Correction (added 07:20 CST, Nov 7)
The first version of this piece, published 21:43 CST on Nov 6, said December hike odds were "71.1%, up only 0.2pp from before the release." That was the 21:37 intraday print, not a closing price. By the US close, CME FedWatch had December 25bp hike odds at 77%-78%, roughly 8-10 points above Thursday's 68% handle. The premise of the original argument was wrong. This version fixes the conclusion and the headline. The error stays visible.
The call first
Payrolls blew past consensus, December hike odds rose 10 points in three hours, and gold still fell only 0.4%. What sold off was stop-losses and hope, not a freshly built hike expectation.
1. The 21:30 print: all three lines beat
BLS released the October employment report at 21:30 CST on Nov 6:
| Metric | Actual | Consensus | Prior |
|---|---|---|---|
| Nonfarm payrolls | 165K | 65K (Reuters) / 59K (Dow Jones) | 29K |
| Unemployment rate | 4.1% | 4.2% | 4.2% |
| Average hourly earnings m/m | +0.4% | +0.3% | +0.1% |
| Average hourly earnings y/y | +3.1% | +2.9% | +2.9% |
Two details matter more than the headline.
First, revisions. August was revised from 133K to 186K, September from 29K to 60K — a combined 84K upward. October is not a standalone strong month; it is the third strong month in a row.
Second, the composition is solid. Professional and business services added 70K, health care 42K, transportation and warehousing 25K. This is not a government-hiring or seasonal artifact. Private average hourly earnings reached $38.
2. The first fifteen minutes
- 21:33: spot gold -1.15% at $4,041.89 (CLS).
- Session low $4,025.87, a new low for the day.
- 21:40: spot $4,051.87, -0.82% on the day; session high $4,106.74, prior close $4,085.49 (Tonghuashun).
- 10-year Treasury yield +6.3bp to 4.898% (Wallstreetcn, 21:36).
- Spot silver actually rose 0.66% to $52.04; the gold/silver ratio compressed from 79.4 to 77.85.
- CME FedWatch: December 25bp hike odds at 71.1% at 21:37 — an early read, and it kept climbing after the US open (see correction above).
3. Why a blowout only cost 1%
This is the part worth getting right. Three transmission chains fired at once, but only one of them is macro.
One: the hike was already priced.
The 71.1% at 21:37 was only the first print. By the US close, FedWatch showed December 25bp hike odds at 77%-78% (TradingKey, Invezz and AInvest all logged 77%; Guanghuan Market Broadcast logged 78%; Sina had 72.8% at 21:42). The spread across those prints is sampling time, not disagreement. The direction is one way: up roughly 8-10 points from Thursday's 68% handle.
So the chain — strong payrolls, higher hike odds, weaker gold — actually holds. I called it broken in the first version because I mistook an intraday print for a closing price. The interesting part is not whether the chain holds. It is the transmission ratio: odds up 10 points, price down 0.4%.
Two: what actually moved is the real yield.
The 10-year nominal yield rose 6.3bp to 4.898%. Gold pays nothing; it competes with the inflation-adjusted real yield. Nominal up, inflation expectations flat, real yield up. This is the ballast, and its effect on gold is immediate and continuous rather than a one-off repricing.
Three: $4,061 was the technical home position.
Yesterday I flagged $4,061.05 as the double-top neckline: the two Nov 5 peaks at $4,139 and $4,146.89 were less than $8 apart, and the neckline was the prior low at $4,061.05. Once it broke, two groups were forced to act: longs with stops under the neckline, and trend shorts waiting for confirmation.
A meaningful share of the $36 slide to $4,025.87 came from those two groups, not from macro money.
The counter-case deserves space. If the 165K contains a seasonal overstatement, or if the market realizes over the next two or three sessions that a 25bp hike is fully priced and there is no more to price, gold can retest $4,061-$4,085. The fact that a 10-point repricing bought only a 0.4% decline is itself the argument that the door is still open.
