Saturday, November 14, 2015

Housing, A Series: Part 83 - Mortgage Affordability vs. Rent Affordability

One measure we might look at as a sign of a housing bubble is the ratio of mortgage payments to rent payments.  If households were bidding up the prices of homes as owner-occupiers in an unsustainable speculative bubble, we might expect to see mortgage expenses rise relative to rent payments.

Zillow tracks a measure of mortgage affordability, which is the monthly cost of a 30 year fixed mortgage on the median home, with 20% down, as a proportion of median income.  They have a similar measure of rent affordability, which is the monthly cost of rent on the median home as a proportion of median income.

Here is a graph of the ratio of mortgage affordability to rent affordability, for the US, the US with NY, LA, and SF removed, and for NY, LA, SF, and Washington DC.

Note: scales aren't calibrated.  M/R ratios of cities on the left scale
do not correspond to expected inflation on the right scale.
The main difference between rent and mortgage payments is that rents rise with inflation and mortgage payments are fixed, so in an efficient market this ratio is a rough measure of expected rent inflation.

Nationwide mortgage/rent ratios were about the same in 2005 as they had been since 1990, and rent inflation was about the same in 2005 as it had been since 1990.  Some might want to make the argument that Fed policy was pulling down interest rates, which created the bubble by pushing mortgage rates lower.  I would argue that the Fed has little control over 30 year mortgage rates, and that if Fed policy really was loose, 30 year fixed rates would rise due to inflation expectations.  But, even taking that criticism on its own terms, the Mortgage/Rent ratio reached a low point in 2003.  Even if mortgage rates had been at levels similar to the 1990s, the national Mortgage/Rent ratio would have been within the normal range.  And, by the time the ratio reached the top of the range in early 2006, short term interest rates had reached their high point and monetary policy could not be described as loose.

Mortgage/Rent ratios in the closed access cities are much higher than they are in the rest of the country, and they rose more steeply during the housing boom.  As with so many of the indicators I have looked at, there are two stories here.  There is the supply story in the closed access cities.  Then, there is the rest of the country, where there is little indication of any unusual activity.

I have included Washington, D.C. in this graph.  It is slightly different than the other closed access cities.  In the closed access cities, the cost of housing is capturing economic rents from geographically captured industries.  Incomes are being pushed up by housing costs.  In Washington, the causality is the other direction.  High incomes are pushing up housing costs as households use their rising incomes to bid up preferred housing units.  That means that rent inflation has been high in Washington, but rent expense has remained low as a proportion of income.  In this measure, Washington looks like the other closed access cities, because the high Mortgage/Rent ratio is a reflection of expected future local rent inflation.

In upcoming posts, I plan on comparing permits for new homes to housing affordability.  One way in which the aggregation of these two housing stories led to a mistaken interpretation is that in the closed access cities, prices were being bid up, but this was dominated by existing sales.  New sales were overwhelmingly in the open access parts of the country.  When these data were aggregated, it was easy to believe that the rise in new housing starts was related to households over-borrowing and overspending on housing.  But, when we keep these stories separate, what is clear is that new homes did not have the characteristics of the aggregate data.  New homes were being built in the inexpensive parts of the country at reasonable prices.

Here, as with all the other scatterplots of city housing data, we can see the two stories - the open access story, where population flows between cities with relatively equalized costs, and the closed access story where constricted cities create wild upswings in valuations.  These graphs plot cities based on housing permits issued from 1995 to 2005 (x-axis) and the increase in the mortgage affordability index from 1995 to 2005 (y-axis).  The first graph is based on permits as a proportion of 2000 population.  The second graph plots the cities based on the raw number of housing permits issued.

Just over a million permits were issued in the largest closed access cities.  In those cities, the median mortgage in 2005 claimed from 14% to 30% more of the median income than it had in 1995.  Over 18 million permits were issued in the US during the same period, around 17 million outside those cities.  And mortgage payments for those 17 million new homes only claimed about 2% more of the median income in 2005 than they had in 1995.  On this scatterplot, we can see places like Phoenix, Atlanta, Dallas, and Houston, where these millions of new homeowners were buying homes at relative valuations much lower than the closed access cities.

Just to be thorough, here is a similar graph, using the mortgage/rent ratio that I used in the first graph above.  Where housing can expand, mortgage/rent affordability was stable.  Where costs rose sharply, housing permits were low.  (The three cities that appear to have rising costs in these graphs along with the main closed access cities I have highlighted are, from left to right, Boston, Miami, and Riverside, CA.)

The housing boom in prices was part of a different story than the housing boom in new construction.  I hope to dig a little deeper into this over the next week.


  1. Kevin--I mention you in latest Historinhas post.

  2. TravisV here.

    "Fed Economic Letter: "What’s Different about the Latest Housing Boom?""

    1. Thanks, Travis. I will probably work a response to that into an upcoming post, because from the description at CR, it's almost like they wrote that article as a compendium of all the ways making the wrong assumptions leads us to misread the correlations. When everyone is making the same bad assumptions, they begin to seem like axioms, and what seems like careful scientific work becomes an exercise in question begging. It really drives home how difficult it is to understand complex systems. I'm afraid that we are doomed to always govern ourselves based on our prejudices, because when the context reaches a certain level of complexity, the sausage grinder of causal density ends up leading us to simply restate our prejudices without realizing that is what we are doing.

  3. These financing options may be used for both owners and contractors. This facility might include the price of the land so it’s totally depending on the lending company. You should be cautious since these advances contain a lot more pit falls as compared to any other loan. Therefore, it is preferred that whenever you select a loan company then perform some research. There are a lot of credible companies out there like best Mortgage Companies Houston