A news EA is an expert advisor built to trade the seconds around a scheduled economic release — non-farm payrolls, a central bank rate decision, CPI. The premise is simple: big numbers move price fast, and code reacts quicker than a person. The catch is that thousands of other systems fire on the same tick, and the moment you most want a clean fill is the moment your broker is least able to give you one.
What it is
Most news EAs use one of two approaches:
- Straddle — it places a buy stop above the market and a sell stop below, a few seconds before the release, hoping the spike triggers one side and runs.
- Reaction — it reads the number, or the first burst of movement, then fires a market order in the direction of the surprise.
Both depend on one thing: getting filled near the price you asked for. On a quiet afternoon that is easy. In the half-second after a payrolls print it is not.
Where execution falls apart
Four things go wrong at once when the number lands:
- Spreads gap wide. A pair that sits at 0.6 pips can jump to 8 or 15 pips for a few seconds. Your straddle's stops sit inside that gap and trigger at the worst moment.
- Slippage. A market order asks for the current price and gets whatever is left. Filling 20 pips from your intended entry is normal here, not exceptional.
- Requotes and rejects. Some brokers stop offering firm prices and hand the order back, so the EA either misses the move or chases it.
- Latency. The price your EA acted on is already history by the time the order reaches the server 40–100 ms later.
None of these are bugs in the EA. They are the market it is trading into.
The backtest almost never shows this. Historical tick data rarely captures the spread blow-out and slippage of a live release, so a news EA's tester results are close to fiction. Judge it only on live fills.
A worked example
Your EA straddles GBP/USD before a rate decision with a 15-pip target and a 10-pip stop, at 1 lot (£10 per pip). The backtest showed the buy stop filling exactly at the level.
Live, the spread gaps to 12 pips on the print. The buy stop triggers, but slippage fills you 9 pips above the level. Price runs the 15 pips and the EA closes — except your real entry was 9 pips worse, so you bank 6 pips: £60, not the £150 the tester promised. On the next release the move reverses through the gap and your 10-pip stop fills 14 pips away: −£140. Two trades the backtest scored as roughly break-even have cost you £80. Repeat that across a month and the edge is gone.
What to check before running one
- Does it use a news filter to stand aside when the spread is abnormally wide?
- What maximum slippage does it allow, and does it cancel rather than chase a runaway price?
- Is your broker's execution fast enough, and does it widen spreads punitively around news?
The Karnek note
Karnek watches your live fills, so you can see the real slippage and spread on each news trade instead of the backtested version. It reads straight from your terminal, read-only, and can never place, close or touch a trade.
Written and reviewed by the Karnek Research team. Last updated August 2026.
Educational content only - not financial advice. Past performance does not predict future results. Trading carries significant risk.