In the early phases of the financial crisis, it wasfashionable to argue that the United States’ system of regulation needed afundamental structural overhaul. Differences of opinion between the Securitiesand Exchange Commission (SEC) and the Commodity Futures Trading Commission(CFTC) had obstructedeffective oversightof investment banks and derivativestrading (only the US believes that it makes sense to regulate securities andderivatives separately).
Indeed, the plethora of separate banking regulators had createdopportunities for banks to arbitragethe system in search of a more indulgent approach to capital. Likewise, thelack of a federal insurance regulator had left AIG regulated by the Office ofThrift Supervision (OTS) and the New York State Insurance Department, whichproved to be a wholly inadequate arrangement.
Little has come of these arguments. The Dodd-Frank Act didsucceed in putting the OTS out of its misery, but jealous congressional oversight committees haveprevented a merger of the SEC and CFTC, and nothing has been done torationalize banking supervision. So the American system looks remarkablysimilar to the one that turned a collective blind eye to the rise of fataltensions in the early 2000’s.
One factor that contributed to institutional stasis was the absence ofa persuasive alternative.In the decade or so leading up to the meltdown of 2007-2008, the global trendwas toward regulatory integration. Almost 40 countries had introduced singleregulators, merging all types of oversight into a single all-powerful entity. The movement beganin Scandinavia in the early 1990’s, but the most dramatic change came in 1997,when the United Kingdom introduced its Financial Services Authority (I was itsfirst chairman).
Other countries adopted slightly different models. Afashionable approach was known as “twin peaks,” whereby one regulator handled prudential regulation– setting capital requirements – while another oversaw adherence to business rules. But twinpeaks itself was further subdivided.
The Dutch model brigaded the prudential regulators inside the central bank,while the Australian version was built on a separate institution. Theseintegrated structures seemed to offer many advantages. There were economies ofscale and scope, and financial firms typically like the idea of a one-stop (or, at worst, atwo-stop) shop. A single regulator might also be expected to develop amore coherent view of trends in the financial sector as a whole.
Unfortunately, these benefits did not materialize, or atleast not everywhere. It is hard to argue that the British system performed anymore effectively than the American, so the single-regulator movement hassuffered reputational damage. And the continuing travails of the Dutch banking system – anotherbank was nationalized last month – suggest that it is easy to fall into the gapbetween twin peaks.
The truth is that it is hard to identify any correlationbetween regulatory structure and success in heading off or responding to the financial crisis.Among the single-regulator countries, Singapore and the Scandinavians weresuccessful in dodgingmost of the fatal bullets, while the UK evidently was not. Among the twin peak exponents, the Dutchsystem performed very poorly indeed, while Australian financial regulation maybe considered a success.
Does it matter whether the central bank is directlyinvolved? Many central bankers maintain that the central bank is uniquelyplaced to deal with systemic risks, and that it is essential to carry outmonetary and financial policies in the same institution. Again, it is hard tofind strong empirical support for that argument.
The Dutch and American central banks, with direct oversightof their banking systems, were no more effective in identifying potentiallydangerous systemic issues than were non-central bank regulators elsewhere.Canada is often cited as a country that steered its banks away from trouble,even though they sit uncomfortably close to US markets. But the Bank of Canadais not now, and has never been, a hands-on institutional supervisor. So perhaps the US Congresshas been right to conclude that changing the structure of regulatory bodies isless important than getting the content of regulation right.
Elsewhere, though, a lot of structural change is under way.In the UK, every financial disturbance leads to calls to revamp the system. Therewere major overhauls in 1986, and again in 1997, when the Bank of England lostits banking supervision responsibilities as a delayed response to the collapseof Barings. Nextmonth, it gets them back – and more.
For the first time, the Bank of England will superviseinsurance companies as well. A similar change has been introduced in France,where a new Prudential Control Authority has been created. The Britishand French rarely agree on anything; one is tempted to say that when they do,they are highly likely to be wrong.
It is difficult now to discern a coherent pattern. Certainly, the trendtoward full-service single regulators outside the central bank has slowed to a crawl (though Indonesiais consolidating regulators at present). There is no consensus on the role ofthe central bank: in around a third of countries, it is the dominant player, inanother third it has responsibilities for banks only, while in the remainingthird it is a system overseeronly.
We could see this as a controlled experiment to try toidentify a preferred model. After all, financial systems are not so differentfrom one another, particularly in OECD countries. But there is no sign of aconsidered assessment being prepared, which might at least help countries tomake better-informed choices, even if it did not conclude that one model was unambiguously best. TheG-20, under its current Russian presidency, is now in search of a role. Here isa useful practical task it might take on.