Paper empirically tests the timing of news pricing and market reaction patterns
Get up to speed
Alireza Kargarzadeh et al. released a revised paper that empirically tests when news is priced in the market based on 4.57 million news articles for approximately 3,000 US stocks from 2023 to 2026. The paper uses a compact classifier distilled from a large model teacher to label news with 17 event categories and 5 attributes, clustering them into 1.68 million stock-day events, of which 364,405 are neutral sentiment events as a placebo. This revised version has been accepted by the 3rd Workshop on LLMs and Generative AI in Finance (AI4F 2026), which is held in conjunction with ICAIF 2026 in Milan, Italy.
AI-generated from reports · 3 hr agoUpdated
Coverage timeline
Follow the coverage to understand different sides of the event.
- arXiv q-finBuy the Rumor, Sell the News: When News Gets Priced In
Based on 4.57 million news articles covering approximately 3,000 US stocks from 2023 to 2026, the paper uses a compact classifier distilled from a large model teacher to label news with 17 event categories and 5 attributes, clustering them into 1.68 million stock-day events, of which 364,405 are neutral sentiment events serving as a placebo.
Trend of this event
Not enough consecutive observations yet; trend is not drawn.