Saturday, May 28, 2016

Housing: Part 153 - Price Changes at the start of the bust

You know that thing that happened back in 2006 and 2007 where increasingly predatory loans were going to low income households that couldn't begin to make the payments, and when the Ponzi scheme inevitably came to its end, those low income households started defaulting?

And, then, a domino effect happened, where low income neighborhoods were full of foreclosures that undermined the market, and just made home prices in those neighborhoods collapse even more?  So, the mortgage brokers got out of town with their fees and their bonuses and left lower income households holding the bag?

Remember when that happened?

I do too.

Except, it kind of didn't happen.  (You knew I was going to say that, didn't you?)

Here are scatterplots of zip codes in Los Angeles.  The x-axis is the log median home price in the zip code.  (12 ~ $160k, 13 ~ $440k, etc.)  The y-axis is how the median home price changed in a year.  The scatterplots contain the years 2005-2009.

In 2005, we see the rising prices we all remember.*  But, note what happens in 2006.  Prices start to moderate, but that moderation isn't led by a collapse in low priced homes.  It is high priced homes that were declining first.

By 2007, prices were widely in decline.  But, even by then, homes at the high end were leading the trend.  Remember, by the middle of 2007 the private mortgage market was gone.  Not diminished, basically just gone.  There was no source of credit for unconventional mortgages by then.

It was only in 2008 that low priced homes really dropped.  And, boy did they drop.  Notice that high priced homes were still declining, but the decline was about where it had been in 2007.  Then, high priced homes stabilized in 2009, but low priced homes continued to collapse wildly.

You know what has a stronger influence on high priced homes than low priced homes?  High interest rates.  Because the implied net return (net rent/price) is lower on high priced homes than it is on low priced homes, so the risk free rate effects them more.  There is basically a spread on home equity returns that gets larger as you move to lower priced homes.  In other words, if interest rates cause required returns in housing to go up 1%, high priced homes where the returns go from 3% to 4% will be effect more than low priced homes where returns go from 7% to 8%.  Interest rate policy creamed the high end housing market, and only later, as a lagging effect, did low priced homes collapse.

Of the 22 MSAs I have been looking at (the largest 20, minus Houston, plus San Jose, Portland, and Las Vegas) where Zillow has zip code level data on the trailing 12 month rate of existing home sales and on price levels, in 20 of 22 cities, existing home sales started drying up in high priced homes before low priced homes.  And in 15 of 22 cities prices of high end homes started to fall before prices of low end homes.

This was not a subprime crisis.  It is a misnomer.  It was a moral panic.  What those neighborhoods in 2008 and 2009 needed was a government that would allow banks to sell mortgages to them.  It's 2016, and we still won't let them.

I have included Phoenix and Dallas here too.  It's a common pattern in all city types.

In Phoenix, of course, there was the big price jump in 2005, although there wasn't much difference between high priced and low priced homes.  But, even in Phoenix, high priced homes led the collapse and low priced homes had their crisis well after the collapse of subprime.

In Dallas, there was just 5%, + or -, regular price appreciation, but a lot of low priced zip codes there got to see a couple years worth of 5-20% price drops, too.  Because they needed to learn a lesson.  If we had supported the housing market in 2006 and 2007**, those low income zip codes in Dallas would be full of people who think a house is a safe investment.  That'd be a pretty stupid thing to encourage, wouldn't it?  They are wiser now.  You're welcome, poor Texans.  Job well done, everybody.

*In 2005, LA did see low priced home values rise more quickly than high priced homes, which I think I have gone over here.  I expand on it more in the book.  This only happened in the Closed Access cities (coastal California, etc.).  It didn't happen in the Contagion cities (Phoenix, Miami, etc.)  The reason it happened is because Price/Rent ratios naturally increase as home prices rise, up to about $500,000.  So, this effect was not pushing the high end of Closed Access home prices up, but it was still in effect for most other homes.

