A number of academics and policy makers have favored the forced debt-for-equity conversion bonds or contingent capital. With any regulation, the key is to know strengths and weaknesses. The strengths of contingent capital have been mentioned a number of times here and there (for most recent FT article, see http://www.ft.com/cms/s/0/0310ebf4-4342-11df-9046-00144feab49a.html). Here, I outline some contingencies for contingent capital that regulators must keep in mind.
The Current Proposals
Both the House and Senate bills call for the issuance of contingent capital to be an additional standard for systemically important institutions. In particular, as a part of its effort to Restoring American Financial Stability, the Senate Committee chaired by Chris Dodd has proposed the setting up of a Systemic Risk Council, which besides deciding which financial institutions are systemically important, will govern prudential standards for these institutions for risk-based capital, leverage, liquidity and contingent capital. Of these, the newest prudential standard is the requirement that the systemic institution hold a minimum level of contingent capital, a form of debt that would automatically convertd to equity capital when pre-specified triggers of an institution and system-wide economic health are met.
The Bill charges that "The Council shall conduct a study of the feasibility, benefits, costs, and structure of a contingent capital requirement for nonbank financial companies supervised by the Board of Governors and bank holding companies [deemed systemically important], which shall include-- (A) an evaluation of the degree to which such requirement would enhance the safety and soundness of companies subject to the requirement, promote the financial stability of the United States, and reduce risks to United States taxpayers; (B) an evaluation of the characteristics and amounts of convertible debt that should be required; (C) an analysis of potential prudential standards that should be used to determine whether the contingent capital of a company would be converted to equity in times of financial stress; (D) an evaluation of the costs to companies, the effects on the structure and operation of credit and other financial markets, and other economic effects of requiring contingent capital; (E) an evaluation of the effects of such requirement on the international competitiveness of companies subject to the requirement and the prospects for international coordination in establishing such requirement; and (F) recommendations for implementing regulations."
Overall Evaluation of Contingent Capital as a Systemic Risk Control
By restoring some market discipline and reducing the likelihood of default of financial firms when adverse shocks materialize, contingent capital can be a valuable tool for averting the need to bail out systemically important firms. Thus, it is clearly a good idea, but in our opinion, there are some important limitations that must be borne in mind. These include: (i) its ability to limit ex ante risk-taking and build-up of systemic risk; (ii) its usefulness in dealing with distress when complex contingent and off-balance sheet liabilities characterize a financial firm's balance-sheet; (iii) its relative attractiveness to standard capital and liquidity requirements; and, finally (iv) the limitations to coming up with an international standard of bank regulation tied to contingent capital.
First, the primary purpose of contingent capital seems to ex post avoid a regulatory bailout. Its ability to control bank risk-taking in good times is however limited. In such times, banks can - as they have in the past - take bets on the tail risk of the economy by selling deep out of the money options, such as A.I.G.'s credit default swaps on mortgage and corporate portfolios, Citigroup's selling of under-capitalized liquidity puts to conduits, and large holdings of AAA-rated tranches by investment banks, Fannie and Freddie, and to an extent, even large commercial banks. A property of taking on such tail risk is that the only outcomes possible are "boom" or "bust", and the intermediate region of risky outcomes over which contingent capital might have some bite is essentially rendered rather unlikely or inconsequential. Such tail-risk seeking would likely have to be addressed through means other than pure reliance on contingent capital requirement.
It is important to recognize that the real problem is not between unsecured creditors and bank shareholders, but between the government and uninsured capital providers. While resolution plans can be designed to limit the extent of government transfers to uninsured capital providers, some such transfers will necessarily arise in future, especially if firms experience abrupt distress due to the tail nature of their risks (as explained above). The moral hazard arising from such transfers is best addressed by imposing a fee based on systemic risk contributions of individual institutions. Unless banks are appropriately charged for losses they impose on the system during aggregate crises, they will not internalize these losses. Thus, we recommend that in addition to contingent capital and resolution plans, an explicit fee be charged to banks in good times based on their expected losses and their systemic risk contributions (measured as described in Acharya, Pedersen, Phillipon and Richardson, 2010, "Measuring Systemic Risk", Download PDF).
Second, we believe that contingent capital is not adequate even for containment of ex post distress in all contingencies, and especially in the form it is proposed whereby there will be a one-time conversion of part of a firm's debt into equity. If instead, and depending on how deteriorated the conditions are, there was a requirement of progressive conversion of debt into equity all the way down the capital structure of financial firms, then indeed all firm losses could eventually be passed to creditors. Such progressive conversion could be a part of the firm's "living will" or resolution plan. Nevertheless, we envision several scenarios in which before such a plan can be fully executed, some counterparty risk or large-scale liquidation risk may arise, necessitating receivership or bankruptcy of some form. In other words, we should not rule out yet the possibility that there will be systemic crises in the future that for lack of any other choice involve bailouts of certain systemically important financial firms. Furthermore, some part of bank debt is explicitly insured, and this debt cannot be converted to equity ex post.
Thus, without progressive debt for equity conversion, contingent capital does not fully address the fact that beneath both contingent capital and equity capital of banks lie a significant portion of debt - deposits, secured debt (repos), non-contingent debt of other types, liabilities to derivatives transactions - that will remain explicitly and, in some cases, implicitly guaranteed by governments. The cost of such debt in good times will not reflect the true risks of banks, and as long as this is true, both contingent capital and equity capital will find it desirable to undertake excessive risks at the expense of guaranteed debt (taxpayer money). Moreover, the amounts of contingent capital being considered currently do not appear to be sufficiently large. Consider investment banks that were operating at leverages of 25:1 to 35:1 in terms of debt to equity before many of them collapsed. With such leverage, even a small quantity of abrupt and adverse negative news about assets will be sufficient to wipe out equity capital and the slivers of contingent capital that are currently being talked about. Such leverage ratios need to be prudentially controlled at the outset.
Third, the attractiveness of contingent capital relative to this alternative of a leverage constraint needs to be evaluated. Indeed, another alternative to contingent capital is simply to increase equity or Tier 1 capital requirements, tying them to systemic risk contributions of firms. The usual argument against this is that demandable debt (bank notes or checking accounts) provides far more discipline on bankers - who can alter risks at fast speeds - than equity capital does. It is time to establish the magnitude of this assumed social cost of equity capital. For one, reliance on short-term debt for market discipline comes at a huge social cost of systemic financial fragility. Second, we have offered huge tax advantages to debt. And third, there are better mechanisms available for shareholder governance today than in times when demandable debt grew in fashion.
Put together, these arguments suggest that contingent capital is a part of the big puzzle of rewriting financial sector regulation, and needs to be complemented with other measures. Hence, we endorse the Senate Bill's recommendation that all aspects of contingent capital - its merits and limitations, individually and in relationship to other possible (systemic) risk controls - be carefully evaluated. Every crisis is different and to the extent that contingent capital would not yet have stood the test of time when the next crisis hits, it seems prudent to combine it with other, more direct, risk-control measures such as systemic-risk based fees or direct leverage restrictions or enhanced Tier 1 capital requirements.
Finally, contingent capital is likely to work well in developed countries with well-developed corporate bond markets, but perhaps not feasible elsewhere. From this standpoint too, some leverage restrictions or systemic risk-based capital and liquidity requirements standards are more likely to emerge as international norms evolve.