how to balance the equation???urgent help

[callen]

i m doing the following regression for prediction of returns:
dependend variable:returns
explanatory variables:lagged returns(12 lags), and other lagged financial variables(12 lags)
the problem is that my dependent variable(returns) is statioanry but some of my explanatory variables are not, they are inegrated I(1).
so i have an unbalanced equation since LHS is I(O) and
RHS has I(O)+I(1)..+I(1) for example.

so what should i do?
should i take the first differernce of all the variables that are I(1)?and since I have 12 lags of each variable, i should take the first difference of each lag of each variable?

or should i find a co-integration relation between the I(1) variables?if yes how do i use that in my model?

i would appreciate any suggestions


callen