I have long believed that a long strategy, and short strategy should have a very similar setup. Nearly a mirror image. I figured if I could find a setup that worked for both by just reversing the setup, that it would be more robust and exploit some underlining "rule" of the market. I have since come to believe this is a fallacy. Recently my results have shown that long and short setups need to behave very differently. For instance in a trend following strategy, the short side shouldn't look for as much of a pullback before entering.. whereas a long strategy should... Has anyone else come to this conclusion that the 2 should be vastly different setups?
The problem with this reasoning is that markets are asymmetrical due to positive drift. An example might be to have a long entry signal based on some type of breakout; if you mirrored the strategy on the short side, over time the long signal entry would likely have a better expectation of being successful merely due to market bias. It might make more sense then to think about maybe adding some type of condition that makes one or the other signal more reliable based upon that condition. A regime might be an example of the condition.