Bill Miller
CEO, Legg Mason Funds Management

This conference is a big experiment in behavioral finance. We have collected a big stack of questions from the audience, and I looked through them earlier to sort them into categories. I discovered something amusing. Remember the anecdote that I told at the outset of this meeting about the investor's process that didn't work. His process involved projecting the economy, the market outlook, industries and down to particular stocks. Well, most of the questions are about these very topics! What's your view on the economy? What's your view on the market outlook? What's your view on these various industries? This process does not work, but I will answer the questions anyway. I will also draw some conclusions that also will not work.

Before I begin, however, I would like to talk about the results of the behavioral finance experiments that we tried here at the conference. We gave you all a questionnaire with a number of different behavioral finance questions on it. The first question was about overconfidence. 88% of the people here say that they are above-average drivers - they believe they are better than the other drivers in this audience.

The second question was about loss aversion and timid decisions. We asked how much money you would have to win in order to bet $100 on a coin toss. The average answer was $284. Across all of the populations that have answered this question, the answer is generally two-to-one, or $200. This audience, then, is more risk-averse and has a greater aversion to loss than most people. If that is your perspective, then I would encourage you not to look at the stocks we hold in the portfolio!

The third question asked you to estimate the number of jellybeans in a jar in the registration booth. The actual number was 2,683. The average of all estimates was 2,566. I don't have the final dispersion in front of me, but a preliminary dispersion ranged from 98 (that person must have had retinitis pigmentosa or something!) to over 10,000. This experiment has been done a number of times, and the average of everyone's bet is always far better than most individuals' bets. Because we have a large sample size, the difference between the actual and the average guess was 117 jellybeans. Only 2% of people who guessed (seven individuals) beat the "market" or the average guess. This is a great example of collective problem solving. These experiments show that in every possible venue, markets are incredibly good at coming to solutions to problems like this. This leads us to conclude that when we look at our portfolios, we must think very differently than most other people. That necessarily engenders controversy. As Paul DePodesta talked about with his baseball team, people never remember all of the times your approach was right, but there is always a flood of second-guessing when you are wrong.

The fourth question was a framing question: what business is Amazon.com in? 32% said they were retailers. 16% said they were a bookseller. 14% said they were a seller. 38% of you had a dispersion of other answers. By requiring a one-word description, we forced you to narrow in on a particular category or characteristic. This is an example of how people typically approach complex problems - they try to get the simplest possible description. Those simple descriptions, then, tend to frame whatever issues you are trying to solve and what types of evidence that you are looking for.

The fifth question is the experiment that behavioral finance people call "the beauty contest". We asked you to pick a number between 1-100 which you think is 2/3 of the average of the numbers that everyone else in the room will select. The question asks you to pick a number between 1-100 and take 2/3 of it. Since you know that everyone in the room has the same question, however, you must imagine what everyone else will pick to be 2/3 of the average and take 2/3 of that. In this case, the mean for this audience was 47, which meant that the average person in the audience guessed 67 (which is 2/3 of 100), and took 2/3 of that to get 47. Three people picked the correct answer, 32, which is 2/3 of 47. Sixty-five people guessed 32 or less, which is about 20%. What is interesting about that figure is that 20% is about the percentage of people that can typically beat the market over the long term - this experiment might be accidentally representative of this phenomenon. When they give this test to other groups, there is a correlation between things like SAT and GMAT scores and how people think about this problem. The average Wharton student guesses that the answer is 38. The average Cal Tech undergrad guesses 22. The theory is that the more you think about this problem, the lower your guess should be.

These are all interesting examples of phenomena that occur over and over. I believe that every exploitable anomaly in the market is behaviorally based. This is the only way that sustainable anomalies can be created. These are the anomalies that are not easily arbitraged away.

I would like to start with a story that I hope is not too confusing. People have asked me what the point of this conference is. The idea for this conference is to expose you to people that we have found to be interesting and useful in helping us think about things, or who represent processes or styles of analysis that we think are helpful and useful. The criterion for this conference was usefulness. We did not have a particular objective in mind. This reminds me of one of my favorite stories, even though it might only be mildly amusing.

