Big-Data Measurement of Blockchain Adoption in Listed Firms: Text Mining, Disclosure Quality, and Supply Chain Finance Evidence
Main article
Abstract
This article reframes the empirical study of blockchain in listed firms as a big-data measurement problem. We assemble a nine-year panel of 29,114 firm-year observations covering Chinese A-share companies between 2015 and 2023, and combine annual reports, exchange filings, patent records and management discussion text into a Blockchain Adoption Intensity Index (BAII) constructed through a tokenisation–dictionary–TF-IDF text-mining pipeline. The BAII captures both whether and how deeply each firm has integrated distributed-ledger technology, addressing well-known limitations of binary adoption proxies. Using the Shenzhen Stock Exchange disclosure rating as the dependent variable, we estimate fixed-effect models, instrumental-variable regressions with a regional R&D-intensity instrument, three-step mediation models with bootstrap inference, and Hansen panel threshold regressions on firm size. We find that higher BAII significantly raises disclosure quality, and the effect is amplified in high-tech industries; supply chain finance partially mediates the relationship; and a double threshold in firm size produces a fivefold increase in the marginal effect when moving from small to large firms. The article contributes a reproducible big-data construct, robust econometric evidence, and actionable governance implications for emerging-market regulators.
