I would like to merge data in the following way in Pipeline:
Say today is 2019-03-01, I would like to construct a value factor which is BookValue/MarketValue. My book value has as_of_date 2018-12-31 which is the last quarter end. When I calculate the value factor, I also want to use MarketValue that as_of 2018-12-31 to avoid look ahead bias.
How could I do this in the Pipeline? It is not obvious to me how to merge things for the custom factor.