Several days ago I received an email from Rick Davis from the Consumer Metrics Institute. In his email Rick provided a very interesting analysis regarding consumer spending and the economy.
As I read through the data that Rick presented to me I was impressed with the methodology utilized and how he arrives at some unique forward looking indicators. I started an exchange with Rick and asked if he would be interested in putting together a collection of his findings for my readers. I was happy that Rick accepted the invitation as I believe his research is unique and shines a new light on the claims of a strong economic recovery.
What was the Economy Really Doing … Yesterday?
By Rick Davis, Consumer Metrics Institute
Recent reports of a strengthening recovery are not fully supported by the behavior of consumers on the web. At the Consumer Metrics Institute we measure the depth and quality of web based consumer "demand" on a daily basis, and during this recovery the year-over-year changes in "demand" that we measure actually peaked in August 2009 and have been declining ever since.
In fact, our "trailing quarter" of web based consumer demand slipped into year-over-year contraction on January 15th, and since then we have been plotting the progress of this 2010 contraction event against the profiles of similar events in 2006 and 2008:
(click images for full size)

As you can see from the above chart the 2010 consumer "demand" contraction event is unique: if there is a "second dip" it may very well be unlike anything we have seen recently. Instead of a "call-911" type of event in 2008 or the "hiccup" witnessed in 2006, we may be seeing a "walking pneumonia" type of contraction that has legs.
In contrast to our measurements, on April 30th the Bureau of Economic Analysis (“BEA”) of the U. S. Department of Commerce published their latest reading of the state of the production (or “supply”) side of the U. S. economy. The BEA’s measurements of the economy are substantially “downstream” from the consumer activities that we measure. It simply takes many weeks for changes in consumer behavior to become reflected in production schedule changes at the factories.
Their measurement showed that first quarter 2010 factory activities were growing at a 3.2% annualized rate, equivalent to where our consumer “demand” side “Daily Growth Index” was on November 24th, 2009, roughly 18 weeks earlier. This means that our “Daily Growth Index” of consumer “demand” side activity is now leading the production oriented GDP by 18 weeks:

Compared to the 4th quarter of 2009, the annualized growth rate of the BEA’s official GDP has dropped by 43%. Depending on your point of view this could be interpreted either as a glass that is “half full” or a glass that is “half empty”:
1). The “half full” reading would mean that the GDP numbers confirm that the recovery has at least moderated to a historically normal growth rate. The good news is that this means that “the economy is still growing,” albeit at a historically normal rate. The bad news is that a normal growth rate would only warrant historically normal P/E ratios in the equity markets.
2). The “half empty” reading would mean that the near halving of the GDP’s growth rate confirms that (at the factory level) the economy has finally begun to “roll over” towards a “second dip”. If so, the BEA’s announcement portends even lower readings in the quarters to follow.
At the Consumer Metrics Institute, our measurements of the web-based consumer “demand” side economy support the “half empty” reading of the new GDP data.
A look at our “Daily Growth Index” also shows that towards the end of November 2009 the “demand” side economic activity was dropping so quickly that a two week change in the sampling period would make a huge difference in the numbers being reported. For the calendar quarter the annualized growth rate is the 3.2% reported by the BEA. If the sampling period had shifted to two weeks earlier, the reported GDP number would have been 4.4%, substantially higher. However, if the sampling period had shifted to two weeks later, the GDP growth rate would have been only 2.0%, less than half the reading from only 4 weeks earlier. This is the sign of an economy in rapid transition.
The methodologies used by the BEA when measuring factory production are ill suited to capturing an economy in such rapid transition. In the 4th quarter of 2009 the production side of the economy was topping, causing some consistency in the BEA’s consecutive estimates (5.7%, 5.9% and 5.6% respectively) of the quarter’s annualized growth rate. The first quarter’s production environment was at a much more dynamic spot in this particular economic cycle, and the subsequent monthly revisions by the BEA may be significant.
From our perspective the GDP is only confirming where our numbers were in November, which is (relatively speaking) ancient history. Since then we have seen our “demand” side numbers slip into contraction (on January 15th), and they have recently lingered in the -1.5% “growth” range.
We have long since recorded the “demand” side activity that has been flowing downstream to the factories during the second quarter of 2010. If the GDP continues to lag our “Daily Growth Index” by 18 weeks we should see the 2nd quarter 2010 GDP contracting at a 1.5% clip.
I say “should” because we have observed before that factories are loath to actually contract production until rising inventory levels force them to curtail normal production schedules and furlough staff. We saw this happen during the 2006 “demand” side contraction event, when the GDP production side growth effectively dropped to zero but never went negative. The 2010 contraction however is showing enough persistence that inventories are likely to eventually build to the point where production curtailments must be made.
In summary, our data is telling us that U. S. consumers are very reluctant to take on the kind of debt that they have traditionally assumed when pulling the economy out of previous recessions. Even a recent upturn in our retail index faded once the seasonal impact of the forward shifted Easter holiday had passed. Furthermore, even during the Easter retail up-tick the quality of the transactions was not very high. Big ticket items requiring longer term financial commitments were relatively scarce, and for that reason our Weighted Composite and Daily Growth Indexes did not materially respond.
Our mission at the Consumer Metrics Institute is to measure (on a daily basis) exactly how consumers are leading the U. S. economy. We "mine" nation-wide internet consumer tracking databases on a daily basis for early warnings about the demand side of the economy. Our data is significantly upstream economically from the factories and the products measured by the GDP, putting us far ahead of the traditional economic reports. Perhaps our data is too timely; we are so far ahead of conventional economic measures that our story generally differs (either positively or negatively) from the stories being simultaneously reported by more traditional sources.
Several points about the Consumer Metrics Institute:
1). We are not economists, formally trained or otherwise. We are simply geeks who are analyzing real-time U.S. consumer tracking data in search of macro-economic trends. On-line marketers use the same data to serve up focused ads or to offer customized product suggestions. Why governmental agencies have not realized that the same data is a gold mine of current consumer economic macro tendencies amazes us.
2). This is a revolutionary new daily source of spin-free hard data about the demand side of the economy. It is purely objective data collected daily from millions of on-line transactions by U.S. consumers. It does not involve any governmental sources. It does not utilize ’seasonal adjustments’ (all numbers are year-over-year). It is simply based on real-time U.S. consumer transactions (please see http://www.consumerindexes.com/Overview.pdf for more information).
3). I’m a physicist, so I understand numbers and the importance of monitoring physical systems in real-time — especially if you care about how the physical system is evolving and have some interest in keeping it from crashing. Measuring what was happening last quarter makes no sense (except, perhaps, for academic papers and governmental archives).
4). We’re not professional doom-sayers. In fact, we were wildly optimistic this time last year. We simply report the numbers — which at the moment just happen to be much less auspicious than the mainstream media has been reporting.
The Indexes themselves can be found at http://www.consumerindexes.com/index.html
Thank you,
Rick Davis
Consumer Metrics Institute
—–
I want to thank Rick for putting together this collection of his recent research. You can find more of Rick’s research at Consumer Metrics Institute
###