PROBLEM SET 1 Exercise 1 A demand/supply model is specified by. Pt 0 1Qt 2Yt ut (demand ) Qt 0 1 Pt 1 2 ( Pt 1 Pt 2 ) vt (sup ply ) where P indicates price, Q quantity, and Y income. (1) (2) Discuss the identifiability of the structural parameters ( the ' s and the ' s ) of this model. Derive the AR equations for Pt and Qt , and show that the AR coefficients are identical. Exercise 2 We consider the following AR(2) process: Yt c 1Yt 1 2Yt 2 t t 0, 1, 2, 3,.... where t is WN (0, ) 2 (1) Give the characteristic equation for this process. (2) What is the requirements for this process to be stationary? (3) Show that the stationary conditions on the roots imply that (a) 1 2 1, (b) 2 1 1 and (c) 2 1 Exercise 3 A macroeconomic model for the log of U.S. real GNP postulates that A( L)(Yt 0 1t ) t where the shift polynomial A( L) is given by A( L) 1 1L 2 L2 An OLS regression yields: Yt 0.321 0.003t 1.335Yt 1 0.401Yt 2 ut (1) Determine the estimates of 1 , 2 , 0 , 1 (2) Does the estimates of 1 and 2 satisfy the conditions (a), (b) and (c) listed in Exercise 2? (3) Compute the roots of the corresponding characteristic equation. (4) Does the empirical regression above describe a stationary process? An alternative specification fitted to the same data yields: Yt 0.003 0.369Yt 1 vt What are the roots of the characteristic equation corresponding to this AR process? Exercise 4 Consider the difference equation Yt 1.6Yt 1 0.89Yt 2 0 t 2,3, 4,..... (1) Give the characteristic equation, and find its roots. (2) Give the expression for Yt if Y0 0.8 and Y1 0.9 (3) If t is WN (0, 2 ) and t 0, 1, 2, 3,..... , is there a stationary time series satisfying the AR process: Yt 1.6Yt 0.89Yt 2 t ? Let the time series Yt , t 2,3, 4,5,..... be defined by Yt 1.6Yt 1 0.89Yt 2 t where Y0 Y1 0 and t is WN (0, 2 ) (4) Give the covariance matrix for (Y2 , Y3 , Y4 ) Exercise 5 Let Z n be sequence of uncorrelated real-valued variables with zero means and unit variances, and define the MA-process r Yn i Z n i i 0 for constants 0 , 1 , 2 ,...., r . Show that Y is stationary and find its auto-covariance function. Suppose now that Yn is an auto-regressive stationary process in that it satisfies Yn Yn1 Z n n For some 1 and Z n is as above. Show that Y has auto-covariance function m c ( m) (1 2 ) Exercise 6 Let A and B be uncorrelated random variables, each of which has mean 0 and var iance 1. Fix a number 0, and define the process X n A cos( n) B sin( n) Show that E ( X n ) 0 for all n. Calculate the auto-covariance function for Explain if this process X n is stationary. X n .