Demonstration for "Econometric Estimation with High-Dimensional Moment Equalities" (2016).
Switch branches/tags
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
sensitivity_IV
.gitignore
EKK.m
EKK_reduced2.m
IC.m
LM.m
L_deriv.m
MomentComponents.m
Numerical_implementation.lyx
Numerical_implementation.pdf
README.Rmd
README.md
boost0.m
boost_EKK.m
boosting.m
country_code.csv
gamm_msk.m
h_deriv_EKK_5.m
master_REL.m
master_boosting.m
minimal_REL.R
minimal_REL.lyx
minimal_REL.pdf
probSum.m
trade_data.mat

README.md

This document briefly describes the code that implement the REL and the boosting-type greedy algorithm in my paper

"Econometric Estimation with High-Dimensional Moment Equalities", 2016, Journal of Econometrics, 195, 104-119

via a demonstration with the Chinese trade data.

Prerequisite

To run the code, we need to install MOSEK and hook it up with Matlab. MOSEK is a commercial convex problem solver. It provides free academic license.

Data and Files

  • trade_data.mat: data file for the Chinese empirical application in my paper.
  • master_REL.m: the master function for the relaxed empirical likelihood.
  • master_boosting:m: the master function for the boosting procedure in my paper.
  • EKK: generates the predicted market entry and the sale. It contains the EKK paper's micro model as a nested function.

Environment

The master files work on the author's PC in Matlab 2014b