Wilson, RyanPark, Seyoung2024-01-102024-01-10https://hdl.handle.net/1794/29228Despite growing evidence managers learn information from stock prices that guide their investment decisions, the forms of information that underlie this learning mechanism are not well understood. This paper explores whether information produced by investors’ in-person human interactions is a key form of this information. Using foot traffic based on GPS location data from customers’ smartphones as a proxy for human-interaction-based information in stock prices, I find that investment-q sensitivity increases with foot traffic, consistent with managerial learning from prices increasing with the amount of human-interaction-based information in prices. To mitigate omitted variable bias, I use lockdowns triggered by the COVID-19 pandemic as exogenous shocks to information produced by human interactions. I find a decrease in investment-q sensitivity during the pandemic. The decrease is more pronounced when foot traffic decreases in places where human interactions are most likely to produce new information (e.g., cafés and restaurants) and among local firms, for which human-interaction-based information production was more active pre-pandemic. I further find that the decrease is more marked among young and growing firms, which investors have a comparative advantage in evaluating. Lastly, I show that my findings are not explained by noise trading, financial constraints, managers’ direct acquisition of human-interaction-based information, and local economic conditions. Taken together, I provide novel evidence of human-interaction-based information being a key form of information underlying managerial learning from stock prices.en-USAll Rights Reserved.foot traffichuman interactioninvestment-q sensitivitylockdownmanagerial learningsoft informationtestingElectronic Thesis or Dissertation