What Is a Stock Screener?

Why Screeners Matter

The recurring problem in equity research is idea generation. Where do new candidates come from? Tip-driven sources — message boards, financial television, paid newsletters — place the analyst at the receiving end of someone else’s agenda, and the agenda generally explains more of the recommendation than the underlying numbers do. A screener inverts that arrangement. The criteria are defined by the user; the universe is sorted by the software; and the output is a list whose composition reflects the analyst’s framework rather than a third party’s incentive to publish.

This does not make a screen a substitute for due diligence. The candidates that survive the filter still require chart review, fundamental verification, and an explicit risk assessment before any capital is committed. A reasonable screen eliminates roughly nineteen of every twenty stocks the analyst would otherwise have to consider, which is most of the value it delivers; it does not deliver the twentieth as a finished investment.

The sequence is the right one: screen first, analyse second, decide third. The screener occupies position one in that order and is generally not load-bearing past it. A most-active screen, for instance, will identify stocks attracting the day’s heaviest trading activity, but the longer-term chart and the underlying fundamentals are what determine whether any of those names warrant further work.

Fundamental Screens

A fundamental screen filters on financial-statement data drawn from the balance sheet, the income statement, and the cash flow statement. The common filters and their conventional thresholds are summarised below.

Fundamental screens are most useful to value investors and longer-horizon owners who are looking to identify financially sound businesses available at defensible multiples. A classic Graham-style configuration would specify a P/E below 15, positive trailing earnings growth, debt-to-equity under 0.5, and a dividend yield above 2%. The output is a population of solvent, dividend-paying, modestly priced companies, which is approximately the population the discipline is intended to surface.

An example may help calibrate the output. In late 2011, Moody’s Corporation traded for a stretch in the low-thirty-dollar range at a trailing P/E of roughly 14×, with operating margins above 35% and ROIC well above the cost of capital; on the available record, the company would have surfaced on a moderately strict quality-and-value screen of the period. The screen does not, by itself, render a verdict on whether the rating-agency duopoly was a defensible long-term moat or a regulatory artefact in slow erosion — that is the analyst’s subsequent work. The screen establishes only that the file is worth opening.

Technical Screens

A technical screen filters on price-and-volume behaviour, which is to say the data visible on the chart rather than the data filed with the SEC.

Technical screens are the operating instrument of day traders, swing traders, and momentum traders. A representative momentum configuration might require an intraday move above 5% on volume at least double the trailing 20-day average, the combination being intended as a proxy for institutional accumulation or a discrete catalyst event. A representative oversold-bounce configuration might require RSI below 30 with price holding above the 200-day moving average, on the working assumption that the selling reflects a transient dislocation rather than a durable change in the underlying business.

Simple screens generally outperform elaborate ones for the prosaic reason that adding a tenth condition to a screen of three useful conditions usually filters out the candidates the first three were intended to surface. The conservative discipline is to define the minimum filter set that expresses the strategy and to add further conditions only if the list is unmanageably long.

Combining Fundamental and Technical Filters

The more interesting screens combine the two registers. A configuration that requires both strong earnings growth (fundamental) and a recent breakout above resistance on confirming volume (technical) is attempting to identify situations in which the financial quality and the price action are corroborating one another. The logic is that of triangulation: two independent data sources pointing the same direction carry more evidentiary weight than either source carries alone, in approximately the way replication strengthens a finding in academic research.

A representative combined screen for growth-with-confirmation might read as follows:

• Earnings growth > 25% year-over-year
• Revenue growth > 15% year-over-year
• Market capitalisation between $200M and $5B
• Price within 5% of the 52-week high
• Average daily volume > 300,000 shares
• RSI between 50 and 70 (trending without being overbought)

The output of such a screen on a typical session is in the range of ten to thirty names, which is roughly the upper bound of what an individual analyst can review by hand in a sitting. The screen identifies candidates; candidates are not decisions, and the conversion from one to the other is the analyst’s work, not the software’s.

