By Bryan McCormick
This morning's preliminary Reuters/University of Michigan Consumer Sentiment Index data was a shocker. None of the estimates came anywhere close to the very sharp drop we saw today.
The numbers were blown out to the downside on every measure, with future expectations, often seen as a leading indicator of the willingness of consumers to spend, coming in at 60.6, versus expectations of 68.4 and a low estimate of 65.
In dissecting the report, I wanted to look at the "headline" number, which is the aggregated index including current and future outlooks. That was also a shocker, dropping to 66.5, from an estimated 74.5. The low forecast for the index as a whole was 71.
At right is a graphic of this index data, shown since the peak of sentiment back in 2007, with a quarterly moving average in orange. We can immediately see the relationship between the sentiment values and the way the stock market behaved.
There has long been a debate about the value of looking at the sentiment data, as it was often seen as a lagging response to the stock market. While high participation rates in the market were the norm, that might well have been the case. As we know, smaller investors have been exiting the market for some time as mutual fund flow data has shown, fleeing for the perceived safety of bonds.
Given the bottoming process in sentiment since the market lows in 2008 and 2009, along with the steady rise since, I think we have to interpret the data differently. It did a very good job of tracking growing confidence through 2009 and early 2010, and that was in essence a confirming indicator.
In an earlier piece, I had a look at retail sales data in context. In light of the recent decline in sales, I think we can see that consumers are concretely expressing discomfort with spending. The sentiment, and possibly the spending peak, may have both taken place for this cycle.
With today's break of the uptrend in sentiment, we could be in for a rougher ride than many have anticipated to date.
(Chart courtesy of DeepFoo Analytics)
