Bank Capital in DSGE Models

Sammanfattning: This paper analyzes the implications of three different modeling approaches to bank capital for business cycle dynamics in a DSGE model with various real and financial frictions. Two approaches feature exogenous capital requirements but differ in their mathematical form. The third approach is novel to the DSGE literature and the capital requirement for banks emerges endogenously. Important features of the modeling approaches become visible only with higher-order Taylor approximations. We find that, relative to exogenous capital requirements, the endogenous capital requirement amplifies modestly business cycle dynamics and leads to opposite effects in some financial variables. In all models, short-run dynamics of bank capital are mainly driven by bank profits and in some cases of surprisingly strong magnitude. Surveying the literature we do not find any consensus on these effects and that most have not yet been tested empirically. We discuss major model results in light of basic empirical evidence and outline important aspects for future empirical work.