The new home sales top-line number announced yesterday by the Census Bureau and the Department of Housing and Urban Development fell way short of housing analyst expectations. So, what happened? Compared with a strong June 2017 number, and an even stronger July 2016 number, the seasonally adjusted annual rate of 571,000 new home orders in July was a sharp surprise to the negative.

Most analysts focus, for perspective's sake, on two areas of yesterday's data print. One is the "noise" in the data. The other is "the tell" in regional detail that may [or may not, since we're merely talking about one month's performance] reflect the beginning of a price mix shift to more attainable payment levels for would-be home shoppers.

Let's look first at the "signal vs. noise" point of focus, the headline data point of 571,000 seasonally-adjusted, which is a 9.4% decrease sequentially from June's total, and an 8.9% year-on-year decline.

Note that the comparisons are to both the strongest month in new home sales thus far for 2017, and to the strongest month, period, for the whole year in 2016.

What's more, the Census, HUD, and the Commerce Department have readjusted--dramatically upward--the NHS tallies for April, May, and June of 2017, so there's no reason to believe July's first-print level will get similar treatment in revision.

The not-seasonally-adjusted total number of homes sold in July 2017, notes Calculated Risk's Bill McBride, was 49,000, compared with an NSA mark of 54,000 in July 2016.

Now, if you look at the revisions to the new home sales data over the months of April, May, and June of this year, which have added 46,000 new home sales to those three months, adjusting initial estimates significantly.

A revision consistent with those of the past few months would more than eliminate the 5,000 unit difference in not-seasonally-adjusted sales separating a very-strong July 2016 from a disappointing July 2017.

What's more, through the first seven months of 2017, new home sales are running 9.2% ahead of 2016's level for the same period.

So, the other key theme in first-blush observations about yesterday's new home sales release emerges out of looking at the regional detail. National Association of Home Builders economist Michael Neal has commentary that exposes an interesting, possibly very significant wrinkle in the data. Neal writes:

The monthly declines took place in the regions of the country that are furthest ahead over the first seven months of 2017 relative to the first seven months of 2016. After reaching their second highest Post Recession level of 630,000 in June, sales of new homes fell to 571,000 in July. Three of the four regions of the country experienced declines: the Northeast (-23.8 percent), the West (-21.3 percent) and the South (-4.1 percent), but the Midwest region recorded a 6.2 percent increase in sales. Over the past 12 months, sales nationwide fell 8.9 percent, reflecting decreases in the Northeast (-13.5 percent), the Midwest (-12.7 percent) and the South (-11.7 percent). Meanwhile, sales in the West are 1.4 percent above their level 12 months ago.

Here's where the wrinkle comes in, and where we see a tipping point that will impact sales unit volume as long as builders can leverage their capacity to deliver more homes. Zillow senior economist Aaron Terrazas writes:

The median seasonally adjusted price of new homes sold in July increased 2.7 percent from June to $320,100 and, though it was shy of the recent high of $326,100 reported in May, it is up 6.5 percent from last July. The pace of new home price growth has slowed in recent months: Looking at the centered 12-month moving average of year-over-year median sale prices, new home prices are now growing at an annual pace of around 1.4 percent compared to a pace of around 5.2 percent this time last year, and a pace of around 7 percent in July 2014. This primarily reflects the efforts of many builders to build more affordable homes, a market segment that is both attractive to more buyers and more difficult to address.

Without doubt, builders will be able to sell homes that price in at the $250k level, vs. the current median price of $320k. The challenge has been leveraging resources to expand supply at that lower level and deliver it, rather than simply selling it.

Many builders are fast at work, ramping up their entry-level product and community counts, but they're on a learning curve that doesn't happen overnight. Once more of them hit a operational comfort zone with those lower-priced new homes, it's off to the races.