by Rick Davis
We continue to analyze our data in an effort to understand what recent on-line consumer behavior can tell us about the current U.S. economy. The changes in our consumers’ behavior since late in the second quarter are just now beginning to show up in the widely reported (but substantially lagging) retail and housing surveys, and judging from the multi-month persistence of the changes in our data we can expect that more lagging surveys will continue to look good (at least from a year-over-year perspective) for at least the next quarter. Meanwhile we can also expect that the broader economic data (which is yet even more lagging) will continue to deteriorate, following the downward tendencies of our data during most of the first and second quarters of 2011.
The nature of these changes remind us of several things that we learned in 2008 and 2009:
— In general, consumer behavior substantially leads the overall economy — particularly if governmental purse-strings are restrained from creating short term distortions.
— Actual on-line consumer behavior also substantially leads “confidence” survey data. This signals some counter-intuitive psychological aspects, as the initial upswing in actual consumer spending generally precedes the onset of a true comfort zone — meaning that much touted confidence surveys are (from at least our perspective) more lagging than their proponents would lead you to believe.
— Consumer spending behavior can de-couple from consumer fundamentals for at least a while. Spending can rise even as unemployment stagnates, real income shrinks, and credit stress becomes more acute. Selective consumer spending is a form of self medication for a collective state of psychological depression, as Hollywood learned well during the 1930s.
— Consumer spending can increase even in the face of continued deleveraging, especially if a portion of that deleveraging is in the form of loan repudiation. And spending on relative trinkets does not necessarily negatively impact the overall credit profile of a consumer, even if some of those “trinkets” comprise major appliances or even replacement autos — which incidentally may also be among the best ways for a household to begin resuscitation of a compromised credit rating.
— Consumer spending is sensitive to consumer inflation expectations. There can be cross-currents between short and long term expectations, with actual short term disinflation experiences generating looser budgets at the same time that longer term expectations encourage moving consumption forward.
— In this U.S. economy not all consumers have shared the pain. Demographics have been particularly brutal this time around, arguably more so than 75 years ago. And the demographic sword has sliced and diced U.S. consumers in complex ways, with overly simple filters on age, education or ethnicity missing the true intricacies of the economic pain. The demographics of where you were in the employment food chain mattered more, as did when you went to college or when you formed a family and purchased (or last refinanced) your home.
The changes that we are seeing now are highly reminiscent of what we saw start in early November 2008, at least a quarter ahead of the official bottom of the “Great Recession” and some 127 days before the equity markets turned.
We start with some of the old standards before looking at other charts we haven’t highlighted recently. Our Weighted Composite Index has been moving in rarefied air for some time now, although the peaks from very early October remained some 4% below the record highs set in mid-August:
This slight decline in the Weighted Composite Index can be seen towards the right margin of our chart of the weekly average values for the past year:
as well as the monthly averages for the past four years:
The mathematics of our trailing 91-day Daily Growth Index, on the other hand, is still being impacted by the dramatic rise seen in our daily indexes during July and August, and we have had to re-engineer the scales of its charts several times to capture the almost ballistic nature of its course — which should start to moderate (i.e., “top off”) in the near future:
(Click on chart for fuller resolution)
(If a chart is not visible above, please click here to see this commentary as a Web Page.)
To keep all of these year-over-year views of on-line consumer demand in perspective, we have recently introduced two charts which translate the year-over-year data into “absolute” demand charts (in which the average consumer demand level during 2005 is shown as the reference level of 100). First, a chart of the past 60 days of the daily values for this “absolute” demand:
(Click on chart for fuller resolution)
And below is the chart for the monthly averages of the Absolute Demand Index for the past four years:
(Click on chart for fuller resolution)
Both of those charts clearly show that despite the recent year-over-year surge in consumer demand, the “absolute” demand is still nearly 10% below where it was (on average) in 2005. Furthermore the 48 month chart (shown directly above) highlights the full extent of the contraction event that we were monitoring during 2010 — something that has yet to fully register in either the overall economic data (because of the distortions caused by 2009’s Federal fiscal “ARRA” and two rounds of Federal Reserve monetary “QE”) or the equity markets (because of the ability of the elite large-cap corporations to shield their earnings from U.S. consumer pain). Our guess is that the full extent of that contraction has not been completely avoided yet; it is likely to be merely another item being “kicked down the road.”
The Usual Suspects
Our Weighted Composite Index is (as its name implies) weighted towards the economic impact of the transactions that we capture, using the BEA’s NIPA tables (from our base year of 2005) for the weighting. As a consequence (and as you might imagine) the automotive and housing sectors are the most heavily weighted among discretionary consumer expenditures — as they have been for the bulk of the past 100 years. To understand what is driving the behavior of our Weighted Composite Index it is useful to look at the recent history of the sector indexes for those segments of the economy:
Both areas are showing substantial strength, which is broadly supported by the sub-indexes within each sector — as exemplified by our charts for the domestic and luxury brands of autos:
and the charts for loans for newly purchased homes and refinancing of existing homes:
And the housing sector is represented by more than just home purchases, with the home improvement stores showing recent strength as well:
Taken together these charts tell us a great deal about the reasons for the rise in our year-over-year data — which must always be interpreted in the context of the abysmal numbers from this time last year.
And Now For Something Completely Different …
Some of our other charts show signs of self-medicating indulgences on the part of our consumers. We have always been intrigued by our Recreational Apparel chart, which may highlight some gender differences in the spending:
This can be contrasted with on-line consumer demand for high-end digital cameras, which are certainly not showing signs of renewed life — and may just be a reminder that the expansive conversion to digital cameras was last decade’s phenomenon:
And we may be seeing some loosening of budgets controlled by younger males:
As long as we remember that these numbers are year-over-year against generally pathetic year-ago reference points, we can also see that the travel industry has at least stabilized, if not bottomed:
Finally we would be remiss if we didn’t revisit one of our all-time favorites, our Personal Finance index. This chart is a composite of (inverted) consumer demand for products that they seek when confronted with severe financial stress (e.g., credit counseling services, bankruptcy attorneys or foreclosure specialists). The index is inverted, so that higher values signify less stress and lower values indicate greater stress:
In July and August our consumers were feeling pretty good about themselves, although we can never be sure exactly why. But nevertheless that confidence likely translated into the behavioral changes that caused the sharp rises we have seen in our composite indexes. In retrospect the chart also explains to some extent why our other indexes were so pathetic during the first and second quarters of 2011.
That said, we take some caution from the fall-off in this chart over the past eight weeks. Maybe the de-coupling between consumer spending and the underlying fundamentals of household economics can only last so long …