Why Did That Stock Move? How Explained Price Moves Work

The short answer

An explained price move is a close-to-close move of 3% or more, up or down, paired with a written explanation of what caused it. Distilla builds each one by ingesting news on the company, extracting dated events, summarizing earnings when a report is involved, and searching social media for sentiment and explanations, then combining them into one account that links back to its sources. This is called price attribution: the link between what happened to a company and what happened to its stock.

For an AI agent, three things make this different from asking a web search:

•           More reliable. The explanation is tied to dated events, not to whichever article ranked first, so the same question gets the same answer.

•           Built for history. You can ask why a stock moved on any past day, and the answer is matched to what actually happened that day.

•           Built for breadth. An agent can pull every big move across a market in the last day or week, or every move in a company or sector over 52 weeks, and look for patterns. With web search alone, that is not practical.

Why web search gets “why did it move?” wrong

“Why did this stock fall?” is the question investors ask an AI agent most often, and it is where web search is least reliable. There are four specific reasons.

1.         Many agents cannot see social media. A lot of the explanation for a move, and the sentiment around it, is discussed on social platforms. Many agents with web search do not scan them at all.

2.         Search ranks by the wrong criteria. An agent usually reads the top few search results. Those are ranked on authority, popularity and search optimization, not on how relevant the article is to that stock on that day. The best explanation may not be on page one.

3.         Search has no sense of time. Web search is not built around timestamps. Ask about a move on a specific past date and the agent has to work out which article describes which day. That is where it breaks down, and where it is most likely to hallucinate or mix up the date, the event and the cause.

4.         Articles are vague about dates. Many pages do not state clearly when they were written, and often cite an earlier or later event as the reason for a particular day’s move. An agent that takes them at face value ends up with the right story attached to the wrong day.

The result is answers that are biased toward the loudest story, out of date, sometimes wrong, and different from one run to the next. For an investor deciding whether a drop is a buying opportunity or a warning, that is not good enough.

The fix is a data layer that has already done the matching. In the four layers of financial data an agent can use (market data, financials, narrative, and company knowledge), price attribution belongs to company knowledge. Market-data servers tell you what the price did. Only a company-knowledge layer tells you why. As of 6 October 2026, we found no other vendor offering dated, linked explanations for every large move across five markets.

How Distilla builds a price move explanation

Distilla works after the close, not during the session. Every trading day it collects each market’s closing prices, and any stock that closes at least 3% above or below the previous close is sent for an explanation automatically. The same threshold applies in all five markets it covers. At that size, ordinary day-to-day noise is mostly filtered out, while the moves investors tend to ask about are still caught. Intraday swings that reverse before the close are not flagged.

For each flagged move, the explanation is built in four steps:

1.         Ingest the news. Distilla reads news articles on the company, its competitors and its sector around the date of the move.

2.         Classify and extract events. Each article is turned into dated company events. Distilla tracks 100+ event types, including earnings reports, guidance changes, analyst rating and price-target changes, management changes, deals, partnerships and legal actions. About 50,000 new company events are added each week.

3.         Summarize earnings, if there was a report. When a move follows results, the explanation picks out the figures and management comments the market reacted to. Whether the quarter beat is only part of it.

4.         Search social media for sentiment and explanations. This captures how investors were reading the news, and explanations that formal news coverage misses.

The pieces are then combined into one explanation that links back to the events it relied on. Each event carries a date. That is what makes it checkable: if an explanation says a stock fell because the CEO left, you can see the dated event behind that claim and go to the original announcement.

As of 7 October 2026, Distilla holds 438K+ explained price moves across the US, China, Japan, Hong Kong and Korea.

A real example: why MongoDB stock fell on 28 September 2026

MongoDB (MDB) closed down about 18.5% on 28 September 2026. We asked Claude, with Distilla connected, what happened.

Claude with Distilla connector, asked why MongoDB fell on 28 September 2026.

The cause was clear. CEO Chirantan “CJ” Desai resigned with immediate effect to join Meta Platforms as Chief Enterprise Platform Officer. It was MongoDB’s second CEO change in under a year, and it came with no warning. Former CEO Dev Ittycheria returned on an interim basis.

Two details in the answer are worth noticing. MongoDB reaffirmed its Q3 and fiscal 2027 guidance the same day, and the stock fell anyway, which tells you the market was pricing leadership risk rather than a change in the numbers. The departure also landed just days before the company’s scheduled Investor Day.

Behind that explanation sit the linked events: the CEO departure, the interim CEO appointment, a competitive-dynamics event on Meta moving into the enterprise data platform market, and rating reactions from Stifel and BofA, both of which kept their ratings.

What you can do with it: screening and pattern study

Explaining one move is useful. The larger value is asking about many moves at once, because every move is dated and labeled with a cause. That makes price moves something an agent can search, group and count.

Scan the market. Ask for every big move in the last day or the last seven days, with the reason for each, and have the agent group them into themes. On 18 September 2026 Distilla recorded 50 moves of 3% or more across four Asian markets, about 21 of them chip-related. A pattern like that is visible at once when the moves are explained and grouped, and invisible when you look at them one by one. See researching Japan, Hong Kong, Korea and China stocks with AI.

Study one company or sector. Ask for every price move in the last 52 weeks for a company, or for a whole sector, with causes. You see what actually moves the stock: earnings, guidance, management, or the sector. That shows you where its risk comes from.

Find signals. Once moves are labeled, you can ask which kinds of cause are followed by a continued move and which tend to reverse. In one study of last earnings season, 141 of 186 stocks that fell after beating estimates pointed to guidance as the explanation, not the results.

