Total Loss-Absorbing Capacity

Chapter 5 considers the role of TLAC—namely, CoCos, contingent write­down debt, and subordinated debt—as essentially buffers that are imperfect substitutes for equity but have a lower cost.

A substantial body of literature applies to TLAC. Much of it concerns the disciplinary role of subordinated debt. There seems to be little recogni­tion in the literature, however, that the mobilization of nonequity TLAC in a crisis could cause problems of its own. With respect to CoCos, the most likely holders are hedge funds, which would be fleet of foot in an incipient crisis, causing a plunge in CoCo prices as well as the stock prices of the banks in question (Persaud 2014). A taste of this phenomenon occurred in early 2016, when concerns arose that one of the largest European banks (Deutsche Bank) would be unable to make a coupon payment on contin­gent write-down debt because certain triggers appeared imminent (see ap­pendix 5A). As for outright subordinated debt that is eligible to qualify as TLAC, virtually none of the literature seems to focus on the fact that the writing down of subordinated debt is equivalent to a haircut in a default. Yet the whole purpose of regulatory structure should be to avoid the bank­ruptcy of banks and hence any need for default and haircut. Once again the problem is a shock to confidence in the banking system if the point of bank­ruptcy is reached. Much of the political pressure for TLAC seems to turn on the desire to avoid taxpayer bailouts of banks, but countenancing bank­ruptcy of banks for this purpose instead of increasing equity and avoiding bankruptcy constitutes a risky strategy.

On the disciplining role of subordinated debt, Ashcraft (2008) finds such a role before the Basel accord but not thereafter. He attributes the change to the Basel requirement that tier 2 subordinated debt cannot include restrictive covenants. Evanoff, Jagtiani, and Nakata (2011) find that for banks, spreads on subordinated debt are indeed related to risk (as indicated by nonper­forming loans and other measures) and argue that the stronger relationship found for the subsample of banks with recent issuance means that a man­datory requirement for subordinated debt should not be rejected because past signals have seemed too noisy to justify such a change. Calomiris and Herring (2013) state the case for CoCos in terms of imposing strong incen­tives on management to recapitalize rather than waiting too long.

In the area of TBTF, the literature has tended to use the “ratings uplift” difference between “support” and “stand-alone” ratings of banks by major rating agencies as a measure of the implicit TBTF subsidy. Marques, Correa, and Sapriza (2013) find that bank “z-scores” or distance to insolvency ratios (return on assets plus capital-to-assets ratio, normalized by standard devia­tion of return on assets), are negatively related to the TBTF subsidy as mea­sured by the ratings uplift. Afonso, Santos, and Traina (2014) similarly find risk taking as measured by impaired loans or net charge-offs to be signifi­cantly related to the ratings uplift. An important caveat to this literature, however, is that some studies find that by 2013 the bond spread advantage of TBTF firms had swung to a disadvantage, suggesting that reforms such as living will resolution plans required of large banks in the United States and the prospect of new mandatory bail-ins in the euro area's banking union reforms may have already substantially reduced the TBTF subsidy.

On economies of scale, Wheelock and Wilson (2012) and Hughes and Mester (2013) find important economies of scale in banking.

Chapter 5 closes with a test for the experience of US banks in the Great Recession. It examines the change in net earnings as a fraction of assets from 2006-07 to 2008-10 for 16 TBTF banks with assets of more than $100 billion and compares it to the change for the next 34 largest banks. If the largest banks had engaged in excessive risk taking as a consequence of the TBTF subsidy, one would have expected them to have incurred substantially larger losses than the next tier of banks. But a simple test for significant dif­ference of means fails to show any difference between the two groups. The implication is that the portrait of excessive risk taking may have been over­done, and correspondingly that the largest banks may place a greater weight on “charter value” (emphasized in earlier literature) relative to TBTF profits than generally recognized.^[7] It may also be appropriate to pay closer atten­tion to economies of scale, another controversial aspect of the literature.