In theory, yes some of the coefficients in the "master regression" equation should change between bull and bear periods, in practice, I have not done that yet and the paucity of data from the 1966/82 period makes that a little difficult to back test. Nothing is "perfect".
Market Data 
Markets 
AI
