Nationalizing Bank Losses

Tyler Cowen has a new argument against nationalization:

Say that banks are in the red by $2 trillion for ever and all eternity. Taking over the banks simply means that the government picks up these losses as owner. Government ownership makes it less likely, not more likely, that bank creditors will "take a haircut."

There's a certain amount of truth to this: nationalizing banks means nationalizing their net losses. But on the other hand, overpaying for their toxic assets means nationalizing their gross losses.

The problem with Tyler's argument is that he's dealing in extremely small probabilities -- specifically, the chances that the creditors of too-big-to-fail banks will end up taking a haircut. And I'm not even sure his statement is true.

On the one hand, it's possible that the government will impose a haircut on those creditors when it nationalizes the banks. On the other hand, how else can bank creditors take a haircut? To-big-to-fail banks, by definition, can't default: the government won't let them. And they're too big to have their assets transferred to a larger institution, as happened to WaMu.

So unless you have a plan for fixing the banking system which clearly does involve bank creditors taking a haircut, I'm not sure this argument carries much weight. Meanwhile, nationalization has lots of good arguments in its favor: it give the government future upside, it allows the government to split the banks into small-enough-to-fail chunks, it minimizes the risk to the dollar, it doesn't involve sticking your finger in the air and coming up with a hard price for highly-illiquid assets, and so on and so forth.

The one argument no one is making in favor of nationalizing the banks is that doing so will somehow make the banking sector's losses go away. It won't. But that's not a reason not to do it.