<SPECIAL FEATURE>
May 10, 1999
Interview with Legg Mason Focus Trust Manager
Robert G. Hagstrom
Part 1 Continued
DW: One thing I wanted to ask about is risk in terms of "beta." There's a riddle that Warren Buffett included in a letter to shareholders [1992], which was one of Abraham Lincoln's favorites. It goes: "How many legs does a dog have if you call his tail a leg?" The answer is: "Four, because calling a tail a leg does not make it a leg." In quantifying risk in terms of beta, are we calling beta a leg?
Hagstrom: I think so. In Chapter 2, "The High Priests of Modern Finance," we go through that and ask, "How did we come up with this whole idea of modern portfolio theory?" It looks like it came up because we screwed it up in the first place. The money management business basically blew it. We walked into 1972-1973 like a Lamborghini at 190 miles per hour and just blew it up. Everybody looked around like "what happened?" Some of these academic types said "well, we'll take a swipe at it." And we had no clues, so we signed onto that.
That's my very anecdotal analysis. You go through Bernstein's Capital Ideas and it basically reads to me like nobody had a clue, so "let's just do what these guys are proposing and see if that works." And their whole game, it seems to me was, wide diversification creates smooth rides and reduces downside risk. And what everyone was upset about in 1973-1974 was that you lost 60-70% of your money. So these guys said, "We've got a way in which this won't happen again." Everyone was saying, "I'll sign up for that!" Money managers were signing up to sell that, because that's what their clients wanted. Along the way, it seems like we got ourselves into a situation where we can't beat the market because modern portfolio theory just basically limits those opportunities.
Beta does not define risk, in my book, but we don't have any other way. Jack Treynor has an article this month in Financial Analysts Journal, and I talked with Jack about this. He's basically trying to disassociate himself from beta. He's saying that he really doesn't think this stuff explains risk at all. He wants to go to an economic risk model. He wants to define risk from an economic base, which is what Warren Buffett tries to do. Of course, how do you quantify this, how do you model it, how do you get it to a decimal? It's easier to come up with a beta with stock prices than it is to try to quantify whether Coca-Cola has more economic risk than Gap. They both had the same beta at one time, but which one has more economic risk? It's not an easy concept to get to a decimal. Obviously, with a focused portfolio, the risk is not beta, it's economic risk.
We've taught for 25 years that "bounce" is bad and you don't like it. And we're saying the only way to get high excess returns relative to the market is to accept bounce. Then the academics come back and say "well, of course you have to take more risk." They all have high betas, high standard deviation portfolios. They're saying, "Nothing new here. To get high returns, you have to have high risk." And I'm saying, "You guys defined it wrong in the first place. Let's let Warren define it differently, which is [that] risk has to do with certainty in management to allocate capital, the certainty with which the business can continue to proceed. That's risk, in my mind. Smooth is more pleasant, but it doesn't get you a higher number than Jack Bogle's [S&P 500 index fund] number does at the end of the period.
When we go out and talk to clients, the first thing we say is you have to be somewhat psychologically ready for some of this, because we'll look brilliant one quarter and sloppy another, which will have nothing to do with what we're doing in the portfolio -- it's just that the market's sunshine moves from one point to another.
The second half of the book is really more influenced by Bill Miller. We've got Charlie Munger in there, we've got Bill on complex adaptive systems, we've got [mathematician and author on gaming] Ed Thorp on probability. Bill introduced me to [mathematician and information theory inventor] Claude Shannon and [mathematician and gaming theorist] J. L. Kelly. Now we've got these 10-15 stocks -- how do we now optimize this? It's weighting the probabilities. There's the analogy of the card game -- if you bet one dollar on every hand, how much money would you end up with compared to a person who could alter their bet based upon the draw of the cards. Considering they are both equal players, the guy who can alter the bet, based on probabilities, should end up with money.
