Explain why adding a lagged dependent variable and lagged independent variable(s) to the model eliminates the problem of the first-order autocorrelation. Give at least two reasons why this is not necessarily a preferred solution.

5-31 : Let us consider a regression equation in a matrix form of the type:

T×1 (T×(k+1)) (k+1)×1) T×1 Y =X ???? +????

We may expect that ????t = ????????T-1+vt , where t {v } is a white noise process. Discuss the generalized least square method to estimate the model. Discuss the maximum likelihood method to estimate the model.