Suppose a bank is “too big to fail”. It’s got some bad assets that the government wants to buy so the bank becomes liquid again and people are willing to lend to it. But the assets come in different qualities and the government would like to buy them at different prices to minimize the loss to the taxpayer. It might want to pay a low price for the really bad stuff and a medium price for the medium bad stuff etc. If it knew the quality of the assets , no problem – you can just pay different prices for different qualities. But if the bank knows the quality it would try to palm off bad asset as a medium quality asset to get the better price. The standard solution to this is to use inefficiency to “screen” different types of assets. For example, the government says it is willing to buy a lot at the low price but less at the high price.
You have to set the quantity traded carefully so there is no incentive to sell the bad assets at the medium price as the amount the government would buy is too small to make it worth it.
All well and good it seems but remember this bank is “too big to fail”. So, here’s what can happen: The bank sells bad assets to the government pretending they are medium assets. It keeps the bad assets the government does not buy. If they tank, guess what, as it’s too big to fail, the bank can dump the assets on the government anyway. So, in the end, this scheme does not work, the government ends up buying medium and bad assets at the medium asset price.
I haven’t worked this out, but it seems to be what when banks are too big to fail this is what’s going to happen whatever you try to do: you end up paying high prices for bad stuff and there’s nothing you can do about it. (Morally speaking, I’m replicating an old argument of Dewatripont and Maskin’s on the “soft budget constraint.”) The banks are happy as they get lots of surplus and the taxpayers pay a high price for liquidity. The fact that the government has a social motive, saving the financial system, makes it impossible to eliminate adverse selection.
One solution might be just to find out the quality of the stuff you’re buying directly by auditing the asset value carefully. Of course, you might have to rely on the bank for information and then they manipulate it and we’re back where we started.
What then is fair to taxpayers and saves the financial system? Some equity ownership in the banks for the taxpayer. Otherwise, all the surplus goes to the banks.

4 comments
Comments feed for this article
March 23, 2009 at 7:34 pm
dw
i am sure they would try and probably succeed too! until some body starting checking the numbers and data, probably about the time the government tries to sell or is. then depending on how the economy is going at that time, either it gets swept under the rug or some one ends up going to jail
March 24, 2009 at 3:24 am
Le plan Geithner de rachat des actifs “toxiques” : quelques liens « Rationalité Limitée
[…] restent sceptiques. Voir notamment ce que raconte Paul Krugman (ici, ici, là, ou encore là). Sandeep Baliga souligne quant à lui le fait qu’il existe, quoiqu’il arrive, un problème irréductible de […]
April 1, 2009 at 2:11 am
LE PLAN GEITHNER EXPLIQUÉ EN QUELQUES LIENS… « Libertes & Internets
[…] restent sceptiques. Voir notamment ce que raconte Paul Krugman (ici, ici, là, ou encore là). Sandeep Baliga souligne quant à lui le fait qu’il existe, quoiqu’il arrive, un problème irréductible de sélection […]
March 21, 2014 at 3:46 am
Abdus
1 million per pesorn is 300 trillion dollars of raw, naked hyperinflation. Infusing more money into the economy is not the goal and it will fix nothing. You have to get rid of the fraud and that means prosecution on a massive scale. You have to let the malinvestments collapse, prices and wages correct like you did in 1920-21(which was much steeper than the great depression) and like you refused to do during the great depression(which dragged the depression on until 1946).