That is the Gennaioli, Ma, and Shleifer paper being presented at the NBER macro conference, pdf here. There are two key points to this paper. First, actual data on the expectations of corporate CFOs have predictive power for investment, even when Tobin’s Q is measured. Second, expectations about future earnings are not rational in the Lucasian sense. I’ll update with remarks in the comments section of this post as the discussion proceeds (feel free to leave comments on my comments), again note that I am not allowed to attribute specific comments to individuals other than the presentation itself.