I have been plugging away at a couple of my first notebooks & backtests this weekend. Trying to stay in the fundamental/value realm to familiarize myself with the platform to start.
The overall approach was to try to determine expensive companies and cheap companies. I decided to utilize P/E ratios to label companies by taking the overall market (Russell 2000) P/E ratio. Then we can label the companies as cheap if they are below the market average, and expensive if they are above.
The results are fairly poor, but I believe that it could be due to the structure of the strategy. By failing to control for other factors such as sectors, growth estimates, and structural differences, comparing companies to the overall market could be a bad place to start.
Any recommendations are gladly welcomed. I will be posting a backtest of the strategy as well.