When estimating a two-country DSGE model using the following command, the Dynare output shows excessively large values in some indicators. The command parameters are as follows:
@#if NONLINEAR_KALMAN_BM
use_univariate_filters_if_singularity_is_detected = 0;
estimation(
datafile=Estimation_Indicator,xls_sheet=Estimation_Indicator_Kor_lndiff,
mh_replic=2000,
lik_init=2, kalman_algo=2,
mode_compute=6,
mh_nblocks = 4,
mode_check,
plot_priors =0
) PY PC;
@#endif
Abnormally Large Indicators Include:
1、EIGENVALUES:
Modulus Real Imaginary
5.18e+16 5.18e+16 0
9.943e+16 9.943e+16 0
2.213e+17 -2.213e+17 0
3.566e+17 -3.566e+17 0
3.754e+17 -3.754e+17 0
3.997e+17 3.997e+17 0
1.226e+18 1.226e+18 0
1.818e+18 -1.818e+18 0
1.909e+18 1.909e+18 0
2.003e+18 2.003e+18 0
2.22e+18 2.22e+18 0
3.758e+18 3.758e+18 0
4.485e+18 -4.485e+18 0
6.584e+18 6.584e+18 0
7.931e+18 7.931e+18 0
8.933e+18 8.933e+18 0
1.318e+19 1.318e+19 0
7.941e+19 -7.941e+19 0
2.087e+20 2.087e+20 0
3.929e+20 3.929e+20 0
6.688e+28 -6.688e+28 0
2、Initial value of the log posterior (or likelihood): -3.295366937638081e+17
3、Final value of minus the log posterior (or likelihood):329536693763808130.000000
4、Fval obtained by the minimization routine (minus the posterior/likelihood)): 329536693763808130.000000
5、Log data density [Laplace approximation] is -329536693763808130.000000.
6、Log data density is -329536693763808190.000000.
My Questions are :?
一、What should be the normal range for these indicators? Where can I find relevant reference materials?
二、Is it possible to use debugging methods to identify which equations or variables may be causing this? What are the key functions or techniques for debugging such issues?
三、What is the framework for evaluating the robustness of a DSGE model?
I would greatly appreciate any guidance from experts and professors.
Key Dynare Output Results:
Starting Dynare (version 5.4).
Calling Dynare with arguments: none
Starting preprocessing of the model file …
Found 121 equation(s).
Evaluating expressions…done
Computing static model derivatives (order 1).
Computing dynamic model derivatives (order 2).
Processing outputs …
done
Preprocessing completed.
EIGENVALUES:
Modulus Real Imaginary
4.817e-18 -4.817e-18 0
1.982e-17 -1.982e-17 0
2.383e-17 -2.383e-17 0
1.261e-16 1.261e-16 0
1.465e-16 -1.465e-16 0
1.563e-16 -1.563e-16 0
3.632e-16 3.632e-16 0
4.43e-16 -4.43e-16 0
4.541e-16 -4.541e-16 0
9.569e-16 -9.569e-16 0
1.755e-15 -1.755e-15 0
5.937e-10 -5.937e-10 0
6.08e-10 6.08e-10 0
6.628e-10 6.628e-10 0
6.772e-10 -6.772e-10 0
5.514e-09 -5.514e-09 0
5.514e-09 5.514e-09 0
0.5092 0.5062 0.05451
0.5092 0.5062 -0.05451
0.706 0.706 0
0.7251 0.7157 0.1162
0.7251 0.7157 -0.1162
0.7462 0.7462 0
0.7498 0.7498 0
0.7635 0.736 0.203
0.7635 0.736 -0.203
0.7697 0.7524 0.1622
0.7697 0.7524 -0.1622
0.7699 0.7699 0
0.7746 0.7519 0.1863
0.7746 0.7519 -0.1863
0.7915 0.7915 0
0.8 0.8 0
0.8 0.8 0
0.8 0.8 0
0.8 0.8 0
0.8 0.8 0
0.8 0.8 0
0.8024 0.8023 0.01574
0.8024 0.8023 -0.01574
0.8077 0.8077 0
0.8099 0.8099 0
0.8167 0.8167 0
0.8913 0.8738 0.1762
0.8913 0.8738 -0.1762
0.9184 0.9184 0
0.9317 0.9283 0.07898
0.9317 0.9283 -0.07898
0.95 0.95 0
0.9606 0.9606 0
0.9685 0.9685 0
0.9807 0.9807 0.00808
0.9807 0.9807 -0.00808
0.9828 0.9828 0
0.9828 0.9828 0
0.9915 0.9915 0
1.005 1.005 0
1.008 1.008 0
1.019 1.019 0
1.031 1.031 0
1.058 1.058 0.0004901
1.058 1.058 -0.0004901
1.074 1.072 0.06504
1.074 1.072 -0.06504
1.122 1.115 0.1293
1.122 1.115 -0.1293
1.14 1.14 0
1.245 1.22 0.2525
1.245 1.22 -0.2525
1.32 1.271 0.3557
1.32 1.271 -0.3557
1.334 1.334 0.01376
1.334 1.334 -0.01376
1.34 1.34 0
1.439 1.439 0
1.953 1.953 0
5.18e+16 5.18e+16 0
9.943e+16 9.943e+16 0
2.213e+17 -2.213e+17 0
3.566e+17 -3.566e+17 0
3.754e+17 -3.754e+17 0
3.997e+17 3.997e+17 0
1.226e+18 1.226e+18 0
1.818e+18 -1.818e+18 0
1.909e+18 1.909e+18 0
2.003e+18 2.003e+18 0
2.22e+18 2.22e+18 0
3.758e+18 3.758e+18 0
4.485e+18 -4.485e+18 0
6.584e+18 6.584e+18 0
7.931e+18 7.931e+18 0
8.933e+18 8.933e+18 0
1.318e+19 1.318e+19 0
7.941e+19 -7.941e+19 0
2.087e+20 2.087e+20 0
3.929e+20 3.929e+20 0
6.688e+28 -6.688e+28 0
There are 41 eigenvalue(s) larger than 1 in modulus
for 41 forward-looking variable(s)
The rank condition is verified.
