Systemic Risk and Resolution Issues

Share

Chairman Frank, Ranking Member Bachus, and members of thecommittee: I am grateful for the opportunity to talk to you today.My name is John Cochrane, and I am a professor of finance at theUniversity of Chicago Booth School of Business. I am here only inthat capacity. I represent no firm, industry, organization, orparty.

I salute you for taking on these difficult issues, which arevital to the economic health of our nation.

The big picture

We are in a cycle of ever larger risk-taking, punctuated by everlarger failures and ever larger bailouts. This cycle cannot goon.

First, we cannot afford it. This crisis strained ourgovernment's borrowing ability. There remains worry of a flightfrom the dollar, and default through inflation. We will probablyescape that fate, but the next, bigger crisis may be beyond evenour government's prodigious resources. That would be acalamity.

Second, the bailout cycle is making the financial system muchmore fragile. In a crisis, it is forgivable to stem the tide todayand worry about moral hazard tomorrow. But now it is tomorrow, andunless we deal aggressively with moral hazard, the next tide willbe a tsunami.

Financial market participants expect what they have seen andbeen told: no large institution will be allowed to fail. They arereacting predictably. Banks are becoming bigger, more global, moreintegrated, more "systemic," more "interconnected" and more opaque.They want regulators to fear bankruptcy as much as possible, andthey will succeed. These actions will make the system less stable,not more so.

We need the exact opposite. We need Wall Street to reconstructthe financial system so that as much of it as possible can fail,with pain to the interested parties, but not to the system. Ourtask is to write the rules so they do it.

Policy

There are two competing visions of policy to achieve this goal.In the first, large integrated financial institutions will beallowed to continue pretty much as they are, with the implicit orexplicit guarantee that they will not fail, but with the hope thathigher capital standards and more aggressive supervision willcontain their risks and forestall failure.

In the second, we think carefully about the minimal set ofactivities that cannot be allowed to fail and must be guaranteed.We commit not to bail out the rest. Private parties needto prepare for their failure, and monintor and discipline theircounterparties to avoid it. We fix, where possible, whateverproblems with bankruptcy law cause regulators to fear it.

I think the second approach is more likely to work. Thefinancial and legal engineering used to avoid regulation andcapital controls last time were child's play.

Powerful authority is attractive ex-post to mop up. Alas, itsets up incentives which makes the system more fragile in the firstplace. Too large to fail must become too large to exist.

Resolution authority

The issue of a "resolution authority" is on your minds so let memake a few comments.

A "resolution authority" offers some advantages. Currentlyregulators feel they must bail out creditors to keep them fromexercising their claims in bankruptcy court. A resolution authorityallows the government to impose some of the economic effects offailure - shareholders lose their equity, debt holders lose valueand become the new equity holders - without actual bankruptcy.

However, nothing comes without a price. First, much of the"systemic effect" regulators seem to fear is exactly thatcounterparties will lose money, or that subsidiary contractualclaims (such as CDS contracts) will be triggered. So, it's notobvious that regulators will use this most important provision.Second, the reason people buy debt in the first place is that theyknow they can seize assets in the event of default. If theauthority is tough with creditors, firms will substitute to shortterm debt and other "runnable" liabilities, making the system lessstable.

The FDIC is a useful model, for its limitations as wellas the rights. A successful resolution authority, whichdoes not just morph into a huge piggybank for Wall Street losses,needs similar strong and clear limitations.

  • The FDIC applies only to banks. We know who they are - and thatthe FDIC cannot "resolve" anything else.

A "resolution authority" must come with a similar clear-cutstatement of who is subject to its authority - and most importantlywho is not, and therefore must be allowed to fail, yes, wereally mean it this time! If, as in the Administration's proposal,resolution authority applies to bank holding companies, we must allclearly understand that does not mean investment banks,hedge funds, insurance companies, and automobile manufacturers -and no last-minute change of legal status either, please.

  • Deposit insurance and FDIC resolution comes with seriousrestriction of activities. An FDIC insured bank can't run aninternal hedge fund.

