News filter for scalpers: when I stand aside vs when I still take A+ setups
Posted: Sat Sep 05, 2026 12:56 pm
Scalping into high-impact news is usually a tax paid to spreads and slippage. My default on XAUUSD and USD majors is stand aside around the calendar, not “trade the spike.” Exceptions exist, but they are rare and written down before the week starts — never invented mid-release because a candle looked tradable.
How I filter:
1. Mark the week’s high-impact events in advance (CPI, NFP, FOMC, major central-bank decisions, and any USD release my broker’s gold typically widens on). Medium-impact items get a lighter filter; high-impact gets a hard window.
2. Hard blackout: typically 5–15 minutes before the release through 5–15 minutes after, adjusted to how violent that event is for my feed. FOMC-style events get a wider stand-aside than a routine data print.
3. Soft filter after the window: I may resume only when spread has normalized to my max gate and M15 structure is readable again. First impulse after the print is not automatically a setup.
4. Intraday “surprise” headlines: if I am already in a trade, I manage risk by plan; I do not add. If I am flat, I do not chase the first M1 explosion.
When I stand aside completely:
- Any high-impact USD event inside my planned scalp session unless it is on the exception list.
- When I am already near my daily soft stop. News is an efficient way to finish a limit I was trying to respect.
- When spreads are already elevated going into the event. The platform is telling me liquidity is thin; I do not need a second opinion from Twitter.
When I still take an A+ setup (exceptions — rare):
- Pre-defined only: for example, a level that already mattered on H1/M15, with bias clear before the event, and entry only after spread normalization and a reclaim/rejection that fits the playbook — not during the first chaotic seconds.
- Reduced size relative to normal risk (often half).
- No averaging, no second attempt if the first post-news idea fails.
- If the exception checklist is incomplete, the answer is flat. Ambiguity means stand aside.
What this is trying to prevent is the classic scalper story: “I only wanted the volatility,” followed by a fill far from the intended price and a stop that never had a fair chance. Volatility without a fill model is not edge.
Practical discipline that helps me keep the filter:
- Calendar on the desk or a single sidebar — not a surprise mid-trade.
- Max-spread gate automated or checked every entry; news days fail that gate often.
- Journal tag: “news-adjacent” so weekly review shows whether those trades earn their stress.
I do not need to catch every gold move. I need to keep process intact across a month of calendars. Standing aside is an active decision. Taking an A+ after conditions normalize is also active — and it should feel boringly pre-committed, not heroic.
If you scalp news by feel, write your blackout windows and exception criteria once, then follow them for two weeks. Compare expectancy and emotional tags inside vs outside the window. Most of the time the data argues for more standing aside, not cleverer spike entries.
How I filter:
1. Mark the week’s high-impact events in advance (CPI, NFP, FOMC, major central-bank decisions, and any USD release my broker’s gold typically widens on). Medium-impact items get a lighter filter; high-impact gets a hard window.
2. Hard blackout: typically 5–15 minutes before the release through 5–15 minutes after, adjusted to how violent that event is for my feed. FOMC-style events get a wider stand-aside than a routine data print.
3. Soft filter after the window: I may resume only when spread has normalized to my max gate and M15 structure is readable again. First impulse after the print is not automatically a setup.
4. Intraday “surprise” headlines: if I am already in a trade, I manage risk by plan; I do not add. If I am flat, I do not chase the first M1 explosion.
When I stand aside completely:
- Any high-impact USD event inside my planned scalp session unless it is on the exception list.
- When I am already near my daily soft stop. News is an efficient way to finish a limit I was trying to respect.
- When spreads are already elevated going into the event. The platform is telling me liquidity is thin; I do not need a second opinion from Twitter.
When I still take an A+ setup (exceptions — rare):
- Pre-defined only: for example, a level that already mattered on H1/M15, with bias clear before the event, and entry only after spread normalization and a reclaim/rejection that fits the playbook — not during the first chaotic seconds.
- Reduced size relative to normal risk (often half).
- No averaging, no second attempt if the first post-news idea fails.
- If the exception checklist is incomplete, the answer is flat. Ambiguity means stand aside.
What this is trying to prevent is the classic scalper story: “I only wanted the volatility,” followed by a fill far from the intended price and a stop that never had a fair chance. Volatility without a fill model is not edge.
Practical discipline that helps me keep the filter:
- Calendar on the desk or a single sidebar — not a surprise mid-trade.
- Max-spread gate automated or checked every entry; news days fail that gate often.
- Journal tag: “news-adjacent” so weekly review shows whether those trades earn their stress.
I do not need to catch every gold move. I need to keep process intact across a month of calendars. Standing aside is an active decision. Taking an A+ after conditions normalize is also active — and it should feel boringly pre-committed, not heroic.
If you scalp news by feel, write your blackout windows and exception criteria once, then follow them for two weeks. Compare expectancy and emotional tags inside vs outside the window. Most of the time the data argues for more standing aside, not cleverer spike entries.