Friday, September 26, 2008

Game Theory: Why the Bailout Won’t Work

Lets assume for the time being that there are only two banks; Bank A and Bank B.

The media / political pundits would have you believe the likely outcome of the bailout is the top-left box in which both Bank A and B sell risk assets to the Treasury. In this case, the result is a more regulated banking industry, with imposed limits to salary, but importantly markets clear.

Click for larger table:

HOWEVER, it is in BOTH banks interest to deviate from that.

Why? Simple. If Bank A (or B) believe the other is selling their risk assets to the Treasury; they will each be better off holding on to theirs.

Why? If the other bank sells and they hold, markets will still clear (in theory) and the bank that holds onto their risk assets can sell at the new market prices. This results in increased market share as they:

*Can pay more for talent
*Are less regulated
*Don’t have the stigma of selling to the Treasury (think of what selling portrays to the market)

This is even worse in the “real world” as all banks have the incentive to wait for other banks to sell risk assets to the Treasury to clear markets.

The likely result? The bottom right box in which no bank sells voluntarily and markets remain frozen. While there were many problems with the initial plan, at least there was a 100% incentive to sell the assets.

Update: Thanks for the link Yves from Naked Capitalsim. Her take on my post:

Brilliant, except it assumes that assets sold to the Treasury will be lower than the prices they will later fetch. We think that's completely wrong; the intent is to overpay relative to current market prices, and with real estate and the economy headed south, these assets are certain to trade at even lower prices for a very long time. Plus banks will sell the stuff where they think Treasury is overpaying the most, and hang on to those assets that they think have the most upside. But more of this sort of reasoning is badly needed.
I get where she's coming from. My assumption above was that markets clear if the "other" bank were to sell to the Treasury eliminating the need for all banks to sell to the Treasury. For more on Yves opinion that this is in fact a capital injection (i.e. the Treasury plans to intentionally overpay, click here).

While I agree that the Treasury will pay above market prices, it is my opinion that market prices are artificially low due to many technical factors and a lack of global balance sheet for risk assets (i.e. everyone is selling, not buying). I feel expectations of 10-15% returns based on today's marks, even with very conservative assumptions, are reasonable.

In thinking more about it, they would NEED to pay more for them than currently marked. Unless a bank was insolvent, why would they clear assets yielding 10-15% to make room for new loans yielding 8%? (it also doesn't hurt that the Treasury just needs to beat their unbelievably cheap financing rate - currently under 4% for 10 year bonds).

That being said, I think there is still a huge incentive for each bank to wait for others to clear the market at these higher prices (and people tend to do what's best for themselves vs. shareholders). In addition, while this in theory could clear the frozen credit markets, I think any package of this size should also impact the housing problem, which this does not.