One question outstanding for me right now is how much slippage will impact my results. I don't have a sense of how big an effort it will be to solidly characterize its impact. I'm not sure how well its effects are represented in simulation and what the best practices are for incorporating it into one's testing. I presume it depends on the ticker and that people collect slippage results from prior time spans and use that as an estimate for the now.
So there's some question about whether I can stay entirely in simulation during development, but it seems possible. Like if I record slippage data from live and then pipe that into my code and tack it on top of the simulation data I'm getting to determine expected profitability.
We shall see...
Comment