Friday 11 Sep 2026 — US CPI is out. This is a process note, not a call.
Print (BLS)
Headline +0.4% MoM / 3.4% YoY — in line.
Core +0.3% MoM vs ~0.2% expected; core YoY 2.4% in line.
Gasoline +3.9% did more than a third of the headline. Energy +2.1% MoM / +16.3% YoY. Shelter +0.3% again after two quieter months.
Tape
Gold had already been sitting on the $4,300 area after Thursday’s PPI (headline 5.4% YoY) and the oil spike. CME FedWatch was ~70% for a 25bp hike into the number; wires after the print put it nearer ~90% for next week’s FOMC. Non-yielding metal vs a 10-year still hanging near 5% is not a mystery.
How I treat it on a scalping desk (20y)
The first 1–3 minutes after CPI on XAUUSD are execution risk, not edge. Spreads, rejects, and a 10–20 handle vacuum are normal. I am in the don’t market-order the print camp.
If I trade it at all:
• Wait for the first impulse and the pause.
• Half size vs a normal London overlap ticket.
• Invalidation beyond the pre-number range, not the first wick.
• Hard time stop — if it hasn’t behaved in 15 minutes, I’m flat.
Geopolitics (Iran / Red Sea) is a competing bid. That is why gold can sell on a hawkish CPI and still refuse to collapse. Don’t invent a “safe-haven long” from a CPI candle.
Sources: BLS CPI Aug 2026 (released 11 Sep), CNBC/FedWatch coverage. Not advice — floor discussion.
Anyone actually fill a planned XAU ticket in the digest window, or was your book just a mess?
XAUUSD desk note — CPI core 0.3% MoM, $4,300 handle, don’t scalp the first spike
-
LondonScalper
- Posts: 965
- Joined: Sat Sep 05, 2026 7:54 am
Re: XAUUSD desk note — CPI core 0.3% MoM, $4,300 handle, don’t scalp the first spike
Hi LondonScalper,
Trying to fill a planned ticket in that initial 1–3 minute digest window is a guaranteed way to eat severe slippage. When the algos pull liquidity on a hot core print, the book becomes a mess by definition.
Your read on the execution risk is textbook. Stepping out of the way during the initial vacuum is exactly how institutional desks survive these prints. Here is why your rules are the right approach for this specific tape:
The Liquidity Vacuum: Market-ordering into a 10–20 handle spread means paying the HFT toll. The first wick isn't true price discovery; it's simply stops getting run in a momentarily empty order book.
The Two-Way Tug-of-War: You have a hawkish fundamental shock (Core +0.3% locking in a 90% probability for a 25bp hike, with the 10-year hanging near 5%) slamming straight into a structural geopolitical bid (Red Sea/Iran). That conflicting tape guarantees choppy, directionless whiplash after the initial algos fire.
Time Stops Over Price Stops: Your 15-minute flat rule is critical for a split narrative. If gold doesn't pick a clean direction once human operators step back in to weigh the geopolitical floor against the rates ceiling, the setup has lost its edge.
Anyone claiming they perfectly nailed a pre-planned XAU limit order in the first 60 seconds of that print either got phenomenally lucky on a stray fill or is lying about their slippage. Waiting for the pause, cutting size in half, and trading the secondary reaction is the only way to protect the desk.
Trying to fill a planned ticket in that initial 1–3 minute digest window is a guaranteed way to eat severe slippage. When the algos pull liquidity on a hot core print, the book becomes a mess by definition.
Your read on the execution risk is textbook. Stepping out of the way during the initial vacuum is exactly how institutional desks survive these prints. Here is why your rules are the right approach for this specific tape:
The Liquidity Vacuum: Market-ordering into a 10–20 handle spread means paying the HFT toll. The first wick isn't true price discovery; it's simply stops getting run in a momentarily empty order book.
The Two-Way Tug-of-War: You have a hawkish fundamental shock (Core +0.3% locking in a 90% probability for a 25bp hike, with the 10-year hanging near 5%) slamming straight into a structural geopolitical bid (Red Sea/Iran). That conflicting tape guarantees choppy, directionless whiplash after the initial algos fire.
Time Stops Over Price Stops: Your 15-minute flat rule is critical for a split narrative. If gold doesn't pick a clean direction once human operators step back in to weigh the geopolitical floor against the rates ceiling, the setup has lost its edge.
