geithners-fed-reserve-pic.thumbnail.jpgTreasury Secretary Geithner has pushed back his announcement of the details of his plan to stabilize the banks to Tuesday, wishing not to throw another controversy into the middle of Congress’ debate on the stimulus.

The details of Geithner’s plan are as important as the details of the stimulus.  This crisis was precipitated by problems in the financial sector and the contraction of credit makes a real recovery difficult, if not impossible.

According to Bloomberg, Geithner’s plan may involve the following:

The Treasury may increase its stake in lenders that are judged short of capital, the people said on condition of anonymity. Should extra taxpayer funds result in majority ownership by the government, officials would then decide whether to liquidate the institutions, place them into receivership or retire the companies’ assets over time, they said.

Sounds good, but the question here is how much equity they will require for money.  In the past, money given or loaned has been much larger than the required equity and the equity taken has been in preferred non-voting shares.  If Treasury is willing to go dollar for dollar and to take voting shares, then this is a good idea.  At the very least, if the government winds up with 51% ownership then it should insist on having the majority of the board members and using those board members to appoint its own CEO.  That may not be official receivership, but it amounts to nationalization, which is only reasonable if the government has spent more than half the stock value of the company.  

In a second key feature of the plan, the Federal Reserve will likely expand what is now a $200 billion program to revive consumer loans, according to two people briefed on the talks.

Again, not a bad idea, but if the Fed continues only to support the secondary market and rely on banks to make consumer loans when the banks are seeking to reduce their exposure, this may not work very well.  The Fed may well need either to loan directly or to have the FDIC take over a major bank and use that bank to do the loaning.  Another good idea is to rewrite mortgages:

Treasury Secretary Timothy Geithner has told Democratic lawmakers that banks tapping a U.S. rescue fund will be required to modify mortgages and help borrowers avoid foreclosure, according to a person at a briefing.

The problem here is that many problem mortgages were securitized and are not owned by the banks.  Banks can’t modify those loans without permission from the owners of the securities and that permission has proved very difficult to get.  This is the gordian knot of this financial crisis.  Unless Congress is willing to write legislation to allow such modification without owner permission, most problem mortgages cannot have their terms reduced without exposing banks or the government to lawsuits. Then there’s the question of what to do with bad assets on the banks’ books:

Officials are also considering ways to deal with the toxic assets clogging banks’ balance sheets, where the debate has moved away from the creation of a so-called bad bank, which some lawmakers have called too costly. Guaranteeing the securities may offer a cheaper option.

Bad banks weren’t necessarily a bad idea, as long as they were also combined with nationalization, so that taxpayers had all of the upside.  A bad bank without nationalization exposed taxpayers to all the risk and very little of the upside, even if the government did take some form of equity.  If the government takes assets off the banks for more than they’re worth without also taking control, it’s both a bad bank, and a bad idea.

The question with insurance is setting rates.  It is impossible to set rates accurately without doing a proper audit of banks’ assets.  If you do set rates accurately, then the savings for banks are minimal because insurance done right doesn’t save money, it either shifts the burden to a larger group or it shifts the savings in time.  If banks that don’t have a lot of bad assets are forced to pay higher premiums than their own books would require, you spread the pain amongst the entire sector, and that might work.  Otherwise there are no real savings for the banks that need it worse.  (To give an oversimplified example: if I think there’s a 50% chance of default on $100 loan, I have to charge you $50.  Take 10 loans like that and charge $500, and where’s the savings for the bank?)  Insurance works before things go bad, when we don’t know who exactly is going to take the losses and no individual insured is massively more likely to take the loss, or it works if you’re just willing to brute force socialize the risk.

So either Geithner has to charge less for insurance than it should be worth, or he has to make healthy financial institutions pay for weak ones.  And if he doesn’t do real underwriting and auditing of the books of banks, even if he wants to, he won’t be able to set prices accurately.

What this means is that "insuring" losses could very easily turn into an open ended guarantee that might as well just be a gift.

(For more details on how to do insurance and bank valuation right, see this article.)

Overall these ideas aren’t bad ones, but the devil is going to be very much in the details.  Do these things correctly and they could help a great deal, refuse to really take equity for money or to change laws with regards to securitized mortgages and it may do much less than one would hope, at very great expense.