Thursday, August 11, 2016

Housing: Part 169 - The subtle misdirection of false positives

I think we had a moral panic about mortgages a decade ago and we imposed a recession on ourselves as a result. This creates a sort of positive feedback loop, similar to, say forming an opinion about a mistreated ethnic group because they frequently lash out against their mistreatment - "See.  They are a bunch of animals.  That's why you can't treat them any better."  We closed down the mortgage market, imposing desperation on low priced housing market, and now we say, "See. We were right.  This is why you can't let people have mortgages.  Too dangerous."

There are so many facts that appear to support the bubble story, which viewed more carefully, really don't.

Here is a recent post at calculatedrisk remembering the rise of housing inventory in 2005 that caused him to begin to call the end of the housing boom:  Bill McBride, understandably, sees this as a vindication of his predictions about unsustainable housing markets at the time.  I don't blame him.  How could he come to any other conclusion?

The links in the post refer to rising inventory, and it looks like the examples are from Orange County, CA.  That is interesting.  We are supposed to believe that an expansion of mortgage credit to marginal homebuyers eventually ran out of steam, where the last suckers in the Ponzi scheme that was pushing home prices up finally were so incapable of paying their mortgages that defaults started piling up, leading to falling demand and falling prices where homes had been overbuilt.

Yet, Orange County is hardly the first place that would come to mind if we think of (1) a surplus of homes or (2) credit constrained households.  The reason Orange County was an early source of excess inventory is because the Fed was sucking cash out of the economy to counteract the housing boom, and since the boom was based on fundamental supply issues, that attempt at a solution was perverse.  So, while housing markets in low income parts of L.A. were still quite strong, and while low income households were still streaming out of L.A. to places like Phoenix for lack of housing in those neighborhoods, places like Orange County were starting to feel the pinch of a liquidity constrained economy.

Here is a graph of the change in housing turnover in LA.  In 2005 and 2006, turnover was falling in the high priced zip codes.  The slowdown started at the top.  There is a similar pattern in prices.

In 2005, when inventory started to build, and even in 2006, there was a significant amount of out-migration from the zip codes with lower incomes because of a lack of housing.  Eventually, the liquidity crisis hit the mortgage markets that were facilitating that migration, and those neighborhoods succumbed, too.  So, missing this subtle aspect of the period - that the problems started at the top - leads to a sense of support for a story that may be mostly backwards.

I don't have detailed enough data on housing inventory to confirm the turnover and price data with inventory data, but the Census Bureau has some inventory data, broken down by region.  They have a measure of the median asking price of vacant homes.  In the next graph, I have plotted the ratio of the median price of vacant homes to the median price of new homes sold.  While this isn't nearly as geographically detailed as the Zillow data above, it does tell the same story.  As the boom aged and became a bust, the inventory of vacant homes appears to have been weighted toward higher priced homes, even into 2008.  This was not the case in the Midwest.  It was mainly in the West, the Northeast, and in the South.

16 comments:

  1. Mortgages are a piece of cake for anyone in the upper-half of the income distribution. They are begging you to borrow. The lower half has absolutely awful income trajectories. I think the market is exactly right for restricting lending until we see some mechanism to smooth out income inequality. Anything political doesn't count as it's subject to change overnight.

    ReplyDelete
    Replies
    1. Income growth is ok right now. Within MSAs, income growth is pretty similar across all incomes, especially the bottom 4 quintiles. The reason real incomes at the bottom are stagnant is because they have high rent inflation because the policies you are supporting have prevented the housing stock from expanding for a decade.

      Delete
  2. I was in Reno in 2005. Inventory, according to a real estate salesman, exploded to 3000 homes by mid 2005. I told the RE agent that this was a bubble, too much supply, and she agreed. So, clearly, Reno was expensive, but not like Orange County. I think that in open access cities like Reno, inventory increased far quicker and earlier than in closed access cities like LA and SF. I once lived in Orange County. Maybe it is a closed access city now but I always felt people were speculating on real estate there and pushing prices up back in the 1970's. Seems like Orange County was always a place for easy money, maybe because a lot of lenders were located there.

    The last chart is clear, the West was the center of easy money lending, and mainly in the San Diego, and Orange County, and out of control in the Central Valley of California, Phoenix, and and in Reno and Las Vegas. However, Florida was also in this mix.

    But that chart shows the midwest had just a little increase in price. Why did the Fed destroy the economy of the midwest to stop a bubble that was rampant in 4 states, and modest in a few others states? It makes no sense.

    The Fed set the risk on mortgage bonds too low, with the Gaussian Copula adopted by the Fed from David X Li. The formula was doomed to failure, yet adopted at Basel.

    So, clearly, the Fed mispriced risk, then took it all down in procyclical fashion, killing all the sound loans, all the good wages, everything. It did it in a calculated manner, IMO.

    The Fed caused the financial crisis, and then destroyed the middle class they convinced to play along through typical procyclical behavior.

