Housingwire cites Zerohedge? I like it.
But a deeper dive into the new home sales data shows that November’s increase over October isn’t quite as promising as it appears.
The folks over at Zero Hedge spotted a potentially troubling trend of the new home sales data being continually revised down, making each month’s figures look not nearly as positive as first reported.
For example, the initial estimate for October’s new home sales data was a seasonally adjusted annual rate of 495,000, so the revised data released Wednesday showed a downward revision of 25,000.
With November’s new home sales data printing at a seasonally adjusted annual rate of 490,000, November’s new home sales are actually lower than October’s initial total.
But comparing the November’s initial estimate to October’s revised total is not an apples-to-apples comparison, but a further look at the year’s data shows why November’s data could look a whole lot different in about a month when December’s data is released.
As Zero Hedge noted, the last five months have each been revised downward from the initial projections.
September’s initial report showed a seasonally adjusted annual rate of 468,000, but the current data from the Census and HUD shows that the revised total for September is actually 442,000, a decrease of 26,000 over the initial total.
The difference between August’s initially reported totals and the revised figures are even more significant.
When September’s initial report came out, the initial estimates had new home sales collapsing 11.5% from what was then August’s revised total of 529,000. But since then, August’s total has been revised even further.
August’s initial report showed at a seasonally adjusted annual rate of 552,000, but the current revised total is 507,000, a decrease of 45,000 from the initial report and a decrease of 22,000 from what was reported in October.
So the actual decrease from August to September was 84,000, not 65,000 as it was initially thought to be.
October, September and August were not the only months where the new home sale data has been revised down.
July’s initial report showed at a seasonally adjusted annual rate of 507,000, but the current revised total is 500,000, a decrease of 7,000.
June’s total was revised down from 482,000 to 469,000, a decrease of 13,000.
May’s total was revised down from 546,000 to 513,000, a decrease of 33,000.
And April’s total was revised down from 517,000 to 508,000, a decrease of 9,000.
So, for the last seven months, the average reduction from the initially reported new home sales data is approximately 22,570.
Subtract that average reduction of 22,570 from November’s total of 490,000 and November’s new home sales print at an estimated 467,430, which would represent a decrease of approximately 3,000 from October’s revised total, not an increase as was initially reported.
So, will next month’s report showed that new home sales actually fell in November?
The math certainly seems to show that that is exactly what happened.