Tuesday, October 13, 2015

Housing, A Series: Part 68 - Bias in Home Price Indexes confirms the local supply constraint problem.

In the previous two posts, I have proposed that we can see the effect of persistent rent inflation on home prices by noticing that Price/Rent ratios are higher in cities that have experienced more rent inflation.  The expectation of continued rent inflation, from local supply constrictions, causes Price/Rent ratios in those locations to rise.  I'd like to look at that a little more here.

First, there is the issue that the Price/Rent ratio estimated by the S&P/Case-Shiller National Home Price Index and CPI rent inflation has tended to rise above the Price/Rent ratio estimated with aggregate numbers from the BEA and the Federal Reserve of owner-occupied real estate values and owner-occupier imputed rents.  I have attributed this drift to supply constrictions.  Price/Rent levels are higher in the constricted cities, so new building must happen where supplies aren't constricted.  Where supplies aren't constricted, prices are lower, so the changing composition of housing must cause indexes tracking individual properties to have an upward bias compared to indexes that track all properties.

In fact, this is one factor that Zillow intended to correct with their price index (ZHVI).  And, in fact, the S&P/Case-Shiller National Home Price Index deviates from the ZHVI at about the same scale as it deviates from the price levels estimated by the BEA & Fed data.  This confirms that rising prices are coming from a supply issue.  If rising prices in the cities where housing is in high demand were caused by demand, then marginal new building would be happening in those high priced areas, and the S&P/Case-Shiller Index would be biased downward, because the ZHVI would be populated by new higher priced homes.

This is a core confusion about what is happening in the housing market and the economy in general.  From an unbiased observers perspective, a market with rising prices due to rising demand looks just like a market with rising prices due to constricted supply.  In both cases, the market will be increasingly dominated by buyers willing to pay the higher price.  We are predisposed to attributing cause to high income buyers.  But, in fact, this data indicates that we have constricted supply, and that, on the margin, households are escaping rising home prices by building homes in less expensive locations.

The same pattern shows up in the difference between S&P/Case-Shiller home prices and the average price of both existing and new homes during the boom period.  Average new home prices and the S&P/Case-Shiller index moved together for decades.  If anything, the average price of new homes grew at a slightly higher rate at times.  But, starting in the late 1990s, the new pattern emerged, and new home prices started rising at a slower rate than existing homes.

This is one of many deceiving issues on this topic.  One could interpret this to mean that banks were pushing subprime loans on low income households, and that would explain the lower value of new homes.  But, despite this widespread belief, there was no shift to lower incomes among home owners during that period.

The divergence between the prices of new and existing homes after the bust is because we have reduced access to mortgage credit for middle income households, so that more new homes are being built by higher income households, and this is creating a countereffect against the flight from high priced cities.


  1. Interesting. There is a 2002 Cato Institute study on residential zoning.

    Everyone believes in free enterprise, except where they live.

  2. Historically, the ratio of new to existing home prices was highly cyclical. As conditions in the housing market tightened builders would drop out of the low cost market and because of the nature of the old stats using average prices it would show up in the data as rising average prices. But this did not impact existing homes as much. So the ratio of new homes to old homes would rise. At bottoms the process would reverse as the market improved builders would reenter the low and moderate housing markets and average prices would tend to fall in relationship to existing homes prices. This ratio use to be a very good signal of when home builders stocks would outperform and underperform.

    Given the changed nature of what the major home builders do now I have reservation that the signal can work as well as it use to.

    1. Interesting. Thanks. It does seem like there has been some down market shift recently by homebuilders that has been interpreted as bullish. But, I think you're right. It's a lot more complicated now.