    ReplyDelete
    Replies
    1. The reason inventories built up in Nevada and Phoenix was because the collapsing housing markets in coastal California released the pressure on low income residents there. There was a huge spike in migration in 2005, which fell off as cost pressures in the Closed Access cities retracted.

      Delete
    2. Kevin, do you have a timeline for that immigration? It seems like all of 2005 was seeing a collapse in demand for houses in Reno.

      Delete
    3. Migration data is only annual, and I don't know about Reno. But, in-migration to Phoenix and Las Vegas was very strong in 2005, then weakened in 2006. San Francisco was earlier than most cities, so if Reno had a particular exposure to San Francisco, it could be that migration dropped off earlier there.

      Delete
    4. Thanks Kevin. Keep in mind that besides immigration declines, there was a huge housing boom in Las Vegas in 2005. I remember this was true because I drove through and saw the massive construction. Reno's construction no doubt tapered off sooner. But 93 thousand homes were sold in Las Vegas in 2005, a massive amount. Then it crashed. Here are hard figures. http://lasvegassun.com/news/2010/jun/11/building-boom-recessionary-bust-local-real-estate-/

      Delete
  3. I often read Calculated Risk, but I cannot accept McBride's reasoning on the Housing boom/bust. He uses the data in a "see, I told you so" fashion. He completely misses causation. I am deeply indebted to the research done here at the Whisk, for explaining the causes of our national myopia. Good work, Kevin.

    ReplyDelete
    Replies
    1. Thanks Chuck. The tough thing is I can see why he and others are so convinced. I probably would be if I was them, too. The conclusions were reached because of subtly wrong premises in so many ways, but once the conclusions are settled, they are understandably difficult to unsettle.

      Delete
  4. Great post.

    Also the big fundamental: For better or worse, home buying in America is financed by mortgages. You cut off mortgages, and you will see house prices go down.

    When house prices go down, the banks cut off mortgages.

    McBride should look at commercial real estate, which had a nearly identical plummet to residential. Same problem. Banks cut off lending.

    No, it was not a conspiracy by evil bankers. It is just the way the system is set up.

    ReplyDelete
  5. I think a big hole in the Housing Bubble Theory is that in 2007 that whole crowd was saying that all the ARMs would be in trouble once the rate resets started. I recall hearing it too in The Big Short. The one manager who "saw it coming" just knew the big problem would be once the ARMs started resetting to higher payments. Of course, rates ended up collapsing due to tight money. Yet the Bubble predictors honestly seem to forget that one important aspect of their predictions. And I do mean that. McBride has always impressed me as an honest economist, and I don't recall if he was one of the predictors who made much or little about the pending rate resets in 2007.

    ReplyDelete
    Replies
    1. Yes. Good point. I think McBride did talk about rate resets, but there are some subtle issues here. What really killed the housing market in 2007 wasn't so much rate resets, as you say, it was the fact that liquidity killed the market altogether, so that borrowers couldn't refinance at any rate. If their loan had a recasting date where they had to start amortizing at a higher rate, then payments would go up regardless of rates. I think he was careful about that point. Generally, I think the calculatedrisk blog generally does pretty good work, and where I think he is wrong, it tends to be for subtle reasons like the one above.

      Delete
    2. I'm with you. I'd add only that the 10% unemployment rate / millions of job losses compounded the situation, which I think most people think the reverse. That is, I think it's more that the recession deepened the foreclosure crisis than the other way around.

      Delete
    3. Absolutely. I think we can even blame the initial housing declines on recessionary conditions that began by late 2006. It was a full employment recession because (1) the focus on popping the housing bubble meant that monetary contraction was manifest through falling housing supply and credit supply, making it inflationary and (2) our peculiar capping of populations in the most productive cities meant that during expansion, there was strong migration out of those cities into the rest of the country. So, hires, quits, etc. were levelling off in 2006, but the first result of that downtrend in employment was not unemployment. It was a decline in migration.

      So, I hope to eventually have a strong argument that recessionary monetary and credit policy caused the housing boom to burst, and together these things led to the housing bust in late 2007, which, for continued lack of support, led to the slow descent into recession until the Fed gave the economy one last kick in the ass in Sept.-Dec. 2008, after which the vast majority of the defaults and unemployment came.

      Delete
    4. As Marcus Nunes has said, borrowers could not use their HELOCS in 2008. So they could not refi in 2007 and the second shoe, Heloc Hell, dropped in 2008.

      Delete
    5. But, Kevin, one thing. All the while that NGDP was slowing, the Fed clearly did not want to save an industry with so many bad bonds. It saved big business, with the nice A- rated bonds. But it let the MBSs fail. There were a lot of crap MBSs to be sure. But the Fed could have been the lender of last resort in the RE industry and chose NOT to do so.

      But you can bet it would save the junk bonds that back the oil industry.

      Delete