Stock Screeners Are Good, but Not Fail-Safe. Here's Why

Stock screens may beat chat boards for investment ideas. But that doesn't make them fail-safe.

Anything but. Mark Martinez, TSC's metrics editor, and I performed the same basic "growth at a reasonable price" screen on 10 different sites. Yet our results often varied wildly. What gives?

See Also
What Works Guide to Using Stock Screeners

A few possibilities: The sites use different databases. They update on different schedules. And then there's just human error -- some screeners are so complicated you may end up inputting the wrong things.

There are other screening blind spots -- variations on industry valuation methods, for one. If you're looking for cable companies, for example, don't expect them to emerge from a price-to-earnings-multiple screen. That's "the problem with mechanizing it," says Jeff Bronchick, fund manager with Reed Conner & Birdwell in Los Angeles. "Screens by their inputs will produce slanted outcomes." For cable stocks, you might be better off screening for price to cash flow, Bronchick says.

Manager Michael James at Kuekenhof Capital Management in New Jersey was similarly cautious when we sought guidance on testing screeners. The "screening process should be the first step in a series of many when determining what stocks they want to invest in," James says. "It is up to the investor to do the research, because it is their hard-earned dollars that are being invested."

Say, for example, you do a value screen, looking for a low price-to-earnings ratio or price-to-book value. Maybe Philip Morris will pop up. But the screen won't say a word about the litigation problems that until recently held the stock back.

One possible exception to this research advice: It does not apply if you are investing in a "black-box" style, where you use a screen based on historical back testing and then invest strictly according to the screen's outcome. If that's your plan, it's even more important to have full confidence in your stock screener.

Protecting Yourself from Pitfalls?

There are steps you can take to protect yourself from a subpar screen. First, if current data are a concern, check to see how often a site updates. Marketguide adds new companies each week, the site says. And MarketPlayer is specific about the update schedule for each data item.

If you can't find the data date info on the screener site, Martinez suggests that you take a few factors like price and market cap and compare them to what you'll find on Yahoo!. That at least will tell you if you're in the ballpark.

Data-freshness issues are a big concern of reader Jerry Fording. "I don't like screeners for one important reason: Most, if not all of the information, is old. Some screeners are old by a week or more. When I first started trading I would make decisions based on these screeners and would get burned. Unless you're a long-time holder the information changes too fast to be very useful."

He points to Datek as a notch above, in that it has a screener that updates intraday. (Note: That tool is only available to the broker's account holders.)

Technical Screening

Fording's email brings up a related point: screening for technical factors. When What Works first requested reader comment on stock screeners, the unspoken assumption was that we were looking at screeners for fundamental and quantitative analysis. That is, we were testing screeners that focused on corporate performance measures and financial ratios, not technical factors primarily concerned with price and volume behavior.

But that's what some readers are looking for. James Babka of Berthoud, Colo., near Denver, wrote in to praise TC2000, a software made by Worden Brothers and used by TSC writer Gary B. Smith. "Not a lot of fundamental info but great technical stuff," says Babka. (Gary points out that the tool does have fundamental info as well, though it's strong on technicals.)

Jim Cardin likes TradingMarkets.com for "the day and swing-type traders," he writes. That screener lets you account for things like the "Plus Directional Movement Indicator" and the "Average Directional Movement Index." Any further thoughts on technical screening systems, please send them along to [e-mail].

Book Learnin'

But back to the fundamentals for a moment. The best way to make the most of stock screeners is to get good at stock screening. If you're just learning about the market, sites like Quicken are the best place to start. But once you get your legs, you won't be able to grow much there because the screening variables are limited. MarketPlayer.com, Marketguide.com and MoneyCentral.com are all good places to improve.

Each one offers pre-set screens, with explanations of the goals (value, growth, momentum, etc.) behind the screens. So you can see how screens are developed, and what gets produced. At MoneyCentral, for example, columnist Jon Markman uses the screening tool in his own work, showing readers various ways it can be helpful.

---

If there's more about screening you'd like to share, please email [e-mail]. Also, tomorrow we begin our next topic, real-time portfolio tracking and streaming-quote services. Plus, we go to the mailbag on wireless trading, and bring you What Works According To options pro Scott Fullman.

>To order reprints of this article, click here: Reprints