Understand historical patterns and build strategies. Labeled causes let you separate events that behave differently. In a study of Tencent’s earnings-window drops since 2020, labeled causes separated 7 drops into groups that behaved differently.

None of this works with an agent and web search. Looking up the cause of one move at a time, with no reliable date matching, does not scale to hundreds of stock-days, and the results would not be consistent enough to count or backtest. As of 6 October 2026, we know of no other source that makes it possible.

How to ask in Claude or ChatGPT

Once Distilla is connected to Claude or ChatGPT, you can ask in plain language. A few questions that work well:

•           “Why did MongoDB stock fall on September 28, 2026?”

•           “List every stock that moved more than 5% in the last 7 days, explain each move, and group them into themes.”

•           “List MongoDB’s price moves of 3% or more over the last 52 weeks, with the cause of each. Which were company-specific, and which were the sector or the market?”

•           “Across Adobe’s last 12 earnings reports, what moved the stock: the results or the guidance?”

Naming the stock and the date, and asking for the linked events, gets you the most precise answer. Since explanations are written after the close, the best time to ask about today’s move is once the market has shut.

Limits

Explanations are written by a language model from the linked events and sources, so check the company’s filing or press release before you rely on a specific figure. An explanation reflects the news and events Distilla captured; if something relevant was never published or recorded, it will not appear. Social media sentiment is a reading of how investors reacted, not a fact about the company. Distilla does not explain intraday moves that reverse before the close.

FAQ

Why did my stock go down today?

A stock usually falls because of company news (earnings, guidance, management changes, deals or legal actions), an analyst downgrade or price-target cut, a move in its sector or the whole market, or because there is no identifiable catalyst. Compare the date of the fall with the company’s events, then check whether peers and the index moved the same way. If they did, the cause is probably not specific to the company.

What is price attribution?

Price attribution is the link between what happened to a company and what happened to its stock: which dated events, news and sentiment explain a given price move. Distilla records it for every close-to-close move of 3% or more.

How does Distilla generate a price move explanation?

It ingests news articles on the company, classifies and extracts dated company events, summarizes earnings when a report is involved, and searches social media for sentiment and explanations. It then combines these into one explanation that links back to the events it relied on.

What counts as an explained price move?

A close-to-close share price change of 3% or more, up or down, with a written explanation and the dated events it was linked to. The threshold is the same in every market Distilla covers.

Why 3%?

Smaller daily moves are often noise. A 3% threshold keeps the moves investors usually ask about without burying them in ordinary fluctuations.

Why can’t an AI agent just use web search to explain a price move?

Web search ranks results by criteria that are not news relevance, may not include social media, and is not built around timestamps. When asked about a specific past date, an agent can mix up the date, the event and the cause, and many articles do not state their dates clearly or cite an earlier or later event. The answer can also change from run to run. Distilla ties each explanation to dated, linked events and returns it the same way each time.

Can I look at all the big moves over the last day or week?

Yes. Ask your agent for every move above a size you choose over the last day or seven days, with a reason for each, and have it group them into themes. This is a scan across the whole market, not a lookup of one stock.

Can I analyze a company’s price moves over the past year?

Yes. Ask for all of a company’s, or a sector’s, explained moves over the last 52 weeks. You can see what drives the stock and whether the pattern is company-specific, and you can use the labeled causes for backtesting and strategy research.

How soon after a move is the explanation available?

After that day’s close. Distilla works from closing prices, so a move during the session is explained once the market has closed and the day’s prices are in.

What happens when there is no clear reason?

Distilla records the move as “No company-specific driver” rather than guessing. That usually points to a market or sector move, or to trading with no single catalyst.

Does it cover stocks outside the US?

Yes. Distilla’s core coverage is the most-followed companies in the US, Japan, Hong Kong, Korea and China A-shares.

How do I check an explanation?

Look at the linked events and their dates, then go to the original filing, press release or transcript for any figure that matters to your decision.

Can I use explained price moves in ChatGPT as well as Claude?

Yes. Distilla’s MCP server works with both, as well as Claude Code, Cursor and WorkBuddy. Developers can call the same data through the HTTP API. There is a free plan, and paid plans start at $9.90/month.

Is this investment advice?

No. Distilla is a research tool, not investment advice.

About Distilla

Distilla brings institutional rigor to individual investors and small teams. It is the fundamental research layer for AI agents: why stocks move, not just where. The same data powers an enterprise product, including an agent, used by some of the world’s top funds, and is also packaged as a self-serve product so individuals can work from the same facts. It is an MCP server and HTTP API that gives Claude, ChatGPT and your own agents quantitative data, qualitative research and structured company knowledge in one queryable, read-only layer.

•           Quantitative: financial statements, analyst consensus, valuation history and company metrics, including AI-extracted segments and industry-specific KPIs.

•           Qualitative: earnings-call and filing commentary, 180K+ sell-side research summaries and 11K+ investor podcast episodes.

•           Company intelligence: 1,000+ company knowledge topics, business drivers updated quarterly, company events (about 50,000 added weekly), products, peers and sector relationships, and 438K+ explained price moves of 3% or more.

Core coverage is the most-followed companies in the US, Japan, Hong Kong, Korea and China A-shares; sell-side and podcast summaries extend beyond the core universe. Read-only MCP server at https://api.distilla.ai/mcp, with the same lookups over an HTTP API. Works with Claude, ChatGPT, Claude Code, Cursor, WorkBuddy and Doubao. Self-serve: free plan, with paid plans from $9.90/month. Enterprise: contact Distilla. Distilla is a research tool, not investment advice. Figures are as of 7 October 2026.