01/25/2013 04:27 pm ET Updated Mar 27, 2013

Who Will Carry the Water?

To make the U.S. financial system safe, we need to end the era of big bailouts. In December, the U.S. and the UK took a significant step in that direction, announcing a common strategy for resolving insolvent cross-border financial giants, known as global systemically important financial intermediaries (G-SIFIs).

In spite of this progress and the Dodd-Frank law notwithstanding, the specter of 'too big to fail' still haunts the financial landscape. It is hard to imagine that a future U.S. Treasury Secretary would risk another Lehman crisis by imposing large losses on a behemoth's creditors, let alone sacrificing several financial giants in a systemic crisis. Such doubts tilt the financial playing field in favor of big intermediaries, which receive a subsidy in the form of lower funding costs (and higher credit ratings) due to this perceived insurance.

Cross-border entanglements make it difficult to resolve a G-SIFI without a crisis. When Lehman failed, it had nearly three thousand legal entities in fifty countries. Such cross-border operations trigger complex interactions between the bankruptcy procedures of multiple countries, encouraging a self-defeating grab race for assets by regulators and counterparties alike. Large, unnecessary -- and potentially crisis-triggering -- losses are likely to result.

The U.S. Federal Deposit Insurance Corporation (FDIC) and the Bank of England (BoE) propose a "single-point-of-entry" approach that cuts through this international mess by intervening only at the top level of a G-SIFI -- call it the Financial Holding Company (FHC) -- and allowing all those subsidiaries that are healthy to continue operating as going concerns. Instead of a liquidity scramble and a fire sale that undermines the supply of credit and economic activity, the government receivers gain time to evaluate the assets of the G-SIFI and apportion losses to its creditors (including possibly senior unsecured debt holders), in contrast to the repeated bailouts of 2008 and 2009.

Acting in this way requires the regulators of the host country (say, Britain for a US G-SIFI operating in London) to trust that the regulators of the home country (in this case, the FDIC) will sustain the firm's operations at a global level, rather than favor domestic creditors at the expense of foreigners. As such, cooperation of the FDIC and the BoE is necessary to make the common resolution approach credible, making it a healthy first step.

However, it is far from sufficient. To persuade us today that a future Treasury Secretary really will pull the plug on an insolvent US G-SIFI, the resolution scheme will have to be widely viewed as nearly foolproof -- that is, virtually free of "Lehman risk." While there are many possible concerns, two big ones stand out: (1) the need for large, non-systemic "loss absorbers" at the FHC; and (2) the need for the capacity to deal (almost) simultaneously with vulnerabilities at multiple G-SIFIs, rather than just one.

Under the single-point-of-entry approach, the FDIC as G-SIFI receiver would allocate all the losses from any subsidiaries to the equity and long-term debt holders at the FHC level. To avoid a bailout, the FHC's equity and long-term debt will have to be large enough to cover all the possible losses. Moreover, the holders of those claims cannot themselves be a potential source of systemic risk, say, like the money market mutual fund that "broke the buck" after taking losses on Lehman debt. It would fool no one -- least of all a future Treasury Secretary -- if (shadow) banks held the FHC liabilities of other (shadow) banks. The spillover risks could still bring down the system.

In effect, the holder of these instruments -- equity or debt -- would be selling rainy-day insurance to the G-SIFI. In a full-fledged systemic crisis, involving a shortfall of capital in the financial sector as a whole, the instrument holder would be selling rainy-day insurance to the entire financial system.

That's a lot of water to carry for only a few potential carriers. Private, non-systemic sources of such insurance probably are limited to a subset of triple-A domestic life insurers, pension funds, select mutual funds, and similar, unleveraged deep pockets. (While foreign investors might see a diversification opportunity, they would naturally fear becoming an easy target for a Treasury Secretary eager to recapitalize a defunct financial system.)

It is questionable whether such private sources could supply sufficient insurance at a price that makes the "single-point-of-entry" scheme workable. For example, the Stern Volatility Lab estimates that the systemic risk of large US intermediaries peaked at about $1 trillion in 2008, only about 20 percent less than the then-prevailing net worth of all U.S. commercial banks.

In theory, a very high price and limited availability of such insurance might compel G-SIFIs to internalize the costs and disassemble themselves into less-systemic parts. That would be a good outcome, but the net effect may be little different from the simpler approach of establishing very high equity capital requirements for G-SIFIs, or even from a direct effort to break them up into non-systemic morsels. It is a very positive step that U.S. and UK authorities have entered a mutually supportive pact, but the worry remains that policymakers who have failed to hike capital requirements sharply or to insist on the breakup of G-SIFIs might also underestimate the necessary scale of FHC loss absorbency, thereby undermining the credibility of the resolution scheme.

Perhaps over time, the FDIC and the public can get a handle on the size of the loss absorbers at the FHC that would be necessary to make its new resolution approach credible. The annual living wills mandated by Dodd-Frank -- the roadmap written by each G-SIFI for safely disassembling it when approaching insolvency -- might be helpful in this regard. But the living wills do not take account of systemic spillovers, say, if a G-SIFI failure occurs during a widespread shortfall of capital. That puts the burden of loss analysis squarely on the regulators.

Turning the Dodd-Frank law into a credible resolution mechanism for G-SIFIs is still a work in progress. The FDIC (and their BoE colleagues) deserve much credit for advancing this discussion in a way that underscores the scale of safe loss-absorption capacity that is needed. But until we know where that capacity would come from, and at what price, we are still a long way from ending too big to fail in a credible fashion.