Realtors Say Housing Is Rosy -- Fancy That

Consider this a Business Press Maven suppression effort. Don't ever, like Tuesday's Wall Street Journal, get lulled into using a particular trade association's data as standalone. Think tactically. Data from trade associations, no matter how convenient and digestible they make their statistics to the business media, must always be considered with a special degree of caution.

But The National Association of Realtors, which has caught plenty of reamings before, is known as one of the most partial. It has, for example, been busily marketing the real estate industry recovery since the moment its troubles began. So if you believe its statistics are givens, you'll have the wrong idea on everything from builders like LennarLEN and Toll BrothersTOL to the economy at large.

They Just Don't Get Durable Goods!

Which brings us right to the edge of Tuesday's blast zone. Here is the headline of the Journal story -- a front page job no less -- that summarizes its flawed thesis well: "Wave of Foreclosures Drives Prices Lower, Lures Buyers: Oversupply Triggers Lenders' Fast Sales; Mr. English Bids."

Got it? Real estate is back. The chief piece of evidence came in the fourth sentence. It involved, you guessed it, a single, solitary piece of data from the National Association of Realtors.

Here is the sentence featuring the offending statistic: "On Monday, new data suggested that pressures like these are starting to drive prices low enough to attract some buyers back into the market. Sales of previously occupied homes jumped 2.9% in February from the month before, the National Association of Realtors said, the first increase since July."

Hmmmm ... month over month data. That's risky stuff, using data that can be measuring nothing more than a seasonal shift in buying. At its extreme, this would mean trying to divine meaning from, say, the retail industry's November vs. its December. Headline: "Oh. My. God. Happy Times." Or its December vs. its January. Headline: "Oh. My. God. Terrible Times."

In reality, it's all naturally occurring seasonal shifts. No cause for celebration or alarm, which is not to say that, under isolated and very controlled circumstances, monthly data can't tell us something about the near-term direction of business. But you need two things.

Number one: you need a statistic on what that month-over-month growth looks like historically. Any up number can look good without setting it beside history. This is especially true (see: retailing, December) considering seasonality in many businesses. Number two: you need to compare that February not only to January but also to the previous February. Then make a judgment.

Somewhat incredibly, and quite unforgivably, the Journal failed to do either of these things. The Business Press Maven will do it for them, the lazy little devils.

Year-over-year February sales were down nearly 24%, which hardly seems to paint a picture of a market that is luring back buying. But since the Journal used the word "lure" and is talking about a near-term development, let's look at those January to February numbers in the proper perspective. Remember: we need to see whether there is a typical seasonality to them. And guess what! There is. Over the past four year, February sales have been more than 7 times greater than January sales. That means this year was less than half as good.