Stock Screener

A stock screener is a database tool that lets an investor specify quantitative criteria and returns the stocks that satisfy all of them. The criteria are drawn from the figures a data vendor maintains for each company: the share price and market capitalization, ratios such as the price-to-earnings ratio and dividend yield, growth rates of revenue and earnings, balance-sheet measures such as debt relative to equity, industry classifications, and measures of trading activity. A screen turns a market of several thousand listed companies into a short list, in seconds, on the investor's own terms.

That is the whole of what it does, and it is worth being precise about the limits. A screen ranks companies by numbers someone else computed from reported figures, using definitions the investor may not have chosen and often cannot see. It is the first step in fundamental analysis, which studies the business behind the numbers, or a way to apply a technical rule to many charts at once; it is not a substitute for either. The SEC's guidance on researching investments starts from the disclosures companies must file, and a screen is a way of deciding which disclosures to read.

What the filters are made of. The SEC's own description of stock types gives the vocabulary most screens use. Growth stocks "have earnings growing at a faster rate than the market average" and rarely pay dividends; income stocks "pay dividends consistently"; value stocks "have a low price-to-earnings (PE) ratio," which "may reflect the fact that they have fallen out of favor with investors for some reason"; blue-chip stocks are "shares in large, well-known companies with a solid history of growth." A second way to categorize, the SEC says, "is by the size of the company, as shown in its market capitalization," into large-cap, mid-cap and small-cap, with the smallest called microcap and the lowest-priced called penny stocks. A screener turns each of these descriptions into a threshold: an earnings growth rate above some percentage, a dividend yield above some level, a price-to-earnings ratio below some number, a market capitalization above some floor. It then adds filters the SEC's list does not mention, such as debt-to-equity, profit margins, return on equity, the number of analysts covering the stock, and technical measures such as whether the price is above its moving average or how volume compares with its recent average.

Every number in a screen carries a definition, and the definition decides the result. The clearest case is the price-to-earnings ratio. The SEC's glossary defines earnings per share as "the earnings for the past 12 months" divided by shares outstanding, so a screen built on that definition compares price with history. Most screeners also offer a forward ratio built on analysts' estimates of the coming year, and the price-to-earnings ratio page explains why the two can differ substantially for the same stock on the same day. A company with negative earnings has no meaningful ratio at all, and screeners handle that differently: some report it as zero, some as blank, some as a very large number, each of which sorts differently. Dividend yield may use the last four payments or the most recently declared payment multiplied by four, which diverge whenever a dividend was just raised or cut. Sector and industry labels come from whichever classification scheme the vendor licenses, so a company can be a technology stock in one screener and a consumer company in another. And every figure is as of a date: a screen run today on last quarter's balance sheet is describing a company that has since reported again, or has not yet reported the quarter in which something changed.

What a screen cannot see. A screen reads results, not reasons. The value stock page describes the failure mode: a low price-to-earnings ratio can mean an overlooked business or one whose last year of earnings included a gain that will not recur, whose customers are leaving, or whose balance sheet is strained, and "both look identical on a screen that ranks by price against past earnings." A high dividend yield can mean a generous payer or a share price that has collapsed ahead of a dividend cut. A low debt-to-equity ratio can reflect prudence or a company that cannot borrow. The screen has done its job by producing the list; the investor's job begins with the question the screen cannot ask, which is why. A screen also cannot know what is not in the database: a pending lawsuit, a customer concentration, a management change or an accounting restatement all live in the filings, and the SEC's guidance on reading a 10-K and an 8-K is the reference for finding them.

Verifying what the screen returns. The SEC's page on using EDGAR describes it as free public access to corporate information that lets you "research a public company's financial information and operations by reviewing the filings the company makes with the SEC." The annual report on Form 10-K carries audited financial statements, risk factors and management's discussion of results; the quarterly Form 10-Q carries unaudited statements and updates; Form 8-K discloses material events between them. A figure that surprised you in a screen, a very low ratio or a very high yield, is a reason to open the 10-K and find out whether the vendor's number reflects a fact about the business or a fact about the data. For funds rather than stocks, the SEC's Researching Investments page points to FINRA's Fund Analyzer as the tool for comparing fees and expenses, a reminder that a stock screener is one tool among several and not a general-purpose research service.

Where screens earn their keep. Used well, a screen enforces discipline: it applies the same test to every company, it surfaces names an investor would never have thought to look at, and it makes an investment thesis explicit enough to write down as criteria. The value investing page notes that a rules-based screen is one of the two things that go by that name, and the same is true of dividend and quality strategies. The discipline cuts both ways. A screen that returns forty stocks with the same attractive ratio may have found forty companies with the same problem, and an investor who buys the list has concentrated in a factor, not diversified across businesses.

The investor chooses a universe (an exchange, an index or all US-listed stocks), sets one or more criteria as thresholds or ranges, and runs the screen. The tool tests every stock in its database against every criterion and returns the survivors, usually sortable by any column. The investor then reads the survivors' filings to learn what the numbers mean, discards the ones where the screen was fooled by the data, and investigates the rest.

A hypothetical example. Devon screens all US-listed stocks for a market capitalization above $2 billion, a price-to-earnings ratio below 10 and a dividend yield above 4 percent, and 23 companies pass. One is Harlow Industrial, trading at $40 with reported earnings per share of $5.00 over the past twelve months, a ratio of 8 ($40 divided by $5.00). The 10-K shows that the year's earnings included a one-time gain of $2.50 a share from selling a division. Recurring earnings were $2.50 a share, so the ratio on the business Devon would actually own is 16 ($40 divided by $2.50), above his screen's limit. Harlow passed because the vendor's earnings figure was correct and incomplete at the same time.

Another survivor, Pell Utilities, shows a yield of 6 percent at a price of $20, implying an annual dividend of $1.20. Its most recent 8-K, filed after the vendor's last update, announced a cut to $0.60 a year. The yield Devon will actually receive is 3 percent ($0.60 divided by $20), below his threshold. Two of the 23 names were artifacts of the data; the screen was still worth running, because the other 21 are companies he had not looked at before, and the ratios that flagged them are a reason to read on, not a conclusion.

Pros

• It applies the same criteria to thousands of companies at once and surfaces names an investor would never have found by browsing.
• It forces an investment idea to be stated as numbers, which makes it testable and repeatable.
• It is fast and, at most brokerages and financial sites, available at no cost, so the expensive step, reading the filings, is spent only on the survivors.

Cons

• Every figure rests on a definition the investor did not choose, and small differences in how earnings, dividends or sectors are defined change which stocks pass.
• Data is as of a date. A screen can return a company whose situation changed in a filing the vendor has not yet processed.
• A screen reads results and cannot read reasons, so it is the tool most likely to hand an investor a value trap dressed as a bargain.
• A list of stocks that share one attractive ratio may share one hidden problem; buying the list concentrates rather than diversifies.