Gold broke $4,000 on the close, and fear had nothing to do with it. The stock market did.
Spot gold ran to 4,066.81 overnight, then slid to 3,982.43 and settled near 3,987.24, down 1.35% (snapshot 07:16 Beijing time, 12 November 2026, Tonghuashun). At the same hour, the Dow was up 0.49% at 52,486.49, the Nasdaq up 1.6% at 25,466.19, and the S&P at 7,619.38 (close 11 November, Wind).
Put those two facts side by side and the character of the move is clear: capital did not run to a safe haven. It ran into equities.
The numbers first
Spot gold: open 4,038.78, high 4,066.81, low 3,982.43, close 3,987.24, down 54.58 or 1.35% (Tonghuashun, 12 November 07:16). Intraday range 84.38.
December COMEX gold: settled 4,000.50, down 29.70 or 0.74% (Kabutan citing MINKABU, 12 November 03:58). Settlement print 3,992.00 (Dow Jones Market Data, 11 November 14:04 ET).
One confusion worth clearing up: spot fell 1.35% while futures fell only 0.74%. That is not a contradiction. Futures marked at 14:04 ET; spot kept trading into the Asian morning. Two slices of the same decline.
Dollar index: 102.908, up 0.66% (Xinhua Finance, 12 November 06:10). It closed at 102.23 the day before, climbing from 102.15 to 102.9 in two sessions.
Treasuries: 2-year 4.742%, up 18.46bp on the day; 10-year 4.996%, up 11.17bp, hugging 5% (11 November market wrap).
Equities: Dow 52,486.49 (+0.49%), Nasdaq 25,466.19 (+1.6%), S&P 7,619.38 (Wind, 11 November close). Note that Nvidia had fallen 5.4% the previous session, 9.6% over two days, and tech still flipped green.
The chain, link by link
First, CPI closed the door on cuts. October headline CPI came in at +0.4% month over month against 0.3% expected and 0.6% prior; core printed +0.3% against 0.2% expected and 0.3% prior, the third consecutive 0.3%. The word that matters is "consecutive". That is stickiness, not noise.
As a result, a December hike is fully priced, with some participants still betting on 50bp, probability in the 10% to 13.6% range (Chinanews, 12 November 07:04). Direction matters here: hike, not cut.
Second, the dollar got repriced. Two sessions took it from 102.15 to 102.908. Gold is quoted in dollars, so a firmer dollar mechanically discounts the metal.
Third, real rates rose and gold bled first. Gold pays no coupon. With the 10-year hugging 5% and the 2-year up 18bp in a day, the opportunity cost of holding bullion rose. Nobody had to panic for that to happen.
Fourth, and this is the one most commentary missed: the money did not leave, it relocated. In a genuine liquidity crisis, equities fall with gold, because bullion is what gets sold last. Overnight was the opposite. Gold was not a refuge. Gold was the ATM.
Add easing geopolitical tension, and the safe-haven premium deflated on top of all that (Chinanews, 12 November 07:04).
Why stocks up, gold down is the more worrying version
I wrote yesterday that 4,000 was the level. It broke on the close.
How it broke matters. Panic selling has a signature: fast down, fast bounce, because what is being sold is emotion. This was different. Gold first pushed to 4,066.81, pulling in late longs, then ground down to 3,982.43 and closed at 3,987.24, only 4.8 dollars off the low. That is distribution with bids, not a flush.
Put differently: the market was not selling gold. It was swapping gold for equities.
So the question I care about is not whether 4,000 broke. It is who was standing next to it when it broke. Standing next to it was a Nasdaq up 1.6%.
Being honest about the other side, two counter-scenarios deserve air:
Counter one: jobless claims weaken tonight. The prior print was 189k for the week ending 31 October (FRED, updated 5 November), an extremely low number. A jump above 200k would soften the December hike pricing, pull the dollar back, and open a rebound toward 4,050 to 4,066. That would be a repair, not a reversal.
Counter two: central bank demand has not gone anywhere. WGC data shows 321 tonnes bought in Q3, up 42% year over year, and 675 tonnes over the first nine months, up 6%. That bid does not vanish because of one CPI print. Which is why the 3,800 to 3,900 zone should have real buyers, and why a slide there will not be frictionless.
Neither counter changes today's read. Both keep me from pressing shorts into 3,984.
Data and levels
Data (Beijing time, 12 November):
21:30 US weekly jobless claims, prior 189k (FRED, week ending 31 October). The hardest variable of the day.
23:00 US October existing home sales.
01:00 next day US weekly EIA crude stocks.
15:00 UK Q3 GDP, indirect for gold but relevant through the dollar cross.
Levels:
- Resistance: 4,000, the psychological line, now flipped from support to resistance; 4,050, first rebound target (StockWire X, 12 November 00:00); 4,066.81, this morning's high; 4,080, the 4-hour Kijun line (Traders Union, 12 November 01:08).
- Support: 3,960, the bull-bear divide named by both Myfxbook and StockWire; 3,919, the 4 November low; 3,903.48, the double-top measured target (height 122.56, neckline 4,026.038, my own calculation on 11 November); 3,860, the next-leg trigger flagged by WGC.
My read is simple: 3,984 is a watch zone, not a buy zone.
If gold cannot reclaim 4,000, 3,960 gets tested. What happens after that test decides whether this is a pullback or a trend change.
Where do you stand on tonight's claims print? If it stays below 190k, do you buy 3,960 or wait for 3,903?
Data collected 07:16 to 07:20 Beijing time, 12 November 2026. Sources: Tonghuashun, Wind, Xinhua Finance, Kabutan, Dow Jones Market Data, FRED, StockWire X, Traders Union, Myfxbook, WGC. Structural projection, not investment advice, no return guarantee.
Related
Check out more trading insights in the Trading Journal.