Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
The Japanese banking system provides a distinctive platform for the examination of the long-lasting effect of problem loans on efficiency. We measure technical efficiency by modifying a translog enhanced hyperbolic distance function with two undesirable outputs, identified as problem loans and problem other earning assets. Our unique database allows us to distinguish between bankrupt and restructured loans to investigate the underlying causality between these loans and efficiency. From the flexible panel vector autoregression specification, primary results reveal that bankrupt loans have a positive impact on efficiency related to the “moral hazard, skimping” hypothesis, with the causality originating from bankrupt loans. In contrast, findings for the relationship between restructured loans and efficiency support the “bad luck” hypothesis.