However, we can prevent the problems in housing from spiraling out of control. We can do this by providing liquidity to the system, in whatever form is most effective. Ideally, the Fed would find a way to provide liquidity where it's needed, but not where it's not. In other words, find a way to throw cash into the problem areas of the financial system without creating any more inflation pressure than necessary.
That's exactly what the Fed is trying to do with this new Term Auction Facility. I think James Hamilton has it right, in that it's nothing more than the discount window with another name and an uncertain interest rate. The idea is to give banks a temporary avenue for raising liquidity at a time when liquidity is dear. It's nothing that banks couldn't do through the discount window anyway, but without the stigma.
Is there something special about year-end? Take a look at this graph of 1-month LIBOR...
Notice what happens when the 30 day window ticks over year end?
Nouriel Roubini, whose writing is always eloquent and bearish, makes the point that "you cannot use monetary policy to resolve credit and insolvency problems in the economy." Absolutely right. But you can use monetary policy to prevent fear from creating illiquidity, and the illiquidity from creating bank failures. So perhaps Roubini and I would agree that the TAF won't help make subprime borrowers current on their loans. But it can help make sure that banks can deal with elevated levels of delinquency and foreclosure.
Steve Waldman calls this a bailout. And the way he's defined "bailout" I guess he's right. But to me that term is sensationalist and not meaningful in this context. All the Fed is doing is creating an alternate discount window with mostly the same terms as the original. The shock and horror that the Fed could take AAA-rated CDOs as collateral is misplaced. They could take the same collateral at the discount window. Those who are complaining about the collateral rules are suggesting that the new vehicle should have more stringent terms than the exiting vehicle.
Anyway, Waldman says it's a bailout because the Fed is offering loans that other banks wouldn't be willing to offer. Here again, any bank who went to the discount window would borrow under the same terms. Does the continued existence of the discount window constitute a bailout? Put another way, banks would never use the discount window if financing was available elsewhere at similar costs. So any time a bank uses the discount window, it's a bailout by Waldman's definition. This is why I am loathe to use that term.
Now let's consider how a loan of this kind would work.
- Bank gives collateral to the Fed.
- Fed gives cash to the bank.
- 28 days pass.
- If the bank is still in business... Fed gives collateral back and bank pays cash back plus interest.
- If bank is out of business, Fed keeps the collateral and prints money to make up the loss.
As Yves Smith puts it, it really doesn't matter if the collateral is an old couch so long as the bank remains in business. That's how repo works. The guy who pledged the collateral remains the one on the hook for any market value differential during the repo term. So while the Fed may be willing to take less-than-stellar collateral, the real risk to the Fed is that the borrower bank is out of business in 28 days.
Anyway, I'll leave you with one final thought. We know that there are many banks under stress. Some may wind up belly up. Some may just need a capital infusion. Some may have losses, but have plenty of capital. The problem is that it's difficult for banks to know who falls into what category. So they are hesitant to lend to other banks period. Any bank with excess capital understandably doesn't want to risk it on banks strapped for cash. And in fact, banks who have taken on some losses but are fundamentally sound, are finding it difficult to get short-term funding. What the Fed wants to avoid is fundamentally sound banks suffering from a lack of liquidity. I believe that's all the TAF is designed to address. Is it possible some bad banks will use the TAF and stick the Fed with bad bonds? Maybe, but the odds are low that a bank would go from apparently sound to bankrupt in 28 days. It could happen, but the odds are low.
And if the Fed has to take on losses on one bank to save 20 more, that's a trade Bernanke will do every time.