I've updated my Consumer Metrics Institute (CMI) charts through the latest data.
For those unfamiliar with these data series, here is a link to the Institute's website. Their page of frequently asked questions is an excellent introduction to the service. See also the Institute's March 22 commentary, The BEA's Third Estimate of Fourth Quarter 2010 GDP.
The charts below focus on the 'Trailing Quarter' Growth Index, which is computed as a 91-day moving average for the year-over-year growth/contraction of the Weighted Composite Index, an index that tracks near real-time consumer behavior in a wide range of consumption categories. The Growth Index is a calculated metric that smooths the volatility and gives a better sense of expansions and contractions in consumption.
The 91-day period is useful for comparison with key quarterly metrics such as GDP. Since the consumer accounts for over two-thirds of the US economy, one would expect that a well-crafted index of consumer behavior would serve as a leading indicator. As the chart suggests, during the five-year history of the index, it initially lived up to that expectation. Actually, the chart understates the degree to which the Growth Index leads GDP. Why? Because the advance estimates for GDP are released a month after the end of the quarter in question, so the Growth Index lead time has been substantial. However, over the past several months the correlation has disappeared. One speculation is that the Federal Reserve intervention with the rumor and subsequent reality of QE2 has stimulated the economy for the time being, but not the consumer.
Leading Indicator of the Market?
Has the Growth Index also served as a leading indicator of the stock market? It seemed to be during its first years of existence. The next chart is an overlay of the index and the S&P 500. The Growth Index clearly peaked before the market in 2007 and bottomed in late August of 2008, over six months before the market low in March 2009.
The most recent peak in the Growth Index was around the first of September, 2009. Since its peak, the Growth Index declined dramatically, entering contraction territory in mid-January of last year. The market showed signs of correcting in early 2010, which would approximate the lag for the earlier reversals. But that wasn't to be the case.
The CMI Growth Index contraction appeared to have bottomed in early October 2010 and started reversing, but 2011 has seen a renewed contraction. In contrast, the market continued its rally to late April 2010, corrected during the summer months, and then returned to the neck-snapping velocity of the spring 2009 rate of recovery.
Theoretically the notion that discretionary consumption leads the market seems reasonable. But the disconnect since early 2010 calls undercuts this assumption. On the other hand, we are living through some unusual economic times. It remains to be seen if the disconnect between the CMI Growth Index and the market is to some extent a result of the Federal Reserve's quantitative easing (illustrated here).
The next chart compares the contraction that began in 2008 with the one that began in January of this year. I've added annotations for the elapsed time and the relationship of the contractions to major market milestones.
Does the CMI Growth Index Provide A Comprehensive Snapshot of the U.S. Consumer?
In January Rick Davis, the founder of the Institute issued an important report on interpreting the index data: Reflecting Back on 2010 [Download PDF]. Davis essentially concludes that the index data is skewed toward a demographic of consumers who were more vulnerable to the economic downturn than the population as a whole.
In sum, we can't take the CMI Growth Index as an indicator of the consumer economy as a whole, but it clearly offers important insight into the health of the consumer. In fact, it is a demographic segment on which the future of the economy will become increasingly dependent. It seems difficult to imagine a sustained business cycle and long-term economic recovery while the CMI demographic sweet-spot sputters in a year-over-year contraction.