A Three-Layer Risk-Sharing Architecture for Voluntary Carbon Markets: Data and Replication Code
Description
This repository provides the replication data and code for “A Three-Layer Risk-Sharing Architecture for Voluntary Carbon Markets: Insurance, Reference Prices, and Digital Emergency Rebasing” (Jin, 2026). The replication package covers three complementary evidence blocks. E1 uses monthly registry–project-type data derived from the Berkeley Carbon Trading Project’s Voluntary Registry Offsets Database (November 2014–June 2026) to examine common and segment-specific variation in voluntary carbon credit market activity using principal component analysis, permutation tests, Horn parallel analysis, and leave-one-registry-out robustness checks. E2 uses California Air Resources Board allowance-auction data to examine the observed relationship between auction reserve prices and market-clearing settlement prices as an institutional analogue for policy-linked price referencing. E3 implements a reproducible energy-storage-system (ESS) stress-test framework that varies physical loss severity, verification delay, measurement overstatement, and recovery duration to evaluate stale-claim exposure, emergency rebasing, and verified recovery. analysis_data.xlsx contains the analysis-ready inputs and reported results, while analysis_code.py reproduces the E1–E3 analyses. A separate Berkeley VROD collector script reconstructs the analysis-ready registry panel from the original registry workbook. Variable definitions, empirical specifications, and the scope of inference for each evidence block are documented in the paper and online supplement.