Non-farm payrolls, US CPI and FOMC decisions move Gold $30–80 in minutes, and they move it in both directions before choosing one. A strategy that is fine 99% of the time can take its worst loss of the year in the three minutes after 08:30 New York. Whether to let an EA trade through that depends on what kind of EA it is — and on knowing exactly when the events are, in your own timezone.
What actually happens at 08:30 ET
- Spread widens 3–10× for 30–120 seconds. A 20-point Gold spread becomes 100–200. Any position opened or closed in that window pays it.
- Slippage: stop orders and market orders fill tens of points away from the requested price. Stop-losses still work, but at a worse price.
- Price gaps: the first tick after the release can be $10–20 from the last one. Grid levels get skipped; a pending order intended for $4,010 fills at $4,022.
- Whipsaw: the initial spike frequently reverses within fifteen minutes. Breakout logic gets triggered both ways.
Which events matter for Gold
In rough order of impact: FOMC rate decisions and the press conference (14:00 / 14:30 ET), US CPI (08:30 ET), non-farm payrolls (first Friday, 08:30 ET), PCE inflation, ISM PMIs, and the Fed chair speaking anywhere. ECB and BoE decisions matter through EURUSD and GBPUSD more than Gold. Everything on the calendar marked "low" can be ignored by an EA; it is the red-folder US events that change the day.
Times are the trap: 08:30 ET is 18:00 IST during US daylight time and 19:00 IST in winter. FOMC is 23:30 / 00:30 IST. A calendar that shows local time removes the arithmetic — ours shows IST by default and your own timezone when you open it.
Grid and basket EAs: the case for pausing
A grid's risk is a fast one-directional move that opens every level before any can close. NFP and FOMC are exactly that move. If the EA has a news filter, use it. If not, the manual version: fifteen minutes before the release, stop new entries (most EAs have an input for it, or switch off Algo Trading for the chart) but leave open positions and their take-profits managed. Closing an open basket into a news spike is usually worse than holding it — you realise the drawdown at the widest spread of the day.
Breakout EAs: the opposite decision
A breakout system built for volatility expansion may want the news move — but only if it was designed for it, with a stop that is respected by the broker and position sizing that assumes slippage. Test it in the Strategy Tester with real ticks across a year of NFP Fridays before deciding. A breakout EA that has never seen a release in testing has not been tested.
How a news filter works
- The EA reads an economic calendar (MT5 has one built in; some EAs fetch a feed) and blocks new entries from N minutes before to M minutes after each event above a chosen impact level.
- Typical defaults: 30 minutes before, 30 after, high-impact only, currencies USD for Gold.
- It never closes existing positions — that is a separate, riskier feature that you should switch off unless you understand it.
- Check that Allow WebRequest is enabled for the calendar URL if the EA fetches one (Tools → Options → Expert Advisors).
If the EA has no filter
- Keep the week's red-folder events in front of you (a calendar in your timezone) and pause manually: right-click chart → Expert Advisors → Remove is drastic; toggling the chart's Allow Algo Trading is enough.
- Wider basket stop-loss than you would otherwise choose, sized in the deposit (see the deposit guide).
- Avoid starting a new grid cycle within the hour before an event.
- Log what happened. After three NFPs you will know whether your EA and broker handle them or not — and that is worth more than any rule of thumb.
