These days there are so many sentences to ponder

If you’re running an insolvent bank, and you get a slug of equity from
Treasury, your shareholders will thank you if you use that equity to
take some very large risks. If they pay off and you make lots of money,
then their shares are really worth something; if they fail and you lose
even more money, well, there was never really any money for them to
begin with anyway.

That’s Felix Salmon: read the whole thing.  Read this too.  Here is Megan McArdle on the pooling equilibrium.  Here is a good article on how Paulson "sold" his plan to the bankers.  And here are yet some more sentences to ponder:

So it in the end, we have what is basically an economic loan, but structured
in a way to game bank capital adequacy requirements. What strange times we live
in when Treasury and the Fed have to engineer a deal to circumvent their own


In Preferred Stock=Subordinated Debt, by Adam Levitin, he asks "And how fast do you think Goldman will use 5% Treasury dollars to buy back Buffett's 10% stock...?"

I would imagine that one option for the banks that don't really need this capital injection is to signal their strength by giving it back. I understand that dividend increases are not possible under the terms, but what about share buy-backs?

I.e. could Wells Fargo, who believe that they don't need the cash to strengthen their capital ratios, and whose share price fell when the news came out, use all or part of the money to buy back their (now undervalued?) shares?

Would not this send the same signal as not accepting the money in the first place?

Sounds like Majumdar and Radner, Linear models of economic survival under production uncertainty, first article in Economic Theory, 1991. Basically, investors considering two "gambles" with the same mean may opt for the one with HIGHER variance if they will be "ruined" shortly. In a sense, think of two lotteries, 50/50 of $1/-$1 and 50/50 of $100,000/-$100,000. If you have little money this period, and you need a decent bit to survive, no need to take the 50/50 bet of $1/-$1 because you will be "ruined" next period any way. May as well take the 50/50 bet of $100K/-$100K.

Noneconomist here, so forgive any naivete, but doesn't the Salmon excerpt ignore expected return?

Suppose Bank's books are at -100. Bank is given a loan of 50. Bank can invest the 50 either on (1) a 25% chance on a return of 800, (2) a 10% chance on a return of 1800, or (3) a 1% chance on a return of 10,000. On Salmon's logic, (3) would be the course most rewarded by shareholders -- it's a big-risk tack with the possibility of high returns.

But the expected returns are, respectively, 200, 180 and 100. So Salmon seems to be arguing that shareholders will reward the choice least likely to get Bank back in the black. What am I missing?

Comments for this post are closed