Algorithm Reactance in Online Investment Communities
Prof. Zhiya ZUO
Associate Professor
in Business Intelligence and Analytics
FBA, UM
Date: 14/04/2026 (Tuesday)
Time: 13:00 to 14:00
Venue: FBA Lobby
Abstract
Algorithms are being increasingly adopted in online investment communities (OICs) to predict stock performance for OIC content consumers (i.e., investors). Nevertheless, such predictions pose an identity threat to OICs’ key content producers (mainly non-professional human analysts), who forecast stock performance through their content generation in OICs (i.e., analysis articles). We investigate how analysts adapt their content generation in response to the introduction of algorithmic predictions in OICs, as well as these responses’ subsequent impact on their stock forecast accuracy. Drawing on identity control theory and coping theory, we theorize algorithm reactance as analysts’ coping responses under algorithmic identity threats, which manifest in two forms, namely, algorithm differentiation and productivity boost. We further delineate the contingent roles of algorithm performance and analyst expertise/reputation for the two forms of responses, respectively. Applying a counterfactual design and a difference-in-differences design, we test the reactance responses using a unique dataset from Seeking Alpha, an OIC that rolled out algorithmic predictions in May 2019. Our results attest to nuanced algorithm differentiation responses—analysts exhibit convergent differentiation with a slight deviation from, but in the same direction as, the algorithm when the algorithm is accurate; by contrast, they exhibit divergent differentiation by opposing algorithmic predictions when the algorithm performs poorly. Moreover, analysts boost their productivity in producing analysis articles, and this effect is stronger among those with greater expertise and those with higher reputation. Finally, we find that algorithm differentiation improves analysts’ stock forecast accuracy, whereas productivity boost diminishes it when the boost was achieved at the expense of analysis depth. Our study reveals novel algorithm reactance responses in a human–algorithm interaction context and offers managerial implications for algorithm-empowered OICs.
Speaker
Zhiya Zuo is an Associate Professor in the Department of Accounting and Information Management at the University of Macau. He received his Ph.D. in Information Science from the University of Iowa in 2019. His research interests primarily focus on digital platforms, Web3 technologies, and business analytics. His work has been supported by the Hong Kong Research Grants Council and the National Natural Science Foundation of China.
All are welcome!
