How to Fix the Resolution Problem of Large, Complex, Nonbank Financial Institutions
Click here for a printer-friendly version of this article
In the wake of the ad hoc bailouts of AIG and Bear Stearns, and the bankruptcy of Lehman, policy makers seeking to avoid having to make the choice between taxpayer-financed bailouts of risk takers and disruptive liquidations are struggling over whether and how to improve the resolution process for nonbank financial institutions.
Reform must create a means to transfer the control of assets and operations of a failed institution in an orderly way, while ensuring that shareholders and creditors of the failing firm suffer large losses. Those outcomes are essential if the resolution of the failure is to avoid significant disruptions to third parties, avoid bailout costs to taxpayers, and restore the market discipline that avoids too-big-to-fail moral hazard.
Two approaches to addressing the problem have been suggested: (1) bankruptcy reform tailored to the needs of nonbank financial institutions (an approach favored by most Republicans in Congress), and (2) the creation of an administrative resolution authority (which is favored by most Democrats). Each approach has strengths, but each also has weaknesses. A hybrid approach – bankruptcy reform with a resolution authority loophole – may be best.
Bankruptcy reform should address several problems in existing law as applied to complex nonbank financial institutions: (1) ensuring clarity regarding the jurisdiction over assets (deciding which countries’ courts have control over which assets); (2) avoiding potential gridlock in financial networks resulting from the inability of the failing firm’s creditors to liquidate positions they hold in the failing firm, which they were relying on for their own liquidity needs, and (3) avoiding the gaming by creditors’ who have laid off their default risk using credit default swaps (CDS) – the problem arises, for example, if a creditor has a larger short CDS position than the amount of debt the creditor holds in the failing firm, which gives the creditor an incentive to use his voting power in bankruptcy to sabotage an efficient renegotiation plan (the recent CIT bankruptcy is pointed to as an example of this problem).
Bankruptcy reforms could address these three problems by (1) establishing “living wills” of financial firms, approved in advance by regulatory authorities and governments in the relevant countries, so that locations of assets and jurisdictions are clear, (2) improving the rules governing payments of maturing obligations to limit damage to third parties from frozen assets (e.g., using conservative valuations of the institutions’ assets, the bankruptcy judge could permit debtor-in-possession financing by private lenders or the Fed to fund payments to creditors of fractional amounts of maturing obligations in an amount less than a conservative estimate of the ultimate recoveries on those obligations, on the condition that creditors receiving fractional payment relinquish their claim to additional payments in the future), and (3) establishing new voting rules for creditors in bankruptcy that would better encourage efficient renegotiation, for example, by using net creditor exposures (debts less CDS hedges) to determine creditors’ voting rights in bankruptcy.
Some advocates of an alternative to reformed bankruptcy, based on administrative resolution (a resolution process in which the government control the allocation of losses rather than a court), oppose the bankruptcy reform approach because they worry (in my view, without a reasonable basis) that reforms might not be sufficient to ensure a speedy and orderly transfer of assets and control during a bankruptcy.
A more plausible version of the argument in favor of administrative resolution would be grounded in political risks; even if bankruptcy would provide an adequate resolution mechanism, if reformed bankruptcy is not deemed to be sufficiently orderly by excessively risk-averse politicians and regulators, then the political process may resort instead to ad hoc resolutions (like the bailout of AIG). Vesting authority in a regulator with access to public funds, while ensuring that administrative resolution does not insulate creditors of a failing institution from significant loss, might give resolution a better chance of achieving an orderly transfer of assets with low social costs.
But resolution has a big downside. Government officials with access to public funds likely will be too quick to spend those funds on bailouts even when there is little risk to the financial system, either because of special-interest pressures, or because policy makers will be excessively risk-averse about spillover effects.
A hybrid bankruptcy/resolution approach could avoid both the risk of the pure-bankruptcy approach (the encouragement of unconstrained, ad hoc bailouts) and the risk of the pure-resolution approach to reform (the tendency to use public funds too frequently to bail out failed institutions’ creditors). That hybrid approach would create a pre-defined resolution process as a backstop to a reformed bankruptcy process. Having a backstop in place would substantially reduce the chance of ad hoc bailouts, and limit the use of resolution to moments of legitimate concern over systemic risk.
Under a hybrid approach, the new bankruptcy process would govern nonbank financial institutions unless strict conditions were satisfied to trigger the use of administrative resolution and its costs. The resolution process should be specified to (1) impose 100% haircuts on stockholders and significant haircuts on creditors and counterparties, who should expect losses similar to what they would suffer in bankruptcy; placing caps on losses may make sense to quell systemic concerns of “domino” effects, but such concerns do not justify avoidance of significant loss, the prospect of which makes creditors and counterparties careful when dealings with high-risk firms, (2) require ex post funding of the costs to taxpayers by the largest 100 financial institutions, who would presumably benefit from any reduction in systemic risk resulting from the proper use of the resolution authority, up to some fraction of those institutions, and (3) require both that Congress vote to allow the use of the resolution authority (since taxpayer funds would backstop privately funded protection), and that a (value-weighted) majority of the financial institutions that would pay the costs of the resolution also vote in favor of using the resolution authority. Those votes would ensure that the resolution authority would only be used when the reformed bankruptcy process truly is unable to resolve failed institutions effectively.
A time-honored principle of incentive-compatible bailouts is that government should take a senior stake in support of a coalition of private sector firms, who bear the first tier of losses during bailouts. That approach underlies the successful resolutions of the Paris Bourse in 1882, and London’s Barings crisis in 1890, both of which were implemented at the behest of privately owned financial institutions that pledged their own resources to avoid potential systemic consequences from the failure of a nonbank financial institution. If the creation of some form of resolution authority loophole is desirable, it should be, and can be, crafted in a way to substantially limit the risk of taxpayer loss and the adverse incentive consequences that accompany it. Government can and should rely on the self-interested behavior of market participants to prevent the institutionalization of too big to fail.
Charles W. Calomiris is Henry Kaufman Professor of Financial Institutions at the Columbia University Graduate School of Business and a Professor at Columbia’s School of International and Public Affairs. His research spans several areas, including banking, corporate finance, financial history, monetary economics, and economic development