We investigate whether two heuristics, the peak-end rule and herding, lead to cognitive biases in the index of consumer sentiment published by the University of Michigan. Both affect respondents' assessment of changes in their financial position over the past year. Consistent with the peak-end rule, respondents rely more on extreme detrimental monthly changes during the year than to changes over the whole year. We rule out that these extremes proxy for risk. The evidence for irrational herding consists in a too strong relationship from expectations about the future of respondents interviewed in a first round to assessments of the past by respondents interviewed in a second round. Both results show that cognitive biases can be found in a key macro variable and outside more controlled environments. They also indicate that the behavioral component of the sentiment index may offer another explanation for its relevance, next to news or animal spirits.