Economic studies of the costs and benefits of capital requirements for banks may usefully be divided into five categories.
The first includes what might be called heuristic seawall studies. These more informal approaches (hence “heuristic”) often appeal to the general proposition of Modigliani and Miller (1958) that equity can replace debt with no adverse profit implications for the firm, such that the very notion of an optimal capital ratio would be flawed.
They tend instead to consider capital ratios in earlier historical periods, magnitudes of losses by banks in the Great Recession, and cursory calculations of enormous output losses in banking crises.Studies in this category tend to generate what may be called a “seawall” estimate of how high the capital levee needs to be to protect the economic coastline from the infrequent severe storm. Such estimates calculate the amount of losses banks would experience in a crisis and then add that amount to a core amount of capital that the authors believe banks should retain even at the depth of a banking crisis in order to estimate the total capital banks should hold in normal times.
A second category involves transient analyses. These studies focus on possible costs and benefits in the period in which banks adjust to higher capital requirements. The central message of these studies is that if banks were required to reach sharply higher capital standards overnight, their equity prices would fall as a result of the flood of new issuances, with possible adverse effects for the economy as they scaled back activities. The policy implication of these studies tends to focus on phasing in higher capital requirements over time.
The third category is more formal modeling of banking crisis costs in comparison with the long-term economic costs of higher capital requirements. The analysis of chapter 4 is in this category. This category represents calibrated optimization based on the measurement of the marginal benefits and cost of additional bank capital.
A fourth category includes dynamic stochastic general equilibrium (DSGE) models. These models examine such effects as overinvestment in risky projects by banks in response to the subsidy provided by deposit guarantees, compared to eventual scarcity of capital formation if capital requirements are raised too high.
A fifth category concerns extension of bank regulatory requirements to include assessment of total loss-absorbing capacity (TLAC). It represents a variant of the seawall approach that focuses on ensuring that taxpayers do not foot the bill in banking crises, that banks instead go bankrupt if necessary, with the costs borne only by their shareholders and creditors. This approach tends to be premised on the proposition that banks engage in risk shifting spurred by subsidized deposit guarantees.
This chapter examines the literature on the two approaches most germane to the main calculations of this study: the seawall and calibrated optimization approaches. Appendix 2A surveys the literature on transient effects and DSGE model estimates. Chapter 5 examines the literature on TLAC and the analytical issues it involves, including economies of scale in banking and panic dynamics for the quasi-equity securities involved (contingent-convertible [CoCo] debt, contingent write-down debt, and subordinated debt).