The real cost of a flat-before-news rule

[Amanhasnoname]

Flattening five minutes before every red-folder event looks like free insurance. It is not free, and most people never add up the bill.

Start with frequency. Trade several currencies and releases touching them can show up 15 to 25 times a month. Each time a rule forces a close, any open winner gets cut early and any open trade that was about to work gets killed. If the average forced close gives up even 2 pips of expected gain, 20 events a month is 40 pips of lost expectancy, which for a scalper can be a meaningful slice of the month.

Then there is the other side. A tail event on a held position can cost 15 pips or more instead of the planned 5, and it only has to happen a few times a year for the rule to pay for itself. So the rule is not wrong, it is just unpriced.

Two alternatives are worth testing against it: cutting size by half and moving the stop to breakeven, or trailing tight. Neither removes the tail but both soften it while keeping some of the open profit.

The test is simple. Over a few months, record what each forced close would have made if left alone, then compare against the realised cost of the surprises that were avoided.

Has anyone compared a flat rule against halving size over a real sample of releases?