By Arpit Gupta and Christopher Papagianis
The Dodd-Frank bill failed to address the key issues of prudential regulation that sparked the crisis. Instead, the FDIC was granted unlimited bailout capacity moving forward. Policymakers need to fundamentally rethink the value of regulatory capital ratios and discretionary bank interventions. Market-based metrics of bank performance that guide the recapitalization process may provide a far more robust and durable mechanism for preventing and handing future financial crises.
Charles Calomiris and Richard Herring have created a useful guide to thinking about Basel III and the Dodd-Frank bill. They note that the Dodd-Frank bill failed to address the key issues of prudential regulation that sparked the crisis and instead granted the FDIC unlimited bailout capacity moving forward. To help remedy this situation – and to help prevent the next crisis – they suggest adding Contingent Convertible ("CoCo") securities to the regulatory toolkit.
Calomiris and Herring identify the key failures in the regulatory system – that bank risks were not clearly identified, and that banks were not recapitalized in a timely fashion. Internally, banks did not face sufficient incentives to address the risks on and off of their balance sheets, and ultimately taxpayer dollars were used to cover the capital gap generated by banking losses.
While measures to increase bank equity are welcome, the traditional approach of higher capital ratios has repeatedly proved insufficient in the past. Higher capital ratios do not ensure that banks have sufficient equity relative to risk, or that banks replenish capital during times of crisis.
Rather (or in addition to) simply encouraging more bank equity, Calomiris and Herring suggest changing the type of bank equity. CoCos are securities that typically pay out like bonds, but can be quickly converted ("triggered") into equity during a crisis. If the trigger is set by some market price – say, the equity value – banks will receive an infusion of equity at the moment when the market judges their situation to be dire.
This design is valuable both from the point of view of bank corporate governance and supervisory regulation. Because a future CoCo conversion would dilute existing shareholders, managers would face stiff market discipline from shareholders (as well as CoCo holders) to ensure proper risk management. Meanwhile, the knowledge that banks will be certain to receive equity infusions in the event of a crisis reduces the need for regulatory forbearance or taxpayer bailouts for financial firms. A key problem in the period preceding the crisis was that bank managers were loath to raise equity while share prices were low because the lower the share price, the more dilutive a given capital raise would be for existing shareholders. CoCo bonds remove this discretion from managers and force common shareholders to take a more active interest in risk management practices.
Calomiris and Herring argue that regulators would be better off figuring out ways to ensure that unsecured investors face bounded losses – enough to ensure that banks do not entrench their too-big-to-fail status. Contingent Convertible ("CoCo") securities can be used to supply additional lending capacity during times of crises while providing a market indicator of risk. In addition, Calomiris suggests using loan interest rates when measuring loan default risk, instead of risk models. Such an approach would have required banks to budget more capital for high-interest subprime loans from 2004-2007.
Calomiris and Herrings' comments echo those of Andy Haldane, Executive Director of Financial Stability at the Bank of England, who recently released a speech given at this year's AEA conference in Denver. The speech reviewed the ways regulators failed to adequately monitor the capital base of major banks; and how market-based solutions such as Co-Co securities can help fix the problem.
Haldane provides a number of tables that powerfully make his case:
This chart illustrates the capital positions of "crisis" banks – those that failed or required government assistance – relative to those banks that were not comparably distressed. These two different types of banks held comparable amounts of Tier 1 capital, both around the 8% threshold mandated under Basel I. Yet the amount of capital held was not informative about the probability of a future bank failure.
Indeed, this degree of capital frequently proved to be insufficient to handle subsequent losses. As the following chart illustrates, the actual degree of losses proved to be greater than designated capital for several major banks:
Even under the new Basel III capital requirement proposal, most of the banks that subsequently failed (during the crisis) would have passed their capital tests with flying colors. The accounting or regulatory measure of capital proved inadequate and often misleading.
Importantly, market performance did prove to be a strong indicator of the future failure or success among banks. As the following chart illustrates, markets valued banks that would fail in the future consistently lower than their more successful peers, and the difference between the two widened beginning in 2006. Other measures of bank market performance – market capitalization relative to the book-value of debt or equity – show similar results.
The dotted line for a "5% threshold" represents Haldane's suggested trigger point for conversion of contingent capital. Under his plan, banks would be required to assume additional capital that would be converted once the trigger was breached. Had such a rule been followed during the financial crisis, troubled banks would have been recapitalized before the Lehman Brothers' failure (indeed, it's conceivable that this bank failure may have been prevented entirely), while non-crisis banks would have been spared conversion.
The use of such market-based rules for inserting additional equity is a topic that e21 has discussed repeatedly (here and here). To the conventional list of positive features, Haldane adds a few more comments:
1. Prevention of regulatory arbitrage. As Haldane notes, simply assessing the regulatory capital ratio of a bank requires hundreds of millions of calculations – increasing the odds that model error or bank efforts to hide risk will result in an inefficiently low capital ratio. The forces of regulatory arbitrage and increased financial complexity have prevented regulators from succeeding in the tasks of adequately proscribing capital ratios for banks.
Using convertible capital requirements tied to market triggers solves this problem. Rather than relying on complex model-based risk assessments of capital that failed in the last crisis, banks would instead have a straightforward market-based procedure behind additional capital requirements. Instead of trying to convince regulators, they would instead face market discipline. And, signs that a trigger might be pulled would motivate investors to better monitor banks.
2. A credible trigger mechanism. The Dodd-Frank financial regulation bill provides for a new resolution authority to wind up failing banks. But it leaves the timing of interventions up to the discretion of the regulator. Having a concrete market trigger ensures a rule-based point of intervention to raise capital, based on informative market signals of bank solvency. Regulators have an exceedingly difficult time assessing when a bank is insolvent and generally wait until a liquidity crisis strikes before taking action. This discretion tends to increase ultimate resolution costs.
In order to advance CoCos, Hardlane makes a simple suggestion – pay 50% of bonuses and dividends in the form of CoCos rather than cash. Had this suggestion been followed from 2000-2006, British banks would have entered the crisis with a capital base 3 percentage points higher. In addition to the gain in capital base, this requirement would also further discipline risk-taking by management. Calomiris also notes that a CoCo requirement would incentivize management to issue capital early, to prevent conversion.
The comments from Calomiris, Herring, and Haldane are a welcome corrective to the Basel III/Dodd-Frank prescriptions that envision more of the same. Instead, policymakers need to fundamentally rethink the value of regulatory capital ratios and discretionary bank interventions. Market-based metrics of bank performance that guide the recapitalization process may provide a far more robust and durable mechanism for preventing and handing financial crises.
Arpit Gupta is a Research Coordinator at Columbia University, where he focuses on consumer finance, real estate, and banking. Christopher Papagianis is the Managing Director of e21's New York Office and was Special Assistant for Domestic Policy to President George W. Bush.