I've taken three months of price return data for two instruments and calculated a β\beta between the two using the formula β=Cov(x,yVar(y)\beta = \frac{Cov(x,y}{Var(y)} with the goal of estimating what the percentage change in instrument yy should be based on what the percentage change in instrument xx is. I have been applying this by multiplying β\beta by the percentage change of xx to determine a beta-adjusted percentage change for yy , but I am wondering if I should actually multiply the derived β\beta by the current percentage change of yy instead. Could anyone shed some light on whether or not I am properly applying this β\beta to arrive at an expected percentage change for yy ?