Intangible Benefits

IN NOVEMBER 1999 I wrote one of my best columns — at least if you measure success by the performance of the stocks I selected. To illustrate the relationship between research and development and share prices, I put together a list of nine companies that looked cheap based on an indicator I call price/R&D ratio. Since then, that portfolio has gained nearly 50%. Two companies (Cognos (COGN), Mentor Graphics (MENT)) more than doubled. Two others (Mark IV Industries, Shared Medical) were taken over at fat premiums. And in a very difficult market, only two are way down.

As you might expect, this is an exercise I want to repeat. So this month I'll review the logic behind that stellar portfolio, and I'll choose a new group of stocks that meet similar — maybe even better — criteria today. I'll also update you on the colorful debate about R&D and other so-called intangible assets.

Our accounting system, developed by a 15th-century Venetian monk named Luca Pacioli, is a weighing machine for "real" assets. Plants, invoices, inventory — things you can touch — go on the balance sheet. Anything else (whether it's lab work on a new drug or tuition so employees can learn Web programming) is an expense, subtracted immediately from earnings.

This worked for centuries. But now critics say it's out of sync with the modern world, that using old-style accounting in a New Economy makes it harder for managers and investors to make smart decisions. No one has done more to argue this case than Baruch Lev, a professor at New York University's Stern School of Business. His solution is to beef up balance sheets to include not just R&D but also brand-value, and what he calls "organizational" assets — things such as Dell's Web-based distribution system.

Valuing all this would be tricky, the professor admits. But the payoff is better accountability. Suppose a company spends billions to develop a new product. Today, the cost is deducted over time. If there's a flop, no one knows how much was wasted. In Lev's world there would be an embarrassing writeoff.

Research Budget
These 10 stocks are cheap by any measure — conventional or unconventional

Balance-sheet issues aside, investors want to know how to tell the difference between an R&D blockbuster and a dud. One way, it turns out, is to study patents. Francis Narin, who co-authored a paper with Lev, is a pioneer in this field. He has a patent-citation database and tracks things like how often a company's research is mentioned in new patent applications and scientific papers. Narin's work indicates that there is a link between these references and stock performance.

There are several hurdles here — perhaps the biggest being that there are only about 300 domestic public companies in Narin's database. But he has ambitious plans and has applied for his own patent on a citation-based stock-picking methodology. Narin's company, CHI Research, is already marketing the data to professional investors.

Even though I'm sympathetic to Lev's reform proposals, I'm more comfortable with the R&D-related stock-picking strategies of another professor: Josef Lakonishok at the University of Illinois. His LSV Asset Management (with partners from the University of Chicago and Harvard) now runs $7 billion, invested in mostly out-of-favor stocks. Since the firm started in 1994, it has posted an average annual return of 17.9% — significantly better than the typical value-oriented portfolio.

I applied lessons from Lakonishok's research (which appears in a soon-to-be-published paper written with Louis Chan and Theodore Sougiannis) to select stocks for my successful 1999 portfolio. And their results indicate that the most tempting R&D opportunities are on the bargain shelves. After studying 20 years of R&D data on thousands of companies, the economists discovered a significant relationship between R&D and market value. This is what I call price/R&D ratio, and low numbers mean that a company is trading at a modest multiple of its R&D budget. Research indicates that over a three-year period such stocks beat the market by six percentage points annually.

To create this month's table of stocks, I limited my search to 1,700 companies with market values of above $1 billion — as I did in 1999. I also used Market Guide for Windows, powerful screening software that includes hard-to-find R&D data. Then I mimicked Lakonishok's approach by zeroing in on the bottom 20% of the price/R&D universe. That left me with just over 100 stocks — fewer than you might expect because many companies (banks and utilities, for example) don't spend enough on R&D that they need to disclose it.

I made my final cut by looking only at companies where insiders have recently been buying stock. At most companies, insiders who sell outnumber insiders who buy 5 to 1, but I wanted companies with more buyers than sellers. As Lev's research shows, insiders tend to be especially savvy when their companies do lots of R&D.

My finalists are the 10 companies in this month's table. They're a diverse lot and high-profile enough that I won't discuss them separately. But pay attention to several things: These stocks aren't just cheap in terms of my price/R&D benchmark. Nearly all of them are bargains based on more conventional value measures such as price/earnings, price/book and P/E to earnings growth. And in recent weeks they've attracted significant insider buying. In short, I'd be delighted to recommend them even if they weren't an attempt to replicate one of my most successful portfolios ever.