Jim, I’d agree with you when you agree with Matt Yglesias on the point that the most valuable (long-term, I’d add) benefit that certain large (“systematically significant”) financial institutions got from TARP was investors’ and creditors’ knowledge that the government would stand behind those institutions in the event of a future crisis. As Yglesias says, the banks cannot give this “immensely valuable implicit federal guarantee” back. He adds that the government cannot take it away either. I’m not so sure about the latter, but if the government were to do so, it would have do so explicitly — and very carefully indeed.
Yglesias is right: The guarantee is immensely valuable; it’s also potentially immensely expensive. It’s hard not to think that it shouldn’t come with some cost to those institutions that benefit from it. That cost should primarily be regulatory and designed to reduce (to acceptable levels) the risk that those institutions will again become such a burden (even if only “temporarily”) to the taxpayer.
It’s worth adding, however, that wherever possible this guarantee should be explicit, rather than implicit. One of the root causes of the financial crash was the uncertainty over which financial institutions were (or were not) guaranteed one way or another by the government, and on what terms: This led to the mispricing of risk all round — with consequences we now know all too well.
It’s better to spell out specifically (by name and/or financial criteria) and to what extent what institutions are guaranteed by the government — and what are not. Another advantage of this approach, of course, is that those institutions that are not guaranteed can be subject to far lower levels of regulation. We shouldn’t want to crush innovation — but we should also do our best to make sure that innovation does not crush us.