In Many Metros, Impact of Housing Downturn Depends on Value of Home

Talk to any local real estate agent about sales trends and it doesn’t take long before the conversation turns to differences between sub-segments of the market.  The most common rubric for segmenting the market is probably based upon price.  Since its inception, Zillow has produced the Zillow Home Value Index (ZHVI) for various price tiers of a local market (see below for methodology).  Deconstructing a market by price tier can yield important insights into the internal dynamics of a local market.

For example, take the first figure below, depicting the ZHVI for three price tiers in the San Francisco metropolitan area.  A couple interesting observations are immediately obvious.  First, the housing recession in San Francisco following the dot-com bust (in 2000-2002) was led by the top-tier, or most expensive, homes.   Annualized depreciation for the top tier reached around 8%, while the middle tier dipped just slightly below zero percent annualized appreciation and the bottom tier of homes never got below 8% positive appreciation.

This pattern was exactly reversed in the current downturn that began in 2006.  This time around, the bottom tier of homes saw the largest depreciation rates and the top tier of homes saw the smallest depreciation rates (negative, to be sure, but still higher than the bottom and middle tiers of homes).

As one might expect, these differences between price tiers of homes play out geographically as well. In the second figure below, we see, for each city in the San Francisco metro (represented by the blue diamonds), the magnitude of the decline in home values. The Y axis of the graph shows the percentage change in home values between their peak value and their current value, and the X axis shows the ZHVI of the city (cities with higher ZHVI tend to have more expensive homes).  Noticeably, there’s a pretty distinct, positive relationship between the typical value of homes in a city and the amount of decline in value that homes in the city have experienced.  That is, cities in San Francisco with higher priced homes have, in general, seen less total depreciation than cities with lower priced homes.

To get specific, the city represented by the top-right blue diamond on the chart is Hillsborough, which has a current ZHVI of $2.5 million and has seen a 4% decline in home values since the peak in home values.  On the other end of the spectrum (in the bottom-left area of the chart) is Pittsburg (in Contra Costa County) where the current value of homes is $178,200 and home values have fallen 63% since their peak.  With this relationship, homeowners living in cities where homes are generally worth $500K have seen their home values decline by 17 percentage points more than homeowners in cities with homes worth $1M (a predicted peak-to-current decline of 33% in the former versus 16% for the latter).

Of course, not all markets are San Francisco.  In the Seattle metro, the performance of homes in the three price tiers has been much more similar over time than is the case in San Francisco (see the third figure below).  Seattle peaked almost two years after San Francisco, annualized depreciation has not been nearly as high as that of San Francisco, and the price tiers have roughly tracked each other closely on the way down.  When we look at home value declines by city across the Seattle metro as we did for San Francisco (see the fourth figure), we see a completely different picture than in San Francisco.  Specifically, there’s no real pattern between the typical home value of a city and the amount of decline experienced in real estate values since the peak. In other words, high-priced and low-priced areas of the Seattle metro region are doing about the same, in general.  Some high-priced areas have held their values well while other high-priced areas have not.  Same for low-priced areas.

The Washington D.C. metro is another region like San Francisco where there is a fair degree of difference in performance across the three price tiers (see the fifth figure nearby).  In the case of Washington, the top tier homes initially dropped more in the early period of the current housing recession, but then stabilized somewhat in 2007. Thereafter, the bottom tier of homes experienced higher rates of annualized depreciation in home values than the top tier.  All three prices tiers are currently seeing quickly falling rates of depreciation (i.e., they are close to being flat on a year-over-year basis).

Looking at the geographic disparities in home values in Washington (see the sixth figure nearby), we can see a more muted version of the pattern we saw in San Francisco in which cities with higher valued homes have, in general, held their value better than cities with lower priced homes.  One of the cities in the top-right of the figure is Chevy Chase, Md. (with a ZHVI of $792,800 and a 15% decline from peak value) and one of the cities in the bottom-left is Manassas, Va. (with a ZHVI of $181,700 and a 53% decline from peak value).  In Washington, homeowners living in cities with $300K homes have seen their home values decline by 10 percentage points more than homeowners in cities with $600K homes (a predicted peak-to-current decline of 35% in the former versus 25% for the latter).

Other cities that have shown differences in home value changes by price tier (like San Francisco and Washington) include New York, Los Angeles, Chicago, Philadelphia, Atlanta, Detroit, Boston, Phoenix, Riverside, Minneapolis-St. Paul, San Diego, St. Louis, Tampa, Baltimore, Denver, Sacramento, Cincinnati, Las Vegas, and San Jose, Calif. Cities that have not shown strong differences by price tier (like Seattle) include Miami-Fort Lauderdale, Portland, Ore., and Orlando.

One last interesting thing to note; in cities where you do find performance differences by price tier, some but not all of this phenomenon is related to the other pattern we’ve noted here at Zillow, regarding areas closer to the urban core holding their home values better than areas more on the periphery of the metro area (true in some but not all metro areas).

Notes about the ZHVI by Price Tier

We compute the price tier data as follows.  First, we assign all homes to one of three price tiers based on the percentile ranking of a home’s current Zestimate value.  Percentiles are computed for each metropolitan region separately.  We then compute the median Zestimate value of homes assigned to each price tier for each prior time period (keeping a home’s price tier assignment constant over time).
Dr. Stan Humphries is a real estate economist and real estate expert for Zillow. Stan is in charge of the data and analytics team at Zillow, which develops housing market data for most major metropolitan statistical areas in the U.S., and provides economic research for current real estate market conditions. He helped create the algorithms for the popular Zestimate® home value and the Zillow Home Value Index (ZHVI).