Skip to main content

by Mish

Jurow supplied detailed, repeat-sale statistics of the top 125 metro regions in the US in an Excel spreadsheet from RealtyTrac.

Housing Price Increases Less Than You Think

Let’s kick off the discussion with Jurow’s article Housing Price Increases Less Than You Think on Advisor Perspectives.

My independent analysis of the data including a huge skew caused by California follows.

Over the last 7.5 years, the Case-Shiller national home price index has increased 24.9% on a cumulative basis. But I have argued in numerous articles that that figure is grossly overstated. A new RealtyTrac report supports my claim, and shows the actual number is only 16%. Let’s take a good look at this report to find out what has really occurred in major housing markets around the country.

In April 2016, RealtyTrac published its US Home Sales Report for the first quarter of 2016. This included a detailed study of 125 housing markets in which a minimum of 300 sales were closed in March. RealtyTrac analyzed all sales for which they had data on the previous sale of that property.

Here is what they found. The median percentage gain for all 125 metros studied was a 16% gross gain (before commissions) from the time of the previous purchase of the house. Home sellers in all metros covered by the report owned their property for an average of 7.7 years. This represents an annual gain of only 2% per year.

In 15% of all these metros, sellers, on average, sold their property for less than what they paid. There was no recovery at all in those markets. West Coast metros showed the largest percentage gain over the previous sale of the home. Take a good look at the following table showing the best and worst metros.

The three housing markets with the highest profit percentages are all major California metros. This is no accident and not a surprise. More than a year ago, 40% of all the outstanding bubble-era non-prime mortgages in California had already been modified, a percentage much higher than any other large state. This percentage has risen steadily from 17% in early 2011. Because of these modifications, the overall delinquency rate is much lower in California.

The result has been the complete collapse of foreclosures in California – from 30,000 at the peak in August 2008 to a mere 2,000 in August 2016 according to the highly-regarded California Real Property Report. The removal of so many of the lowest priced homes from the market, has artificially inflated both Case-Shiller and median sale prices in California.

Housing demand has been stimulated in California by two major factors. One is the employment boom in Silicon Valley due to the tremendous growth of five Internet giants – Apple, Google, Amazon, Facebook and Netflix. The other is the huge influx of wealthy buyers from China looking for a safe haven for their money. This has caused the high-end markets in both the San Francisco and Los Angeles metros to soar.

Jurrow also provides a detailed explanation as to why Case-Shiller is wrong. His discussion is well worth a look.

Mish Analysis

There are a couple of minor errors in Jurrow’s presentation.

Starting with the most problematic, I cannot verify the article’s opening sentence: “Over the last 7.5 years, the Case-Shiller national home price index has increased 24.9% on a cumulative basis.”

A couple of charts will help explain that problem, while showing various ways in which one can twist numbers.

Case-Shiller Since February 2012

Image placeholder title

That chart shows a 37.6% rise in the Case-Shiller National Home Price Index.

But wait a second. That chart is from February of 2012. The opening line says “Over the last 7.5 years, the Case-Shiller national home price index has increased 24.9% on a cumulative basis.”

The RealtyTrac data goes to March of 2016. If we go back 7.5 years from March 2016, we see this.

Case Shiller March 2016 vs. August 2008

Image placeholder title

RealtyTrac says there is an average of 7.7 years (7 years 8 months) between sales.

The above chart dates back 7 years 7 months as the mid-point between the opening sentence and 7.7 years. No matter how I jiggled that timeline, I could not come up with 24.9%. Possibly there is a place, but I could not locate one.

The two preceding charts assume that Case-Shiller reflects reality, which Jurrow disputes. Let’s resume with that thought.

Problem Two

Image placeholder title

Chart duplicated for ease of discussion.

The above numbers accurately reflect RealtyTrac data (rounded to the nearest percent). Howerver, the chart skips Denver, and Denver is one of the top 10 metro areas in the nation.

According to RealtyTrac data, Denver has price appreciation of 42%. Houston, at 26%, would drop off the list.

