Rumor: Government to Buy Bad Mortgage Securities

More Barney dollars!!

This story is so much more entertaining when you tell it. You've been an oasis these past few weeks, Dr. Kling. Thanks!

This part, I don't get:

According to policy research firm the Stanford Group, such a setup would allow the government to refinance borrowers in the loans owned by the government, thereby lowering the risk of their defaulting and eventually boosting the price of the mortgage security in which those loans are packaged.

Refinancing all those loans is going to require considerable up front costs. More importantly, though, is it really that likely to make the loans more attractive? If a mortgage backed security is unmarketable because no one feels they can predict the default rate on it, is the security automatically more attractive at a lower interest rate? The default rate might be lower, sure, but how much lower?

The costs seem sure to outweigh the benefits here.

Professor Kling, your ability to cut through the BS, right to the heart of the matter is breathtaking.

It would seem logical to un-bundle these CDO's, and repackage them as loans of like types. By interest rate. Loans that have been performing, those that are current, those in default. Those that are occupied vs. unoccupied, those with equity from those that are upside down. Then repackage them. Then auction them uninsured. Once you tell people what they're buying -- you may find a market. You would also get a really clear picture how bad this really is.

I laughed out loud at this post, went back, read it over, and laughed out loud again.

If Kling could be this consistently humorous and on point he might just be a Louis Rukeyser for the new millenium!

Being a furriner, I know of this Barney Frank fellow only by the comments of you bloggers, but he reminds me a lot of Fred Flintstone's dim wee chum.

What Treasury is proposing is a more sophisticated (I hope) version of what I had come up with the other day. The basic problem we are facing is that no one trusts a balance sheet anymore, and that is shutting down banking activity. My idea was to swap out mortgage-backed securities for treasury securities (at a discount)to restore liquidity. Pumping additional money into the system wasn't working because it's not addressing the trust issue. Once trust can be restored (hopefully), normal banking and commercial activity will start again. Plus, it may have the added benefit of disentagling the financial institutions from the CDS's written on MBS's.

The problem with MBS's these days is not just that they are hard to value; it's that for many purposes that are being valued at zero.

Your argument is logically inconsistent. If it is impossible to value these securities, then there can be no adverse selection. Shake-E does not have a better window into the ultimate recoverable value of the underlying mortgages than Bailie Mae just by virtue of having held a CDO for the past year.

The heterogeneity of these securities does make it more difficult to design the sort of reverse auction you outline, but that's not necessarily an insurmountable problem.

The other major issues with all these "hard-to-value" securities is that many may be not only valued at zero but may, in fact, be worthless. Once you get into lower tranches, especially CDO, there may be no value there.

Once the government owned the securities couldn't they practice some forbearance on them?
Maybe a better way for the government to spend $200 billion is to offer to pay off some % of all existing mortages. Perhaps adjust slightly for subprime or distressed areas.

The whole scheme actually makes good sense to me. As Norman Pfyster points out, the current problem is that nobody wants to buy any of these things, because as Alex points out, some of them may actually be completely worthless. So how to purge the market of the worst garbage, and set some kind of expectation of a floor under the rest of the market?

This proposal seems designed to do that. Anybody with a completely worthless asset would presumably be happy to unload it on the goverment for maybe a cent or two on the dollar. Of course, they'd rather unload it for twenty cents on the dollar, but then they risk being underbid by people with just-as-worthless assets who are willing to admit their value more frankly.

So if all goes well, then everything whose value is less than x percent of its face value (for some value of x determined by the auction) gets bought out at some probably-somewhat-generous-but-not-exorbitant approximation of its proper value, and everything that's left can be reasonably expected by potential buyers to be worth at least x percent of its face value. A few bold souls can then start offering x percent on the dollar for what's on the market, and something like a normal market can (we hope) slowly start up again.

The big danger, it seems to me, is that if enough people are in denial about the real worth(lessness) of their assets--or believe that others are--then they can in effect collude to overvalue them. The result is that the government overpays, a "false bottom" is formed under the market, and the crash replays itself when investors lured back into the market later discover that it hasn't been cleared of garbage after all. Perhaps some additional conditions on the auction could alleviate that concern. Anyone have any ideas?