I think that was one of those well-reasoned arguments you spoke of elsewhere. Flame alert!
I can't disagree with your personal view of testing. I will simply observe that a robust system gives you the flexibility to select a comfortable stop at the expense of a lowered expectation relative to a looser stop. Guilty as charged. Most of the systems I have developed have optimal regions in both stop loss and profit stop so that this is relatively painless. However, I recently developed a system whose expectation dramatically increases with much wider stops, to the point that it eliminates all losers. If someone else wrote that sentence I would opine "Then you don't have the right entry conditions!" But there it is. Since all of my good systems have two to three losers for every winner, I am motivated to think carefully about wider stops. My systems also occasionally exhibit eith or nine losses in a row, so I am motivated just for sanity's sake to figure out how to avoid the trap of being drawndown when a system fails.
I can't disagree with your personal view of testing. I will simply observe that a robust system gives you the flexibility to select a comfortable stop at the expense of a lowered expectation relative to a looser stop. Guilty as charged. Most of the systems I have developed have optimal regions in both stop loss and profit stop so that this is relatively painless. However, I recently developed a system whose expectation dramatically increases with much wider stops, to the point that it eliminates all losers. If someone else wrote that sentence I would opine "Then you don't have the right entry conditions!" But there it is. Since all of my good systems have two to three losers for every winner, I am motivated to think carefully about wider stops. My systems also occasionally exhibit eith or nine losses in a row, so I am motivated just for sanity's sake to figure out how to avoid the trap of being drawndown when a system fails.
