Preliminary June 2012 Michigan Consumer Sentiment Below Expectations

by Doug Short, Advisor Perspectives/

The University of Michigan Consumer Sentiment Index preliminary number for June came in at 74.1, a 5.2 point drop from the May final 79.3 and the weakest reading in in six months. Today’s number was below the’s consensus forecast of 77.

See the chart below for a long-term perspective on this widely watched index. Because the sentiment index has trended upward since its inception in 1978, I’ve added a linear regression to help understand the pattern of reversion to the trend. I’ve also highlighted recessions and included real GDP to help evaluate the correlation between the Michigan Consumer Sentiment Index and the broader economy.

To put today’s report into the larger historical context since its beginning in 1978, consumer sentiment is about 13% below the average reading (arithmetic mean), 12% below the geometric mean, and 13% below the regression line on the chart above. The current index level is at the 23.4 percentile of the 414 monthly data points in this series.

The Michigan average since its inception is 85.0. During non-recessionary years the average is 88.0. The average during the five recessions is 69.3. So the latest sentiment number of 74.1 puts us below the midpoint (78.7) between recessionary and non-recessionary sentiment averages.

The indicator can be somewhat volatile. For a visual sense of the volatility here is a chart with the monthly data and a three-month moving average.

For the sake of comparison here is a chart of the Conference Board’s Consumer Confidence Index (monthly update here). The Conference Board Index is the more volatile of the two, but the broad pattern and general trends are remarkably similar to the Michigan Index.

And finally, the prevailing mood of the Michigan survey is also similar to the mood of small business owners, as captured by the NFIB Business Optimism Index (monthly update here).

The trend in sentiment since the Financial Crisis lows had been one of slow improvement, but it topped out in February of last year at 77.5 and plunged to an interim low of 55.7 in August. The steady rise since the August trough has been encouraging. But today’s report raises the possiblity of reversal of the trend, especially in light of the weakening employment situation, both in the unemployment rate and initial unemployment claims.

Caveats on the Use of University of Michigan Consumer Sentiment

This survey is quantitatively derived from a fairly complex questionnaire (sample here) via a monthly telephone survey. According to Bloomberg:

This release is frequently released early. It can come out as early as 9:55am EST. The official release time is 10:00. Base year 1966=100. A survey of consumer attitudes concerning both the present situation as well as expectations regarding economic conditions conducted by the University of Michigan. For the preliminary release approximately three hundred consumers are surveyed while five hundred are interviewed for the final figure. The level of consumer sentiment is related to the strength of consumer spending. Please note that this report is released twice per month. The first is a preliminary figure while the second is the final (revised) figure.

This is a survey, a quantification of opinion rather than facts and data. The question – does sentiment lead or truly correlate to any economic activity? Since 1990, there seems to be a loose general correlation to real household income growth.

Related Articles:

All Posts on Consumer Sentiment

2 replies on “Preliminary June 2012 Michigan Consumer Sentiment Below Expectations”

  1. Let’s all “weigh in” here” regarding Sentiment Surveys: Beyond providing content publishers with stuff to publish and overanalyze, what good do they serve? Have you ever, or have you ever know anyone to ever, alter purchasing decisions as the result of reading that others opine differently as to the economy’s tone or direction than that instantly held by a would-be contributor to the economy?

    Some information is great, far to much of it, repeated ad nauseum, becomes overload-overkill!

  2. Thom,
    I am of two minds here.

    1) i agree with you because sentiment is a reaction to events, not a driver to events.
    2) we try to validate data every way we can. sentiment is a data validator.

    if the data was saying things were getting better – and sentiment was falling, it is saying the data is not telling us the whole story. with the data currently saying we are not in a recession, and sentiment at recession levels – do you believe the data or sentiment?

    i believe the average Joe is in a recession – and the data is not being presented in a manner which tells the whole story.


Comments are closed.