The great Austrian genius Ludwig Wittgenstein was teaching a class in the foundations of mathematic at Cambridge University. One of the people who decided to attend the lectures in 1939 was Alan Turing, the father of computing. Turing was known even then as a mathematical genius, and he wanted to hear what this great philosophical genius had to say about the foundations of mathematics. During his lectures, Wittgenstein would ruminate and say that "Turing would think this" or "Turing would say that." Finally Wittgenstein asked Turing directly to discuss the role of contradiction in mathematics. "Why is contradiction important in things like proofs?"

This puzzled Turing. "Well, it's part of the meaning of 'proof'. You cannot have a proof if there's a contradiction. If you have a contradiction, you don't have a proof."

Wittgenstein said, in essence, "So what?" Turing was now even more puzzled, and tried to explain it again. Wittgenstein asked him, "Why can't I just think of it like a stop sign? When I'm working on something and run into a contradiction, all that it tells me to do is to stop and to go in a different direction."

Turing argued that it wouldn't be mathematics if you thought of it that way. Wittgenstein continued with his train of thought. "If it's just a stop sign, I could choose to just ignore it and go straight through the stop sign."

Turing is getting very frustrated. "No, you cannot just ignore a contradiction in mathematics. If you ignore a contradiction in mathematics, then you'll get a nonsensical or wrong answer."

Wittgenstein countered that the answer would not be wrong, just different.

They went back and forth in this way for awhile, until Turing finally says, "If you were to use mathematics that contained contradictions in order to do something like building a bridge, then the bridge might fall down."

Wittgenstein said, "Yes, it might. That would be the sense of the contradiction, wouldn't it? Isn't that why it would be important not to ignore it, and not because it has anything to do with mathematics or proof?"

Turing thought for a moment and said, "Oh, I see your point."

In response, Wittgenstein got very angry and said, "I have no point!"

That is the theme of my remarks here at the end of the conference. I have no point that I will try to get across to you! This conference has no point, either, except that we tried to do something useful.

Lee Cooperman talked earlier about the length and depth of the bear market. This quote from T.S. Eliot expresses what it was like as a money manager or professional investor to go through the bear market. Since you were all invested with us, this probably captures the essence of what you were feeling as well.

"Where is there an end of it, the soundless wailing…
Where is there an end to the drifting wreckage…

There is not end, but addition…
Years of living among the breakage
Of what was believed in as the most reliable…

We had the experience but missed the meaning…"
-T.S. Eliot, "The Dry Salvages"

I want to highlight two words from this quote - "belief" and "meaning". Part of my comments will focus on the role of beliefs and the notion of meanings. What do words like "value" or "value investing" really mean?

I showed this quote to Chris a few months ago, and he recommended that we use a Calvin and Hobbes cartoon to express the same thing. In the cartoon, Calvin tosses up a water balloon, runs to catch it, and it explodes on his head. He walks away, sopping wet, and says, "How can something seem so plausible at the time and so idiotic in retrospect?" This is what Paul DePodesta and Michael Mauboussin would call "bad process and bad outcome".

William James said this about belief: "Beliefs are man-made… They are a conceptual language we use to write down our observations of nature… Ideas become true in so far as they help us get into satisfactory relation with other parts of our experience… A belief is true and has 'cash value' if it helps us get from one place to another. Truth becomes a 'verb', not a 'noun'."

We look at things in the context of usefulness. We try to be pragmatists when it comes to information and truth. There are three theories of truth: a correspondence theory of truth, a coherence theory of truth, and a pragmatic theory of truth. Most people in their natural lives are naïve correspondence theorists when it comes to truth and reality. These people tend to believe that there is a reality out there, and that something is true if it corresponds to reality. The statement "I am in a room with a lot of people" is true because, in reality, I am in a room with a lot of people. This certainly sounds like a reasonable view of truth, but it happens to be wrong. Why is this wrong? Isn't there a reality and doesn't the truth correspond to it? Philosophers have thought about this for a long time. We get the information we have about "reality" by mediating it through our senses, but our senses can make errors. How can we tell that our senses are not deceiving us? This led them to the coherence theory of truth.

The coherence theory of truth admits that we don't ultimately know what reality is. We can, however, make all of our beliefs logically cohere so that we don't believe things that our contradictory. This theory encourages us to put all of our beliefs into a coherent whole. That's pretty good.

The pragmatic theory of truth goes a step further. We want our beliefs to cohere in order to be useful, but why should we care about coherence at all? Usefulness is the only trait that matters. As William James says, truth then becomes a verb instead of a noun. You want truth to get you from one place to another.

