Replication Data and Code for "Supplier Sharing and Input-Price Incidence: An AI-Chip Supply-Chain Application"

Published: 17 August 2026| Version 1 | DOI: 10.17632/szts7rn9ty.1
Contributor:
Hyungjin Kim

Description

This replication package contains code, public data extracts, and generated numerical artifacts for "Supplier Sharing and Input-Price Incidence: An AI-Chip Supply-Chain Application." The MATLAB workflow reproduces the structural calibration, counterfactuals, topology exercises, and numerical validation. The Stata workflow reproduces the descriptive customs-data figures, event-time profiles, tables, robustness exercises, and audit files using the exact included public destination--HS10--month Korean export extract. The package also contains provenance documentation, a manuscript-to-code crosswalk, software requirements, and SHA-256 checksums. Licensed third-party calibration files are not redistributed; the canonical parameter cache and accepted result structures needed to reproduce the reported structural calculations are included.

Files

Steps to reproduce

Unzip the archive into a writable directory and follow README.md. For the empirical results, install the listed Stata community packages and run code/stata/TrassEmpiricalMaster.do with the package root as its argument. For the structural results, set MATLAB to code/MATLAB, add that directory to the path, run the test suite, and then run run_canonical, run_equilibrium_validation, and run_qualification_ramp. The exact public analysis inputs are included, so no API key is required. Software versions, output locations, exhibit crosswalks, provenance, and checksums are documented in README.md, DATA_SOURCES.md, MANUSCRIPT_CROSSWALK.md, and SHA256SUMS.txt.

Categories

Economics, Econometrics, Industrial Organization, International Trade

Licence