Direct Answer

Building a stock screen means translating one strategy into objective, ordered filters that can be applied consistently to any security in a defined market, in eight steps: define the objective, establish the universe, add tradability requirements, add strategy-defining filters, add confirmation filters sparingly, add exclusion rules, rank the results, and manually inspect every finalist. Running the universe layer first is what keeps a screen built for liquid large caps from returning microcaps or thinly traded foreign securities. The output is a candidate research list rather than a trade signal, so every finalist still needs manual inspection before any decision.

Key Takeaways

  • The process starts with a one-sentence objective specific enough to make filter choices obvious, "find good stocks" gives no direction, but naming the market, liquidity requirement, trend requirement, and strategy type does.
  • The universe layer (country, security type, exchange, price, market cap) should run first, so a screen built for liquid large caps doesn't return thinly traded foreign securities or microcaps.
  • Average dollar volume, average daily share volume multiplied by average share price, is more useful than share volume alone, since a stock trading a million shares at $2 has very different liquidity than one trading a million shares at $100.
  • Adding many confirmation filters can create a brittle screen that returns very few results and fits past examples more closely than future opportunities; a useful test is whether you'd reject a stock specifically for failing just that one condition.
  • A screening result is a candidate, not a completed trade idea, every finalist should be manually inspected before any decision, since aggregate filter values can differ from individually verified data.

How Do You Build a Stock Screen?

Building a stock screen means translating a trading strategy into a set of objective, ordered filters that can be applied consistently to any security in a defined market. The process has eight steps: define the objective, establish the universe, add tradability requirements, add strategy-defining filters, add confirmation filters sparingly, add exclusion rules, rank the results, and manually inspect every finalist.

The screen should express one strategy clearly. A screen built for a long-term value investor uses fundamentally different conditions than one built for an intraday momentum trader, and combining the two produces a result that is coherent for neither.

See the stock screening overview for the distinction between a screen, a scan, and a ranking system before building a screen for the first time.

Step 1: Define the Exact Setup

Write one sentence describing what the screen must find. The sentence should be specific enough to make the filter choices obvious.

Weak objective: Find good stocks.

Better objective: Find liquid U.S. stocks in established intermediate- and long-term uptrends consolidating near their 52-week highs for possible multi-day continuation setups.

The second version specifies the market (U.S. stocks), the liquidity requirement (liquid), the trend requirement (uptrends), the pattern requirement (consolidating near highs), and the strategy type (continuation). Every filter chosen in later steps should connect back to this sentence.

Questions to answer before writing filters

  • Is the strategy intraday, swing, position, or long-term?
  • Does it seek momentum, value, growth, income, reversal, or breakout behavior?
  • How long is the expected holding period?
  • How much volatility is acceptable?
  • Does the strategy require a catalyst or earnings gap?
  • Which securities are explicitly prohibited?
  • What would make a result unsuitable even if it passes all filters?

Step 2: Establish the Stock Universe

The universe defines which securities can appear in results before any strategy conditions are applied. This layer should run first because a screen designed for large liquid U.S. equities should not return thinly traded foreign securities, closed-end funds, warrants, or microcaps.

Illustrative universe for a liquid U.S. swing-trading strategy:

Country: United States
Security type: Common stock
Exchange: Nasdaq, NYSE, or NYSE American
Price: At least $10
Market capitalization: At least $1 billion

These values are teaching examples, not universal requirements. A small-cap momentum strategy would use different conditions. A long-term value approach might remove the price floor entirely. The universe should match the strategy's eligible market, not someone else's default settings.

Market capitalization is calculated as share price multiplied by shares outstanding. The categories, micro-cap, small-cap, mid-cap, large-cap, have no universal definition and can differ across data providers. Use numeric boundaries rather than category labels in a saved screen.

Step 3: Add Tradability Requirements

A chart can look attractive while the stock remains difficult to trade. Tradability filters remove securities that may create practical problems at the entry or exit.

A person using a smartphone app to trade stocks with a laptop displaying market data.
Photo by StockRadars Co., via Pexels

Possible requirements for a swing-trading strategy:

Average daily volume: At least 500,000 shares
Average dollar volume: At least $20 million
Share price: At least $10

Average dollar volume, average daily share volume multiplied by average share price, is more useful than share volume alone when comparing differently priced stocks. A stock trading one million shares at $2 generates approximately $2 million in daily dollar turnover. A stock trading one million shares at $100 generates approximately $100 million. The share volume is identical; the dollar liquidity is not.

