Hi, I'd like to create an algorithm such that:
Every month do:
Filter universe to 1B market cap or larger.
Find normalized_diluted_eps averaged over 3 years for each stock.
Filter out any stocks for which that 3 year avg eps is negative.
Find price/(3 year avg of normalized_diluted_eps ) for each stock. Lets call that earnings_to_price.
Find the sum of all earnings_to_price. Call that eps_total.
For each stock find earnings_to_price/eps_total. Call that weight.
Then for each stock buy or sell until the weight percentage of our money is in that stock.
Questions:
Is it possible to find normalized_diluted_eps avgeraged over 3 years with the pipeline? If so what does window_length need to be? Is window_length always measured in days?
Using the pipeline I know that it's set up at the beginning of every day right? Is there a way to set it up at start of each month instead?
Is there anything about this algorithm that seems hard to do? Something else I'm missing?
I've attached the algorithm so far.