Abstract We analyze the profitability of government-owned banks’ lending to their owners, using a unique data set of relatively homogeneous government-owned banks; the banks are all owned by similarly structured local governments in a single country. Making use of a natural experiment that altered the regulatory and competitive environment, we find evidence that such lending was used to transfer revenues from the banks to the governments. Some of the evidence is particularly pronounced in localities where the incumbent politicians face significant competition for reelection.

Surprisingly little is known about the business cycle dynamics of leverage. The existing evidence documents that target leverage evolves pro-cyclically either for all firms or financially constrained ones. In contrast, we show that, on average, target leverage ratios evolve counter-cyclically once cyclicality is measured comprehensively, accounting for variation in explanatory variables and model parameters. These counter-cyclical dynamics are robust to different subsamples of firms, data samples, empirical models of leverage, and definitions of leverage. There is a fraction of 10–25% of firms with pro-cyclical dynamics whose characteristics are consistent with counter-cyclical dynamics for loss-given-default and probability of default.