Quantitative Methods in Macroeconomics

PhD Course, Fall 2002

Problem set 1 is at the end of Lecture Notes 1. A number of students have asked to hand it in later. This is ok, but I recommend that you try to hand it in on September 5.
You need to have access to a computer with Matlab, Gauss, or Octave. I use Matlab, and I recommend that you do too, if possible. The SSE has a site licence for Gauss (see IT Support pages). For information about Matlab and Gauss, see Paul Söderlind's homepage.
You need an equation solver for problem set 2. If you have the optimization toolbox for Matlab, you have fsolve. There is also a version of fsolve in Octave. If you do not have the optimization toolbox, you can e.g. use csolve.m (written by Chris Sims).
Exercise 3 due on Tuesday September 24, 10.15.
Exercise 4 (in Lecture Notes 3) due on Tuesday October 1, 10.15.
Exercise 5 due on Tuesday October 8, 10.15.
No lecture notes will be published for lectures 7-8.

Pdf files

Exercise 3
ex3.m
func_r.m
grad.m
hessian.m

Exercise 3 a (pdf)
deaton.m (Exercise 5, approach (a))

Growth model: linear-quadratic method (LN 2)
rbc_ql.m
func_r.m
grad.m
hessian.m

Growth model: discrete state-space method (LN 3)
rbc_discrete.m

Growth model: piecewise linear decision rule method
rbc_decision.m

Coconut model (LN 3)
coconut.m

Code to generate Markov chain that approximates an AR(1) process
tauchen.m