Free Screener Tools

Several screeners are available to retail accounts at no charge:

• Yahoo Finance Stock Screener — basic fundamental and technical filters with limited customisation; adequate for the standard P/E, market-cap, sector, price, and volume parameters.
• MSN Money Stock Screener — a competent free option with pre-built screens organised around value, growth, and income strategies.
• Brokerage screeners — most online brokers provide built-in screens for account holders, generally with a wider filter set and more current data than the free public-web tools. A most-active screen, in particular, is a quick way to identify stocks with unusual trading activity for further review.

The free tools have predictable limitations — delayed quotes, fewer filters, restricted alerting — but for an analyst learning the workflow they are sufficient to generate a working candidate list on a daily basis. (On a Bloomberg terminal the equivalent function is EQS, exposed with deeper data and more filter primitives, but the conceptual machinery is essentially the same as what the retail screeners surface; the differentiator is data breadth and currency, not sophistication.)

Building Your Own Screen: A Procedure

The recommended procedure is to start narrow and expand only as required by the output volume.

1. Define the strategy. The filters should follow from the strategy, not the other way around. A value investor hunting cheap securities, a momentum trader looking for breakouts, and a swing trader looking for oversold reversals will all use different filter sets, and the filter sets must be specified before they are coded.
2. Set three to five filters. Begin with a price range, a minimum-volume condition, and one or two strategy-specific criteria. Resist the impulse to add ten conditions; the screen that returns no results returns no information.
3. Run the screen daily. Conditions change. A name that fails to qualify on Monday may qualify on Wednesday, and the discipline of running the same screen across many sessions is itself part of the analysis.
4. Review each result manually. The chart, the news flow, and the earnings calendar all require human inspection. The screen produces candidates; the analyst produces a decision.
5. Maintain a watchlist. Many candidates will be “almost ready” — awaiting a pullback, a confirming print, or a catalyst — rather than immediately actionable. The watchlist is where they are stored for ongoing review.

A representative weekly workflow at the buy-side desk level looks like this: a Sunday-evening screen surfaces eight to twelve names; most are discarded by mid-week after fundamental review; the surviving one or two are added to a watchlist and revisited monthly. The screen is the input to a process, not the process itself, and the discipline of the procedure is what produces the work product.

Common Screener Mistakes

The recurring errors among new users are predictable enough to enumerate.

• Over-specification. Eight or ten simultaneous conditions generally returns an empty list. Three to five is the working range for most strategies.
• Ignoring liquidity. A $2 stock with 500 shares of average daily volume cannot be entered or exited at posted prices. A minimum-volume filter (around 200,000 shares per day for most retail strategies) is structural rather than optional.
• Skipping follow-up analysis. The candidate list is an input to research, not the research itself. Every name on the list still requires chart review, news verification, and a position-sizing decision before capital is committed.
• Curve-fitting. Building a screen around the characteristics of a single past winner and assuming the screen will surface comparable winners going forward. The historical record does not, in most cases, support that assumption; broad strategy-driven criteria generally outperform criteria reverse-engineered from one example.
• Ignoring the regime. A breakout-momentum screen tends to perform well in a bull market and to produce frequent failed signals in a bear market in which breakouts roll over. The screen parameters should be reviewed against the broader market trend rather than left static across regimes.

From Screen to Trade

The screener occupies the first position in a longer sequence:

1. Screen — run the filters and generate the candidate list.
2. Chart review — inspect each candidate’s chart for the intended setup.
3. Fundamental check — verify there is no imminent earnings report, pending SEC filing, or other obvious red flag.
4. Risk assessment — specify entry, stop, and target before the trade is placed (risk/reward ratio).
5. Execute — place the order with the appropriate order type.

The function of a screener is to compress the universe to a workable size. What it does not do is substitute for the judgement that distinguishes a workable candidate from an investable one, and it is generally a mistake to ask it to.