Actionable insights — Gold and Silver's Selloff Was an Overreaction

1. Strip the seasonal and non-cyclical sectors out of a payroll print before you trade it

1. Never act on the headline number. Open the sector table and rank the contributions to the month's gain.
2. Flag every line that is calendar-driven or non-cyclical rather than a read on demand: school-year education hiring, seasonal retail/hospitality, government, weather-affected construction, strike returns, census-type one-offs.
3. Subtract them and re-derive the underlying private-economy gain. That residual is the number the cycle actually produced.
4. Compare the residual — not the headline — to what the market had already priced. If the residual matches the prior expectation, the print delivered no new information, and any price move against it is positioning, not repricing.
5. Cross-check the household survey (the unemployment rate) for confirmation: a genuinely hot payroll month should show up there too.

• The share of a month's gain coming from education, government and food services; the gap between the headline and the ex-seasonal residual; whether next month's revisions take the seasonal lines back out again.

2. Discount any large move made in thin, pre-holiday liquidity

1. Before interpreting a sharp move, check the calendar first — the session before a long weekend, month/quarter-end, the last trading day before a major holiday, the August/late-December desks.
2. In thin books, the same order flow travels much further: algorithmic programs reacting to a headline keyword face no depth to absorb them, so the print overshoots the information.
3. Classify the move: liquidity event or repricing? A repricing survives the return of normal volume; a liquidity event "exhausts itself."
4. Do not add to or cut a position into the illiquid session. Wait for the first full-liquidity session to see how much of the move holds — that is the real vote.
5. Treat a headline-driven overshoot in an asset you already wanted to own as an entry, not a signal to reduce.

• Session volume versus the 20-day average; whether the move holds through the first two normal-liquidity sessions after the holiday; whether the move is confined to the futures/paper market while physical demand is unchanged.

3. Ask whether the new datum actually disturbs the driver that set the price

1. Name, explicitly, the thing the asset was trading on before the shock — the specific policy signal, spread, or flow that produced the prior move.
2. Ask a single question of the new data: does it change that driver? Not "is it good or bad news" — does it move the variable that is actually setting the price.
3. If the driver is intact, the prior thesis is intact, and the move is noise around an unchanged setup.
4. Only revise when the datum reaches the driver itself — here, a report strong enough to force the Fed off a hold.

• FOMC-member language between the print and the meeting; whether a datum shifts the market-implied path for the meeting date, versus just moving the asset.

4. Test "overheating" against the real wage, not the nominal one

1. When a jobs number is called hot, go straight to average hourly earnings, year-over-year — the wage line is where an overheating labor market has to show up first.
2. Subtract the prevailing inflation rate. Nominal wage growth on its own says nothing; the sign of the real wage says whether the consumer is gaining or losing.
3. A negative real wage is a squeezed consumer, not a wage-price spiral — the condition that argues for holding or easing, not tightening.
4. Pair it with the unemployment rate over a multi-quarter window: a rate sitting in the same narrow band for years is a labor market with no signal in either direction.
5. Conclude on the policy variable, then trade the monetary hedge off that — not off the payroll headline.

• The AHE-minus-CPI spread each month; whether unemployment breaks out of its multi-year band in either direction; the funds rate's time-on-hold.

5. Read a policymaker's or politician's demand as a tell about the data behind the banner

1. When an official celebrates a data release, note what they ask for in the same breath.
2. Check the request against the textbook response to the data as described: easing accompanies weakening economies, tightening accompanies hot ones.
3. A mismatch between the description and the request is information — the person with the better view of the underlying data is behaving as if it is weaker (or stronger) than the version they are selling.
4. Weight the behaviour over the commentary, and use it as corroboration for your own decomposition rather than as a standalone signal.

• Officials praising a print while pressing for accommodation; agencies revising a celebrated number down; the gap between the political framing of a release and the subsequent revision.

6. Size the round-trip off the last identical setup, not off a forecast

1. Once a move is classified as headline-driven noise, don't forecast a price — find the most recent instance of the same setup in the same asset.
2. Match on mechanism, not magnitude: same asset, same kind of trigger (a hawkish speech, a headline print), same absence of a change in the underlying driver.
3. Use that precedent's recovery for the shape and timing of the expected reversal — specifically, what the market had to "process" before the selling stopped.
4. State the invalidation: the precedent fails if the driver itself moves (here, if the Fed actually shifts off hold).

• How many sessions the prior analogue took to round-trip; whether the September FOMC delivers the hold Waller signalled; a second data print that confirms the seasonal read.