#4000-level

Gold lost 4,103 last night. Then 4,029. The low was 4,013. It trades at 4,048.74 now (09:44 Beijing time, 11 Nov, London spot via Tonghuashun). What knocked it over was not gold's own story. The 21:30 CPI print: headline +0.4% m/m versus 0.3% expected, core +0.3% versus 0.2% expected. The part that matters is that core has now printed 0.3% three months in a row. There is no year-on-year number, because the shutdown cost the BLS its 2025 base, so the monthly path is all we have, and three straight 0.3s is a harder problem than one hot print. After the release, 10-year Treasury yields touched 4.94%, the 2-year 4.73%, and DXY closed at 102.24. A 25bp December hike is now fully priced, while 50bp odds jumped from zero before the data to somewhere between 10% and 13.6%. Waller was blunt: 25bp alone puts the Fed's credibility at risk. On the chart, the 9 Nov high at 4,148.60 and the 10 Nov high at 4,146.70 form a double top, and last night's candle confirmed it. Daily range: 133 dollars. Price is still below the 20-day Bollinger mid at 4,059 and has not reclaimed it. I do not think 4,000 breaks cleanly on the first attempt. Oil is down 15% this month with WTI near 86, which leaves room for November CPI to cool; and equities are selling off, Nvidia down 9.6% in two sessions, so that money has to sit somewhere. How I am playing it: 3,970 to 4,000 is the buy zone with a stop below 3,955; 4,080 is where I would look to sell a bounce, because the OANDA order book shows the heaviest resting offers there. If 4,000 cannot be reclaimed on a close, the next shelf is 3,858, the 38.2% retracement of the entire 2015-2026 advance, per WGC. Are you long or short, and at what price? Give me one number and I will use it as the case study in tomorrow's review. Structural read, not investment advice.

Gold gave up 4103 last night. Then 4029 went too. The low was 4013, and right now it’s at 4048.74 — 11-11, 09:44 Beijing time, Tonghuashun London spot gold. This wasn’t really about gold itself. It was the 21:30 CPI print. Headline month-over-month came in at 0.4% versus 0.3% expected. Core was 0.3% versus 0.2% expected. Worse, core CPI has now printed 0.3% for three straight months. The year-over-year number isn’t available because of last year’s government shutdown, so this report can only be judged on the monthly figures. One hot print is noise. Three months without cooling? That’s a different problem. After the data, the 10-year U.S. Treasury yield touched 4.94%, the 2-year hit 4.73%, and the U.S. dollar index closed at 102.24. A December 25-basis-point hike is almost 100% priced in. The probability of a 50-basis-point move went from zero before the release to 10%, then 13.6%. Waller put it bluntly: hiking by only 25 basis points is gambling with the Fed’s credibility. Technically, the 11-09 high at 4148.60 and the 11-10 high at 4146.70 formed a double top. Last night’s bearish candle confirmed it, with a daily range of $133. Price is still below the 20-day Bollinger middle band at 4059 and hasn’t reclaimed it. I don’t think 4000 breaks cleanly on the first try. Oil is down 15% this month, with WTI near 86, which gives November’s CPI room to pull back. U.S. stocks are also falling — Nvidia is down 9.6% in two days — and the money coming out has to park somewhere. Here’s my approach. The $30 zone from 3970 to 4000 is a buy-the-dip area, with the stop loss below 3955. I’ll only consider shorts on a rebound toward 4080, because the OANDA order book shows the thickest sell orders at 4080. If it can’t close back above 4000, the next level is 3858 — the 38.2% retracement of the entire 2015–2026 rally identified by WGC. Are you holding a long or a short? What’s your cost? Give me one number, and I’ll use it as an example in tomorrow’s review. Structural projection, not investment advice, and no return guarantee.