Let's say we have a time series for an illiquid future and we would like to replicate this time series using two time series for liquid futures using daily rebalancing. What would be a good approach to do that?
My idea was to use linear regression to find the proportions of each liquid future. Then I would maintain these proportions throughout by daily rebalancing one future and then maybe do a weekly rebalancing on the second one. Does this make sense?