NON-PERFORMING LOANS IN THE EURO AREA:DOES MARKET POWER MATTER?
Maria Karadima
Athens University of Economics and Business
Helen Louri
Athens University of Economics and Business and London School of Economics HO/EI
ABSTRACT
As consolidation in the banking sector has increased impressively in the wake of the global financial crisis, the question of the impact of market power on bank risk has become topical again. In this study we investigate empirically the impact of market power as evidenced by concentration (CR5 and HHI) and (lack of) competition (Lerner indices) on the change in NPL ratios (ΔNPL). We use an unbalanced panel dataset of 646 euro area banks over the period 2005-2017. Since the distribution of ΔNPL is found not to be normal but positively skewed, we employ a penalized quantile regression model for dynamic panel data. We find conflicting results which are in line with the argument that more concentration does not always imply less competition. The results suggest that competition supports stability when NPLs increase but concentration enhances faster NPL reduction. In addition, we find that the effect of bank concentration is stronger in periphery euro area countries while the effect of competition is enhanced in banking sectors with higher foreign bank presence. Finally, bank competition is more beneficial for commercial banks in reducing NPLs than for savings and mortgage banks, while commercial banks are more prone to creating NPLs than the other two bank types. A tentative conclusion of our study could be that post-crisis consolidation facilitates the faster reduction of NPLs, while as the situation normalizes competition discourages the growth of new NPLs. Policy makers should take such findings into account by encouraging consolidation especially in periphery countries but also inserting competition in the banking sector through either regulating anti-competitive behavior or inviting new and/or foreign entrants.
Keywords: Non-performing loans; Competition; Lerner index; Concentration; Quantile regression; PQRFE estimator; PIVQRFE estimator
JEL Classification: C23; C51; G21
Correspondence:
Helen Louri
Athens University of Economics and Business
76, Patision str.
104 34, Athens
e-mail: elouri@aueb.gr