The Reality Behind The New Home Sales Number

Tyler Durden's picture

Yesterday's "better than expected" New Home sales served as the "good news" pre-market boost to send futures ramping higher once again, if not enough to cause a fresh all time high.

Here is what really happened when one spreads the numbers, courtesy of Mark Hanson's housing blog.

This New Home Sales data are virtually worthless due to what I call "the laws of small numbers".
 
That is, unlike Existing Sales that averages about 400k a month, New Home Sales averages under 35k, with the off-season winter months when the seasonal adjustments are the greatest in the high 20k's.  In fact, as a percentage of total US house sales they have never been lower at about 7%.
 
This presents a problem when the Census Bureau counts 'sales', as  it rounds up to the nearest thousand.
 
Bottom line, in the CB data below, last month it says a rounded-up 34k houses sold. Last Jan they say 32k sold.  With an 8% RSA, this is flat YoY at best.   Also note, the volatile South was responsible for all the 2k gain plus another 2k. 

Rounding up...  See data below.  If all of the 4 regions were in this morning's New Home Sales print were rounded down to the nearest thousand by the Census Bureau vs up, it would subtract 4k sales, or about 12%.  Even with the massive January seasonal adjustments, this would result in a SAAR headline print of 428k, or flat YoY vs the up 10% reported.   If only the South was flat YoY like the other regions, the same thing would occur. 
 
This is a shame, because New Home Sales have always been a great way to track "end-user" demand, as this segment isn't driven by investors like the Existing resale market.  This leaves the only real way to gauge the true, fundamental health of the "end-user" housing market NAR's existing sales and the first-time buyer cohort.    
 
1)  Census Bureau New Home Sales NSA beat last year by 2k houses with ALL FOUR regions rounded up to the nearest thousand.

In other words, with an 8% RSA, there is no change YoY.  Moreover, if the volatile Southern region was flat or lower YoY like the other regions, the headline 10% YoY gain would have been flat YoY. 
 
2)  I am exiting my long builder trade reco before Existing Home Sales was released last week into this squeeze...that 10% was too easy, I am a chicken, and housing far too volatile in here to fall in love.