Posts filed under “Data Analysis”
I was stunned to hear a few pundits proclaim yesterday’s rally was due to the 10AM announced
real estate contract signings. That ignores 1) the data; b) its source; and iii) the actual futures trading in advance of the actual data release.
Let’s put aside for the moment the source of the data — the National
Association of Realtors (NAR) — and delve into the data itself:
The Pending Home Sales Index, based on contracts signed in February, stood at 109.3 – down 8.5 percent from February 2006 when it reached 119.4, but is 0.7 percent higher than a downwardly revised reading of 108.5 in January. Earlier, mild weather caused the index to spike at 113.3 in December."
So from January to February, real estate contract signings to
buy existing homes appear to have increased 0.7%. This is measured by the NAR’s "index of signed purchase
agreements." Year-over-year, the contract
signings dropped 8.5%, hardly an encouraging data point, but that small fact didn’t seem to slow down the woefully misinformed pundits one bit. The monthly data was an improvement from January, which suffered a month-to-month drop of 4.2%.
So what does this small monthly gain mean? Let’s go to the NAR release:
"An index of 100 is equal to the average level of contract activity
during 2001, the first year to be examined and the first of five
consecutive record years for existing-home sales. There is a closer
relationship between annual changes in the index and actual market
performance than with month-to-month comparisons. As the relatively
new index matures and seasonal adjustment factors are refined, the
month-to-month comparisons will become more meaningful over time." (emphasis added)
Bloomberg notes that “The existing-home sales report is
based on a sample of about 40 percent of transactions in the multiple listing
service used by real estate agents, while the pending-sales index [the one we are presently discussing] covers about
20 percent.” That’s right — this 0.7% monthly gain is extrapolated from one out of five contracts. And as the NAR notes themselves, the year-over-year changes is much closer to reality than the month-to-month measure.
So back to our punditry: What does this data — a major negative, according to its source — have to do with yesterday’s rally?
Most likely nothing.
The inimitable Bill King points out that if any of the T-Heads bothered checking the overnight futures activity, they would have seen very clearly that the markets were being bought up way prior to the 10 am news release:
"A cursory look at a SPM chart reveals that the rally started just before Europe opened in the
wee hours of Tuesday. Yes, that is the window when people tend to commence gaming of the SPMs.
The second wave of the rally commenced about 7AM ET. The third wave of the rally started just before the 10AM ET release of the Pending Home Sales number."
The following chart is courtesy of King, who points out all the fun was had long before the 10 am release
SPMM tick chart – Arrow is release time of Pending Home Sales number.
(Time is CT in above chart).
"After the early rally, stocks traded sideways, amid excruciating ennui, for the balance of the session.
The rally, due to lack of later follow through and vigor, appears to be the handiwork of someone, or more, that exploited this week’s high absenteeism and lack of enthusiasm.
Ain’t high finance grand?"
Grand, indeed . . .
Pending Home Sales Show Effects of Weather, Possible Subprime Impact
NATIONAL ASSOCIATION of REALTORS
WASHINGTON, April 03, 2007
U.S. Economy: Pending Sales of Existing Homes Unexpectedly Gain
Bloomberg, April 3 2007