worning: Your prior allows for negative standard deviations for structural shocks. Due to working with variances, Dynare will be able to continue, but it is recommended to change
your prior.
In initial_estimation_checks (line 190)
In dynare_estimation_1 (line 159)
In dynare_estimation (line 118)
In msDSGEPF.driver (line 1450)
In dynare (line 278)
Initial value of the log posterior (or likelihood): -3.295366937638081e+17
==========================================================
Change in the posterior covariance matrix = 0.031489.
Change in the posterior mean = 0.0040347.
Current mode = 3.295366937638081e+17
Mode improvement = 0
New value of jscale = 7.1913
==========================================================
Change in the posterior covariance matrix = 0.00058172.
Change in the posterior mean = 0.0016971.
Current mode = 3.295366937638081e+17
Mode improvement = 0
New value of jscale = 1.9598
==========================================================
Change in the posterior covariance matrix = 0.0014272.
Change in the posterior mean = 0.0064741.
Current mode = 3.295366937638081e+17
Mode improvement = 0
New value of jscale = 1.7928
Optimal value of the scale parameter = 1.7928
Final value of minus the log posterior (or likelihood):329536693763808130.000000
MODE CHECK
Fval obtained by the minimization routine (minus the posterior/likelihood)): 329536693763808130.000000
RESULTS FROM POSTERIOR ESTIMATION
standard deviation of shocks
prior mean mode s.d. prior pstdev
barSigmaPY 0.1000 0.1000 0.1798 norm 0.0500
barSigmaPC 0.1000 0.1000 0.1818 norm 0.0500
Log data density [Laplace approximation] is -329536693763808130.000000.
Estimation::mcmc: Multiple chains mode.
Estimation::mcmc: Old mh-files successfully erased!
Estimation::mcmc: Old metropolis.log file successfully erased!
Estimation::mcmc: Creation of a new metropolis.log file.
Estimation::mcmc: Searching for initial values…
Estimation::mcmc: Initial values found!
Estimation::mcmc: Write details about the MCMC… Ok!
Estimation::mcmc: Details about the MCMC are available in msDSGEPF/metropolis\msDSGEPF_mh_history_0.mat
Estimation::mcmc: Number of mh files: 1 per block.
Estimation::mcmc: Total number of generated files: 4.
Estimation::mcmc: Total number of iterations: 2000.
Estimation::mcmc: Current acceptance ratio per chain:
Chain 1: 36.25%
Chain 2: 37.2%
Chain 3: 34.7%
Chain 4: 34.15%
Estimation::mcmc: Total number of MH draws per chain: 2000.
Estimation::mcmc: Total number of generated MH files: 1.
Estimation::mcmc: I’ll use mh-files 1 to 1.
Estimation::mcmc: In MH-file number 1 I’ll start at line 1001.
Estimation::mcmc: Finally I keep 1000 draws per chain.
MCMC Inefficiency factors per block
Parameter Block 1 Block 2 Block 3 Block 4
SE_barSigmaPY 6.295 4.115 12.041 5.542
SE_barSigmaPC 7.174 5.318 5.834 3.295
worning: estimation:: MCMC convergence diagnostics are not computed because the total number of iterations is not bigger than 2000!
In McMCDiagnostics (line 127)
In dynare_estimation_1 (line 500)
In dynare_estimation (line 118)
In msDSGEPF.driver (line 1450)
In dynare (line 278)
Estimation::marginal density: I’m computing the posterior mean and covariance… Done!
Estimation::marginal density: I’m computing the posterior log marginal density (modified harmonic mean)… Done!
ESTIMATION RESULTS
Log data density is -329536693763808190.000000.
standard deviation of shocks
prior mean post. mean 90% HPD interval prior pstdev
barSigmaPY 0.100 0.0999 -0.1915 0.3719 norm 0.0500
barSigmaPC 0.100 0.0865 -0.1813 0.3781 norm 0.0500