The Bureau of Economic Analysis will issue the 3rd quarter GDP (advance estimate) on October 29th. Would you like to know now what the estimate will be? I think we all would and perhaps we can get a glimpse thanks to Dr. Davis at Consumer Metrics Institute (CMI). If you're on the fence about whether the U.S. (world) economy is recovering or contracting, the following data just might help you decide.
Dr. Davis looks at what he considers to be the most important factor
in measuring real time growth, consumer spending, which as we all know
has accounted for about 70% of economic activity. He measures consumer
Internet spending which, he believes, yields the most current indicator
of what consumers are doing. In fact he literally measures what happened
yesterday, not 5-week old BEA data. Further, he says his data leads GDP
by 18 to 20 weeks.
Here is an example of his most recent Growth Index:
As you can see from this graph, their lead time has been fairly consistent.
Dr. Davis has developed a method to capture current Internet buying
data from major retailers. His method is proprietary and well thought
out. He then weights the data according to a good’s impact on GDP. He
weights his data according to the same weighting used in the BEA’s
National Income and Product Accounts (NIPA).
This weighting gives more importance to goods that have a bigger impact
on GDP. Dr. Davis claims his Weighted Composite Index is very accurate
with about a 20-week lead time.
Which goods he measures is interesting. He collects data reflecting
discretionary spending rather than the daily necessities such as rent,
food, and gasoline. He admits that there is a great deal of volatility
in his data, but he smoothes it out using 91, 183, and 365-day averages.
He not only measures the level of spending but the duration of any
“extended deviation” from the norm. This allows him to construct
“Contraction Watch” which shows how spending as a leading indicator is
expanding or contracting during a cycle. The following table is based on
a 91 day average:
Dr. Davis says that the current measures of economic growth used by the
Bureau of Economic Analysis (BEA), the NBER (National Bureau of Economic
Research, a private organization that determines the dates of cycles),
and the Conference Board use outmoded data.
If you go back to the old formula used to measure GDP:
GDP=C+I+G+(X-M) [GDP = private consumption + gross investment +
government spending + (exports − imports)], Davis would say that it is
more accurate to look at discretionary consumer spending versus putting
more weight on manufacturing (important back in 1937 when the formula
was constructed).
The following is an excerpt from the CMI website...
Our own headline for the end of September would be that both our Weighted
Composite Index and our Daily Growth Index have turned down again. In
fact, our Daily Growth Index has now reached a level exceeding the
lowest level recorded during the "Great Recession of 2008". [Please see
the above Contraction Watch chart.]
…[T]he "Great Recession of 2008" had a total of 793 percentage-days
of contraction over the course of 221 days, whereas the current 2010
contraction has already exceeded 690 percentage-days -- already over 87%
as bad. And the 2010 contraction has already lasted for 261 days, 40
days longer than the entire 2008 event. Additionally, within the past
week the 2010 event has reached levels of daily contraction worse than
anything recorded in 2008.
But looking ahead, should the 2010 event recover from its bottom
exactly like the 2008 event did, it would still experience nearly
another 490 percentage-days of contraction before ending -- resulting in
a grand total of 1180 percentage-days of contraction for the 2010
event, fully 49% more severe than the "Great Recession of 2008."
Our concern with the above argument is the implicit projection of the
blue line "recovering" in a manner similar to the upswing seen in the
2008 contraction. The 2008 "recovery" was aided by stimuli unlike
anything currently pending. And the 2008 "recovery" started with a
background level of 6% unemployment. Furthermore, we don't expect any
consumer driven U.S. economic miracles at least for the next 30 days --
until the results of the 2010 U.S. midterm elections are known. That
should add nearly 150 percentage-days of contraction to the above totals
before any "recovery" upswing could plausibly start, placing the
economic consequences of the 2010 contraction at least 1.67 times the
pain suffered in 2008.
Looking at the chart above, the striking difference between 2008 and
2010 is the implied longevity of the current event. Projecting forward,
we will probably see another 30 days of political "Fear, Uncertainty and
Doubt" ("FUD") pushing the blue line laterally to the right. And when
the blue line eventually starts back up, we face the real possibility
that the plateau visible in the left half of the chart's blue line is
the new consumer "norm," reflecting the realities of a deleveraging U.S.
consumer. If that is true, the economy's "800 pound gorilla" will have
gone on a serious diet.