Bank credit allocation and productivity: stylised facts for Portugal

Abstract

Purpose The linkages between credit allocation and productivity have particular relevance in Portugal. This study aims to investigate whether credit extended by the Portuguese banking system has been allocated to the most productive firms within each sector.

Design/methodology/approach With a data set covering 95% of total outstanding credit to non-financial corporations recorded in the Portuguese credit register, the authors investigate whether outstanding loans by resident banks to 64 economic sectors have been granted to the most productive firms. First, the authors estimate a baseline, reduced-form model of credit reallocation, where the parameter of interest gives the response of total credit granted to each firm to its level of productivity. Second, the authors assess how this response is affected by the share of credit allocated to unproductive firms. Third, the authors redo the analysis with credit granted to each firm by each banking group, instead of by the entire banking system, so that bank indicators can be taken on board.

Findings The authors find evidence of misallocation, which reflects the joint effects of credit supply and credit demand decisions taken over the course of time, and the adverse cyclical developments following the accumulation of imbalances in the Portuguese economy for a protracted period. In 2008–2016, the share of outstanding credit granted to firms with very low productivity (measured or inferred) was always substantial, peaking at 44% in 2013, and declining afterwards with the rebound in economic activity and the growing allocation of new loans towards lower risk firms and away from higher risk firms. Furthermore, the authors find that misallocation is associated with slower reallocation. The responsiveness of credit growth to firm relative productivity is much lower in sectors with relatively more misallocated credit and when banks have a high share of such credit in their portfolios.

Originality/value Banking system distortions are often mentioned as potential or likely culprits for capital misallocation, but they are not empirically analysed with credit data. The ability to explicitly analyse bank credit and link it to variables pertaining to both firms and banks is a novel feature relative to most previous studies, which largely rely on firm-level or sectoral data alone.