NFP preview: 65,000 jobs and gold's $4,147 decision line

Gold spent Thursday proving it can rally, then failing to keep it. Spot printed $4,146.92 before the London open — the highest since 20 October — and traded $4,100.10 at 21:00 UTC (gold-api), $47 under the high and still about 1% better on the day. Friday decides whether that was a breakout attempt or the top of a three-day squeeze. The October Employment Situation lands Friday 6 November, 13:30 UTC — 08:30 EST, 21:30 SGT.

Consensus is 65,000. The forecast range is 55,000 wide.

Consensus is +65,000 payrolls, 4.2% unemployment, +0.3% m/m average hourly earnings and a 34.4-hour workweek, per FactSet (Morningstar). The Reuters survey matches the headline with private payrolls at +64,000 (macronotebook); the WSJ panel is softer at +59,000 with wages at +2.9% y/y (MarketScreener/WSJ).

The dispersion matters more than the median. House calls run from +45,000 at KPMG to +100,000 at Pantheon, with Deutsche Bank at +55,000 (KPMG, Kiplinger, Seeking Alpha) — against a series that printed +29,000 in September, about a third of what economists expected, then revised July from +21,000 to −10,000 (Morningstar).

Wages can move December on their own. September CPI ran 3.7% y/y against average hourly earnings of $37.81, up 3.0% — real pay negative again, with KPMG expecting a thirteenth consecutive month of falling real wages on Friday's print (macronotebook, KPMG). A 0.4% monthly wage print is the only genuinely hawkish input available on Friday.

The rate backdrop the number lands on

The Fed held at 3.75–4.00% on 28 October by 10–2, with Cook and Hammack dissenting for a hike, after one hike in September (Investing.com). December odds have been the week's least stable number: 81% at 13:16 GMT (Kitco/Reuters), 79–82% through the US morning (Investing.com), 66% by 19:05 GMT (CNBC). Fifteen points of a quarter-point hike left the curve in six hours while gold finished $46 under its high. Bonds pricing less tightening and gold not getting paid for it is the most useful divergence to carry into the print.

Yields did the work: the 10-year at 4.831% after 4.888% on Wednesday, the 2-year at 4.50% — 14 and 13 basis points out of the belly and front end in two sessions (Investing.com). Against 3.7% CPI that is a crude real yield near 1.1%: restrictive enough that this rally stops at moving averages.

The level grid before 13:30 UTC

Thursday's bar was $4,061.40–4,146.92, a $85.52 range or 1.08 × ATR(14) of $79.20, closing $4,111.24 — mid-range, the least informative close of the week (Investing daily). A second read puts the overnight low at $4,057.50 (Investing.com), so the floor is a band, not a tick.

Support: $4,075.02 is the fast average gold has ridden since Wednesday (Capital Street FX); $4,061 is Thursday's low and the first real shelf; $4,022 is the stop in this week's published plan; then $4,000.

Resistance: $4,146.92 is the rejection high. Above it October's shelf stacks at $4,177.59, $4,190.57 and $4,198.84, then the averages at $4,203.81 and $4,209.90 — the $4,187–4,227 band where the 50- and 100-day sit, and the ceiling this rally has to clear to stop being a bounce (Capital Street FX, Investing.com).

Gold ran $276.75 from the 30 October low of $3,870.17 to Thursday's high, up 7.15% in five sessions, two of them wider than $150. The 2 October payrolls ran a $100.37 day range and closed $35.90 lower on a report well below forecast (Investing daily, mrkts) — range expansion without direction.

Three scenarios, three different trades

Hot — 100,000+, wages 0.4%, unemployment 4.1%. The curve reprices toward Thursday morning's 80% December hike; two-year yields lead and the dollar follows. Gold loses $4,061, and the trade is a retest-and-reject toward $4,022 with $4,000 as extension.

In line — 55,000–70,000, wages 0.3%, unemployment 4.2%. The consensus box and the least tradeable outcome. Gold stays inside $4,061–4,147, the range compresses into the close, and the correct position is none.

Soft — 40,000 or less, or unemployment 4.3%. December odds below 66% and falling, the 2-year through 4.50%, and gold takes $4,146.92 with $4,177–4,199 as the first zone. The $4,203.81–4,209.90 wall is where the trade gets re-underwritten: a close above $4,227 restarts the $4,400 conversation, and a rejection there is the board's cleanest fade.

What says the read is wrong

If a hot print cannot break $4,061 and gold reclaims $4,075 inside the hour, the dollar is not confirming the rate move — the print is liquidity, not information. If a soft print spikes into $4,177–4,199 and closes the five-minute back under $4,147, the rally was already positioned and the fade is the trade instead. If the 2-year and the dollar disagree with equities in the first ten minutes, follow the 2-year.

The execution note that decides the day

Spread widens to several dollars in the release minute, and stops placed before 13:30 UTC fill at the wick rather than at your level. The desk's reaction dataset holds the clean version of that lesson: on September PCE, gold was +$5.79 at five minutes and −$0.51 at thirty — a round trip inside half an hour. Wait for the five-minute close, size at half normal risk, and let the first hour print the level.

Payrolls is a coin flip with a 55,000-wide forecast range attached. Friday's print is the last top-tier input before the 10 November CPI. The NFP hub collects the desk's payroll coverage, and the claims piece explains why consensus sits at 65,000 and not 150,000.

Sources: BLS Employment Situation schedule, Morningstar, Kiplinger, CNBC, Investing.com; spot and daily bars from gold-api and Investing daily.

Not financial advice. Every level here is the desk's own published work, not a recommendation; gold can move $80 on this release and slippage can exceed any stop distance quoted. Past reactions do not predict this one.