The profitable-default frontier of revolving consumer credit: reproducibility package (code, derived data, figures)

Reproducibility package for the study The profitable-default frontier of revolving consumer credit: a cross-country map and its regulatory ceiling. The study defines the profitable-default frontier p* = (r − c)/(r + 1 − R) — the highest borrower default probability at which revolving consumer lending still breaks even — and maps it across 29 economies using public data: ECB MFI interest-rate statistics, U.S. Federal Reserve G.19, Banco Central do Brasil SGS, BIS household debt-service ratios and policy rates, and the World Bank Survey on Interest Rate Controls 2019. Key findings: Brazil's frontier stands an order of magnitude beyond the OECD (p* up to ≈0.83); household debt-service increments are heavy-tailed (q ≈ 1.26, robust to serial and cross-sectional dependence); and two pre-registered nulls show that the frontier does not predict realized household debt service and that 2019 interest-rate caps bound the ceiling of the frontier without setting its level. The package contains the full Python pipeline, the derived analytical tables and result files, the five figures, the pre-registration, the results ledger, and SHA-256 checksums. Raw inputs are not redistributed; every script points to its public source and can be re-run to regenerate the derived outputs.