BETHESDA, MD (Oct. 11, 1999) -- In the past few weeks, I've been writing about Pathfinder companies -- those that are opening up communications that are bringing the world together. Because of the work these companies are doing with the Internet, it doesn't matter whether I am writing this column at Fool HQ in Alexandria, Va., at my home in Atlanta, Ga., or where I am right now -- at my sister's kitchen table.

In my previous columns, I looked at how we could select these companies, and the work I've done so far hasn't made me very happy. I looked at past financial data for a number of companies and found no significant correlation on the returns in five years from the date that information appeared. For a more definitive description of the work I did, please read my post here in the Drip board.

It does appear that buying "what you know" works well, but there is no way to quantify this. Also, I've known of companies that made products that I loved that were very poor investments. I would like to develop a good way to screen our Pathfinder companies so we can stay with the potential winners. As my post on the Drip board shows, there may be some correlation between return on equity and net margins with future returns. My sample size was relatively small, so I'm going to revisit this question with a much larger number of companies and see what I get.

There is a serious problem with buying a company on the basis of "what you know" -- it's pretty easy to convince yourself that a company is a good investment against all logic. Admitting a mistake is hard, especially after you've put a lot of money into it. We need to come up with a way to rationally evaluate companies and eliminate the bad ones. Let me take a shot at a "Buy What You Know" screen:

  1. Are you familiar with the product the company makes?
  2. Is there a wide demand for that product?
  3. Is the company a leader in making that product?
  4. Can you explain in five minutes or less to somebody totally unfamiliar with the company why you think that it is a good investment? (This comes from Peter Lynch.)
Now, buying stock in a company like Lucent Technologies (NYSE: LU) is a stretch on this screen for many of us. I probably won't be buying a "Stinger ATM-based DSL Access Concentrator" from them anytime soon. However, look at the growing demand for Internet access, and perhaps most importantly, the demand for broadband. We know that Lucent is a leading provider of the equipment needed for this. To be sure of an investment like Lucent, I recommend spending time on the company's website and reading the latest 10-Q or 10-K filing with the SEC (available in easy-to-read format at Fool Quotes & Data.)

For Lucent, you can check out its site at To get the SEC filings for other companies, check in our Quotes area under the "Info" column, click on "More Info." From there, you go down to "Looking Back" and click on "SEC Filings." You could also just click on the "Financials" link when you quote any stock, and that takes you to the SEC link. The SEC 10-K is the annual report, and the 10-Q is the quarterly report. Check on whichever of these is the most recent. Read carefully "Management's Discussion and Analysis of Results of Operations." This is a pretty blunt assessment of future opportunities and risks, and lacks any corporate hype the marketing guys always seem to throw in. It's not exciting reading, but very important.

Criterion number 4 is the most important. The reason for buying the company should NOT involve mental gymnastics. Imagine you are up on a podium talking to 100 people -- no, let's make this a tougher audience. Imagine you are explaining to your mother (or my sister -- she's worse) why you are buying this stock. Your explanation should not elicit a sarcastic "uh-huh" and not make her eyes roll.

Next week, I'm going to revisit financial measures to screen companies using a much larger sample. Also, on October 19th I am scheduled to visit Scientific-Atlanta's (NYSE: SFA) headquarters. If you have any questions that you want me to ask while there, please e-mail them to me at

[To discuss this column, please visit the Drip message boards linked in the top right of this page.]