Tuesday, May 31, 2016

Housing: Part 154 - The black hole at the middle of the moral panic

We had a moral panic about housing and finance built on deep prejudices.  As is usually the case with moral panics, there was a source of discomfort that began the process.  This mixed with a healthy dose of true anecdotes that confirmed our prejudices, a healthy dose of imagination about other things that were happening, and wads of attribution error.  As I have been working through the narrative of the period, it has occurred to me how ad hoc and broad the attribution error was that we used to build the narrative.

First there were greedy and powerful GSEs but they were cut back by 2004, so then there were fraudulent subprime predators and unrealistic homebuyers but there weren't any marginal new homeowners after 2004, so then there were rabid investors but even they weren't utilizing enough mortgages to meet investor demand by 2006 and 2007, so then it was the banks who were greedy for AAA and irrationally bidding up synthetic CDOs.  By the way, has it occurred to anyone that while the crazed banks were "reaching for yield" by buying up those AAA rated securities in 2006 and 2007 that the Fed had short term rated pegged at 5.25% and the yield curve was inverted?  I would like to flush the whole concept of "reaching for yield" down the toilet.  It is simply a direct rhetorical application of theory by attribution error, and it's useless.  But, for those who use the term, is there any context where it is inapplicable?  Fed officials today talk about how we need to raise rates because there is speculative buying in asset markets as investors "reach for yield", but when they were sucking the economy dry with rates at 5.25%, banks were supposedly "reaching for yield" by leveraging up AAA securities then, too.

And, how in the world can the narrative hold on that blames Fed accommodation and CDOs for the bubble and bust when the CDOs were mostly constructed during a time when the Fed Funds rate was at 5.25% and the yield curve was inverted - and, by the way, real GDP growth was declining.  Is there any evidence of loose money except for (1) the Taylor Rule and (2) the prices of homes themselves?  It is a circular argument.  Take home prices out of the picture, and what evidence is left that the Fed was ever particularly loose?  Yet, the Fed is basically the capstone boogeyman that we use to explain the hodgepodge of dupes and villains that explain the bubble.

Think about it.  Low income buyers, small time real estate investors, bank asset portfolio managers, investment bankers underwriting securitizations, mortgage originators, housing advocates.  None of these groups really have anything in common.  They have different motivations, they show up at different points in the timeline.  Yet they all had to go crazy, in turn, to explain a demand-based bubble.  So, the irresponsibly loose Fed is the core actor in our stew of attribution error that explains all the other actors in our ad hoc attribution error drama.

This madness is so deep that even today, the public conversation seems to mostly revolve around how the Fed is propping up the economy.  What in the world would the graph above need to look like to pop this intellectual bubble?  "Reaching for culprits" seems to be a more accurate description of our time.

Here is a measure of residential mortgages outstanding, by holder type.  We call the period from 2000 to 2007 the "subprime bubble" or the "subprime crisis".  Moral panics do not require facts.  After 1999, subprime was basically just growing with the mortgage market in general.  The market share gain in subprime is barely perceptible.  And, between Ginnie Mae and subprime, combined market share was nearly cut in half.

There are several papers that purport to show that defaults were unrelated to borrower or loan characteristics.  Even the original securities downgrades by the rating agencies note the fact that borrower and loan characteristics didn't seem explanatory.  And, I had thought, well, even though there must have been a lot of dicey loans, it looks like it was a lack of liquidity that led to prices falling and then defaults rising.  But, at this point, I think it isn't just that the dicey mortgages didn't trigger the first defaults.  It's that there wasn't even a surge of highly leveraged, risky borrowers.  By 2007, we had undercut the owner-occupier market enough that Alt-A loans had become more prevalent, so there were a lot of young loans held by investors that were especially prone to a default crisis triggered by an unprecedented price decline.

Notice that Alt-A loans grew mostly in the 2005-2007 period.  In the contagion and Closed Access cities, prices were already beginning to fall in early 2006.  Ownership rates were falling.  High priced neighborhood sales and prices fell earlier and more sharply than low priced neighborhoods.  Households were already being defensive.  These investors were buying into markets where rent inflation was rising and home prices were stabilizing or falling.

In 1995, about 15% of mortgages were Ginnie Mae mortgages, by 2006, about 16% of mortgages were Alt-A or subprime.  The average homeowner was slightly younger, but had no less income than they had had in 1995.  But, these private issuers utilized different models than the public programs had.  What did we expect?  Of course they did.  2004, and really even 2005, cohorts of those mortgages performed pretty decently, in spite of the fact that those buyers had to handle unprecedented price volatility.  Defaults were highly correlated with the timeline.  Originations that happened right before prices collapsed saw very high defaults.  Cities with the same underwriting but without the price drop tended not to have dramatically higher defaults.

So, the premise determines the conclusion.  If the price drop was inevitable, then it was inevitable, and if it wasn't, then it wasn't.  The moral panic is its own evidence.  The fact that the yield curve was inverted for nearly two years before the collapse happened.  The fact that home prices in the high default cities had been collapsing for more than two years before major investment banks started to collapse.  These facts don't matter, because the thing was fated to happen, even if we needed to be fate's handmaidens to make sure it did.  It was the only responsible thing to do.  And it's their fault.  We just can't all agree about who "they" is.


  1. "So, the premise determines the conclusion."--Kevin Erdmann.

    Oooh, I like that line. Gonna steal it someday.

    I seem to recall a few years back there was also a "the Clintons caused the bust" rationale. Something along the lines that Clintons strengthened the Community Reinvestment Act (CRA) and that forced banks to lend on inner-city housing, and then the force-fed urban markets collapsed.

    And, in fact, I can remember then-President George Bush speaking on national TV, in the eighth year of his Presidency, and explaining that the cause of the 2008 bust extended back to rule changes made "10 years ago."

    Actually I detest the CRA and most forms of regulations on lending and development, including property zoning.

    But "the CRA did it" was another one of those convenient arguments that fit political narratives post-bust. The lefties have similarly deluded explanations.

    I am glad Keven Erdmann, evidently free of political concerns, can attempt to understand what really happened. IMHO Erdmann is more right than anybody else I have read.

    What is sad is that due to a combination of local property zoning, and federal nooses on lending, we are depriving millions of Americans of better living arrangements. Maybe tens of millions.

    We may also be starving leading industries of labor.

    But hey, let's talk about how bad the other party is, and transgender toilets.

    1. Thanks Ben. And amen.

      Yeah, the CRA stuff is interesting. At first glance, it seems really important regarding bank merger approvals, etc. Yet there is no sign in the survey of consumer finances that there was any shift in homeowner qualifications. Yet, if somehow the SCF isn't seeing something that really did happen, that could only mean that millions of marginalized households were able to buy homes in the 1990s right before they suddenly doubled in value. An unalloyed success. But as Marc Andreessen says, there can be no good news.

  2. Great post. Reminds me of The Big Short - a key character says all the mortgages will default once interest rates RISE. We all know what happened. NGDP collapsed.

    @BenjaminCole. Good point about the 10 year old laws. Even if those laws are the cause, six years of controlling Congress and the Presidency but leaving the laws in place kinda means something.