The banking crisis that shocked the US economy and many other major economies in 2008-09 revealed the need for stronger banking systems.
Equity capital of banks is the bulwark against bank failure, because it provides a cushion to absorb losses without defaulting on debt and provoking a panic. The postcrisis Basel III reforms accordingly call for higher bank capital as the centerpiece of strengthening the international banking system.
Better supervision, including through the use of stress tests, is an important complement to higher capital.This study examines whether the capital reform has been correctly gauged or gone either too far or not far enough. The framework of the analysis is economic optimization. Higher bank capital provides a benefit to the economy by reducing the probability of banking crises and accordingly reducing the chance of severe recessions with output loss. The cost to the economy from imposing higher capital requirements is that there is a resulting increase in the average cost of capital to banks as they shift from low-cost debt to high-cost equity capital. As banks pass along these higher costs in the form of higher lending rates—and there is some spillover to higher lending rates in nonbank finance (such as corporate bonds)—the cost of investing in physical plant and equipment increases, reducing the expansion in the stock of productive capital in the economy. The optimal amount of bank equity capital is that amount at which the marginal benefit to the economy from avoiding banking crises equals the marginal cost to the economy from lower formation of capital stock available to cooperate with labor in the production process.
Some economists have argued that banks should hold extremely high levels of equity capital because there is a theoretical reason why there would be no increase in capital cost at all. This theory is the Modigliani-Miller (1958) theorem of capital structure irrelevance, which states that the sourcing of a firm's capital as between debt and equity makes no difference whatever to the average cost of capital.
The reason is that as investors perceive less risk in the firm as it deleverages from debt and moves to equity, they will provide equity at a cheaper unit cost than before. The starting point for an analysis of optimal capital requirements, then, is to test whether and to what extent the Modigliani-Miller (M&M) theorem holds for banks.Chapter 3 conducts tests for US banks on this question. It concludes that slightly less than half of the M&M offset attains in practice. As a consequence, some cost is passed along to the economy when higher bank capital requirements are imposed, even if this cost is only about half of what would be predicted if the M&M effect were ignored altogether.
With an estimate of the M&M offset in hand, chapter 4 estimates the optimal level of equity capital. It first estimates a benefits curve, relating the value of crisis losses avoided to the level of bank capital. It then estimates a corresponding cost curve for the economy, relating output lost as higher capital requirements are imposed, based on the higher cost of capital to the economy. The cost relationship turns out to be an upward-sloping straight line. The optimal capital ratio occurs where the slope of the benefits curve equals the slope of the cost line, such that marginal benefits equal marginal costs.
The central finding is that although Basel III has made major improvements, it has not gone far enough. For the large international banks that dominate the banking aggregates, the new requirements are set at 9.5 percent of risk-weighted assets. Risk-weighted assets represent only a little more than half of total assets, because holdings such as OECD sovereign debt enjoy a zero risk weight and other assets with strong collateral (such as mortgages) also have relatively low risk weights. The Basel III target corresponds to capital of slightly more than 5 percent of total assets.
In contrast, the analysis of chapter 4 finds that the optimal capital ratio is 12 to 14 percent of risk-weighted assets (or 7 to 8 percent of total assets), with the higher end reflecting a conservative 75th percentile instead of the median outcome in the possible outcomes under differing assumptions.
The Basel III capital requirements thus need to be increased by about one- third to reach optimal levels.This study also touches on several major issues surrounding these policy questions. Chapter 5 examines total loss-absorbing capacity (TLAC), which adds other categories of finance to equity in gauging whether the bank could absorb losses. I believe that nonequity TLAC is a poor substitute for equity, because it is precisely the kind of instrument that will experience panic runs when bank stress develops (and some of it, simple subordinated debt, could not be mobilized without entering into bankruptcy resolution). The chapter also addresses the potential problem that the Dodd-Frank reform in the United States and the Banking Recovery and Resolution Directive in the European Union have gone too far in shifting from lender-of-last-resort capacity to forced bail-ins that, although designed to protect taxpayers, may exacerbate panics and do more harm than good to the economy.
Chapter 6 addresses recent research that contends that there is too much finance already in the advanced economies and that it is choking off growth. If this diagnosis were correct, a growth benefit of curbing finance through increasing capital requirements would need to be incorporated into the optimal capital calculations. Instead, the chapter finds that the too much finance diagnosis is a case of statistical illusion.
Finally, the study also provides a critical review of the literature on bank capital requirements with respect to economies of scale and the issue of too big to fail (TBTF).