News filters: which economic-calendar events actually move your pairs

Most automated systems choke during high-impact news events. It is tempting to write a blanket rule that stops trading whenever there is a red-folder event on ForexFactory, but that is a blunt instrument. You end up missing the actual trend because you were scared of a temporary spike.

Start by categorizing events by their historical volatility impact rather than the site's default rating. Central bank interest rate decisions, CPI prints, and Non-Farm Payrolls usually have legs, but many others create noise that dies within minutes. If your strategy is mean-reversion, news is your enemy. If you are trend-following, you might actually want to be in the market when the initial break happens.

To backtest this honestly, you need to track the 'slippage cost' of your news filter. Many people forget that by avoiding a bad trade, they also miss the good ones that occur immediately after the volatility settles. Run your backtest twice: once with the filter enabled and once without. If the net profit difference is negligible, the filter is just adding complexity without reducing your risk.

Here are three things to check before adding a filter:

• Does the asset class actually react to the specific news type?
• What is the average duration of the volatility spike?
• Does the strategy have a wide enough stop-loss to survive the noise without the filter?

Filtering too much is a common trap because it makes the equity curve look smoother in the past. In reality, you are just curve-fitting your way out of reality. Does your current strategy actually perform better during news, or are you just assuming it is safer to stay out?