I would like to hear other's opinions about overfitting, because I want to avoid being guilty of it, but at the same time I want my model to be as good as it possibly can be. So here goes... In my mind, choosing stocks like Apple, Netflix, Google, and Amazon would be overfitting because past performance of those stocks isn't a guarantee of future results.
HOWEVER, in the algorithm I am working on now, I use a 5% trailing stop loss. I tried backtests at 4%, 5%, 6% and 7%, but settled on 5% because it had the best backtest result. In your opinion, is this overfitting? The idea of a stop-loss is completely reasonable, and I used the tools available to me to figure out where that number should be.
So am I overfitting, or being reasonable?