Results from a Private Survey of Consumer Conditions and Expectations: October, 2019

I have mentioned in other posts, here and elsewhere, that I am regularly running private surveys of consumer conditions and expectations. This is used as an input to proprietary “macro” trading I am involved in, and the outputs of those models will be kept confidential.

However, the data can also be processed to output metrics that are similar to those discussed in market moving data, such as the University of Michigan survey which is due to be released today at 10 am EST.

It’s very interesting to me to see how my data aligns with other people’s data, and that is a critical calibration point for me. It affects whether I view my measures to be “relatively true,” i.e. a measure that is only informative about the sample of the population who answered the survey, or “globally true” — by which a mean a genuine estimate of the unknown, but fundamental, state of the American consumer of which myself, the University of Michigan, and others, all make separate measurements.

Ex post addendum: the University of Michigan released the values of 95.5 for its own Index of Consumer Sentiment and 2.3% for long-term inflation expectations at 10 am EDT, 10/25/2019. These numbers are consistent with mine (although the monthly sentiment model I used here needs more data to converge, the confidence limits are very broad).