You assume they are random. However, IF the trades follow something even as bad as 60% profit, 20% b/e, 20% loss, and the win moves are statistically significant, then this is not a problem. Keep in mind, many trades are actually on the extreme, and thus a scale out approach in the cases when...