5. Compare, screen and follow the money

Screening the market (EQS)

EQS keeps only the stocks in its stored universe that meet all your rules (valuation, growth, margins, liquidity, ownership and more) and ranks the matches.

  • 4 min
  • 3 questions
  • Lesson 5 of 5

Why you would care

Which companies fit what you are looking for? There are thousands of listed companies. Nobody reads them all. A screen turns "cheap, growing, not too indebted" into a short list you can study one by one. It is the first step of many investment processes, and the place where a sloppy rule wastes hours.

The idea from scratch

A screen is a set of criteria (rules). Each criterion is a field, a comparison and a value:

  • P/E between 0 and 25
  • Revenue growth ≥ 10%
  • Market cap ≥ 10B

All criteria must be true at the same time (they are joined with AND). Then you rank the survivors by one metric, for example cheapest first.

What a screen cannot do

  • It only sees numbers it has. A company with a missing value cannot pass a rule on that value.
  • It sees the covered universe: the listings in its stored daily snapshot, not every stock in the world.
  • It does not know why a number is odd: a low P/E may come from a one-off gain (chapter 04).

The percentile column

Next to the chosen metric, a percentile ranks each company within the whole covered universe: 99 means higher than 99% of covered companies.

  1. Covered universe: daily snapshot
  2. Criterion 1
  3. Criterion 2
  4. Criterion 3
  5. Matches ranked by your metric
  6. Open each one and read

See it in Gloom

Gloom screenshot: EQS, Results tab
EQS, Results tab: the default screen (US-dollar companies above 10 billion of market cap) ranked by market cap, with each value's date. Historical example.

open ittype EQS. Tabs: Results, Criteria, Saved. Keys: a add a criterion, m change the metric, s save the screen, o open a company, x export all matches.

  1. Metric Market cap: the ranking column, here biggest first.
  2. PCTL 100: these companies are in the top percentile of the covered universe by market cap.
  3. AS OF: the date of the value. Values shown in amber are stale (older than usual), like JNJ's 2026-09-19 row.
  4. PRICE, CHG%, P/E: context columns. -- means not available (no P/E when earnings are negative or missing).
  5. Duplicates are real: GOOGL and GOOG are two share classes of Alphabet; BRK.A and BRK.B of Berkshire Hathaway.

Fields you can screen on include valuation (P/E, P/B, EV/EBITDA), growth, margins, ROE, returns (1W to 1Y), distance from the 52-week high, beta, estimate revisions, analyst upside, short interest, insider and 13F activity, implied volatility rank, and social attention. Thresholds accept k, M, B, T (10B = ten billion).

Practice and recap

Try it3 tasks
  1. In the screenshot, which company has the highest P/E in the visible rows? (TSLA, 345.0.)
  2. Write three criteria for "large, cheap, growing". (Market cap ≥ 10B; P/E between 0 and 20; revenue growth ≥ 10%.)
  3. Why put "P/E between 0 and 25" rather than "P/E below 25"? (To exclude negative P/Es from companies with losses.)
Common mistakes4 mistakes
  • Treating the match list as a buy list. It is a reading list.
  • Forgetting missing data: a company without a value is silently excluded.
  • Screening on one period's number that includes a one-off.
  • Duplicating share classes and thinking you found two companies.
Check yourself3 questions
  1. How are several criteria combined?
  2. What does a percentile of 95 mean in the results?
  3. Why can a great company be missing from your results?
Answers
  1. With AND: a company must pass all of them.
  2. Its value is higher than 95% of the covered universe for that metric.
  3. It may lack a value for one criterion, be outside the covered universe, or fail a rule because of an odd period.
Words in this lesson6 words
screen
A filter that keeps only stocks meeting your criteria.
criterion
One rule: a field, a comparison and a value.
metric (ranking)
The column used to order the matches.
covered universe
The set of listings the screener has data for.
share class
One of several kinds of shares of the same company (GOOGL and GOOG).
stale value
A number older than it should be.

Educational material about reading market data, not investment advice.