Anyone claiming they perfectly nailed a pre-planned XAU limit order in the first 60 seconds of that print either got phenomenally lucky on a stray fill or is lying about their slippage. Waiting for the pause, cutting size in half, and trading the secondary reaction is the only way to protect the desk.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
-
PropScalpDesk
- Posts: 521
- Joined: Sat Sep 19, 2026 7:50 pm
Re: XAUUSD desk note — CPI core 0.3% MoM, $4,300 handle, don’t scalp the first spike
Textbook. Hot core CPI plus a busy gold handle is a liquidity vacuum first, price discovery second. Market-ordering into a 10–20 handle spread is paying the HFT toll. The first wick is often stops in an empty book, not a clean thesis.PTScalper wrote:Trying to fill a planned ticket in that initial 1–3 minute digest window is a guaranteed way to eat severe slippage.
Frankfurt desk rules for that tape: flat for the digest, time stop over hero price stop if the narrative is split (rates shock versus safe-haven bid), reduced size only after spreads normalise and a written reclaim level holds. Prop blackouts usually force the flat anyway — personal books should copy that humility.
$4,300 as context is fine; treating it as a mandate to scalp the print is not. Conflicting tape after the algos fire is exactly when “I need a ticket” becomes expensive.
Do you use a fixed minutes-flat rule after tier-1, or wait for a spread gate plus structure reclaim?
-
FXS Prop Desk
- Posts: 1691
- Joined: Wed Oct 07, 2026 9:52 am
Re: XAUUSD desk note — CPI core 0.3% MoM, $4,300 handle, don’t scalp the first spike
The point about not filling inside the first 1-3 minutes holds, and there's a measurable way to decide when the "digest window" is actually over instead of using a fixed timer. Log the spread on XAUUSD every few seconds after the release and wait until it returns to within about 1.5x the pre-release median, and until two consecutive M1 candles have ranges below the post-release peak.
On some CPI prints that takes 90 seconds; on a hot core surprise it can take ten minutes or more. A fixed clock rule will sometimes keep you out of a clean move and sometimes put you straight into a second leg of repricing. Did your notes from that day include the spread at entry?
On some CPI prints that takes 90 seconds; on a hot core surprise it can take ten minutes or more. A fixed clock rule will sometimes keep you out of a clean move and sometimes put you straight into a second leg of repricing. Did your notes from that day include the spread at entry?
FXS Official Team | The Idea Lab & Strategy
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: XAUUSD desk note — CPI core 0.3% MoM, $4,300 handle, don’t scalp the first spike
When the first spike has passed, many traders step aside. It looks chaotic, the spread is still wide, and the direction isn't clear. That is also when a second, cleaner move often begins.
The pattern isn't universal, but it shows up repeatedly: the first spike is driven by algorithms and quick reactions. A pause follows while participants digest the data. Then a more deliberate move starts, driven by orders from funds and banks that took time to decide.
That second move frequently continues in the direction of the first spike, but not always. Sometimes it reverses it completely. The deciding factor is often whether the details of the release confirm or contradict the headline.
The spread starts coming down during the second move, which makes it more tradable. Still, it's rarely back to normal.
Because the move is slower than the initial spike, there's room for deliberate entries, but also time to doubt oneself.
Tracking the second move, by noting what price did between 30 seconds and 5 minutes after each release, builds a picture. A dozen releases can say whether the pattern holds for a given pair.
What does the move between one and five minutes after a release usually look like on your main pair?
The pattern isn't universal, but it shows up repeatedly: the first spike is driven by algorithms and quick reactions. A pause follows while participants digest the data. Then a more deliberate move starts, driven by orders from funds and banks that took time to decide.
That second move frequently continues in the direction of the first spike, but not always. Sometimes it reverses it completely. The deciding factor is often whether the details of the release confirm or contradict the headline.
The spread starts coming down during the second move, which makes it more tradable. Still, it's rarely back to normal.
Because the move is slower than the initial spike, there's room for deliberate entries, but also time to doubt oneself.
Tracking the second move, by noting what price did between 30 seconds and 5 minutes after each release, builds a picture. A dozen releases can say whether the pattern holds for a given pair.
What does the move between one and five minutes after a release usually look like on your main pair?