What was so bad about the bailouts? Everything, except that they sort of worked, at least as a short-term patch-up and a bid for time. But that time is running out, and we should now start thinking about the next crisis, and the next - and how to mitigate what cannot be avoided in the post-TARP era.
The really offensive thing about the bailouts was the prevailing sense of adhocracy - that Congress and the White House and the Treasury and the Fed were more or less making things up as they went along. This bank got rescued, that one didn't. This firm got a bailout on generous terms, that one got the pillory. Dick Fuld got vilified, Tim Geithner got made Treasury secretary.
It didn't have to be that way: We have a pretty good system for regulating traditional banks and, when necessary, for taking over failed banks and "resolving" them - taking care of depositors and sorting out losses among creditors and shareholders. The Federal Deposit Insurance Corporation is one of the few players in the recent crisis that have acquitted themselves reasonably well. No American depositor lost a dime from his savings account, checks cleared, and everyone's ATM card kept working. The FDIC works as well as it does because there is not much adhocracy in its approach - terms and practices are defined in advance, and its operations are prefunded through insurance premiums charged to the banks whose deposits it insures.
But we also have a shadow banking system: a menagerie of hedge funds, structured-investment vehicles, non-depository investment banks, and other intermediaries that shuffle money between borrowers, lenders, and investors outside of traditional banks. Before we can get our economy fully un-TARPed, un-Fannied, and un-Freddied, we need an FDIC-style resolution authority that can do for the shadow banking system what the FDIC does for banks: police safety and soundness and, when necessary, take troubled institutions into custody and disassemble them in an orderly manner.
Some free-marketers will protest that such a resolution authority promises to be just another failed federal regulator, that we should "let markets work." But the bailouts have proved beyond any doubt that "too big to fail" is a durable feature of Washington's thinking about finance - the reality is that an immaculate free-market solution is not in the works. It's rather a question of what sort of regulation we are going to have and who is going to be doing it. We don't expect the new resolution authority to be perfect, but if its powers are well defined and reasonably insulated from electoral politics, it could prove as useful as the FDIC at stemming panic and containing spillovers into the real economy.
The new authority probably should be under the jurisdiction of the Federal Reserve, though its activities and the Fed's traditional monetary-policy functions should be walled off from each other. Why the Fed? It has a great deal of financial expertise and knowledge at its disposal, and it is not headed by a cabinet secretary with an eye on the next election. The Fed's haughty independence, for many a source of irritation and suspicion, is in fact its great virtue. It has made its mistakes - keeping interest rates too low for too long, and thereby helping to inflate the housing bubble - but an obsession with short-term politics is not one of them. The FDIC has enough to do, and neither Treasury nor Commerce nor any other cabinet agency should be trusted with the broad powers that any effective resolution authority would have to command.
The institutions that make up the shadow banking system are a diverse and complicated lot: If traditional banking is a game of checkers, this is 3-D chess on dozens of boards at the same time. It is therefore likely that the regulators will lack the expertise to establish appropriate, timely resolution programs for the complex institutions they are expected to govern. The solution to that problem is found in Columbia finance professor Charles Calomiris's proposal that every TBTFI - Too Big to Fail Institution - coming under the new agency's jurisdiction be required to establish and maintain, in advance, its own resolution plan, which would be subject to regulatory approval.
Such a plan - basically, a pre-packaged bankruptcy - would make public detailed information about the distribution of losses in the event of an institutional failure - in other words, who would take how much of a haircut if the bank or fund were to find itself in dire straits. This would be a substantial improvement on the political favor-jockeying that marked the government's intervention in General Motors, for instance, or the political limbo that saw Lehman doing nothing to save itself while waiting to be rescued by a Washington bailout that never came. The authority's main job would be to keep up with the resolution plans and, when necessary, to execute them.
Like the FDIC, the new resolution authority should be prefunded, its day-to-day operations and its trust fund underwritten by insurance premiums charged to the institutions it oversees. This in itself might have a useful dampening effect: Institutions not wishing to fall under the resolution authority's jurisdiction, thereby becoming subject to the expenses and inconvenience associated with it, would have an incentive to moderate the size and complexity of their operations, which would be a good thing in many cases. Unlike TARP, the authority's trust fund should be treated as what it is - capital backing an insurance program - and restricted by statute from being used as a political slush fund. Being funded by the financial institutions themselves, it would not be subject to the whims of congressional appropriators.
Taking a fresh regulatory approach would give us the opportunity to enact some useful reforms at the same time. At present, capital requirements - the amount of equity and other assets financial firms are required to hold in proportion to their lending - are static: X cents in capital for every $1 in, for example, regular mortgage loans. This makes them "pro-cyclical," meaning that, during booms, banks suddenly find themselves awash in capital as their share prices and the value of their assets climb, with the effect that they can secure a lot more loans with the assets they already have on the books. But the requirements are pro-cyclical on the downside, too:
During recessions, declining share and asset prices erode banks' capital base, hamstringing their operations and making financial contractions even worse.
Instead, we should use counter-cyclical capital requirements: During booms, the amount of capital required to back each dollar in lending should increase on a pre-defined schedule, helping to put the brake on financial bubbles and to tamp down irrational exuberance. During downturns, capital requirements should be loosened on a pre-defined schedule, to facilitate lending and to keep banks from going into capital crises for mere accounting reasons. But these counter-cyclical capital requirements should begin from a higher baseline: The shadow banking system exists, in no small part, to skirt traditional capital requirements, and its scanty capital cushions helped make the recent crisis much worse than it had to be.
One other aspect of the FDIC that should be incorporated into the new resolution authority: automatic triggers. The FDIC Improvement Act ensures that the agency has relatively little regulatory discretion: If a bank fails to satisfy certain standards, the FDIC is not only empowered to move in and resolve it, but required to do so. Likewise, the resolution authority should have relatively little leeway in its operations. More than the FDIC, perhaps, due to the variety and complexity of the institutions it will be expected to oversee - but not much more. What is most important is that its rules, processes, and standards be well defined in advance - before the next crisis, and the next opportunity for the ad hoc shenanigans that made TARP the hate totem it is.