Three Flavors of the Economic Crisis
Varieties and Alternative Definitions of the "Financial Crisis" Revisited...

Hoisted from Comments: Risk Controls at Fannie Mae During the Mid-2000s

Ken Houghton:

Three Flavors of the Economic Crisis: You are, as usual, too kind to Mr. Cowen's predilections. I'm still trying to figure out if that's not having followed through the history of the mortgage market or deliberate obtuseness, but the benefit of nagging doubt goes with the former.

You're letting Cowen use numbers to confuse and conflate. "It's not denied that the mortgage agencies were guaranteeing about half of all U.S. mortgages right before the crisis (Yet somehow they had not so much to do with the crisis?)" implies that those mortgages were equally shite with those issued by Countrywide/BofA, Bear/EMC, Lehman/GKW, Wells Fargo, The Big C, and others.

They weren't. Despite Daniel Mudd's best efforts, the risk controls at Fannie (at least) were still somewhat in place.[1] Also--fortunately, since it was offsetting Mudd's desperate "if Wells and Countrywide are doing it, why aren't we? There are fees to be made!" attitude (which made him ridiculously rich, while that git at the NYT tries to blame the man long gone, but that's a side issue)--the credit reserves required of Fannie and Freddie during that period were rather higher than the banks. (Strangely, no one--not even the people who talk about how TX was "saved from the RE bubble" because of regulations put in place after the last RE bubble--talks about that.)

Tell you what: pari passu and sight unseen (random selection), you take $100MM initial face non-GSE MBSes from 2004 and I'll take the same Face Value bonds from the GSEs, and we'll see who is still closest to a positive NPV.

If you play that same game with 2005 and 2006 vintage MBSes--no cherry-picking, just random--you'll find the gap tightens but the premium remains with the GSEs.

So if you want to say "Fannie and Freddie still controlled half the market and there was a problem that was not strictly a subprime problem," your next dependent clause better be "although most of the problem was and is sourced to those non-GSE-originated securities."

Which does, of course, explain one of the major reasons Fannie/Freddie are still problematic, even ignoring the politics. The Fed bought many of those non-GSE-issued MBSes from the banks and stuffed them into Fannie and Freddie. This makes sense if you're trying to shore up bank balance sheets--we can discuss whether that is the same as "recovery the economy"--but it also means that the banks who made bad loans/purchases now appear more solvent, and Fannie and Freddie appear less so--not through business acumen, but because the deck is rigged.

I cannot believe Tyler Cowen does not know this has happened. That he perpetuates such counterfactual horse puckey on his unsuspecting readers (and even gets Buce or you to go along with him a ways) is therefore presumptively out of malice, not ignorance.

[1] I presume, based on the data, that they were in place at Freddie, too.