4. The chart: $4,061 flips from support to resistance
Levels, re-ranked:
| Level | Price | Source |
|---|---|---|
| Session high | $4,106.74 | Asian session peak |
| Prior close / open | $4,085.49 / $4,085.68 | Gap reference |
| Key resistance (former neckline) | $4,061.05 | Nov 5 low + double-top neckline |
| Spot | $4,051.87 | 21:40 CST snapshot |
| Session low | $4,025.87 | Post-data low |
| Minor support | $4,021 | Lower edge of prior consolidation |
| Major support | $4,000 | Round number |
Measured move: $4,061.05 − ($4,146.89 − $4,061.05) × 0.5 ≈ $4,018, which lands on the $4,021 shelf. The session low of $4,025.87 is $4.87 away from $4,021 — the pattern target is at the door but not filled.
Spot sits $9 under $4,061. That is the entire conflict right now: a retest is underway, and the neckline is directly overhead.
5. Three setups, each with invalidation
Setup A: failed retest, another leg down.
Price returns to $4,058-$4,062, gets capped, and closes under $4,061 — the break is valid.
Entry: short near $4,060, but only after a 15-minute or hourly close below $4,061. Do not chase the first bounce candle.
Stop: $4,072, above the retrace high.
Targets: $4,025.87 first, $4,021 second.
Reward/risk: to $4,021 it is (4,060 − 4,021) / (4,072 − 4,060) = 39 / 12 ≈ 3.2:1, which clears the bar.
Invalidation: any hourly close above $4,072, or a reclaim of $4,085.49.
Setup B: false break, reclaim $4,061.
The post-data break was a wick; price recovers above $4,061 in the later US session and holds.
Entry: long the retest at $4,055-$4,058 after $4,061 holds.
Stop: $4,045, twenty dollars above the session low.
Targets: $4,085.49 first, $4,106.74 second.
Reward/risk: to $4,106.74 it is (4,106.74 − 4,058) / (4,058 − 4,045) = 48.74 / 13 ≈ 3.7:1, which clears the bar.
Invalidation: a slide back under $4,040, or a break of $4,025.87.
Setup C: chop between $4,025.87 and $4,061.
Data is out, direction is not, and the market waits for the Nov 10 CPI. The range is only $35 wide, so no directional bets — trade only the edges at half size.
Getting swept repeatedly is normal here. The goal is not to make money; it is to avoid donating capital to post-data noise.
6. The silver divergence is not noise
Spot silver $52.04, +0.66%; the gold/silver ratio compressed from 79.4 to 77.85.
Gold down 0.82% while silver is up 0.66% tells you two things.
First, this is not a liquidity shock. In a real risk event, gold and silver fall together because everyone needs cash. Not today.
Second, the gold pressure is rate-driven, while silver is being held up by its industrial leg. If silver starts falling too and the ratio pushes back above 79, that is the systemic signal — and then gold will not stop at $4,021.
Silver's own levels: support $51.177 (session low) and $51.10; resistance $52.366 (session high) and $53.47.
7. A note: the market already knew about the hike
Yesterday I read the $4,146.89 upper wick as both sides refusing to carry risk into the print. Today's $4,061 break is the other half: after the print, there is no need to carry risk at all.
Put together, the most honest number in this report is not 165K. It is that 10 points of hike odds bought only a 0.4% decline. The hike is still coming — gold just refuses to pay for the same message twice.
Gold did not sell off on a hike. It sold off on the last bit of wishful thinking.
The real variable is the Nov 10 CPI. If October CPI comes in soft, the real-yield chain loosens and $4,061 can be reclaimed. If CPI is hot as well, $4,021 and $4,000 are not the destination — they are a stop along the way.
One question: if the Nov 10 CPI is soft and $4,061 gets reclaimed, does today's candle count as another shakeout?
Data first collected 21:40 CST on Nov 6, 2026; revised and re-verified 07:20 CST on Nov 7. Closing quotes: spot gold $4,066.11 (-0.47%, high $4,106.74, low $4,025.87, prior close $4,085.49), spot silver $51.952 (+0.48%) from Tonghuashun, read 07:21 Nov 7. New York close $4,071.39 with a 5.06% weekly gain from Wallstreetcn, 06:15 Nov 7. COMEX gold $4,079.50 (-0.83%, +4.63% weekly), COMEX silver $51.97 (-0.78%, +5.46% weekly) from NBD, 07:12 Nov 7. Event: BLS October employment report, 21:30 CST Nov 6. Hike odds: CME FedWatch — 68.7% Thursday, 71.1% at 21:37, 77%-78% by the US close. 10-year yield: Wallstreetcn, 21:36.
Structural read, not investment advice.