I generally shy away from sentiment surveys as almost a rule because I believe they are more often than not misleading. The various indices of consumer confidence fall into that categorization, especially over the past few years. According to many, consumer confidence is back to where it was during the housing bubble even though by nearly every other statistic, particularly those related to consumers and their spending behavior, is nowhere near the comparison (and often becoming much worse). Like JOLTS and its Job Openings series, however, there may be some use to them if in more limited fashion.
The University of Michigan Index of Consumer Sentiment absolutely surged between July 2014 and the end of that year. It was completely out of line with everything except the BLS numbers that were suggesting the "best jobs market in decades." At the time, I wondered half-heartedly if it was nothing more than a grand experiment in rational expectations, diving almost into the realm of conspiracy (really confirmation bias through trend-cycle bias) where economists "colluded" to rig the employment data in order to see if they couldn't get the recovery to show up by mere happy feelings alone. Obviously, it didn't work if that was what was tried.
Like so many other things, UofM's Index peaked in January 2015 and has been unevenly lower ever since. In the latest update for October 2016, the index fell back to 87.9, the lowest since last September (87.2) in the immediate aftermath of the first break of "global turmoil"; and significantly less than that early 2015 high of 98.1. Again, like JOLTS both Job Openings and Hires, this index of consumer "feelings" also appears to register "something" that changed at about the same time, making it more likely that this "something" was real even if the index itself might be misleading about the context inside which that shift occurred.
Apart from also suggesting the economic shift, the UofM data contains perhaps more relevant and worthwhile information as to why. Even though these kinds of survey-based data sets might be of dubious construction, you still have to pay attention to at least when true outlier positions appear. In terms of this consumer "sentiment" data, that means surveyed inflation expectations. As noted last month, both short- and longer-term inflation expectations have sunk to unusually low levels.
The downward turn in expectations occurs first in the summer of 2014 with the appearance of the "rising dollar", fully corroborating inflation breakevens concurrently moving against the recovery narrative. The next break lower happened last summer with the appearance of "global turmoil." More importantly, however, despite the constant mainstream emphasis of temporary and "transitory" as well as the imposition of "full employment" at every possible juncture, expectations have only continued to get worse even this year in a clear sign of shaken faith in central bank power and efficacy.
The update for October revised last month's inflation perceptions slightly higher, but replaced them this month with still lower numbers; the 5-year inflation expectation now the lowest in the entire series.
This isn't necessarily important as to suggesting exactly what people expect inflation might actually be in the future, I believe it is more like what I wrote yesterday about the curiously sideways stock market:
It is, I believe, the possible last remnants of QE religion in stocks. Investors were willing to pay up for just the prospects for rapid growth that especially QE3 and QE4 would surely deliver. Like economists, stock holders viewed the arrival of the "rising dollar" with suspicion but still holding on to that QE-positive scenario as if only delayed by it.
The continuation of 2015 processes well into 2016 increasingly threatens those expectations. In other words, a "market" waiting for earnings to "correct" higher fulfilling if belatedly those expectations of QE effectiveness begins to understand that just isn't very likely, then expectations start to turn toward paralysis as to what other piece of P/E might have to instead "correct" to restore much needed balance.
Inflation expectations in this context are also a reflection of increasing doubts as to whether QE will work, or in fact had ever worked. Bringing these two parts of the UofM survey together ,then, suggests as a great deal many other accounts that "something" changed coincident to the "rising dollar" and that more and more people believe monetary policy both didn't stop it and likely won't (or can't).