A strategy can be profitable for months and then give back a large part of it in two minutes after a US jobs report. The cause is rarely the direction of the move — it is spread, slippage and gaps that the EA was never designed or tested for.
- High-impact releases such as NFP, CPI and FOMC widen spreads and cause slippage and whipsaws.
- Backtests rarely reproduce news conditions, so EAs often lose more live than in testing.
- A news filter pauses trading before and after selected events and can close positions.
- Filter by impact and currency so unrelated releases do not stop your EA.
- Gold reacts strongly to US data, making a filter especially valuable for XAUUSD EAs.
- Test results with and without news periods to see what the filter changes.
The events that move the market
| Event | Typical time (ET) | Most affected |
|---|---|---|
| Non-Farm Payrolls (NFP) | 1st Friday, 08:30 | USD pairs, gold, indices |
| US CPI inflation | Monthly, 08:30 | USD, gold |
| FOMC rate decision | 8× per year, 14:00 | Everything USD |
| ECB / BoE decisions | Thursdays, varies | EUR, GBP |
| GDP, retail sales, PMI | Varies | Local currency |
Check a live economic calendar each week: dates move, and US daylight-saving changes shift the times relative to your terminal clock.
What happens in the seconds around a release
- Spreads widen. Liquidity providers pull quotes; a 2-point EURUSD spread can become 20+, and gold spreads can expand several times.
- Slippage. Stop orders fill at the next available price, which can be far beyond your stop-loss.
- Whipsaw. The first move often reverses within a minute, hitting stops on both sides.
- Requotes and rejections on some brokers, leaving an EA with unmanaged exposure.
An EA risks 1% with a 15-pip stop. During NFP the stop fills 25 pips worse — a 40-pip loss. The “1% risk” trade just cost 2.7%.
Why most EAs struggle on news
Backtests usually assume a fixed or historical average spread and perfect fills. They cannot reproduce the order book of a news minute. So the backtest says "profitable", but the live account takes losses the strategy never saw. Scalpers, grid systems and EAs with tight stops are hit hardest.
What a news filter does
A news filter reads the economic calendar and, around selected events, stops the EA from opening new trades. Better filters can also:
- Disable Algo Trading for the whole terminal X minutes before the event and re-enable it Y minutes after.
- Filter by impact (high / medium) and by currency, so a JPY release does not stop your EURUSD robot.
- Optionally close open positions and delete pending orders before the release.
- Show the next event and a countdown right on the chart.
Sensible filter settings
| Strategy type | Pause before | Resume after | Close trades? |
|---|---|---|---|
| Scalper / M1–M5 | 30 min | 30–60 min | Yes |
| Intraday trend | 15 min | 30 min | Optional |
| Swing / H4+ | 5 min | 15 min | Usually no |
| Prop firm account | Per firm rules | Per firm rules | If required |
These are starting points, not rules. Test them on your own history: some strategies lose most on news, others make most of their profit on the volatility that follows.
How to read the economic calendar
Every release shows three numbers: previous, forecast (consensus) and actual. The market reacts to the surprise — the difference between actual and forecast — not to whether the number is "good". A strong jobs report that was even more strongly expected can push the dollar down.
| Column | Meaning | Why it matters |
|---|---|---|
| Impact | Low / medium / high | Filter only what moves your symbols |
| Previous | Last release (may be revised) | Revisions can surprise too |
| Forecast | Analyst consensus | Already priced in |
| Actual | Published number | Surprise = Actual − Forecast |
Why gold reacts so strongly to US data
XAUUSD is priced in dollars and is sensitive to US interest-rate expectations. Inflation, jobs and Fed decisions change those expectations instantly, so gold often moves more in dollar terms than major currency pairs on the same release. For gold EAs a news filter is usually the single most valuable protection.
Test the filter on your own strategy
- Run the backtest normally and save the report.
- Run it again excluding the minutes around high-impact news (or mark news trades in the report).
- Compare profit factor, maximum drawdown and the largest single losses.
- If the filtered version has a similar profit with a smaller drawdown, the filter is protecting you. If the profit disappears, the strategy depends on news volatility — trade it only with realistic spread and slippage assumptions.
Weekend gaps and holiday liquidity
News is not only scheduled releases. Political events over a weekend can open the market far from Friday's close, skipping stop-losses entirely. Around Christmas, New Year and US holidays liquidity drops and spreads widen even without data. Many traders treat these periods like a high-impact event: no new trades, reduced size, or the EA switched off.
Which impact levels to filter
Most calendars rate events as low, medium or high impact. For automated trading, filtering only high-impact events for the currencies you trade is a sensible default. Add medium-impact events for scalpers and gold strategies, which react to more releases. Filtering everything makes the EA idle most of the day and can remove profitable trades.
| Your symbols | Filter currencies | Impact |
|---|---|---|
| EURUSD | EUR, USD | High (+ medium for scalpers) |
| XAUUSD | USD (plus major global events) | High + medium |
| GBPJPY | GBP, JPY | High |
| Multi-symbol EA | All traded currencies | High |
Trading news manually
If you trade releases on purpose, use pending orders with realistic slippage expectations, risk half your normal size, and never hold a position whose stop is inside the typical spread spike. For most traders the edge is to wait for the first 5–15 minutes to settle and trade the structure that forms after.
Weekly news routine for EA traders
Every Sunday or Monday, open the calendar for the coming week and note the high-impact events for your symbols, converted to your broker's server time. Check that your filter covers them, decide whether any positions should be closed before the biggest releases, and plan reduced activity around central bank days. Ten minutes of preparation prevents most news surprises.
Frequently asked questions
Educational content only, not financial advice. Leveraged trading carries a high risk of loss. Past performance does not guarantee future results. Broker and prop-firm rules change — always check the current conditions of your provider.
