Dennis Snower works out the arithmetic behind the Public-Private Investment Program and shows something that we’ve suspected: if the assets are really toxic (the gap between book value and long-term expected value is big), the subsidy just isn’t big enough. He also shows that if the assets are only a little toxic, the government subsidy induces private sector bidders to overbid, making the subsidy bigger than it needs to be.

Snower’s hypothetical asset has an expected value of $50. According to his calculations:

• If the bank has it on its books at $70, the private sector will bid it up to $85 because of the government subsidy. The government would have been better off under the original Paulson Plan (just buy it off the bank at book value, in this case $70).
• If the bank has it on its books above $85, the private sector will not buy it at all and the plan will do nothing.

Now, his asset has different characteristics than the assets out there in the real world, whose expected values are not knowable, let alone known. That may change the analysis, but I doubt it changes the ultimate result.

Thanks to the reader who recommended this.