By JONATHAN LANSNER
THE ORANGE COUNTY REGISTER
Local home prices may not be gaining as fast as we think.
Buyers are certainly paying significantly more for homes today that they were a year ago as the real estate market emerges from its lengthy slumber.
Virtually all price barometers show local buyers paying up. DataQuick’s median selling price, for example, has found eye-catching double-digits gains for an overall Southern California median of $320,000 in February.
Such gains led some observers — and even DataQuick itself — to warn that some recent increases in this price measure are tied to a changing mix of homes sold. That is, while sales prices are up — the value of individual homes may not be rising as swiftly.
How so? A bit of mathematics, for those who can stand it!
For starters, remember that a median is the midpoint of a string of numbers. At times, this measurement can be tugged up or down by a significant swing of activity at either end of the data series. And Southern California’s housing market offers a prime example of the mechanics inside a median.
DataQuick’s Southern California median selling price includes sales activity for all residences — new and old, single-family and condo — in six counties: Orange, Los Angeles, San Diego, Ventura, Riverside and San Bernardino. For argument’s sake, these counties split nicely into three, roughly equally active housing groupings: the pricey, coastal living (Orange, San Diego and Ventura); inland affordability (Riverside and San Bernardino — the Inland Empire); and Los Angeles (the big middle).
A few years back I reviewed a study by Australians researchers who found that a median price trend could be slightly “improved” by adjusting the math for modest demographic (meaning, geographical) changes in buying patterns. I figured I’d try the experiment with local numbers.
DataQuick was kind enough to give me 15 years of monthly sales and pricing results for the six SoCal counties. I tossed that data into my trusty spreadsheet, and created my Geographically Updated Median Price – can you say “GUMP” — by rebalancing the series of selling prices to match the historical share of sales that each county created.
For example, high-priced Orange County residences averaged 15 percent of sales for this period. So by my math, each month Orange County’s price performance accounted for 15 percent of my regional GUMP index to adjust the median for the product mix flaw that crops up at times.
What did my new GUMP benchmark tell me?
In the year ended in February, SoCal home prices averaged 8.4 percent gains – somewhat significantly less than the 11.9 percent gains shown for the same 12 months running by DataQuick’s traditional median.
But does that mean that the past year’s been an outlier — and that recent gains are illusionary due to current trends not easily repeated?
Actually, no. When you look through my key prism — share of sales by county — you see recent sales patterns by county run virtually equal to the way home sales went since 1997. Yes, this year is a normal sales mix.
The oddity was the three years leading up to this recovery period. As real estate hit bottom, an early segment to find buyers was the cheapest end of the market. Financing was available for lower-priced homes — albeit, hard to get – plus investors liked these residences as rental opportunities. That made the Inland Empire — SoCal’s housing bargain — a hot property.
Inland Empire homes represented 33 percent of SoCal homes sold in the three years ending February 2012 – a noteworthy jump from the 29 percent share of sales activity seen in other years.
That increased share of cheaper homes sold — i.e. Inland Empire – around the market’s bottom boosted the losses shown by DataQuick’s traditional median. When that sales trend reversed, a stronger price recovery is revealed by that same median math.
Look at it this way: DataQuick’s SoCal median selling price fell 51 percent from peak to bottom. My own GUMP index that tries to correct for geographic product mix? It fell only 46 percent in the great slide.
Or the flip side: DataQuick’s SoCal median is now up 30 percent off its bottom — a sharper rebound than the 21 percent price reversal shown by my GUMP measure.
Look, to butcher a Forrest Gump line: “Economic indexes are a lot like a bunch of chocolates; you never know what you’ll get.” And a median has its mathematical limits.
Yet if you want to assume that the DataQuick median’s overstating today’s price gains — for example, if you’re a big housing skeptic — then you must accept that this same math “overstated” the previous price decline, too.