There is another problem with that list, more subtle, and more problematic. The above numbers represent a single month, March of 2016 to be precise. The larger cities do not vary much month-to-month, but the smaller cities do, and on occasion, wildly so.

Scroll to Continue


My tables and analysis eliminate that problem by using a 3-month average. For comparison purposes, I also show a 6-month average. Three months would correlate more directly with Case-Shiller methodology.

Spotlight on RealtyTrac Data

Let’s now dive into the data.

The RealtyTrac spreadsheet contains 125 metro areas in which there were a minimum of 300 resales of the same house.

The following tables show those numbers various ways, with a primary focus on the affect California likely has on national numbers, and a secondary spotlight on very weak markets.

RealtyTrac Data for Top 20 Case-Shiller Metro Areas

Image placeholder title

Cities with an “*” are in the Case-Shiller top 10 list. The 3-month unweightedaverage price gain in the above list is 22.6%. Excluding three California cities, the average gain drops to 18.6%.

Chicago is the only top 10 city that is negative. Cleveland, a top 20 city is also negative.

RealtyTrac Data of Bottom 20 Metro Areas

Image placeholder title

The above table lists the bottom 20 out of 125 metro areas in the US. 18 of those 20 have negative returns.

Note that Chicago is in the Case-Shiller top-10 and Cleveland in the Case-shiller top-20. The other regions are not exactly small areas.

Congratulations (of sorts) to Rockford, Illinois. It is the third largest city in Illinois, and the blue ribbon winner for the worst real estate market in the country. Rockford Illinois, Cleveland Ohio, and Mobile Alabama were the only double-digit price losers.

RealtyTrac California

Image placeholder title

The above table shows every California region that RealtyTrac analyzed. In sharp contrast to the sorry list of bottom 20 losers, the unweighted average gain (treating Visalia the same as San Francisco) in California is 40.4%. The median gain is 38.3%.

The next table is very long. It shows all 125 regions.

Realty Trac All 125 Regions

Image placeholder title

Spotlight California

  • California has 5 of the top 7 performing metro areas
  • California has 7 of the top 10 performing metro areas
  • California has 15 of the top 38 performing metro areas

3-Month Average and Median Price Appreciation

  • Including California the average unweighted gain not counting kickbacks and commissions, was 16.2%.
  • Excluding California the average unweighted gain not counting kickbacks and commissions, was 13.8%.
  • Including California, the median unweighted gain not counting kickbacks and commissions, was 13.8%.
  • Excluding California, the median unweighted gain not counting kickbacks and commissions, was 10.1%.

Those are numbers from repeat sales. Unweighted means “not population adjusted, not time-price adjusted”. All of the areas are treated equally without regards to the number or price of actual sales.

Some of the numbers vary widely month-to-month, even in the same city. For example, Flint shows a 1.2% decline in the the latest month but a 31.9% gain last month. Flint is a good example of why it’s best to use a 3-month average.

In contrast to Flint, Rockford is uniformly bad. Flipping houses in Rockford has not been a winning move this entire recovery. The original spreadsheet shows Rockford has been negative every month since July of 2010. The median and average seller in Rockford regrets the purchase.

It’s important to note that all of this data reflects price gains or losses since the last sale of the same house, not monthly gains or losses.

The average holding period is 7.7 years but likely varies widely, another reason to use 3-month averages.

Assuming the data is reasonably accurate, if one subtracts kickbacks, sales commissions, maintenance and taxes, the median and average home seller made little, if anything, when selling. Excluding California, the numbers look much worse.

California Skew

California alone added 4.8 percentage points to the average sales price increase and 3.7 percentage points to the median sales price increase.

Use of unweighted numbers when calculating averages skews the results. However, population adjusted results of each region, median and average timeframes between sales in each region, and how close to the geometric center of each region each sale occurred are stats not readily available.

Results Vary Widely But Mission Accomplished

RealtyTrac repeat-sales data strongly suggests the housing rebound was weaker and far more uneven than widely reported.

Nonetheless, it’s important to step back and consider the mission:

For further discussion of the above tweet, please see …

Mike “Mish” Shedlock