Well, how are portfolios designed? 100-200 stock portfolios have 1/2% to 1% of assets in each stock. They're not equally weighted events. One of these is a better stock than another but you're not weighting them according to their probabilities. You've already got that suboptimized. I already know that a 100-stock portfolio with 1% in each stock is totally suboptimized from a mathematical standpoint, and we try to draw that out.
I also wanted to throw in the psychology of investing and complex adaptive system discussion, because the one thing that Warren doesn't do very well is that he just basically says you can't predict the market. I always thought it was fascinating to try to figure out why we can't. The concept of complex adaptive systems finally helped me understand why we can't. The mathematics just aren't invented yet to explain the agent-based system with continuous feedback loops. Once I went out to Santa Fe [the Santa Fe Institute], I said OK, I've got no problem with not predicting the market. Now I understand why. Instead of just saying it can't be, well, I always thought there might be a way. Now I'm pretty convinced that if it's going to be discovered, it's not going to be discovered on Wall Street, it's going to be discovered out in Santa Fe or at a think tank by a Nobel physicist.
DW: One of the great ironies I found in your book was that John Maynard Keynes, one of your focus portfolio investors, thought up a national income accounting equation, which tries to sum up in a neat algorithm what happens in the complex system of a national economy.
Hagstrom: It was dicey to throw Keynes in there, because he has a mixed record as to his abilities with economics and with investing, as I understand he had a tricky time being an investor, too. When he hit the Chest Fund period, he seemed to have gotten it all together, but in periods before that I think he was a little sloppy, and there was this period where he absolutely had no idea what he was doing and he blew up a lot of money.
DW: When it comes to economics, a lot of people think it's a hard science, that you can quantify everything. Ultimately, you're trying to quantify human behavior. People exhibit patterns, but you can't always make the linear projection from past data.
Hagstrom: I agree. I didn't really have a firm grasp of that until I went to work for Bill. In the last three years, in going to Santa Fe with Bill, and his becoming a partner in my limited partnership, he said, "Robert, you've got to start to think of this stuff." I guess Bill got turned on to this about the late 1980s, early 1990s, but really geared up in it in the 1993-1994-1995 time period. By the time he was really up on the curve, he said, "Let me take you by the hand, I'm going to show you something."
And I went "Oh my God." So I came late, but it's clear that what's going on at Santa Fe is really helping Bill think in very worldly ways. So now the trick is, we've got the stock selection process, the portfolio management process, and now [we need to] let Bill help us understand how to apply this to technology, how to understand complex adaptive systems, how to understand the Internet and how it affects business. I think Bill's absolutely right. It's not that [valuing Internet and technology-oriented] companies can't be done, it's just that you have to work with different models and think about things differently. Mike Mauboussin's influence is strong here, as well.
Buffett's not saying that it can't be done, but that he doesn't feel competent enough to do it at the level where he thinks he has a high win rate. I think he's being modest; he's got a Rolodex to kill for. He can look at a bank or an insurance company and say, "There are probably only five guys on the planet that can do it better than me, and I can be number one or two." Then he sees technology and he knows that there are many people that can do it better than he. In his rational mind, he doesn't want to play a game where he doesn't think he's going to be one of the best at it. So he says, "I'll pass."
That's how a rationalist behaves. They don't want to enter into games where they don't believe they're operating at the highest level. Why bother? It would be irrational to do that. I just wish Buffett would get up and get to it. It's not that he doesn't have the skills, he just hasn't gotten around to it. That's why we went to work with Bill [Legg Mason acquired Robert Hagstrom's fund advisory firm in June 1998]. Bill could fill in that piece, which is, "How do we take the Warren Buffett Way to the next level, how do we take this methodology and figure out technology and things like America Online?"
Bill's the one that took the 2x4 and slammed me upside the head and said, "Would you please sell Disney and buy AOL?" To say that Bill's been instrumental in our performance and understanding is an understatement. Bill raises our intellectual benchmark. Just as we're looking for ways to raise our economic benchmark in the portfolio, we're looking for ways to raise our intellectual benchmark and he represents that.
</SPECIAL FEATURE>