Managerial Learning from Decoding Noisy Stock Prices: New(s) Evidence from Billions of Internet Article Reads

Abstract

A long literature argues corporate managers learn from stock prices, but a firms’ learning process is challenging to observe. We present a novel test using firm-level readership of financial media articles as a measure of managerial learning behavior. We hypothesize that reading financial media, perhaps alongside other unobserved learning activities, helps managers interpret noisy signals in stock prices. We show that the classic Q-sensitivity of R&D expenditure increases by 26% when firms’ reading of financial articles increases by one standard deviation. This relationship is driven by reading from near the headquarters and by articles likely more informative to managers.

Keywords: big data, managerial learning, market feedback effects, financial news, R&D.

JEL Classification: C55, G11, G23, G24, Q01