    Institutions, or parts of them, that are eligible for"resolution" and consequent government resources must be limited totheir "systemic" activities as much as possible.

  • Deposit insurance and FDIC resolution address a clearly defined"systemic" problem.

Bank deposits are prone to runs, because they have a fixed value(unlike, say, mutual fund shares), and they are redeemed on afirst-come first-served basis. Deposit insurance stops runs butputs the government at risk. FDIC supervision and resolution is thesensible covenant of the most senior debt-holder.

A "resolution authority" must similarly be clearly aimed atspecific, defined, and understood "systemic" problems.

In the Administration's proposal, the legality of asystemic determination is admirably clear, but not thegrounds. All the Secretary and President have to do isannounce their opinion that "the failure of the bank holdingcompany would have serious adverse effects on financial stabilityor economic conditions in the United States." This is an invitationto panic, frantic lobbying, and gamesmanship to make one's failureas costly as possible.

Of course, this obscurity is not a new problem. As an ardentobserver of events in the last tumultuous year, I have heard manydeclarations of imminent systemic risk, but never any clearexplanations.

Fix bankruptcy

This last point is the most important. Before designing aregulatory regime, we have to ask, what problem is it that weare trying to fix, anyway? Once stated, what is the best wayto fix this problem?

Regulators fear "systemic" effects of bankruptcy, but if you askwhat they are, you typically find technical problems that arereadily solved. Some examples:

  • Lehman and Bear Stearns both experienced runs on theirbrokerage businesses. If you own stocks in a brokerage account,there is no more reason you should have to go to bankruptcy courtto get them - or pull them out in a panic ahead of time - than youshould need to go to court to get your car out of the repair shopif the auto dealer fails. Putting a "ring fence" around brokerageaccounts in bankruptcy, or otherwise separating "systemic"brokerage from risk-taking, solves this problem, removing theincentive to run.
  • Many investors found collateral tied up in foreign bankruptcycourts. Others, knowing this problem, "ran," refusing to renewshort term debt even against good collateral. This is easy to fix.Collateral is collateral, it's yours if the other sidedefaults!
  • Money market funds holding Lehman debt suffered a run, sincethey promise steady $1 value. There is no reason money market fundscan't seamlessly trade at net asset value any time the value fallsbelow $1, removing entirely the incentive to run. Money marketfunds are not mom-and-pop bank accounts.

In fact, one healthy effect of Lehman's failure is thatfinancial market participants are already addressing theseproblems, demanding greater soundness of prime-brokerrelationships, clearer treatment of collateral, and rewritingmoney-market fund accounting rules. I don't mean to make light ofthe substantial legal problems. But fixing them will cost a lotless than the hundreds of billions of dollars we are throwingaround in bailouts.

Perhaps then the fear is that losses in bankruptcy will lead tothe failure of other "systemic" institutions down the chain. Butlosses in credit markets are small compared to the losses that thefinancial system absorbs easily in stock markets every day. In anycase, the right answer is to protect the systemically importantactivity downstream, not to bail out losers to restore theappearance of solvency.

Why then is Lehman's failure perceived to be such a problem? Themajor complaint, and the only persuasive argument, ispsychological, not technical: Markets expected the government tobail everybody out. Lehaman's failure made them reconsider whetherthe government would bail out Citigroup. If everyone expects thegovernment to bail out, it has to do so to avoid a panic.

Needless to say, the right answer to this problem is to limitand clearly define, rather than expand and leave vague, thepresumption that everyone will be bailed out.

Bottom line

The major systemic problem last fall was the freezing ofshort-term debt markets. There are many classic remedies to thisproblem, including limits on how much systemic activity can besupported by rolling over short-term debt, (the FDIC won't let abank finance a loan portfolio with overnight debt!) intervention bythe Fed as lender of last resort, and removing uncertainty aboutgovernment action.

We should focus on this question. A broad guarantee that nofinancial institution can fail is not the answer.

John H. Cochrane

Committee on Financial Services
United States House of Representatives