Tuesday: ISM 54.6 held expansion, JOLTS 7.27M openings. Gold lost 2.86 percent.
Three-print 10 AM ET compressed morning corroborated Friday post-Warsh flip. ISM Manufacturing PMI for August at 54.6 (down 1.0 from 55.6, eighth month of expansion); Employment sub-index 51.2 above 50; Prices Paid above 60 for the fifth consecutive month. JOLTS July at 7.271 million openings (+89k), openings rate 4.4 percent; hires fell 294k to 5.054 million (rate 3.2 percent, weakest since February). Case 1 aligned-hawkish signature under the compressed-release framework. Gold -$128 (-2.86 percent) to $4,357, intraday low $4,318 tested the $4,300-$4,400 back-fill zone. 10Y broke through 4.75 percent to 4.80 (+8.8bp). DXY reclaimed 99.42 with third sub-100 session. 2s10s widened to 41bp (not recession pricing). Setup: hawkish-hike moves to 55 percent from 45.
Catalyst check. Tuesday September 1. The three-print 10:00 AM ET compressed morning landed as scheduled. ISM Manufacturing PMI for August printed 54.6, down 1.0 point from July's 55.6, the eighth consecutive month of expansion; New Orders 53.7 (down 3.0 from 56.7), Production 58.3 (down 0.2), Employment 51.2 (down 1.6 from 52.8), Prices Paid stayed above 60 for the fifth month. JOLTS for July printed 7.271 million openings (up 89k from a revised 7.182 million), openings rate held at 4.4 percent; hires fell 294k to 5.054 million (rate 3.2 percent, the weakest hiring month since February); quits 3.1 million and layoffs 1.7 million both essentially unchanged. Construction Spending for July printed +0.4 percent MoM in line with consensus. All three release times verified against the ISM Report On Business calendar and the BLS JOLTS release schedule.
Yesterday's piece called for ISM between 53 and 56 and JOLTS above 7.0 million as the confirmation of the post-Warsh hawkish-hike distribution; both hit. The distribution-flip framework's Week 2 forecast is that data either corroborates or contradicts the flip, and the market re-prices accordingly. On today's data, the flip is corroborated.
The tape
- Gold: $4,357.90, down $128.30 (-2.86 percent) from Monday's $4,486.20. Intraday range $4,318 to $4,509. Session low of $4,318 tested the $4,300-$4,400 back-fill zone the multi-year-high framework's Phase 3 rules identified as the next test lower. The pair closed above $4,350 but the low print is now confirmed below $4,320. Two-session cumulative decline from Friday $4,474 to today $4,357 is roughly 2.6 percent; three-session from Wednesday $4,593 is 5.1 percent.
- 10-year yield: 4.8010 percent, up 8.8bp from Monday's 4.7130. Broke through 4.75 percent on the ISM print at 10:00 AM ET and extended through the London afternoon. The 2-year moved 7bp to 4.396 percent; 2s10s widened another 2bp to 41bp. The flattening reversal from Friday's 27bp is a specific tell: the market is pricing the hawkish flip as growth-supportive, not as recession-inducing.
- DXY: 99.42, up 14 pips from Monday's 99.28. Held above 99 for the third consecutive session. Intraday range 99.16 to 99.58. The ISM+JOLTS print at 10:00 AM produced a 26-pip pop to 99.58 that faded 16 pips into the London close. Second-derivative firming, not the amplified extension of a fresh flip signal.
- Brent CFD spot: $84.72, up 86 cents. Modest firming; no fresh headline.
- USD/JPY: 160.86, up 68 pips. The 161 handle tested in Asia; MOF verbal-intervention watch now live at 162.
- EUR/USD: 1.1548, down 64 pips. Broke through 1.155 for the first close below since August 8; the reversal of last week's post-PCE bid is now confirmed on the daily.
- GBP/USD: 1.3438, down 84 pips.
