In the game of pin the tail on the number, today’s release on new home sales was a big miss. Wall Street conomists had expected a seasonally adjusted annualized number (i.e. seasonal adjustment error times 12) of 520,000. Instead, that number came in at 481,000. Are we surprised?

Hell no. Except this month’s number is just as misleading as last month’s, partly a result of the stupid practice of annualizing monthly seasonal adjustment errors or other one time anomalies in the data. In the big picture, what matters is the actual trend of actual sales and asking the question whether anything has changed. In that regard, the answer is still a resounding “NO” in terms of sales volume, but “maybe” in terms of the price bubble.

If you recall last month, the new home sales number game was a big “beat.”

U.S. New Home Sales Reach Highest Level in Seven Years

blared the Wall Street Journal. It subheaded:

Sales of newly built single-family homes increased 7.8% in February to annual rate of 539,000

Holy cow! This month’s 481,000 looks like a veritable crash. But it wasn’t. And the reversal was both predictable, and predicted.

I wrote last month:

That’s the ticket! Strong demand! Housing is back, America! Low pay, unqualified borrowers are back, and we’re selling over a half million houses annually!

Behind that headline, a bump in southerly migration joined with the usual random noise in February in other regions to send the number reported by the back slapping, self-congratulatory, Washington-Wall Street media echo chamber, to da moon.

As usual, they were annualizing a monthly, seasonally adjusted, abstract impressionist interpretation of loosely estimated reality. In other words they multiplied the seasonal adjustment error plus the huge sampling error that is a feature of the first release of this data, times 12. To its credit, the WSJ did point out in a later paragraph that “February’s advance estimate came with a margin of error of plus or minus 15.2 percentage points.” 15.2%! Are you kidding me! Why are we even discussing this number?

So annualizing the seasonally adjusted advance estimate launched the headline number into outer space, in an arc almost certain to crash back to earth in March and April.

The Commerce Department even has an official warning about the oversized revisions that follow the first release, after they’ve had a chance to look at the actual data as it comes in over the next few months.

“A necessary part of the process of issuing these early data involves the issuance of subsequent revisions. The revisions to new single-family houses sold estimates are primarily the result of the replacement of imputed data with data which are reported in subsequent months. New residential sales have larger revisions than other residential construction series. This is due to the fact that most of the new residential construction survey data are based on a permit being issued. Since many homes have a sales contract signed prior to a permit being issued, an estimate must be determined for these sales prior to permit authorization.”

In other words, they base their estimate for the current month on building permits. That means that if builders who have been sitting on their butts through a frozen winter decide to start a few extra spec homes to get a jump on the Spring selling season, the Census Bureau won’t know that until a few months later. If that’s what happened here, we won’t find out for another month or two as the revisions come in.

And that, ladies and gentlemen, is exactly what happened. Builders pulled a ton of permits in February. The Census Bureau seasonal adjusters and their idiot fans on Wall Street and in the Wall Street captured media read that as a big increase in sales. With the heavy permit pud pulling already done, March was quiet and that looked like weak sales. But the trend is essentially right where it has been since the “recovery”, such as it is, began. The focus on the month to month error prone seasonally adjusted headline noise is insane, but this is what the media reports and what the market pays attention to, at least for a few milliseconds before it goes back to the trend it was on.

The actual reported number of sales in March, per the survey, not seasonally manipulated, was 45,000. That’s the same as February. Usually, March sees a bigger jump, but again, February was unusually active permit wise. The year to year gain was 15.4%, which is right in line with the trend.

New Home Sales Trend Since 2010- Click to enlarge

Before we get too excited, this is only back to 2008 levels, right in mid crash. As a result of that crash, the housing industry no longer plays a significant role in boosting the US economy. Here’s the big picture.

New Home Sales Long Term Trend – Click to enlarge

But that hasn’t prevented a new price bubble from forming. New home sale prices rose by 48% between the bottom of the crash in 2009 and the price high in December 2014. Since then they have backed off a bit. Over the same period, average home size also increased, but not to that degree. The gain in average size from 2009 to 2013, the last available data, was 6.5% to an all time record of 2,598 square feet. So not only has there been a bubble in prices, but in size as well. The question now with both is whether this marks just a pause or the end. It’s too soon to tell. The trend still looks intact, but the surge late last year may have been a bubble blowoff, similar to what happened in new home prices in 2007, months after the broader housing bubble had peaked.

New Home Sales Price Bubble- Click to enlarge

These price increases also belie the idea that there’s been no inflation in recent years. There’s been plenty of inflation, just not the kind that conomists recognize.

Over the course of this bubble, virtually all of the gains in sales have been in the Sunbelt as older people continue to move south or buy a second home there when they retire.

New House Sales By Region – Click to enlarge

The data shows a massive surge in sales in the South and a good sized bump in the West, which includes sunny California and the Southwest. In the Northeast and Midwest sales have remained moribund for 6 years. The entire housing “recovery” story has been based on that southward migration, driven by the demographic wave of boomers retiring or nearing retirement. There’s no organic growth here, just people getting old and tired of freezing their asses off in the winter. They’re buying retirement homes or second homes down south.

One last point–along with this “growth” story, we must thank the almighty Fed for it’s perspicacity in promoting programs to help the average Joe. Here’s an example. It’s the spectacular growth of affordably priced homes for the “middle class” family. Since the Fed started QE and ZIRP sales to middle income families have been cut in half. Remind me again who Fed policy is supposed to be helping?

Thanks Fed! Click to enlarge