The Fed's Mission Impossible by John H. Cochrane

The Federal Reserve last week announced its new "Enhanced Prudential Standards and Early Remediation Requirements" for big banks, as required by the Dodd-Frank law. You have to pity the poor Fed because it faces an impossible task.

The Fed's proposal opens with an eloquent ode to the evils of too-big-to-fail and moral hazard. And then it spends 168 pages describing exactly how it's going to stop any large financial institution from ever failing again.


More capital is at least a step in the right direction. But the Fed's capital proposals don't go nearly far enough. Putting less than one investor dollar at risk for every 10 borrowed dollars seems laughably low when we're guaranteeing the debts. With a 50-50 chance of a banking tsunami coming across the Atlantic from Europe, you wonder why the Fed is allowing any dividends at all.
But there's nothing here to solve the deeper problems. The last generation of smart MBAs got around capital requirements by pooling risky assets into "AAA" securities that had lower risk weights, and then putting those securities in special-purpose vehicles with off-balance-sheet credit guarantees. Voilà! Same risk, no capital. I can't wait to see what they come up with this time. Diligently following risk weights, European banks built capital ratios by selling good loans and keeping "risk-free" sovereign debt.

Will this new round of rules, and greater discretionary supervision, finally stop too big to fail?

The Fed's proposed "credit limits" are a revealing mess. They seem simple and obvious—big banks can't bet more than 10% of their equity on a single counterparty.
But on second thought, it's not so obvious. This wasn't the problem we had in 2008. Banks didn't fail because they lent to other banks. We had a classic run: Investors pulled money from banks that lost a lot on mortgage-backed securities. Yes, banks take too much risk. But they have no incentive to take stupid undiversified counterparty risk.
Credit limits are not so simple either. Suppose you buy a $100 Bank of America bond. OK, you have $100 at risk, though usually there is some recovery in default. But what if they give you $102 of collateral, yet that collateral might be hard to sell or stuck in court for a while? How does a regulator measure that risk? Or what if loans from A to B are funneled through shell company C using derivatives? Ten percent of equity is less than 1% of assets, and a tiny fraction of gross exposures, so measuring it right will matter a lot.
How does the Fed address these problems? Read the 22 pages of overview with 39 separate explicit questions. Translation: Help! We have no idea how to measure and regulate "credit exposure" for modern banks.
The Fed's proposed "triggers" for "early remediation" are interesting attempts to regulate the Fed, not the banks. The Fed recognizes that last time "while supervisors had the discretion to act more quickly, they did not consistently do so." Triggers will force the machinery to action.
Or will they? You're a regulator facing a bank in trouble. If you label it in trouble, you will start a panic in markets. This is the inherent contradiction—your job is to prop up banks, not cause runs. We'll see.
The Fed goes on to a chilling list of "corporate governance" rules, gems such as: "The covered company's board of directors (or the risk committee) must oversee the covered company's liquidity risk management processes . . . [and] determine whether each line or product has created any unanticipated liquidity risk." Well, duh, isn't that what boards do? Why must this be written into federal regulations, with force and penalty of law?
The Fed's proposal exemplifies what a recent editorial in these pages described as Washington's "badly written bad rules." Everything under the sun gets regulated, with no attempt to measure benefits or costs. Sure, as the Fed make clear, Dodd-Frank is to blame, but it could fight back just a bit.
Big picture time. Is any of this going to work?
For 70 years, our government has sought to stop crises by guaranteeing more and more debts, explicitly with deposit insurance, or informally with predictable too-big-to-fail bailouts. Guaranteeing debts gives obvious incentives to gamble at taxpayer expense, so we try to limit risks with regulation. But big banks still have every incentive to avoid, evade and financial-engineer their way around the rules, and they have lots of lawyers, lobbyists and ex-politicians to pressure regulators to use their wide discretion. The government has lost this arms race time and time again. Will this new round of rules, and greater discretionary supervision, finally stop too big to fail?
The depressing scenario is that the six big banks will use this massive regulation as an anticompetitive fortress. We will have the same six big banks 30 years from now, spurred to even greater size with continuing subsidies, cheap Fed-provided financing, the government guarantee, and occasional bailouts. And a financial system as innovative as the phone company, circa 1965.
The only hope I see is that nimble, new small-enough-to-fail competitors will spring up and rebuild the financial system. But this is faint hope in the face of the vast discretionary powers in last year's Dodd-Frank financial legislation and the Fed's rules, which allow the government to step in whenever they decide that a financial risk is "systemically important."

John H. Cochrane is a professor of finance at the University of Chicago Booth School of Business and an adjunct scholar of the Cato Institute.

More by John H. Cochrane

What is not "systemically important?" How I can I build a new financial company that demonstrably causes no "systemic" danger—and is therefore not subject to the Fed's onslaught of regulation, discretionary supervision and "remediation"? How can I assure my creditors that they will receive the legal protections of bankruptcy court, and not be dragged into some arbitrary and politicized "resolution"?
The Dodd-Frank legislation never defines "systemic" or, more importantly, its absence. Under the law, the Financial Stability Council can just "determine" that any company might have "serious adverse effects on financial stability." They can consider any "factors that the Council deems appropriate." The Fed proposes to subject any company to "other requirements or restrictions" if it thinks existing rules do not "sufficiently mitigate risks to U.S. financial stability."
There is nothing to say that a risk to "financial stability" can't be, for example, taking profits away from the big six, or a failure that takes money away from an influential voting bloc. Don't laugh: Life insurance companies were bailed out in 2009 at least in part so they could keep up payments on guaranteed-return retirement products.
The Fed does not propose any such limit to its powers or describe how it will encourage a financial system free of too-big-to-fail firms. The Fed's report has instead a searching inquiry on how it can expand its powers, and how it can begin "designating" and regulating companies beyond the big banks.
If we are going to get out of the guarantee-regulate-bailout trap, we must legally define what is not too big, and what can, will and must—by absence of legal authority—fail. If the government won't break up too-big-to-fail banks, we must at least allow competition to do it.