DP16670 A q-theory of banks
We propose a dynamic bank theory with a delayed loss recognition mechanism and a regulatory capital constraint at its core. The estimated model matches four facts about banks' Tobin's Q that summarize bank leverage dynamics. (1) Book and market equity values diverge, especially during crises; (2) Tobin's Q predicts future bank profitability; (3) neither book nor market leverage constraints are binding for most banks; (4) bank leverage and Tobin's Q are mean reverting but highly persistent. We examine a counterfactual experiment where different accounting rules produce a novel policy tradeoff.