Generative AI for Analysts
Jian Xue, Qian Zhang, Wu Zhu
TL;DR
This paper provides causal evidence on the impact of domain-specific Generative AI (GenAI) on financial analysts' information production. Using FactSet's Mercury launch as a quasi-experiment, it shows AI adoption markedly increases report richness and timeliness, expanding information sources, topics, and analytic methods, while also speeding up issuance. However, forecast accuracy declines by about $59\%$, driven by synthesis costs from balancing more diverse signals rather than deterioration in signal quality. The results reveal a productive yet cognitively demanding AI-enabled regime, with stronger effects for analysts with higher workloads or limited processing capacity, and they carry implications for market efficiency and regulatory transparency around AI-assisted research. Overall, GenAI amplifies information supply and analytical reach but imposes new cognitive frictions in financial information intermediation.
Abstract
We study how generative artificial intelligence (AI) transforms the work of financial analysts. Using the 2023 launch of FactSet's AI platform as a natural experiment, we find that adoption produces markedly richer and more comprehensive reports -- featuring 40% more distinct information sources, 34% broader topical coverage, and 25% greater use of advanced analytical methods -- while also improving timeliness. However, forecast errors rise by 59% as AI-assisted reports convey a more balanced mix of positive and negative information that is harder to synthesize, particularly for analysts facing heavier cognitive demands. Placebo tests using other data vendors confirm that these effects are unique to FactSet's AI integration. Overall, our findings reveal both the productivity gains and cognitive limits of generative AI in financial information production.