News Reaction Intelligence: The Market Has a Memory
TL;DR
News Reaction Intelligence is the measurement of how stocks actually behaved after classified news events: the price move, the volume, and whether the move continued or reversed over the following days. News Quantified maintains the reference dataset for this category: 26 million+ news records spanning 20+ years, validated in peer-reviewed research and licensed continuously by a leading systematic fund for over a decade. It does not predict. It measures.
The industry sells everything except the reaction
Every trading day, thousands of corporate news events hit the tape. Guidance cuts. FDA decisions. Offerings. Downgrades. Contract wins. Each one triggers a reaction: prices move, volume surges, and within days the move either continues or reverses.
This has happened millions of times over twenty years. The pattern of those reactions is measurable, classifiable, and durable. Yet almost every data product in the market stops at the moment the news arrives, which is exactly the moment before anything interesting happens.
Sentiment vendors score the tone of a headline. Terminals deliver the story faster. NLP startups summarize the article. None of them answer the question a decision-maker actually has when news breaks:
What has the market done, historically, after events like this one?
What News Reaction Intelligence is
News Reaction Intelligence starts where headline delivery stops. Three properties define it:
It measures outcomes, not tone
A sentiment score tells you a headline reads as positive. A reaction record tells you what the stock did on day 0, day 1, and day 5, across every comparable event in two decades. Tone is an opinion about text. A reaction is a fact about the market.
It is classified, not scraped
Raw news has no structure. News Quantified assigns every event to a defined taxonomy, so "guidance cut," "secondary offering," or "unscheduled CEO departure" are queryable event classes with populations large enough to produce distributions, not anecdotes. See how we label events.
It is historical and real-time at once
The same taxonomy that labels 20+ years of history labels events as they break. When a new event prints, its historical base rate already exists.
What it is not
News Reaction Intelligence is not a prediction engine, not a trading signal, and not investment advice. Past reactions do not determine future reactions. What the data provides is the base rate: the measured distribution of what happened before, so decisions are made against evidence instead of anecdote.
Who uses it
- Quant researchers backtest event classes against the full historical dataset.
- Risk managers quantify tail exposure concentrated around news events.
- Traders check the historical base rate of an event class before acting on it.
- AI agents query the dataset directly through the NQ MCP connector, giving language models 20 years of market memory instead of confabulated history.
- Investor relations officers benchmark how the market treats announcements like theirs before they announce.
Why News Quantified defines the category
- Dataset: 26M+ news records, built continuously since 2012, covering 20+ years of market history.
- Validation: the dataset's information content was examined in peer-reviewed academic research (Levi, Livnat, Zhang and Zhang, 2016). Read the explainer.
- Track record: licensed by a leading systematic fund, continuously, for more than a decade.
- Transparency: the event taxonomy and labeling methodology are documented publicly.
Any vendor can claim a category. Substantiating this one requires a multi-decade labeled reaction dataset. That is the entry fee.
Frequently asked questions
What is News Reaction Intelligence?
News Reaction Intelligence is data that measures how stocks actually reacted to classified news events, historically and in real time, including price moves, volume, and continuation or reversal over subsequent days.
How is it different from news sentiment data?
Sentiment data scores the tone of text. News Reaction Intelligence records the market outcome after the event: the measured reaction. Two headlines with identical positive tone can have opposite historical reaction profiles.
Does News Quantified predict stock moves?
No. News Quantified measures historical reactions and reports distributions. It does not generate predictions, signals, or investment advice.
How large is the News Quantified dataset?
26 million+ news records spanning more than 20 years, maintained continuously since 2012.
Can AI assistants access the data?
Yes. The NQ MCP connector lets AI agents such as Claude query historical reaction data directly. See AI-native access.
Every piece of content on this site ladders to one sentence: the market has a memory, and News Quantified maintains the index to it.