1. Given the existence of a lender of last resort, *everyone* favors capital requirements of some kind, at least if you prefer to close down bankrupt institutions rather than to let them continue borrowing from the Fed and gambling with the money.
2. It makes sense for some of these capital requirements to be upfront and clear ex ante, noting that any attempt in this direction will be imperfectly realized.
3. Most observers overestimate how effective capital requirements will be. Many a crisis has happened right under their nose and that includes Japan and Bear Stearns, among others.
4. As Megan suggests, regulators are not very good at outguessing the market; read Arnold Kling as well.
5. Regulation can discourage leverage at the margin; the social costs of leverage are higher than the private costs, due to contagion effects and the inability of the lender of last resort to precommit to no bailouts. This need not require selective regulatory intervention or regulatory outguessing of the market.
6. A good deal of regulation requires voluntary compliance on the part of the financial institutions rather than foresight or intelligence on the part of bureaucrats. Banks of course don’t *have* to follow the rules they are given and sometimes they don’t. But for a bank to cross the line of systematic legal disobedience, observable by potential whistle-blowers and the like, is a big step and many of them are not willing to do this. Much of the power of financial regulation derives from this fact rather than from any particular abilities of regulators.
7. We should make capital requirements as symmetric as possible across different kinds of banking activities, at least to the extent systemic risk is present. That said, this still probably won’t prevent the next bubble and blow-up. It will limit some of the damage ex post.
8. It would be nice to have greater use of clearinghouses and netting of positions for the unregulated "shadow banking sector."