Bloomberg has a new story out comparing the investment terms achieved by TARP with those achieved by Warren Buffett when he invested $5 billion in Goldman back in September. The results aren't pretty for the U.S. taxpayer: the government received warrants worth $13.8 billion in connection with its 25 largest equity injections; under the terms Buffett got from Goldman, those warrants would be worth $130.8 billion. (The calculations were done using the Black-Scholes option pricing formula, which has its critics, but which I think is still a good way of estimating the relative difference between similar options.) That's on top of the fact that TARP is getting a lower interest rate (5%) on its preferred stock investments than is Buffett (10%), which costs taxpayers $48 billion in aggregate over 5 years, according to Bloomberg. The difference in the value of the warrants themselves is due to two factors: (1) Treasury got warrants for a much smaller percentage of the initial investment amount; and (2) those warrants are at a higher strike price - the average price over the 20 days prior to investment, while Buffett got a discount to market price on the date of investment.

The comparison isn't a new one - we recommended that TARP emulate Buffett back in October - but Bloomberg's analysis has put the performance gap in striking perspective. Simon has a quote in the article, using the word "egregious," but the really harsh words came from Nobel prize-winner economist Joseph Stiglitz, who said, "Paulson said he had to make it attractive to banks, which is code for 'I'm going to give money away,'" and "If Paulson was still an employee of Goldman Sachs and he'd done this deal, he would have been fired."

Now, to be fair, there are some plausible defenses of TARP. One is that on that fateful October day when Henry Paulson summoned the CEOs of nine major banks to Washington, he needed all of them to accept the deal on the spot, so the terms could not be too punitive. While that may be the case, it doesn't explain why bailouts since then have to be equally generous (since the program is optional, after all) - culminating in the GMAC bailout, where the "warrant" is just the option for the government to lend $250 million more at a slightly higher interest rate. Another defense I have heard is that the plan needed to leave the banks in a situation where they could attract private capital. I have only limited sympathy for this defense, because it's not as if private equity funds are lining up to invest in Citigroup (or any other major bank), even after two rounds of generous bailouts. Finally, there is the oft-repeated mantra that the country doesn't want the government to nationalize banks, and larger warrants would lead to effective government ownership. Here, I think that the clever minds in Washington could come up with a trust-like structure to shield day-to-day operations from too much government meddling (some oversight is arguably a good thing anyway), and a concrete plan for divesting those ownership stakes would go a long way to defusing any worries about creeping socialism.

On balance, I think it's hard to argue that TARP needed to be as generous to banks and their shareholders as it has been. Broadly speaking, TARP recipients have fallen into two categories: those who didn't need the capital but took it because the terms were good, and those who really needed it (like GMAC). If the terms were tougher, the former might not have taken them, but that would be fine; the latter still would have taken the money, because the government was the only place they could get it.

So the question remains: why did Henry Paulson, former CEO of the most respected investment bank on the planet, strike such bad deals for the American taxpayer? I don't know the man, but I strongly doubt that it was because of any conscious desire to enrich his colleagues. More likely, I suspect it was an unconcious product of the conventional wisdom, so strongly rooted these last twenty years, that government involvement is bad and should be minimized at all costs - even to the point of avoiding any possibility that the taxpayer might make money in dealings with private-company shareholders.