The best anecdotal way of looking at these theories of truth is the story of the three baseball umpires. They are asked how they call balls and strikes. The first umpire believes in the correspondence theory of truth and says, "I call them as they are. If they are balls, I call them balls. If they are strikes, I call them strikes." The second umpire believes in the coherence theory of truth and says, "I call them as I see them." The third umpire, a pragmatist, says, "They ain't nothin' until I call 'em."

"For a large class of cases - though not for all - in which we employ the word 'meaning', it can be defined thus: the meaning of a word is its use in the language." (Ludwig Wittgenstein, "Philosophical Investigations")

This tests the word's usefulness. Wittgenstein believed in his 30's that he had solved all of the problems to do with meaning and language in his book called the The Tractatus. He then quit philosophy because he said that he had solved all philosophical problems with basically a picture theory of language. He later realized that he was totally wrong, reentered philosophy, and spent the rest of his life trying to figure out this "use" approach.

Let's think about the use of the word "efficiency" when people talk about efficiency in markets. Richard Roll, one of the gurus of the efficient market hypothesis, said this:

"I have personally tried to invest money, my client's and my own, in ever single anomaly and predictive result that academics have dreamed up. And I have yet to make a nickel on any of these supposed market inefficiencies. An inefficiency ought to be an exploitable opportunity. If there's nothing investors can exploit in a systematic way, time in and time out, then it's very hard to say that information is not being properly incorporated into stock prices. Real money investment strategies don't produce the results that academic papers say they should." (Wall Street Journal, 12/28/00)

The key here is "systematic". Many academic papers find these approaches that they claim would have produced excess profits had you followed them some time in the past. As Roll says, when you try to put these things into practice, it never works. The market is always smarter than these particular strategies, notwithstanding the fact that when you back test them, it appears they should work.

There was some interesting work by a scientist named Blake Lebaron, who was affiliated with the Santa Fe Institute. About ten years ago, he studied a number of technical trading rules. Most market professionals, value investors and finance professors think that technical trading doesn't work. If it did work, then there should be a lot of technical traders who are making a lot of money, but it is very difficult to find many of these people. Blake found that there were some technical trading rules (not very many) that worked. The most powerful of these was Simple Moving Averages. He published this in a very well-known paper. About two or three years ago, a few academics decided to re-look at Lebaron's work in light of new developments in statistics since his time. They wanted to test his methodology using the latest statistical tools. Lo and behold, his methodology and data all checked out! They decided to update his study to today. When they looked at the more recent data, however, they discovered that his technique stopped working about a month after he published it. They went on to test roughly 1,100 technical trading rules and discovered that not one of them produced any excess returns. Markets are remarkably efficient.

This take us back to value investing and what we mean by this term. People tell us that we're not pure value investors because we have some growth stocks in our portfolio. We agree perfectly with Warren Buffett on this issue, so I'll just quote him.

"Whether appropriate or not, the term 'Value Investing' is widely used. It connotes the purchase of stocks having attributes such as a low ratio of price to book value, a low price-earnings ratio, or a high dividend yield. Unfortunately, such characteristics, even if they appear in combination, are far from determinative as to whether an investor is indeed buying something for what it is worth and is therefore truly operating on the principle of obtaining value in his investments. Correspondingly, opposite characteristics - a high ratio of price to book value, a high price-earnings ratio, and a low dividend yield - are in no way inconsistent with a 'value' purchase." ("The Essays of Warren Buffett")

Lee Cooperman is a great investor and one of the most knowledgeable and sophisticated people in investing in the world. Lee associates "value" with the characteristics that Buffett describes in this quote. When Lee talked about the bull market, he said that bull markets don't typically start with characteristics like low dividend yields and high P/E ratios. The conclusion that one draws from this is that there is not much value in the market. I am not disagreeing with that, but I want to point out that the way Lee uses the word "value" is in the more traditional, accounting-based way. If you invest with him, that is the way that he thinks about value.

The next quote appears at the heading of every one of our annual reports. "When we think about the future of the world, we always have in mind its being where it would be if it continued to move as we see it moving now. We do not realize that it moves not in a straight line… and that its direction changes constantly." (Ludwig Wittgenstein, "Culture and Value")

If you take this quote to heart, then your default position in the market is that whatever you are currently experiencing in the market (high prices, low prices, making money or losing money) is going to change. You can count on it. This is one of the very few close-to-certainties that you can find. Because the market looks forward, because the market discounts, and because the market prices reflect, in essence, the data refracted through the decision procedures and emotions of investors, then the market will change as the world changes because it is incorporating new information.

