Cognitive Biases in Consumer Sentiment: the Peak-End Rule and Herding

Abstract

We show that two heuristics, the peak-end rule and herding, generate biases in the University of Michigan's index of consumer sentiment. Both affect respondents' assessment of past changes in their financial position. Conform the peak-end rule, their answers relate more to extreme detrimental monthly changes than to yearly aggregate changes in key financial and macro variables. These effects are stronger for respondents for whom particular variables are more salient. The evidence for irrational herding consists in the answers of second-round respondents being too strongly related to future expectations of first-round respondents. These effects persist when we account for structural differences in sample composition or for the effect of other predictive variables. Our results show that these biases are not restricted to controlled environments. The resulting behavioral component can explain the relevance of sentiment indexes.

Keywords: consumer sentiment, cognitive biases, peak-end rule, herding, feedback loops

JEL Classification: G41, E71, E32