2. Schedule and Readings
Monday 25/02/08: Dynamic Programming: An Introduction. Read Lecture Note 1 and Ljungqvist and Sargent on value function iteration, pp. 85-97.
Friday 29/02/08: Value function Iterations: The Neoclassical Growth Model.
Monday 03/03/08: Incomplete Markets and Firm Dynamics. Read Ljungqvist and Sargent chapters 16 and 17 and Hopenhayn "Entry, Exit and Firm Dynamics in the Long run", Econometrica, 60(5), 1992, pp.1127-1150.
Friday 07/03/08: Calibration and the Optimal Linear Regulator. Read Cooley and Prescott, Chapters 1 and 2, Ljungqvist and Sargent Cahpter 5, King, Plosser and Rebelo, JME 1988, Kydland and Prescott, Econometrica 1982, Hansen, JME 1985.
Monday 10/03/08: Undetermined coefficients. Read Harald Uhlig's notes for his toolkit. Lecture notes 2 and 3 are also posted now.
Friday 14/03/08: 2nd order approximations and Parametrized Expecations. Read Schmitt-Grohe and Uribe, "solving Dynamic General Equilirium ... " which available on Stephanie's webpage and den Haan and Marcet (1990).
Monday 17/03/08: Filtering and Stylized Facts. Read Kydland and Prescott (1991) (Real Facts and a Monetary Myth).
Wednesday 19/03/08: Indirect Inference and Structural VARs. Read Gali (AER, 1999), Christiano, Eichenbaum and Evans (2004).
3. Handouts and Lecture Notes
(a) Slides for Lecture 1 and Lecture Note 1
(b) Slides for Lecture 2
(c) Slides for Lecture 3
(d) Slides for Lecture 4
(e) Slides for Lecture 5 , and Lecture Note 2 and Lecture Note 3
(f) Slides for Lecture 6
(g) Slides for Lecture 7
(h) Slides for Lecture 8
(i) Slides for Lecture 9
(j) Slides for Lecture 10
4. Exercises
(a) Week 1: Introduction to Matlab and GAUSS
(b) Week 2: Assignment 1
(c) Week 3: Assignment 2
(d) Week 3: Assignment 3
(d) Week 4: Assignment 4. You also need to download this file: impulse.out
5. EXAM
Dowload this file for question 18: impulse.mat
6. Software (GAUSS Software by Morten O. Ravn, Matlab Software by Pontus Rendahl)
A. Codes for Value Function Iterations
Value_det.prg (GAUSS code) Value_det.m (Matlab code): Value function iterations using discrete state space for the deterministic growth model
Value_detcheb.prg (GAUSS code) Value_detcheb.m (Matlab code): Value function iterations using continous state space with Chebyshev interpolation for the deterministic growth model
chebsample.m (Matlab code): This code shows how one can approximate a function using a Chebyshev polynomial and how one can compute the derivatives
B. Codes for Heterogeneous Agents Incomplete Markets Economies
aiyagari1.prg (GAUSS code). This solves the Aiyagari (1994) economy the "hard way" with iterations on the capital stock. It's slow but foolproof.
aiyagari.m (Matlab code). This solves the Aiyagari (1994) economy with bisection. It's a bit faster than the "hard way"
aiyagarismooth.m (Matlab code). This solves the Aiyagari (1994) economy with bisection, endogenous grid points and policy function iterations. This is superfast.
C. Codes for Optimal Linear Regulator
Prog1.e (GAUSS code) prog1.m (Matlab code): Computes decision rules and impulse responses for simplest RBC model using optimal linear regulator
D. Harald Uhlig's Toolkit (GAUSS software can be requested from me)
E. Stephanie Schmitt-Grohe and Martin Uribe's Matlab software for 2'nd order approximation
F. Codes for Parametrized Expectations
An elegant Matlab code is available on Lilia and Serguei Maliar's homepages. Notice that the algorithm includes "moving bounds" to rule out explosive paths.
G. Codes for Estimation by Matching Impulse Response functions
impgen and estimation1 (GAUSS codes). First run impgen to generate some artificial impulse responses that you wish to match. Then run estimation1 to see if the estimator gets it right. This software uses the Blanchard-Kahn algorithm to solve for the decision rules.
impulse_matching.zip The Matlab version (save the file and then unzip into your Matlab directory). This software uses the Uhlig software to solve for the decision rules.You generate the impulse responses that you wish to reproduice with impgen. The estimator is in estimation.m which uses qproc.m to define the quadratic form. Other files are from Uhlig's toolbox.
No. of visitors since January 29, 2004: