Behind the Adjusted Real Estate Numbers

It doesn’t take much to recognize back-to-school season, winter season, or allergy season, for that matter.

Stores know when to stock up on pencils and backpacks, and all of us know when to break out the cold-weather gear, or the tissues and Benadryl.

On the flipside, we know not to be surprised when these same cyclical surges come to an end.

We know the purchasing of new backpacks will slow down in late September and that cold-weather and hay fever won’t last forever.

Statisticians know this, too, and they try to adjust for these cycles in data.

The problem is, not everything that happens needs to be adjusted.


Yet it seems that sometimes people can’t seem to help themselves. They end up tinkering with the data, and misrepresenting reality.

Right now, this is happening in real estate.

The Case/Shiller Home Price Index (CSHPI) is being seasonally adjusted to a positive number, and there is no good reason why.

The selling cycles in real estate are well known. One time of year is the high-flying, spring-selling season, when everyone is trying to get their situation figured out before the summer.

Another is the back-to-school season, when people scramble to get moved before school starts in earnrnest.

A minor selling opportunity occurs over the winter holidays, but it’s short. On the flip side, in the dark days of winter, absolutely nothing happens.

If a person simply looked at the number of homes sold each month, it would look like a repeating patternrn of waves, albeit at different heights, each year.

To smooth out the waves in order to make month-to-month comparisons useful, statisticians estimate the percentage of home sales that happened, just because it was high-selling season, or didn’t happen, because it was the dead of winter, and then adjust accordingly.

While a person could make many arguments about how these adjustments are added, few suggest that the adjustments aren’t necessary.

However, these changes are made on the number of sales.

The Case/Shiller Home Price Index doesn’t address how many homes were sold, but only the price at which homes were sold. When it comes to price data, the time of year should not matter.

No one puts his or her home up for sale and then, looking at the calendar, thinks: “Oh, it’s January. I need to lower the price by 3% from what I think I would have gotten last September.”

People who argue for adjusting price data by season typically point to transactions that might not hinge on getting the best price, like foreclosures.

In the months when traditional sales by owners are low, foreclosures will make up a higher proportion of units sold, thereby artificially dragging down prices.

In the months with more traditional sales, the discounted foreclosure sales will have less of an effect on prices.

But this logic is based on the notion that buyers won’t seek out foreclosures in heavier-sales months, and that sellers of foreclosed property won’t try to get the best pricing, no matter what time of year it is.

Neither of those assumptions makes a lot of sense.

While the pros and cons can be argued, the data speaks for itself. The seasonally adjusted Case/Shiller Home Price Index was up 0.8% in January, the latest data available. Unadjusted, the measure was down 0.1%.

This nearly 1% swing for the month of January is important, particularly given that it moved the measure from positive to negative and it belies a trend. On an unadjusted basis, the index has been down for three months in a row, while the seasonally adjusted figure has remained gleefully positive.

As Marcellus in Shakespeare’s Hamlet said: “Something is rotten in the state of Denmark.”

To figure out what measure more closely resembles reality, I looked at research from other groups. A newcomer in the space, Zillow, compiles the Zillow Home Value Index (ZHVI).

Its report for January showed a 0.2% increase in home prices from December, which is above zero but much closer to the Case/Shiller unadjusted number.

Interestingly, Zillow issues its research on a timelier basis than the Case/Shiller, so the February Zillow Home Value Index is already out. Those data showed a further decline to a 0.1% increase.

The trend here is pretty obvious: The increase in home prices has essentially stopped.

With all this in mind, it’s probably wise to discount the seasonally adjusted Case/Shiller figures, even if they paint a rosy picture that so many people want to believe. Instead, look at the unadjusted numbers.

What the more honest numbers show is that after a good run over the last two years, home prices are softening.

If you were thinking of selling a house before the markets roll over, doing so now looks pretty good.


P.S. For another angle on the housing market, make sure to read Harry Dent’s latest research on the subject here. Spoiler Alert! — His title, “This is NOT a Good Sign for the Housing Recovery,” offers you a clue to his point of view.

Follow me on Twitter @RJHSDent