The bid-ask spread is the difference between the highest price a buyer is willing to pay and the lowest price a seller will accept. Wider spreads increase transaction costs and are most visible in low-volume or low-priced stocks. Investor.gov defines stock liquidity in terms of how rapidly shares can be bought or sold without substantially affecting the price, a condition not fully captured by average share-volume filters alone.

For intraday strategies, liquidity requirements are typically stricter and should also include live spread inspection before any order.

Step 4: Add the Strategy-Defining Filters

Strategy-defining filters express the actual trading idea. These conditions should describe the setup, not just the eligible market.

For a trend-continuation screen

Price > 50-day moving average
Price > 200-day moving average
50-day moving average > 200-day moving average
Three-month performance > 0%

For a growth screen

Year-over-year revenue growth > 10%
Year-over-year EPS growth > 10%
Positive operating margin
Positive free cash flow

For a value-quality screen

Positive trailing earnings
Positive free cash flow
Debt-to-equity below chosen limit
P/E below industry or strategy limit
Return on equity above chosen minimum

Do not combine unrelated conditions merely because the platform makes them available. Every filter should have a documented reason that connects it to the original objective statement.

For detailed coverage of technical and fundamental filter options, see the guides on technical stock screening and fundamental stock screening.

Step 5: Add Confirmation Filters Sparingly

Confirmation filters strengthen the main premise without redefining it. For a breakout screen, a primary condition might be price within 5% of a 52-week high, with confirmation filters including minimum average volume, positive six-month relative strength, and price above the 50-day moving average.

The problem with adding many confirmation filters is that each one removes passing candidates. Adding 20 indicators often creates a brittle screen that:

  • Returns very few results on any given day.
  • Fits historical examples more closely than future opportunities.
  • Cannot distinguish between a candidate that barely failed one condition and one that failed many.

A useful test: ask whether you would reject a specific stock specifically because it failed this one condition. If the answer is no, the filter may not belong in the screen.

Step 6: Add Exclusion Rules

Exclusion rules remove candidates with known conflicts that the inclusion filters did not catch. Exclusions are often as important as the strategy filters themselves.

Common exclusions include:

  • Earnings scheduled within the planned holding period.
  • Average volume below the strategy minimum on the current day.
  • OTC-listed securities.
  • ETFs, closed-end funds, and leveraged or inverse products.
  • Recently reverse-split stocks.
  • Active bankruptcy proceedings or delisting notices.
  • Companies with missing or incomplete financial data.
  • Stocks in industries the strategy explicitly avoids.
  • Extremely wide bid-ask spreads.

Not every screener provides reliable real-time data for all of these exclusions. Earnings dates, news, and special corporate events often require a separate check using a reliable calendar, news feed, or SEC filing review.

The SEC's EDGAR database provides free public access to company filings. A Form 8-K is required to be filed promptly after a material corporate event such as an earnings release, a merger announcement, or a significant asset sale.

Step 7: Rank the Results

A binary screen answers pass or fail. Ranking determines which passing candidates deserve attention first.

Momentum ranking example

  1. Highest relative volume (current volume divided by average volume).
  2. Strongest three-month price performance.
  3. Closest distance to a defined breakout level.
  4. Highest average dollar volume.

Quality-growth ranking example

  1. Revenue-growth consistency across reporting periods.
  2. Free-cash-flow growth trend.
  3. Operating-margin direction.
  4. Return on invested capital.
  5. Valuation relative to growth rate.

Avoid combining metrics with incompatible scales without normalization. Adding a 25% revenue-growth rate directly to a $4 billion market-cap figure produces a meaningless score. Rank by one metric at a time, or normalize each input before combining them.

Close-up of a smartphone displaying a stock trading app interface, emphasizing financial analysis and investment insights.
Photo by StockRadars Co., via Pexels

Step 8: Manually Inspect Every Finalist

The screen should end with a review list, not an order ticket. Every finalist requires a manual inspection before any decision is made.

What to review for each candidate

  • Price chart: trend quality, pattern, distance from support, key levels.
  • Current bid-ask spread and available liquidity.
  • Current volume versus average volume.
  • Upcoming earnings, dividends, or scheduled announcements.
  • Recent news and any material catalysts.
  • SEC filings: most recent quarterly or annual report.
  • Sector conditions and broad market trend.
  • Entry location and confirmation trigger.
  • Stop location based on chart structure.
  • Position size based on risk amount and stop distance.
  • Reward-to-risk estimate using the target versus the stop.

