Python package that computes worst-case standard errors (SE) for minimum distance estimators, given knowledge of only the marginal variances (but not correlations) of the matched moments.
The computed worst-case SE for the estimated parameters are sharp upper bounds on the true SE (which depend on the unknown moment correlation structure). For over-identified models, the package also computes the efficient moment selection that minimizes the worst-case SE. Additionally, the package can carry out tests of parameter restrictions or over-identifying restrictions.
Reference: Cocci, Matthew D., and Mikkel Plagborg-Møller (2023), "Standard Errors for Calibrated Parameters", arXiv:2109.08109
Tested in: Python 3.8.18 (Anaconda distribution) on Windows 10 PC with NumPy version 1.24.3.
Other versions: Matlab
-
example.ipynb: Simple interactive example in Jupyter Notebook illustrating the main functionality of the package (also available in HTML format)
-
example_ngm.ipynb: Even simpler example in Jupyter Notebook of calibrating the Neoclassical Growth Model (also available in HTML format)
-
stderr_calibration: Python package for minimum distance estimation, standard errors, and testing
-
estimate_hank.py: Empirical application to estimation of a heterogeneous agent New Keynesian macro model, using impulse response estimates from Chang, Chen & Schorfheide (2023) and Miranda-Agrippino & Ricco (2021), which are stored in the data folder
-
sequence_jacobian: Copy of the Sequence-Space Jacobian package developed by Auclert, Bardóczy, Rognlie & Straub (2021), with minor changes made to the file hank.py
-
tests: Unit tests intended for use with the pytest framework
The Python packages cvxopt and numdifftools are required.
We thank Minsu Chang and Silvia Miranda-Agrippino for supplying the moments used in the empirical application.
This material is based upon work supported by the NSF under Grant #1851665. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.