Economists' Forum (Laurence Kotlikoff, professor of economics at Boston University, and Jeffrey Sachs, professor of economics at Columbia University and director of the Earth Institute): The Geithner-and-Summers Plan (Gasp) to buy toxic assets from the banks is rightly scorned as an unnecessary give-away by virtually every independent economist who has looked at it. Its only friends are the Wall Street firms it is designed to bail out. In an article in the FT, one of us (Sachs, March 23) described the systematic overbidding entailed by the proposal. Others have since made similar calculations, including Joseph Stiglitz and Peyton Young.

The situation is even worse than it looks, however, since the Gasp can be gamed by the banks that own the toxic assets to boost the purchase prices for their bad assets even higher than has been suggested to date.

Suppose that Citibank holds $1bn face value of toxic assets that will pay $1bn with 20 per cent probability and $200m with 80 per cent. The market value is $360m. The Gasp calls on investors to establish a public/private investment fund (PPIF) to bid for the toxic assets. For each $1 that a private investor brings in equity to the PPIF, the Treasury will put in another $1, and then the Federal Deposit Insurance Corporation (FDIC) will leverage the $2 in equity with $12 of non-recourse loans (6-to-1 leverage).

It's easy to show that a risk-neutral and arm's-length PPIF will bid $636m, financed with an FDIC loan of $545m, Treasury equity of $45m and private equity of $45m. (The expected profit to the private investor is a half of 20 per cent of $1bn minus $545m, or $45m. The private investor therefore has a net expected profit of zero.) The PPIF overpays by $276m, which equals the expected loss to the Treasury. The ultimate beneficiaries are Citibank's shareholders and bondholders, whose net worth rises by $276m at the taxpayers' expense. But the outcome could be even more outrageous than this. Citibank can arrange to receive even more than $636m for its assets by setting up its own Citibank PPIF (CPPIF) to bid for its bad assets. The CPPIF will bid the full $1bn in face value for its own toxic assets!

To see this, note that on a bid of $1bn by the CPPIF, Citibank would finance $71m in equity of the CPPIF, the Treasury would add another $71m in equity, and the FDIC would add $857m in loans to the CPPIF. The CPPIF will either break even (20 per cent of the time), or go bankrupt (80 per cent of the time). The CPPIF is therefore a wash-out - with no chance of profits, yet also zero liability.

On the other hand, Citibank gets a sure boost of $1bn minus $360m, or $640m in net worth, for which it pays $71m. Citibank's gain from the CPPIF's overbidding is $569m, which exactly equals the taxpayer's expected loss that is incurred by the FDIC loan and Treasury equity. The real icing on the cake is that Citibank still ends up owning the toxic assets even after the assets are "auctioned", but this time in an off-balance-sheet structured investment vehicle called the CPPIF. The toxic assets revert to the FDIC when the CPPIF goes bankrupt.

It's possible that some fine print of the Gasp would try to preclude explicit hyper-self-dealing of the type just described. But when there is free money on the ground, Wall Street will figure out ways to pick it up. For example, Citibank could arrange to overpay Bank of America for some unrelated securities in exchange for having Bank of America do its bidding at the auction. Indeed, Citibank, Bank of America and other toxic asset owners might join together in a consortium to finance an "arm's-length" PPIF on favourable terms, with the proviso that the PPIF bid for the toxic assets of the consortium. BusinessWeek has reported that "administration officials confirm Treasury may allow such seller financing".

The sad part of all this is that there are excellent alternatives to the Gasp that are vastly more transparent and cheaper for the taxpayers.

The best of these involves separating a weak bank such as Citibank into a "Good Citibank" that holds Citibank's good assets and its deposits, and a "Bad Citibank" that holds the toxic assets, the bondholder debt and the shares of the Good Citibank. The Good Citibank returns quickly to normal business, while the Bad Citibank is eventually liquidated under bankruptcy, with the bondholders and other uninsured claimants getting partial repayments depending on their priority under bankruptcy. The best description of this approach is by Jeremy Bulow and Paul Klemperer.

Over time, we should consider more fundamental reforms, including the idea of establishing limited purpose banking, in which the liquidity services provided by banks are undertaken by institutions with 100 per cent reserve requirements and which, therefore, are immune from runs, panics and reckless gambles. It would be absurd and self-defeating to bear the enormous social costs of the current financial crisis, only to return to the same kind of flawed banking institutions that got us into this mess.

The Geithner-and-Summers Plan should be scrapped. Mr Obama should ask his advisers to canvass the economics and legal community to hear the much better ideas that are in wide circulation.

www.ft.com/economistsforum