The Daily Interview: Being Warren Buffett

With an assiduousness that would make Fox Mulder proud, fund manager Robert G. Hagstrom has been keeping what he calls the "Warren Buffett Files" since the 1980s. They contain every article about and by Buffett, as well as all of Berkshire Hathaway's annual reports going back to the 1960s, and they have helped inform the manager's own stock-picking philosophy in his (FOCTX)Legg Mason Focus Fund.

Robert Hagstrom Director Legg Mason Focus Capital

Recent Daily Interviews

National Association of Realtors' Lawrence Yun

Diane Swonk Bank One

Deloitte & Touche Martha Priddy Patterson

Liberty Funds Jim Blakeslee

B.J. Greenwald & Associates' Burton Greenwald

Here, the author of the newly released The Essential Buffett explains how, like Warren Buffett, he looks past a company's balance sheet to select stocks, and why a liberal arts approach can be useful for picking stocks.

TSC: How did you become so interested in Warren Buffett?

Hagstrom: I started reading about Warren Buffett in the 1980s when I was a stockbroker. I had heard about this rather unusual man who ran a company called Berkshire Hathaway, and I started reading his annual reports.

From there, like a kid following a ballplayer, I began collecting Warren Buffett memorabilia and began the Warren Buffett Files -- any article about him, all of the Berkshire Hathaway annual reports going back to the late 1960s, all of the annual reports of the major companies that he owned and any article Warren wrote.

TSC: Do you actually know Warren Buffett?

Hagstrom: We're not close personal friends but we keep in touch. He's always very aware of what I am writing and of all the things I've written, he's never come back to me and asked me to change anything.

I asked him one time, "What did we get wrong and what could we have done better?" and he came over to me and patted me on the back in a grandfatherly way and said, "Oh, you did just fine" and, in fact, he has given us copyright approval to reprint quotations from his own annual reports.

TSC: So, what is the secret to Warren Buffett's success?

Hagstrom: People ask me that all the time. There are three things. One, he thinks about stocks as businesses. A businessperson is very interested in how a company is valued -- companies with low debt, good profit margins and cash earnings. The second is to run a very concentrated, low-turnover portfolio. And the third thing is to try to protect yourself from some of the speculative forces that seduce so many investors.

You'll find that a lot of people talk the Warren Buffett talk, but they don't walk the Warren Buffett walk. His portfolio is very unusual. It's a concentrated, big, bad, low-turnover portfolio. You need a lot of psychological reinforcement to run a portfolio this way because it's so unusual.

TSC: Buffett is notorious for not liking technology stocks. Do you agree?

Hagstrom: Because we're not in technology, I think people have incorrectly assumed that technology is un-analyzable. That's not true. It's just difficult because there are higher uncertainties in their earnings going forward. But you can compensate for that uncertainty by reducing the amount of technology stocks that you own. You might have a 1% or 2% weighting in something with high unpredictability and a 4%, 5% or 6% weighting in something with a high sense of predictability.

You can also demand a much larger margin of safety: the difference between the stock price and the value.

I had to admit that in my own fund, we held a number of technology stocks that made a lot of money in 1999, but that got clobbered in 2000. Gateway(GTW) was down a lot, AOL-Time Warner(AOL) was down a lot. Those companies hurt us a lot.

But we currently own a Dutch company called Philips Electronics(PHG) and Applied Materials(AMAT).

TSC: How difficult is it to value companies across very different industries?

Hagstrom: Each industry is unique and has its own challenge, but what's amazing about the Buffett approach is that you can use it for all environments.

Predicting cash flows of a consumer nondurable has its own metrics; a Coca-Cola(KO), a Gillette(G) or a Procter & Gamble(PG) are easier to predict than a software company because the software industry is changing so rapidly. There are new entries and new competitors and it's a fast-evolving world with great unpredictability as to future earnings.

TSC: Last year, you wrote a book called Latticework, which spelled out a liberal-arts approach to investing. What do psychology, philosophy, physics and biology have to do with selecting stocks?

Hagstrom: That book originated with Charlie Munger, who is the vice chairman of Berkshire Hathaway. Charlie has long lectured about the need to take a multi-disciplined, liberal arts approach to thinking about investing.

In Latticework, we try to give investors some additional insight into the behavior of markets by looking at them, first from a biological standpoint as opposed to a physical standpoint, like a Newtonian framework. Like living organisms, markets, companies and industries ultimately change. Like a living environment, they evolve, adapt and compete.

In the philosophy chapter, we relied heavily on the philosophy of William James, a famous American philosopher from the turn of the century who fostered the idea of pragmatic philosophy, or using what works.

Now, what you find about investing is that people are not very pragmatic, particularly at the professional money management level. They come up with rigid views of how the world works. They craft a model to that view. And then they hold steadfastly to that view, even when the market ultimately changes.

You have to have a philosophy that allows you to latch onto new ideas and new viewpoints.

For example, 70 years ago, one of the most powerful ways of analyzing stocks was when the stock/dividend yield fell below the coupon yield of a bond. At that point in time, previous evidence showed that stocks had been overvalued relative to bonds, so you sold stocks and bought bonds. That worked from the Depression on, but it quit working in the late 1950s and 1960s as people began to recognize the value of future earnings relative to just current dividends.

From the late 1960s into the '80s, low P/E tended to predict stocks that would outperform and had a high predictive value. Then it quit working in the mid 1980s.

Today, cash/earnings on low levels of capital, or high returns on invested capital, seem to be a very big predictor of successful stocks.

TSC: You built a Warren Buffett model to select stocks through Quicken. Which stocks does this turn up?

Hagstrom: Dell(DELL) and McGraw-Hill(MHP). In my own fund, my top holdings include Citigroup(C), Wal-Mart(WMT), American Express(AXP), Home Depot(HD), AOL-Time Warner, WPP Group(WPPGY), IBM(IBM), International Speedway(ISCA) and Amazon(AMZN).

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