How do you program adaptability?
retained text
I have recently been getting into creating and testing models. The original intent was to remove myself from the picture as I believe emotion adds too much uncertainty to be successful over the long run if you are not able to curb it. Unfortunately, after months of testing I have realized that the way I actually trade is incredibly difficult to actually program. The main reason, my strategy adapts. Any algo will follow strict set rules and will not bend. But there are times where I will ignore a signal if other conditions exist, or make a move that has not officially signaled yet if I have high conviction we are going that way. For instance, I may program my rule to exit within 10% of the next pivot point to capture profit and reduce giveback. But managing the play live, I may hit that 10% barrier and take into account that the last 3 candles were all in my direction and growing in size so I may extend to over the next pivot point. This is just one example. I realize I could program this into the algo, but to be able to do this for every conceivable point of flex that my system would need would require hundreds, possibly thousands of these little rules and some would obviously contradict with each other. At this point, I have several models that I am carrying into paper trading. None of them really capture how I actually trade. But all of them attempt to capture a simplified version of one aspect of how I trade on any given day. As such, it naturally limits large winners (3-10 baggers) where I can hit them at a decent pace controlling it myself. So at this point, my idea to remove myself completely has adapted. I will now use my algos to do the majority of my trading, then set aside a set amount for manual high risk/return style plays. But my question remains. How do you guys get around these obstacles when you are making your algos? Do you end up breaking strats into chunks like I am doing? Or have you found a way to actually build in adaptability to your models?
document fetched 2026-10-29 00:47 · canonical https://reddit.com/r/algotrading/comments/1xhgll7/how_do_you_program_adaptability