The read
Two prints in one window, both leaning the same direction, produced the amplification the compressed-release framework forecasts for a Case 1 aligned session. ISM Manufacturing held expansion at 54.6 with Employment sub-index still above 50 and Prices Paid above 60 for the fifth consecutive month; that combination is the specific signature Warsh's Jackson Hole speech pointed at (labour still tight, prices still sticky). JOLTS at 7.271 million with the openings rate holding at 4.4 percent tells the same story from the labour-market side: openings are still elevated relative to unemployed workers, the vacancy-to-unemployment ratio remains above 1.0, and the Beveridge-curve position is still on the tight side.
The hires-side softness in JOLTS is the one flag worth naming. Hires fell 294k to a rate of 3.2 percent, the weakest hiring month since February. Under the Beveridge curve framework, that combination (openings elevated, hires falling, quits stable) is the specific signature of a frozen labour market: employers are posting jobs but not filling them, workers are staying in place, and the churn that normally reallocates labour across sectors is not happening. That is a different signature from a soft labour market; it is a stalled labour market. For Warsh's hawkish reaction function it is roughly neutral (openings support the tight-labour story; falling hires support a wait-and-see stance). For the September FOMC it does not change the setup materially.
The gold move is the story worth staring at. A $128 drop from Monday's close into a low of $4,318 that touches the $4,300-$4,400 back-fill zone is a Phase 3-to-Phase 4 transition on the multi-year-high framework's clock. The two questions from yesterday's piece (whether Phase 4 is extended decline or Phase 4' rebase-and-restart) are being decided now. The specific signal on today's tape that leans toward extended decline: the intraday session low of $4,318 was not defended aggressively; the pair spent time under $4,325 through the New York afternoon before closing $32 higher. That is a market that is comfortable trading at the lower end of the range, not a market that is aggressively bidding the pair off the low. The Phase 4 extended-decline base rate is roughly 60 percent when the intraday recovery from the session low is under 50 percent of the day's range; today's recovery was 30 percent.
The rates move is the second-cleanest signal on the tape. A 10-year through 4.75 on a print day that ran ISM+JOLTS both firm is exactly the hawkish-Fed pricing the distribution-flip framework calls for; the 8.8bp move is above the trailing 30-day daily average of 4.2bp by about 2x, which is the amplification signature of a Week-2 confirmation session. The 2s10s widening rather than flattening is the specific tell that the market is not pricing a recession from the hawkish Fed; if it were, the front-end would move more than the belly and the curve would invert further.
Setup update
Working thesis from Friday holds and strengthens. Hawkish-hike at 55 percent (up from 45; today's ISM+JOLTS combination gives Warsh the labour-and-prices cover he needs for a September hike). Hold-with-hawkish-language at 30 percent (down from 35). Dovish-hold at 12 percent (down from 15). Dovish-cut at 3 percent (down from 5).
Confirmed if: ADP tomorrow prints above 60k and the July revision holds firm. ISM Services Thursday prints between 53 and 55 (holding expansion). Claims Thursday prints between 205k and 225k (labour market firm but not tight). NFP Friday prints between 100k and 150k with unemployment at 4.2 or below.
Invalidated if: ADP below 20k tomorrow with a negative July revision. ISM Services below 50 (contraction). Claims above 240k. NFP below 50k. Any two of the four would give the dovish-hold tail (currently 12 percent) enough cover to move to a real weight, and would put the flip-fades base rate into play under the distribution-flip framework's 30 percent second-tier resolution.
Watch tomorrow: Wednesday September 2 is ADP Employment Change at 8:15 AM ET (consensus around 47k). ADP is the second-derivative Wednesday-before-NFP anchor; a firm print corroborates the hawkish flip, a soft print puts the flip's underlying labour-market story in question. See the framework piece we published this morning on ISM plus JOLTS as a single signal for the compressed-morning read; the follow-on Wednesday framework on ADP's signal quality in a hike-flip week publishes tomorrow.
Nothing on this site is investment advice or a recommendation to trade. Setups published here are falsifiable hypotheses, not signals.