Econometrics II, Spring 2010
Final Exam
Please answer all four questions.
1. (20 points) What does it mean for a time series to be covariance stationary?
a. A random walk with drift is said to be nonstationary. Explain why in light of
your denition.
b. Su
Econometrics II, Spring 2011
Final Exam
Please answer all four questions.
1. (15 points) Suppose that has a vector ARMA representation,
() = + ()
where is an iid normal random vector with mean 0 and variance For a particular calibration of the models par
Econometrics II, Spring 2009
Final Exam
Please answer all three questions. They have equal weight.
1. Consider a loglinearized present value model,
dt pt = 1 Et
P
j =1
j (rt+j dt+j ),
(1)
where dt+j is dividend growth, dt pt is the log of the dividend-pri
Problem Set 1 Solutions
by Taisuke Nakata
Q1
Solution can be found in Appendix 3.A of Hamiltons textbook. My solution modies his
solution to account for two-sidedness of j and also elaborates his presentation slightly in case
someone needs to see more det
Econometrics II, Spring 2012
Problem set 1
Due in recitation section on Thursday April 5.
1. Consider a covariance stationary process with autocovariance sequence
cfw_ Suppose that the autocovariances are absolutely summable. Show that
=
absolute summabi
Econometrics II, Spring 2012
Problem set 2
Due in recitation on Thursday April 19.
1. This question is inspired by Cochrane ( 1994). On my webpage is Matlab
data le, ps2.mat, that contains data on log real GDP and log real consumption for
the period 1947.
Problem Set 3 Solutions
by Taisuke Nakata
There are two commonly used Matlab codes for solving the rst-order expectational dierence
equations. One is gensys.m by Chris Sims and the other is solab.m by Paul Klein. Both codes
are vailable at the homepages o
Econometrics II, Spring 2012
Problem set 3
Due in discussion section on Thursday May 3.
The following is a simple new-Keynesian model. The variable is the output
gap, is ination, is nominal interest, is a demand shock, is a cost-push
shock, and is a monet