Fading the news spike: is it a fair fight?

[Shadow Trader]

A big number hits and price rockets 25 pips in seconds. The temptation to fade the move is strong because it so often retraces.

The reasoning is sound in places. The first move is frequently an algorithmic reaction to the headline, and humans correct it a few minutes later. The problem is execution. Spreads are huge, fills are poor, and the retracement may start before an order even gets filled.

Then there's the opposite scenario: the spike is the start of a sustained move, and the fade is a lesson in how fast losses pile up.

Those who trade it successfully usually wait, not for the spike but for the stabilization, then trade the second move with defined risk.

Things that tend to separate a decent fade attempt from a disaster:

- Waiting for spread to return close to normal
- Using limit orders, not market orders
- Sizing small enough that a miss doesn't hurt
- Having a time stop in case the move doesn't retrace
- Skipping events where the surprise was enormous

A big surprise tends to produce a sustained move, not a spike and fade. The size of the surprise relative to expectations often says a lot about which one is coming.

Has fading news spikes ever worked consistently in a real account, or only in backtests with perfect fills?

[FXS HongKong Scalper]

The execution point is the key. In our tests, fading the first spike after US data with market orders lost money mainly because of fills, not direction. The idea was right more than half the time, but entering with a 3 to 5 pip spread and slippage ate it.

What worked better was waiting for the first M1 candle to close and the spread to normalise, then using a limit order at a level inside the spike. Fewer trades, but entries were at a fair price. Even then, we skip it on NFP and CPI, where the first move continues often enough to hurt.