Peter Bernstein wrote a piece for the November 1 edition of Economics and Portfolio Strategy. Bernstein is a great economics and investment thinker. He is talking here about the trade and budget deficits.

"It is important to emphasize that the data pouring out unmercifully every minute of the day and night are only a record of the past. The measured imbalances between the private sector savings and investment, government taxing and spending, and the deficit are only a reflection of how each sector has reacted to the ongoing changes in the other sectors. Any resemblance between the data we see and what people anticipated at the beginning of any time period is purely coincidental. [That may be a little too strongly-worded.] The process is a dynamic one, in which outcomes for each player depend upon the decisions made by all the other players, and as a result, almost all outcomes are a surprise and provoke responses of their own."

There are two things that I tell our analysts. First, 100% of the information that we have about any company or any investment reflects the past. 100% of the value of that investment depends on the future. The real question is how the past data connects with the future. The key characteristic of capital markets from an investor's standpoint is what's called nonstationarity - the degree to which the future does not resemble the past and you get novelty, surprise and the other things Peter Bernstein was talking about. People rely too heavily on past data - when the market has done this, you should do that. This is all valuable information and should play a part in your investment process, but most of the time, people are insufficiently aware that there is going to be some degree of nonstationarity there. A key question is not what has happened in the past and how will in relate to the future. Instead, how will the future be different from the past? What does this data tell us about the potential sources of difference?

This last quote from Wittgenstein has to do with framing and thinking about problems:

"Take as an example the aspects of a triangle. This triangle can be seen as a triangular hole, as a solid, as a geometrical drawing; as standing on its base, as hanging from its apex, as a mountain, as a wedge, as an arrow or a pointer, as an overturned object which is meant to stand on the shorter side of the right angle, as a half parallelogram, and as various other things." (Ludwig Wittgenstein, "Philosophical Investigations")

As you move through those various descriptions, you can see how three simple lines can be seen as any one of those particular things. How you see it depends on how you want to see it or how someone has suggested that you see it or how you are trying to use those lines in some set of contexts. The lesson to take from this example is that even such a simple thing as seeing a simple set of lines is totally context-dependent. Framing and understanding the context is critical.

There is a paper that I have not yet read (which I nevertheless think is very good) by Richard Rorty. He is probably now the greatest living philosopher. Rorty is talking about how to think about very broad problems. He says that all problems are problems of description. All of the issues that people have in the philosophical realm tend to be problems of description. Every problem in the history of knowledge can be thought of as a problem of description. If we find it to be the case that string theory does a better job than current physical theory at describing what deep physical reality is like, then all that will mean is that the current description has changed. Darwin or Newton or Einstein offered a different description to use in thinking about nature. Mathematicians offer different descriptions. People have different descriptions of Moby Dick or different interpretations of poetry. What makes his views do controversial is that he says at the end of the day, there is no difference between an interpretation of Moby Dick and Einstein's Theory of Relativity (or any other scientific thing which purports to express reality). The only issue is whether this particular description is more useful than another description. If you ruminate on this for awhile, you would probably come up with the best description of how our investment process is animated and informed by this type of thing.

Now, let's move from this very broad realm to what you might be more interested to hear. What do I think of the market? We are in a bull market. It is not a cyclical bull market and a secular bear market. Although it could be, that's not what I happen to believe because the evidence does not support it. We are in a bull market. We had a triple or almost quadruple bottom at the end of the last bear market. The bull market will be characterized, like most bull markets are, by strong returns at the front and then weaker returns at the back end.

Chip mentioned that many people think that the market is ahead of itself. It is hard for me to see how the "rational" or "intelligent" investor could have a view other than that the market should be "ahead of itself". The reason is that the market looks forward. The market should be incorporating the expectations of the group of investors about what is going to happen in the future. When people say that the market is ahead of itself, the data they are looking at only reports on the past. One of the most remarkable behavioral anomalies that we see is that people take today's data (e.g. the GDP report, the unemployment report) and conclude that the market is getting ahead of itself. The market does not look at today's data. It is looking at the data down the road.

