The Financial Stability Oversight Council is about to decide which nonbanks with assets of $50 billion or more to designate as systemically important. It should use this authority extremely sparingly because it is based on the flawed premise of connectedness.
The Lehman Brothers failure is instructive. Lehman's bankruptcy did not set off a chain reaction of failures of connected financial firms (those with direct exposure to one another). What it did set off was a contagious run on a wide spectrum of financial firms, with no significant exposure to Lehman, because short-term creditors were concerned that if Lehman could fail, other firms might well be next. It was 1933 revisited, as nonbank firms' short-term creditors were not insured.
Similarly, there was no legitimate basis for concern that the failure of AIG would cause the failure of its derivative counterparties, since the exposure of these parties was limited to a reasonable percentage of capital—and that exposure was protected by collateral and hedges, for example by credit default swaps written on AIG.
Nevertheless, the Dodd-Frank Act entrusted the FSOC with the responsibility of designating nonbanks as systemically important nonbank financial institutions subject to enhanced supervision by the Federal Reserve, most likely including some form of higher capital requirements.
Dodd-Frank also stripped away crucial tools for dealing with contagious runs on financial institutions, like emergency Federal Reserve lending (the central bank now needs Treasury approval), and public insurance and guarantees (which was increased and extended to the nonbank sector during the crisis, for a risk-based fee). These tools—coupled with central clearing and new policies designed to decrease banks' reliance on short-term funding—would create an environment where banks and nonbanks, be they large or small, could with very limited exceptions be permitted to fail. Not only would there be no need for SIFI designation, there would also be no need to break up or limit the activities of large banks. Neither would there be a need for onerous capital and liquidity requirements.
The only legitimate remaining concern would be TCTF—too-crucial-to-fail, where a financial institution played a critical role in the financial system, such as through clearing or valuation, and such function could neither continue in bankruptcy nor be quickly transferred to another institution. The "living wills" requirement of Dodd-Frank can address this problem.
Even if the FSOC were to focus on the threat of connectedness, it is hard to see how any threat can come from firms that are not, in fact, exposed to the failure of other financial institutions. Take a traditional insurance company that has no significant derivatives exposure, as opposed to the AIG holding company, which did in 2008. A traditional insurance company does not have significant exposure to any particular counterparty, since its credit exposure is to a wide range of policyholders making premium payments. Furthermore, a traditional insurance company is not exposed to contagion because its funding is not short-term. There cannot be a run on traditional insurance companies. In short, they pose no systemic risk.
Not only is SIFI designation unnecessary to prevent systemic risk, its use may on the contrary prove harmful. First, the designation of nonbank SIFIs has the real potential to create a funding advantage for the designees, since the market may assume that the government regards SIFIs as too big to fail.
Such expectations also create a moral hazard problem. Of course, mere designation as a SIFI does not guarantee public rescue, or even that such an institution would be resolved through the new Orderly Liquidation Authority created by Dodd-Frank, under which Treasury funds are made available to render make a resolution smoother than might otherwise be the case. Nevertheless, the market may well regard SIFI designation as reducing creditor exposure, thus creating a significant funding advantage.