The following regression was estimated to explain the inflation rate in the USA. The data set contains annual observations from 1970 to 2010.
Inft = 2500 + 50*Xt + 0.4Inflt-1
se(B) (5.0) (0.002) R2 = 0.92
where Inft is the inflation rate in a given year Xt is the growth in money supply in the same year
a. Suppose this model represents a Koyck lag structure. Find the 1st, 5th, and 10th lag coefficients on Xt.
b. If this was a forecasting model, exlpain what AR(3) and MA(4) models would look like (separate not ARMA)
c. Now assume you have the same time series data for 4 countries (USA, Canada, England, Japan). Explain what would a fixed effects model look like, and how would it improve on the basic model?.