FINRA notes that high volume and volatility can cause execution prices to differ, sometimes significantly, from the quote visible when an order is entered. Inspecting the live spread and current order-book activity before entry is part of the review, not an optional step. For position sizing guidance after a candidate has passed all reviews, see the stock position sizing guide.

Daily Workflow Around the Screen

Before the session

  1. Review the broad market trend and sector conditions.
  2. Run saved strategy screens.
  3. Remove results with unsuitable current liquidity.
  4. Check earnings and major scheduled events.
  5. Review recent company news.
  6. Inspect charts of remaining candidates.
  7. Rank the strongest by the chosen criteria.
  8. Add qualified stocks to a watchlist with a noted entry trigger.

During the session

  1. Monitor price and volume for the trigger condition.
  2. Confirm the spread is acceptable before submitting any order.
  3. Recalculate risk using the actual fill price and stop level.
  4. Reject trades that no longer offer adequate reward relative to risk.
  5. Record why a trade was taken or skipped.

After the session

  1. Save the screen output and compare expected versus actual behavior.
  2. Identify which filters added or reduced value.
  3. Avoid changing rules based on a single trade result.
  4. Evaluate results over a meaningful sample before adjusting the screen.

If you want to test a screen against historical data rather than running it live, the backtesting cluster covers the process of applying rules to point-in-time data to estimate past performance.

Frequently Asked Questions

What is the first step to build a stock screen?

Write one clear sentence describing what the screen must find, including the strategy type, holding period, and eligibility requirements. A vague objective like "find good stocks" cannot be translated into filters. A specific objective like "find liquid U.S. stocks in established uptrends consolidating near 52-week highs" provides enough detail to choose relevant conditions.

How do you define the stock universe in a screen?

The universe defines which securities can appear in results before strategy filters are applied. Common universe conditions include country, exchange, security type (common stock only, excluding ETFs), share price minimum, and market-capitalization range. A strategy designed for large liquid U.S. stocks should specify those conditions explicitly so microcaps, foreign issues, and funds do not appear in results.

How do you rank stock-screening results?

Ranking determines which passing candidates deserve attention first. A momentum screen might rank by relative volume, then three-month price performance, then distance to a breakout level. A quality-growth screen might rank by revenue-growth consistency, then free-cash-flow growth, then return on invested capital. Avoid combining metrics with incompatible scales without normalization.

What is an exclusion rule in a stock screen?

An exclusion rule removes candidates with known conflicts that the inclusion filters do not catch. Common exclusions include earnings scheduled within the planned holding period, OTC securities, ETFs and closed-end funds, recently reverse-split stocks, and companies with missing or unreliable financial data. Exclusion rules are often as important as inclusion rules.

Should you manually review every screening result?

Yes. Every finalist should receive a manual review of the price chart, current spread, volume, upcoming earnings, recent news, sector conditions, entry location, stop location, position size, and reward-to-risk estimate before any decision is made. A screening result is a candidate, not a trade recommendation. The screen ends with a review list, not an order ticket.

How many results should a screen be aiming to produce?

Few enough that every one can be examined properly, which for most individual workflows means a manageable shortlist rather than hundreds. A screen returning too many is not filtering; one returning zero on most days is either too tight or is describing a condition that rarely occurs. Adjusting thresholds to hit a target count is itself a form of fitting, so the count is better treated as a signal that the design needs rethinking.

Should thresholds be absolute values or percentile ranks?

Absolute thresholds are easy to reason about and shift in meaning as market conditions change, so a fixed level can pass most of the universe in one environment and almost none in another. Percentile ranks adapt automatically and always return a similar proportion, which means they never signal that conditions have changed. Using absolute levels with a periodic review, or ranks with an absolute floor, addresses the weakness of each.

What is the risk of building a screen from a set of past winners?

Assembling conditions that all the admired examples satisfy produces a screen fitted to those specific outcomes, and the conditions will typically include features that had nothing to do with why they worked. It is the screening version of fitting a strategy to history. A screen designed from a stated rationale and then checked against examples is testing a hypothesis; one assembled from examples and rationalized afterwards is not.

How should a screen handle securities that are not ordinary common shares?

Broad universes contain funds, trusts, depositary receipts, preferred shares, units and shell companies, and many will satisfy filters designed for operating businesses while being something else entirely. Most platforms provide a security-type field for exclusion. Where they do not, a manual pass over the results catches them, and the alternative is a shortlist that quietly includes instruments the screen was never intended to describe.

References