** Some will probably react to that by saying, "Oh, so now, the Fed is in charge of making sure nobody ever loses on a real estate speculation?"  No.  It's not.  There is a long distance between micromanaging an asset market and engaging in a multi-year attack on the mortgage market that first exogenously cuts 10% market share from conventional mortgage originators and then creates a liquidity crisis in the mortgage markets that rise up to replace them.***

*** Some will probably react to that by saying, "Oh, so now, the Fed is in charge of making sure nobody ever loses on a reckless MBS with rigged credit ratings?"  No.  It's not....Ah, heck.  It's turtles all the way down, isn't it?****

****As I think through the book, it really is amazing how deeply the premises determine the conclusions throughout this topic.


  1. tighter credit conditions were a feature throughout the entire economy. I see nothing unique to housing.

  2. The Fed destroyed the commercial paper market. It started with the subprime crash in August of 2007 because the commercial paper market was undwriting subprime and pricey alt a, but that destruction spread to everyone since the banks underwrote alt a and rich non prime loans. The Fed could have bought Commercial paper (though technically illegal but doable), and would have freed up banks to continue to lend. But of course, the Fed wanted to liquidate. It is true that the Fed mispriced risk for MBS bonds in the first place. But it also allowed the commercial paper market to be destroyed and then allowed mark to market on top of that! And then houses in non bubble areas were destroyed by layoffs in late 2008. I agree with much of what Kevin says. The Fed played a major role.

    All those subprime and alt a loans ended up on the banks' balance sheet, having been off balance sheet for the easy money period.

  3. There may be...the rest of the story.

    As noted, modern-day commercial Banks basically extend loans on collateral which means property.

    So, ifvthrough regulation and interest rates wec asphyxiate property lending, we will also suffocate the increase of the money supply.

    The only practical solution to housing inflation is more supply....

    1. Thanks, Benjamin. The commercial banks actually underwrote the shadow banks through the SIVS that were off balance sheet. When those were forced back upon the banks, the ability of the commercial banks to engage in real estate was essentially over. The commercial banks continued to lend to businesses, but the CP market which funded the shadow banks went under.

  4. Here's my take. It may change after lunch.

    Agreed, in Phoenix less expensive homes continued to increase in value for several months after home prices had peaked in the rest of the market. Case-Shiller price tier data show this too.

    My working theory is that lending standards declined faster after home sales peaked in 2005. It seems lenders desperately lowered their standards to slow the decline in the mortgage business.

    The expanded pool of buyers who could now get a loan with the new lower standards were lower income people because, by mid-2005, pretty much everyone who wanted to buy a home and who could get a loan had already bought one (or two). These new potential lower income buyers bought in lower income neighborhoods.

    So I THINK we ended up with the riskiest loans being more concentrated in lower income neighborhoods.

    That would explain;

    1) why prices continued to increase in those lower income zip codes after prices had already stopped increasing in the rest of metro Phoenix, and

    2) it would also help explain why price declines were worst in those neighborhoods, they had more of the worst loans.

    1. This does seem like a potential answer, but there are a number of issues that I think work against it. Some include:
      1) The rise in homeownership was concentrated at higher income levels.
      2) Most of the decline at the low end happened after the recession started - it was a lagging effect - and it was fairly widespread - hitting cities like Dallas where all home prices had been moderate.

      I think the particular shape of changing prices is more likely due to national factors that were pushing against housing from the top, and this was pushing households at the middle and top more downmarket. The top end in several metro areas had already experienced problematic declines in prices when the low end began to drop. But, those owners just tend to be less apt to default.

      One interesting issue I have seen and I need to think about more is that even though home ownership was declining, originations were increasingly going to first time homebuyers. There was a strong transition of owners selling and exiting ownership. I think before the default crisis hit, there had been a long period of time where financing had become less available, but many households were still able to sell instead of defaulting. But, I need to look at this more.

    2. I use the Survey of Consumer Finances for ownership by income quintile, but it is only published every 3 years. The Census Bureau publishes ownership rates for households above and below the median income, which is less informative, but this shows ownership declining for the bottom half of incomes in late 2006 and 2007. (Table 17 of the Census homeownership tables)