the breakeven unchanged at 2.35% — by Wednesday’s split, that is the half that raises the opportunity cost of a zero-yield asset. The 10-year real yield at 2.43% is 7bp from marker I1’s 2.50% bar. Over the week, the 2-year went from 4.34% (28 August) to 4.37%: three basis points, through a Jackson Hole aftermath, a JOLTS miss, an ADP miss, a governor’s speech and the largest payrolls beat of the year.
What bitcoin did, put beside what it did on the other seven days like Thursday
Friday’s −1.955% is the 39th-largest daily fall of the 247 recorded in 2026, which is to say an ordinary bad day, not an event. What makes it worth a paragraph is what preceded it. Thursday’s +5.127% was the eighth daily gain of 5% or more this year. The day after each of the other seven: 7 February −1.822%, 26 February −0.744%, 5 March −2.469%, 14 April −0.353%, 20 August +5.354%, 21 August +7.283%, 22 August −1.608%. With Friday, six of the eight next-days were red, with a median of −1.176%; the mean is +0.461% only because the two green ones were the middle of the 19–21 August three-day run, each a 5%-plus day in its own right. Eight observations are not a rule and we do not print one; the honest statement is that a 5% day has more often than not been followed by a give-back of one or two percent, and Friday was that, with a macro reason attached.
The attribution is cleaner than most. The checklist asks what else happened, and the list is short: the payrolls print at 08:30 ET, the hike-odds repricing that followed it, a 0.4% fall in stocks, and Brent crude holding near $95 on a week in which Reuters had it up 7%. The ETF column, which was $730.8 million on Thursday, had not been posted for Friday on Farside’s table at our 06:10 UTC pull — the row was blank — so we cannot yet say whether the cash leg reversed. What we can say is that the leverage leg did: Binance perpetual open interest, which had added 5,065 coins into Friday’s 00:00 UTC snapshot, gave back 4,805 of them by Saturday’s, and the funding rate fell from 9.37% annualised to 1.11%. Today’s markers piece has the decomposition. Strategy’s stock fell 1.39% to $142.80 on 26.3 million shares, per S&P Global data via StockAnalysis, about 0.7 times bitcoin’s fall after moving 3.4 times it on Thursday; the company’s 4,603-coin tranche at $80,318, $4.36 million in profit at Thursday’s close, was $2.95 million underwater at Friday’s. Marker H1 settled on Thursday’s close and stays settled; a settled marker grades the claim, not the argument, and the argument — that the tranche was bought above its window’s VWAP — is unchanged.
What has to hold, and the next print
The Committee meets in eleven days with a hold that one governor has conditioned on a print that has not happened, a Chair who used the word “hike” three times at Jackson Hole, and a labor market that just removed the weakest argument against tightening. Waller’s own text names the August CPI, due Friday 11 September at 08:30 ET, as the deciding variable; his hike branch (“it may not take much acceleration in inflation to nudge me”) sits closer to the current data than his hold branch, and Friday’s report did nothing to move it further away. Between now and then: Monday is Labor Day, markets closed; Tuesday is the earliest Strategy 8-K that would grade marker M1; the Farside row for Friday, when it posts, tells us whether Thursday’s $730.8 million was a day or the start of something.
As standing practice we mark one falsifiable claim on this. J1 is graded and failed; we do not re-set it. N1: the CME FedWatch probability of a hike at the 16 September meeting, as reported by at least two major outlets on the evening of Friday 11 September after the CPI print, is at or above 60%. The reasoning is that the labor data has now closed the “weak jobs” path to a hold, so the CPI alone decides; a print at or above consensus should push the probability through the level it reached on Friday afternoon, and a soft print should not. If it is below 60% on the 11th, the CPI was soft enough to matter and Waller’s hold branch was live. We set it where the answer is unknown. Marker L1 — a Bitstamp close at or above $81,265.00 on 16 September — is unchanged; Friday’s close below it is information, not a grade.
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.