In the March 8th New Yorker, John Cassidy has a long piece about Tim Geithner.  While presenting various points of view, it gives a lot of space, fairly persuasively, to the view that Geithner and other adminstration officials like Lawrence Summers' policy of not taking over shaky financial institutions was the right way to handle the financial crisis.  The "stress tests", forcing those institutions whose financial position was deemed risky to raise new capital, is said to have been the key component.

Some key points, which I'm not expert enough to evaluate decisively off the cuff, but which need to be singled out, are:

1) Takeovers and restructuring would have imposed huge costs on the economy, outweighing the potential benefits.  Cassidy does cites some who claim that the bailed-out institutions are still "wounded" and slow to lend, slowing the recovery ("wounded" is Bob Kuttner's term; Paul Krugman was using the term "zombie banks" back when he was advocating takeovers), but counterbalancing this are the metaphors like "amputating a limb" (from Lawrence Summers) for the takeover-and-restructuring process.  A detailed cost/benefit analysis is missing here, and would be nice to see.

2) Raghuram Rajam of the University of Chicago's business school is appealed to for the point that bondholders needed to take a hit.  It's claimed that while equityholders would have been wiped out in a takeover, bondholders might not have been hit hard enough.  Why this is isn't fleshed out, but I guess it's because in a bankruptcy proceeding, which probably would have been the legal framework for takeovers, there's a definite order of precedence for creditors, with bondholders coming before equityholders (who are generally last).  However, in a restructuring-style bankruptcy, where the aim might be not to liquidate, but to get an enterprise running again, it's not clear to me that such absolute precedence would hold, especially when the government might have some pretty strong powers of persuasion over some of the bondholders (many of whom were presumably in difficult straits themselves).  Perhaps this mainly boils down to the possibility of protracted legal battles...a "special feature" of our system that might make this option more difficult than in other countries where it's succeeded?

As Cassidy says, "Bondholders are supposed to monitor risk-taking at firms they lend to. If they know they can rely on a government bailout, they have little incentive to do so. "  Just what hit the bondholders took under the bailout as it unfolded, isn't clear in the article, though.

3) Cassidy also recognizes that takeovers would have had the advantage of punishing the executives and shareholders of some of these institutions.  A big question is whether they got off too lightly.  That seems to be the public perception.  It is true that existing shareholders may have been diluted when companies were forced to raise capital under the stress test, if this was done by issuing new stock.  But if they didn't pay a high enough price, the take-home lesson for future financial managers is that this kind of mismanagement is profitable for them.  This is a recipe for repetition, unless strengthened financial regulation can help prevent managers from taking these opportunities to profit by risking the stability of the financial system, next time they present themselves.

In light of all this, one should note that Mike Konczal, guest-blogging on Ezra Klein's blog at the Washington Post, claims that Geithner is arguing hard to keep explicit caps on leverage out of the financial reform bill.   He quotes a January letter from Geithner to Rep. Keith Ellison (cc: The Honorable Ben S. Bernanke) opposing fixed numerical caps.  Konczal's a bit over the top in the spin he puts on some of what he quotes.  Geithner's point that "The statutory leverage constraint and detailed statutory risk-based capital requirements for Fannie Mae and Freddie Mac proved to be inadequate to the task of ensuring the safety and soundness of the firms." should hardly be paraphrased as "If you put this in the bill you will be responsible for another Fannie and Freddie"---rather, a good point that *even* numerical caps can't do the whole job in the absence of a serious, committed, regulatory authority analyzing the actual situation.  But the point that statutory limits can help keep things from getting out of control, even if they're not a guarantee, is a good one.  It puts some weight on the right side of the scales when irrational exuberance about leverage is sweeping the financial markets.