Join GitHub today
GitHub is home to over 31 million developers working together to host and review code, manage projects, and build software together.Sign up
This repository contains a Matlab suite to construct weak-instrument robust confidence intervals for impulse response coefficients in Structural Vector Autoregressions identified with an external instrument. See "Inference in Structural Vector Autoregressions identified by an external instrument" by J.L Montiel Olea, J. H. Stock, and M. W. Watso… http://www.joseluismontielolea.com/pa…
Fetching latest commit…
Cannot retrieve the latest commit at this time.
|Type||Name||Latest commit message||Commit time|
|Failed to load latest commit information.|
This README_SVARIV.txt file was generated on 9/18/18 by José Luis Montiel Olea —————————————————————————————————— i) GENERAL INFORMATION —————————————————————————————————— The folders/files: OilSVARIV.m TaxSVARIV.m Data functions MCexercise Output PaperReplication seed Contain .xlsx & .txt files and Matlab scripts, functions, MAT-files, and .eps files used to generate the figures reported in the paper “Inference in Structural Vector Autoregressions Identified With an External Instrument” by José L. Montiel Olea, James H. Stock, and Mark W. Watson. The Matlab scripts OilSVARIV.m and TaxSVARIV.m have been created such that a user can substitute their own data to conduct their own SVAR-IV analysis and generate figures pertaining to their own variables of interest. —————————————————————————————————— ii) HARDWARE/SOFTWARE (specifications and requirements) —————————————————————————————————— All the files have been tested on: * An iMac @3.4 GHz Intel Core i5 (16 GB 2400 MHz DDR4) running Matlab R2018a. * A Macbook Air @2.2 GHz Intel Core i7 (8GB 1600 MHz DDR3) running Matlab R2016b. —————————————————————————————————— iii) RECOMMENDED CITATION —————————————————————————————————— When using this code please cite: “Inference in Structural Vector Autoregressions Identified With and External Instrument”, Montiel Olea, José L; Stock, James H.; and Watson, Mark W; Working Paper Columbia University. —————————————————————————————————— iv) DATA & MAIN FILE OVERVIEW —————————————————————————————————— * OilSVARIV.m & TaxSVARIV.m These scripts generate IRFs and confidence intervals for an oil and a tax SVAR-IV using the methodology described in the paper. In order to illustrate the applicability of these scripts to other settings, these files take different data as input and have differing parameters (i.e lags, variable names, time horizons, etc.) Our goal is that researchers can use these scripts with their own data and parameters to report different objects of interest in SVAR-IV analysis. * Data This folder contains data for the oil-SVAR used in the paper as empirical illustration, as well as a the fiscal-SVAR application in “Marginal Tax Rates and Income: New Time Series Evidence”, Mertens and Montiel Olea (2018); Forthcoming at The Quarterly Journal of Economics. The data have been organized into separate “Tax” and “Oil” subfolders. If you intend to use your own data for your application, we recommend that you place it in a folder within “Data”. * functions This folder contains all the functions used in the analysis. The folder “Inference” contains the main function (MSWfunction.m) used to construct the Anderson-Rubin confidence sets described in the paper. * MCexercise This folder contains scripts for the Montecarlo exercises. MonteCarlo_Script.m can be modified and used for your own applications. * PaperReplication This folder contains scripts and .eps files used to generate Figures 1 & 2 & the Montecarlo figures in the paper. * seed This folder contains the MAT-file generated using the seed in the script. *Output This folder contains all of the saved output generated by OilSVARIV.m and TaxSVARIV.m. The subfolders “Figs” contain all of the figures (.eps files) that were generated in the script, “Mat” contains all of the MAT-files that were generated, and MC contains the figures and MAT-files that were generated and saved from the Montecarlo exercises.