Identification Analysis of DSGE model parameters with Dynare скачать в хорошем качестве

Identification Analysis of DSGE model parameters with Dynare

This video covers the Identification Toolbox of Dynare We'll go through some theoretical concepts and have a look at some examples that showcase both the identification problem in DSGE models but also how one can use Dynare's identification toolbox to uncover and asses these. Time-stamps 0:00 Motivation: Parameter identification (and not shock identification) 3:55 Overview features of Dynare Identification Toolbox 5:24 Example 1: Shapes of likelihood 6:02 Example 2: ARMA(1,1) 11:09 Example 3: Simple forward-looking DSGE model 17:17 Which observables? 18:55 Example 4: RBC model with two kinds of investment adjustment costs (Kim, 2003) 27:23 Identification Problem in Theory 30:42 "Unidentifiability causes no real difficulties in the Bayesian approach" Is there a systematic way to detect identification issues? 32:08 Theoretical lack of identification 33:01 Definitions 34:51 Strength of Identification 35:44 Literature Overview Identification in linearized Gaussian DSGE models 40:07 Linear Gaussian state-space framework 43:58 Diagnostics based on moments 47:32 Diagnostics based on spectrum 50:14 Diagnostics based on control theory for minimal systems Dynare Implementation 55:07 identification command 56:50 warnings 58:28 Tracking singularities 59:27 Example: Point vs Monte Carlo mode 1:03:52 Computational remarks 1:05:18 Weak identification diagnostics Identification strength: Information matrix approach 1:06:16 Idea 1:08:02 Formally 1:09:02 Implementation in Dynare: Strength and Sensitivity 1:10:24 Identification Strength Plots 1:11:25 Numerical Remarks 1:12:24 Example: Investment Adjustment Costs Identification strength: Bayesian Learning Rate Indicator 1:15:06 Idea 1:16:42 Implementation 1:18:36 Example: Investment Adjustment Costs Advanced Option of Identification Toolbox 1:20:38 Point Mode 1:21:19 A Different Sensitivity Measure 1:22:09 Analyzing Identification Patterns 1:23:46 Example: Investment Adjustment Costs identification(advanced) 1:27:46 Monte Carlo Mode 1:28:22 Example: Investment Adjustment Costs identification(advanced,prior_mc=100) Identification of nonlinear and non-Gaussian DSGE models 1:32:16 Idea 1:34:07 Dynare's General Model Framework 1:35:34 Pruning 1:36:36 Univariate example 1:39:38 Pruned State Space System 1:41:20 Identification Diagnostics 1:42:30 Example: Investment Adjustment Costs identification(order=2) 1:43:53 Concluding Remarks