Jeremy Stein offers a reasonable technocratic approach, consisting largely of higher capital requirements for large banks. My problem with this is that any time you have a regime in which government guarantees institutions while trying to protect taxpayers with regulations, the regulatory regime degrades over time. Banks figure out how to game the system. And when they can't game the system, they use political muscle to fix it in their favor. Along those lines, Lynn Stout writes,
Politicians and regulators may act as if they're thinking about reining in giant banks, but this seems to be just a political head-fake. From their perspective, actually breaking up the megabanks would only make fundraising and job-seeking harder. It's much more difficult and time-consuming to raise $5,000 from 20 small banks, than it is to raise $100,000 from one enormous bank. And big banks can't hide. Because they clearly suffer most from threatened regulation, a big bank can't afford to refuse to take a senator's call.
Politicians and big banks have a symbiotic relationship. Sad, but true.
Phasing out deposit insurance as domestic deposits grow beyond a certain size would be far more effective in reducing unnecessary size, and far easier to implement, than blanket size limits.
Sounds like a nifty idea, but how would it work, exactly? Would a megabank only be allowed to insure some of its depositors? Which ones? How would a depositor know if he is insured or not? Could a previously-insured deposit become uninsured if it gets merged into a big bank or if the bank grows too large?
In the end, how could the government possibly allow a small depositor to lose money at a big bank?
I think that one's view of big banks ought to be informed by strong public-choice cynicism. Lots of technocratic solutions work in theory, but in my view none of them will work in practice.