5. Compare, screen and follow the money

Attention: research and social buzz (ATTN, BUZZ)

ATTN ranks the tickers Gloom users are researching against each ticker's usual level, and BUZZ counts the day's posts on X naming a ticker, with the top posts and their stance.

  • 4 min
  • 3 questions
  • Lesson 4 of 4

Why you would care

Is anyone paying unusual attention to this stock? Prices often move after attention spikes: a stock nobody looked at suddenly trends, and buyers follow. Sometimes the attention is the news itself (a viral post); sometimes it is a reaction to news you missed. Attention is a clue, never a reason. This lesson is about reading it without being fooled by it.

The idea from scratch

Two kinds of attention:

  • Research attention

    people opening a stock's pages (overview, financials, options...). It shows what investors are studying.

  • Social attention

    people posting about a stock. On X, posts name tickers with a cashtag, the ticker with a dollar sign: $NVDA.

A raw count means little: a mega-cap is always studied more than a small company. What matters is abnormal attention: today versus the stock's own usual level.

The tool for that is a z-score: how many standard deviations today's value sits above its usual average.

Traps

  • Bots and spam inflate social counts.
  • Privacy rules: research attention is published only when enough distinct people looked (so no one can be identified); small names may show nothing. Missing is not zero.
  • Attention follows price as often as it leads it.
Diagram: Users research a ticker, per hour leads to Published only above a privacy threshold; Published only above a privacy threshold leads to Abnormal score vs the ticker's own usual hour; Posts naming the cashtag leads to Daily count vs 30-day median, top posts, stance; Abnormal score vs the ticker's own usual hour leads to Clue: look for the news; Daily count vs 30-day median, top posts, stance leads to Clue: look for the news.

See it in Gloom

open themtype ATTN for the board, ATTN 6758:JPX for one listing's history; BUZZ TSLA for social posts.

ATTN (keys: Enter open, e evidence, d overview, g chart, n news):

  1. Views: Ranking (most researcher-hours), Abnormal (highest latest-hour z-score), Sectors, Countries, History.
  2. Periods: Now (the latest completed hour), Today (since midnight UTC), Week (the last 168 hours).
  3. Four summary figures: total research hours, how many tickers have a z-score of 2 or more, the top sector and the top country.
  4. Evidence: the publication delay and threshold, the price change and relative volume, and dated news from the same time.

BUZZ TSLA (keys: Enter open, o open the post):

  1. The day's count of posts naming $TSLA against its 30-day median.
  2. The top posts of the day and their stance, a score from -1 (bearish wording) to +1 (bullish wording), also averaged per day.

Practice and recap

Try it3 tasks
  1. A ticker's research hours jump from 30 to 60, with a usual spread of 10. z-score? (3.)
  2. Why compare a stock's attention with its own history rather than with other stocks? (Big companies always get more attention; only the change from normal is informative.)
  3. Name one reason a social spike could be meaningless. (Bots, spam, a meme unrelated to the business.)
Common mistakes4 mistakes
  • Treating attention as a buy signal.
  • Reading a missing value as zero attention (it may be below the privacy threshold).
  • Forgetting the lag: research attention is at least an hour old.
  • Trusting the stance of a few loud posts.
Check yourself3 questions
  1. What is a cashtag?
  2. What does an abnormal (z-score) ranking show?
  3. Why are some tickers missing from ATTN?
Answers
  1. A ticker written with a dollar sign in a post, like $NVDA.
  2. Tickers whose latest attention is far above their own usual level for that hour.
  3. Their hourly activity did not reach the minimum number of distinct researchers required for publication.
Words in this lesson6 words
research attention
How much people are studying a stock's pages.
cashtag
A ticker with a dollar sign in a social post.
z-score
How many standard deviations a value is above its usual average.
abnormal attention
Attention far above a ticker's own normal level.
stance
A score from -1 (bearish) to +1 (bullish) for a post's wording.
privacy threshold
The minimum number of distinct people before a count is published.

Educational material about reading market data, not investment advice.