http://krugman.blogs.nytimes.com/2010/04/12/failure-is-a-failed-strategy
April 12, 2010 Paul Krugman - New York Times Blog
Failure Is A Failed Strategy
One thing that keeps coming up in comments, both here and on my column,
is the widespread belief that all we need to do on the banking front is
(a) break up the big banks, so that none of them are too big to fail
(b) promise not to bail out any banks in the future. That way, the
claim goes, bankers will know that they will face dire consequences if
they misbehave, and market discipline will do the rest.
Dream on.
There are at least three things crucially wrong with this argument.
First, even when banks can fail, bank managers and/or owners have an
incentive to make risky bets; after all, their downside is limited --
at worst, the bank goes under -- while their upside isn't: if they can
earn high profits for a few years, they can walk away with a lot.
Remember that in the S&L crisis of the 1980s, quite a few people made
out like bandits while running their banks into the ground.
Second, a wave of bank runs that brings down many small banks can do as
much damage as the failure of a few big banks. The biggest banks didn't
fail in 1930-1931, when a generalized run on the system began with the
failure of the 28th largest bank in America; nonetheless, the results
were catastrophic. The idea that we can cheerfully let banks fail as
long as none of them is big is just wrong.
In fact, we know what a system in which banks are allowed to fail looks
like: that's how the US banking system worked before the creation of
the Fed. And you know what? It wasn't a smoothly functioning system,
with sound banking enforced by market discipline; it was a system
periodically wracked by "panics" that destroyed peoples' savings and
plunged the economy into recession.
Finally, because that's what really happens when banks are allowed to
fail freely, promises not to bail out banks in the future aren't
credible. Fail to reform finance now, and there will be two, three,
many TARPs in our future.
What is true is that there are bailouts and then there are bailouts.
What has to be protected in a crisis are bank deposits and things like
bank deposits -- basically, bank-created money. Money market accounts
and "repo" -- very short-term loans in which businesses often park
their funds -- have to be protected to avoid 1930-31-type collapses. On
the other hand, bank shareholders and long-term bondholders can be made
to pay a price without collapsing the system.
Now, in 2008-2009 the shareholders were not cleaned out, and the
bondholders left untouched; in part this was a policy decision, but it
was also influenced by the lack of "resolution authority": there was no
clean, well-established route for seizing complex financial
institutions. We can fix that, and deal with future Citigroups (one of
which, given history, is likely to be ... Citigroup) the way the FDIC
deals with smaller banks: protect the depositors, clean out the
shareholders.
But just letting banks fail isn't going to happen -- nor should it. In
practice, talking about doing so is just an excuse to avoid real
reform.