A daily loss limit of two losses instead of a percentage

[FXS London Scalper]

A daily loss limit of 2% of the account sounds reasonable. In practice it often doesn't stop you when it should.

The problem: with 0.5% risk per trade, 2% means four full losses. For many traders the third loss is already trading badly, and the fourth trade is usually the worst of the day.

An alternative is a count: two full losses and you stop. Small losses from time stops or scratches don't count, only trades that hit the full stop.

How to check what it would have done, from your journal:

• Find every day with a third or fourth full loss.
• Remove all trades after the second full loss, including any winners.
• Compare the average losing day and the total, before and after.

What you'd expect: days with a third and fourth loss disappear, and the average losing day shrinks. A cost appears too: some recovery days, where the third trade would have won. The journal tells you whether those recoveries were worth the bad days.

The count often works better because it tracks your state of mind, not a number on a screen. Two full stops in a row is a clear signal; 1.3% down isn't.

Does anyone use a count of losses rather than a money limit? Do you count partial losses?

[Amanhasnoname]

I switched to a count-based limit about six months ago and it's been one of the better changes I've made. My rule is two full stop-outs and I'm done for the day, and scratches or small time-stop exits don't count. Before that I used a percentage limit, and I almost never reached it, because I'd start trading badly well before the number was hit.

When I ran your test on my old journal, the result was clear. Removing everything after the second full loss cut my worst days noticeably, and the few recovery days I lost didn't come close to making up for it. The third trade of a bad day was usually forced, and the fourth was often revenge.

The thing I like most is that it's easy to follow in the moment. Working out percentages while annoyed is hard, but counting to two isn't. I still keep a money cap as a backstop, though I can't remember the last time I needed it.