NON-PERFORMING LOANS AND FINANCIAL CONSUMER PROTECTION POLICIES: EMPIRICAL EVIDENCE FROM A BROAD INTERNATIONAL SAMPLE

This study explores the relationship between policy choices in financial consumer protection (FCP) and non-performing loans (NPLs) in a broad international context, recognising that NPLs negatively affect citizens’ well-being and the social pillar of sustainable development. Nine FCP indices were constructed to capture distinct regulatory choices based on regulators’ responses to the World Bank FICP surveys. The estimation framework extends standard macroprudential models by adding institutional indicators for the general environment, as well as for the corporate and household credit segment, with the latter being represented by the FCP indices. We estimate seven models using an unbalanced dataset of 113 countries. These include a within model, two supplementary cross-sectional models, and four time–cross-section panel models that serve as the main specifications. Two FCP indices show unfavourable associations with NPL dynamics – financial inclusion (moderate) and credit pricing (small) – while three show favourable associations – financial literacy (moderate), credit reporting (moderate), and regulatory capacity (large). Four other indices – information disclosure, affordability, fair treatment, and debt resolution – do not show sufficient evidence of association with NPL dynamics. Financial inclusion shows a concerning association, suggesting a potential trade-off between wider credit access and credit quality, which may be particularly acute under weak responsible-lending requirements. By contrast, the small positive association for credit pricing is economically minor and may be explained by a denominator effect arising from lower pre-COVID NPL levels. Policies on financial literacy, credit reporting, and regulatory capacity show evidence of robust negative associations, indicating favourable outcomes in terms of NPL dynamics for these policy choices. The FCP policies represented by the other four indices may still be favourable for citizens’ well-being through effects not captured in this NPL-focused analysis, which is also constrained by the short and small panel dataset.