THE CREDIT PARADOX IN PERU: WHY CONSUMER CREDIT CONTRACTS DURING AN EXPANSIONARY MONETARY POLICY (2004–2025)

This repository contains the data and replication code for the empirical analysis in the paper. The study documents a structural break in Peru's bank lending channel using 21 years of quarterly data (2004Q1–2024Q4) and formally separates two transmission-weakening mechanisms: AFP extraordinary pension withdrawals (59% of the break) and supply-side NPL tightening (41%). Contents: data/ — Quarterly macroeconomic series (M3, consumer credit, microcredit, reference rate, CPI, GDP, exchange rate, NPL) sourced from the Banco Central de Reserva del Perú (BCRP) and the Superintendencia de Banca y Seguros (SBS). AFP withdrawal rounds manually constructed from official MEF/SBS records (7 rounds, 2020–2024, S/ 63,219M cumulative). scripts/01_vec_cointegration.py — Johansen cointegration tests, VEC estimation, Chow structural break test, sub-period elasticities. scripts/02_panel_fe.py — Fixed-effects panel estimation (Equation 7 and 7'), temporal heterogeneity interactions, institution-level exploratory panel (Appendix C). scripts/03_iv_2sls.py — 2SLS instrumental variables estimation using AFP round-eligibility binary instruments; Hausman endogeneity test; Sargan-Hansen over-identification test. AFP_sources.md — Full documentation of the seven AFP withdrawal rounds with official source references. README.md — Execution instructions and software requirements. Software: Python 3.10+. Required packages: pandas, numpy, statsmodels, linearmodels. Key results: Long-run VEC elasticity β^LR = 0.87 (full sample), declining from 0.92 (2004–2019) to 0.61 (2020–2024); panel FE consumer credit elasticity = 0.68; 2SLS first-stage F = 23.4; Hausman χ²(1) = 4.87 (p = 0.027). JEL codes: E51, E52, G21, C32, G28 Keywords: consumer credit, microcredit, monetary transmission, cointegration, structural break, pension withdrawals, Peru, bank lending channel