Fluharty, Andrew2020-08-252020-08-252020-04https://hdl.handle.net/11299/215156University of Minnesota Ph.D. dissertation. 2020. Major: Business Administration. Advisor: Pervin Shroff. 1 computer file (PDF); 47 pages.I examine the effects of herding by financial analysts on the wisdom of crowds. I identify significant herding of analyst earnings forecasts around forecasts of All Star analysts. Despite the fact that All Stars (and by extension those that herd with All Stars) are more accurate forecasters than their peers, I find that consensus forecast accuracy decreases in the degree of All Star herding, consistent with diminished group performance in the presence of herding as predicted by the popular wisdom of crowds literature (Surowiecki 2005). This is in contrast to prior studies that focus on herding around the consensus forecast and how it affects forecasts of individual analysts. Corroborating my main finding, I also find an increase (decrease) in consensus forecast accuracy in the period after an All Star discontinues (initiates) coverage of a firm. I further find that All Star herding is associated with an increased market response to earnings announcements consistent with herding leading to a greater amount of uncertainty that is resolved by a subsequent earnings release. Overall, my findings suggest that the detrimental effect of analyst herding on the wisdom of crowds leads to a deterioration of firms’ information environments.enaccuracyAll Staranalystsherdinginformation cascadewisdom of crowdsThesis or Dissertation