Dynare Toolbox 4 Matlab Introduction - PowerPoint PPT Presentation

1 / 17
About This Presentation
Title:

Dynare Toolbox 4 Matlab Introduction

Description:

Questions regarding welfare evaluation & risk premium in stochastic environment. ... Stochastic simulations: 1st and 2nd approximation ... – PowerPoint PPT presentation

Number of Views:677
Avg rating:3.0/5.0
Slides: 18
Provided by: aya6
Category:

less

Transcript and Presenter's Notes

Title: Dynare Toolbox 4 Matlab Introduction


1
Dynare Toolbox 4 MatlabIntroduction
  • Alon Binyamini
  • 26 MAY, 2008

2
Roadmap
  • 1st / 2nd order Taylor approximation
  • Technical intuition
  • Economic intuition
  • Dynare
  • Background.
  • Hands on.
  • Procedural issues.

3
Taylor Approximation
  • Intuition
  • 1st order linearization.
  • 2nd order
  • In 1st order we employ the derivative at a.
  • Better to employ f a0.5(x-a).
  • So, lets linearize the function f .

4
2nd order technical intuition
Begin with 1st order
But, instead of f (a), take
Substituting instead of f (a), we get
5
2nd order economic intuition
  • Non-linear utility
  • Diminishing marginal-utility
  • gtgt The direction also matters (not only the size)
  • gtgt Risk aversion
  • gtgt Second moments of shocks also matter
  • The more we can deviate, the lower is
    E(u).
  • Non-linear technology
  • Diminishing marginal-productivity
  • gtgt Likewise

6
1st or 2nd order approximation?
  • 1st order
  • Only first moment matters (0)
  • Certainty equivalence
  • 2nd order
  • With risk aversion or concave technology
  • Second moment also matters
  • No certainty equivalence
  • Questions regarding welfare evaluation risk
    premium in stochastic environment.

7
Dynare background
  • Collection of Matlab files.
  • Solves and estimates nonlinear-dynamic models
    with RE.
  • By several people from CEPREMAP

8
Dynare background 2
  • Steady state
  • Alternative optimization algorithm (not only
    fsolve)
  • Stability analysis (Blanchard Khan, 1980)
  • Deterministic simulation
  • Earlier version Fair and Taylor (1983).
  • Now Newton-type algorithm (Collard and Juillard
    2001) all equations simultaneously.

9
Dynare background 3
  • Stochastic simulations
  • 1st and 2nd approximation
  • An eigenvalue method of undetermined
    coefficients.
  • State space representation (policy)
  • Impulse responses
  • Moments
  • Estimation
  • Maximum likelihood
  • Bayesian estimation
  • Kalman filter

10
Hands on RBC problem
  • Simple RBC model (two shocks)

11
Hands on RBC FOCs
  • The first is a standard Euler-condition (FW
    looking)
  • Note on the time-impatient shock.
  • The second is the standard labor-supply
  • Note on income and substitution effects

12
Hands onwhat Dynare needs 2 know
  • The three constraints.
  • The two FOCs.
  • All together
  • 5 equations (fw looking).
  • 5 variables.
  • Note on the capital time-index.

13
Hands onthe model in Dynare syntax
  • c(-sigmaC) zb(1)/zb beta c(1)(-sigmaC)
    (alpha y(1)/k 1 - delta) // Euler
  • psi lsigmaL csigmaC (1-alpha)y/l //
    Labor supply
  • y zy k(-1)alpha l(1-alpha)
    // Technology
  • k i (1 - delta) k(-1)
    // Investment
  • y c i
    // Sources
  • zb zb(-1)rhoB exp(zbe)
    // Impatient-shock process
  • zy zy(-1)rhoY exp(zye)
    // Tech-shock process

14
Hands onJust do it!
  • Dynare
  • Download the Dynare toolbox (version 3.065)
  • Save it in the Matlab-work directory
  • Set path to dynare_v3/matlab
  • Our model
  • Download RBC.mod from the course website.
  • Save it in a new directory
  • In the command window dynare RBC

15
Hands on
  • Example

16
Procedural issues
  • Set path
  • Web site
  • Courses and Conferences

17
The End!
  • 10 Q
Write a Comment
User Comments (0)
About PowerShow.com