Resumen Most empirical studies in the banking literature assume that the alternative profit function is linearly homogeneous in input prices. We show that such an assumption is theoretically unwarranted and that its use may yield misleading results. We use Koetter et al. (Review of Economics and Statistics 2012; 94(2): 462-480) as a benchmark to showcase how empirical results can be sensitive to the linear homogeneity assumption. Contrary to Koetter et al., we find a positive relation between market power and profit efficiency when this assumption is dropped. This relation is slightly weakened after the wave of intrastate and interstate deregulation but not enough to support the 'quiet life' hypothesis. © 2014 John Wiley & Sons, Ltd.