Why a normal bad day can hit the daily loss limit

[Amanhasnoname]

Take a scalper with a real, modest edge: 40 trades a day, 55 percent winners, wins and losses both 1R. Expectancy is +0.1R per trade, so the average day is +4R. Standard deviation per trade is about 1R, which makes the daily standard deviation about 6.3R (the square root of 40, times 1).

Now set a daily loss limit of 5R, for example 1 percent risk per trade and a 5 percent cap. A day closing at -5R sits about 1.4 standard deviations below the +4R average, which is roughly a 7 to 8 percent chance. That is around one day in thirteen, so a 20-day month should contain one or two limit days with the edge fully intact. Touching the limit intraday and then recovering is more likely still.

The independence assumption is generous too. Real losses cluster by regime, so bad days can be worse than this.

The fix is not a better entry, it is geometry. Halve the risk per trade to 0.5 percent and the same 5 percent cap becomes 10R, about 2.2 standard deviations below average, roughly 1 to 2 percent of days.

What ratio of daily limit to risk per trade are you running, and how many limit days a month before it was worth questioning the edge?