On Tuesday and Wednesday the Federal Reserve Bank of Atlanta hosted its annual Financial Markets Conference, titled this year Up From the Ashes: The Financial System After the Crisis. Much of the first day was devoted to conversations about rating agencies and their role in the economy, for better and worse. The second day was absorbed by the issues of too-big-to-fail, macroprudential regulation, and regulatory reform.
One theme that ran throughout the second day’s conversations related to the two aspects of regulatory reform highlighted by Chairman Bernanke in his recent congressional testimony on lessons from the failure of Lehman Brothers (LEHMQ):
“The Lehman failure provides at least two important lessons. First, we must eliminate the gaps in our financial regulatory framework that allow large, complex, interconnected firms like Lehman to operate without robust consolidated supervision… Second, to avoid having to choose in the future between bailing out a failing, systemically critical firm or allowing its disorderly bankruptcy, we need a new resolution regime, analogous to that already established for failing banks.”
Though those two aspects of reform are in no way mutually exclusive, there is, I think, a tendency to lean to one or the other as the first most important contributor to avoiding a repeat of our recent travails. To put it in slightly different terms, there are those that would place greatest emphasis on reducing the probability of systemically important failures and those that would put greatest emphasis on containing the damage when a systemically important failure occurs.
“…the best chance for durable reform is to start with the assumption that failure will happen and construct a strategy for dealing with it when it does…
“In a world with the capacity for rapid innovation, rule-writers have a tendency to perpetually fight the last war…
“I am not arguing that … the ‘Volcker rule,’ derivative exchanges, trading restrictions, or any of the specific regulatory reform proposals in play are necessarily bad ideas. I am arguing that we should assume that, no matter what proposed safeguards are put in place, failure of some systemically important institution will ultimately occur—somewhere, somehow. And that means priority has to be given to the development of resolution procedures for institutions that are otherwise too big to fail.”
At our conference this week, University of Florida professor Mark Flannery expressed concerns that, placed in an international context, a truly robust resolution process for failed institutions may be tough to construct:
“In principle, a non-bankruptcy reorganization channel for SIFIs [systemically important financial institutions] makes a lot of sense. But the complexity of SIFIs’ organizational structures introduces some serious problems. Not only do SIFIs operate with a bewildering array of subsidiaries… but they generally operate in many countries. Without very close coordination of resolution decisions across jurisdictions, a U.S. government reorganization would likely set off a scramble for assets of the sort that bankruptcy is meant to avoid. Rapid asset sales could generate downward price spirals… with systemically detrimental effects. Second, supervisors would have to assure that SIFIs maintain the proper sort and quantity of haircut-able liabilities outstanding. Once a firm has been identified as systemically important, this may be a relatively straightforward requirement to impose, but there remains the danger that ‘shadow’ institutions will become systemically important, before they are properly regulated. (This is not a danger unique to the question of resolution.)
“I conclude that the international coordination required to make prompt resolution feasible for SIFIs is a long way off, if it can be achieved at all.”
Not an encouraging note, and the point is very well taken. Flannery concludes that we would be better served by focusing on changes that lie on the “avoiding failure” end of the reform spectrum: standardized derivative contracts, tying supervisory oversight to objective market-based metrics on the health of SIFIs, limitations on risky activities, and higher capital standards.
As I noted above, I am certainly not hostile to these ideas, and the answer to the question “should reform strategies be rules-based or resolution-based?” is surely “all of the above.” But even if it will take a long time to develop better resolution procedures to address the types of problems that emerged in the past several years, I strongly argue that development of such procedures are necessary for the long-term, and work on these procedures should begin. And here, I have a relatively modest proposal, returning to my remarks:
“…there is a pretty obvious way to vet proposals that are offered. We have a couple of real-world case studies—Bear Stearns, Lehman, AIG. One test for any proposed resolution process would be to illustrate how that plan would have been implemented in each of those cases. This set of experiments can’t be started too soon, and I think should move it to the top of our reform priorities.”
Whether it be the specific provisions of reform bills winding their way through Congress or the “living will” idea championed this week by the Federal Deposit Insurance Corporation, I think we would do well to let the stress testing of those proposals begin.