National Housing Stats and Lost in Translation


Photo credit: By chokola

The National Association of Realtors® (NAR) released their home sales statistics for July. This data is for all property types and reflects homes sales nationally. And According to NAR, the national median existing home price was up 9.4% July over July, while the number of existing home sales was up 2.3%.

As my local news ticker (I call him “Steve”) was reporting NAR’s numbers, real time, as they were reported to him by a perky morning news anchor, I was reminded that national statistics are just that – national. They are relatively meaningless to the couple three blocks away that is trying to sell their house, and they are meaningless to the buyer in my backseat trying to pin a value on the home they just visited.

So, just for giggles, I ran the July numbers for a few San Diego I-15 Corridor Zip codes to see how we stacked up. (Information is from the Sandicor Multiple Listing Service, all property types, July 2012 versus July 2011. Data is deemed reliable but not guaranteed, blah, blah, blah.)

Zip Code Homes Sold Change Median Price Change %Detached Sales 2012 %Detached Sales 2011
92131 (Scripps Ranch) +14% +1.2% 68% 67%
92128 (Rancho Bernardo) +57% +28% 58% 48%
92127 (4S Ranch) +13% +28% 80% 73%
92126 (Mira Mesa) +15% -6% 60% 71%

You can see that the overarching, “Everything is just ducky, thanks,” message from NAR generally holds true in these San Diego communities, with Scripps Ranch most closely tracking NAR’s national numbers. But, the numbers do vary widely.

The increases in numbers of homes sold are no surprise to those of us in the trenches. We’ve had quite the summer rally. What surprised me was the magnitude of the price changes in Rancho Bernardo and 4S Ranch. But this is where you have to consider the dangers of lumping property types.

The one thing that stands out is that we are comparing apples and bananas in 92128 and 92127. A higher mix of single-family sales in these areas, of course, resulted in a higher median sale price. In Scripps Ranch, where the percentage of detached versus attached sales remained relatively unchanged, so also did the median sale price. And in Mira Mesa where attached homes represented a bigger piece of the pie in 2012, the median price dropped.

The point then, assuming you are still awake, is that broad-brush statistics are fun and great, and can give us a feel for what’s going on. But, statistics alone lack soul. In order to know how your market is doing, you can’t rely on NAR, on Case Shiller, or even on my little MLS exercise without human intervention. Housing numbers require interpretation, and only a real human being entrenched in your local real estate market can provide the context.

Which brings me to the Zestimate, Zillow’s now-infamous “estimated market value.”

Last week, Steve and I represented three clients in closed transactions. Next week, we have four more closing. And while a discussion of Zestimates slapped onto the end of a discussion of NAR’s housing data may seem only loosely tangential, this was a good time for me to compare Zestimate accuracies – before they pick up the tax recordings and the Zestimates reset to sale price. And the comparison is relevant because it underscores the dangers in relying on data without soul, without human intervention to provide context. (Yes, real estate agents are, generally speaking, human.)

For our three closings last week, Zillow hasn’t picked up the recordations yet. I should also note that two of these were our listings, and we no longer gift our listings to Zillow. Similarly, in the case of the one buyer side, that listing agent does not syndicate to them. In other words, Zillow does not reflect these homes as for sale or having been listed, but they do provide a Zestimate for each.

Address Zestimate Sale Price Difference Days on Market
10555 Arbor Park Place, 92131 $582,989 $651,000 +12% 5
7696 Andasol Street, 92126 $429,472 $457,000 +6% 4
10528 Stony Ridge Court, 92131 $627,500 $689,000 +10% 3

You can see that the reality of our market was just a little off the opinion of the Zestimate algorithim.

As for the homes closing next week, all our listings, here is what the differences will look like.

Property Zestimate Sale Price Difference Days on Market
#1 $275,884 $300,000 +9% 4
#2 $669,258 $735,000 +9% 3
#3 $908,023 $960,000 +6% 11
#4 $369,421 $352,000 -5% 5

With the exception of one little condo, the Zestimates were all significantly lower than the buyers’ opinions of value, and it is only the latter that counts. You may think that 9% sounds close enough, but $25,000 – or $65,000 – sounds like a lot of money to me. And it is a lot of money to our clients.

Finally, and as a preemptive strike to those who might claim that our selling clients are being somehow being wronged by not having their listing data outsourced to third party syndicators like Zillow, I threw in the bonus market times. And in the case of all seven homes, I should point out that there were multiple offers – on every single one.

NAR is right, at least where our San Diego market is concerned. We do seem to be turning a corner. Our shortage of inventory coupled with continuing favorable interest rates is not an insignificant factor. We’ll have to wait and see how rising interest rates, a Presidential election, a fiscal cliff, or other external factors might affect us moving forward. But if you want the real skinny as it relates to you – your neighborhood and your home – look not to national stats or a mysterious home valuation program. Talk to somebody in your hood. Your mileage may vary.



Get your Instant Home Value…