There is a tendency among young analysts who have just gotten their CFA to make this kind of error. They will project a stock's performance over the next six or twelve months based on today's data. They go through elaborate calculations to peg the valuation of the stock in the future. They use the wrong time horizon for their valuations, though, especially if they are using accounting-based valuations. Younger people invariably look about a year into the future, give an outlook, and discount it back into the price. I tell them that I don't want to know the outlook for the next twelve months. I agree with Ben Graham that the market is not unaware of the outlook for most businesses over the next six to twelve months. The real question is my judgment is not what the market thinks will happen over the next twelve months. Instead, in twelve months, what will the market be looking at twelve months hence? That is the viewpoint that will determine whether or not a stock is attractively priced today. It has nothing to do with the next twelve months, unless it is really surprising.

If we buy the stock today, then each day we move forward, the headlights of the market move forward one day (let's call it). The return that I earn over the next twelve months is the difference between the market's expectations for the first twelve months relative to the expectations it will have twelve months hence. You are looking at an expectation set change one year forward.

If you look at the S&P 5001 over the last twelve months, it is up 20% or more (as of last week). The earnings expectations twelve months ago for this last year were about $54. The actual earnings of the S&P for that period were about $52. The earnings were actually lower than the market expected a year ago, but the market is up 21%.

Look at individual stocks as well. People say that a stock declines because it misses market expectations. In some fundamental sense, that is true. This phenomenon can sound counterintuitive to you. If I told you that IBM was supposed to make $5 over the next year and they really made $3, you would expect the price to drop. Their price might go up 50%, though, if in twelve months the market expects IBM to make $7 over the subsequent year.

This represents a very important analytical offset, but most people don't pay enough attention to it. There is a variant to this idea that we also tell our analysts: "If it's in the newspapers, it's in the price." By the time you read about it in the newspapers, the market already knows about it. There will be some ongoing frictional discounting - the evidence indicates that it takes a few days for the prices to fully reflect new information. As far as we're concerned, however, this is just noise, and we are not interested in trying to take advantage of it on behalf of our investors. Anything you see in the paper will be in the price. What will not be in the price is what is not discounted. People's expectations change slowly. That is what we're seeing right now in the stock market. We are seeing analysts' and investors' expectations changing slowly. The economy is getting stronger. Earnings estimates are being exceeded consistently. Companies are raising guidance. Stocks are extremely firm. It looks to us like the path of least resistance for profits is considerably higher over the next twelve months. What is discounted in terms of the path for profits is lower than what is likely to happen. That would tend to indicate that the market should be very firm, all things being equal. This should hold true as a provisional hypothesis. If in twelve months something really bad is going on, then things could be lower, no matter what the actual results are.

This all sounds highly complicated with a lot of feedback loops. It is. So how do we think about these issues, and how do we put this thinking into our portfolios? It seems to me, thinking probabilistically, that the best place to be is probably equities. If equities are going to surprise us, it seems more likely that they will surprise us on the up-side, at least if fundamentals are any guide.

It does not appear that the Fed is in any hurry to tighten interest rates. An economist named David Malpass of Bear Stearns expects the Fed to start tightening in the first quarter of next year, and then have a series of rate hikes throughout the year. He says this is because the reflation process is underway and the economy is a lot stronger and the Fed will be forced to tighten. On the other hand, Paul McCauley, the economic strategist at PIMCO (a competitor to our guys at Western), believes almost the exact same economic scenario but does not believe that the Fed will tighten at all. We agree with Paul that the weight of the evidence points in this direction. How do we factor this view of the economy into our portfolio? We do not. If we tried to conform our portfolio to an economic forecast, then we would be doing very badly. We would not be having a conference, and we probably would not have a job if that's how we ran our portfolio.

We are asking what the evidence shows us about the Fed's likely course of action. My evidence includes the statements of the Fed governors, especially Chairman Greenspan. He reiterated the other day that the probability of an unwelcome fall in inflation, while minor, still exceeds that of an unwelcome rise in inflation. That statement has appeared in the speeches of Chairman Greenspan and several other Fed governors. That statement is just about as unequivocal as you can get from the Fed that there is no tightening on their minds. When that statement disappears or is changed, then we can begin to think about new evidence for tightening. It always surprises me how much people talk about what the Fed "should" be doing or what they think the Fed is going to do. The Fed is pretty clear in most cases, if you look at their words without preconceived notions.

Here is some of the evidence that we're seeing. Productivity is going through the roof. The economy is expanding far more rapidly than anyone expected. The GDP forecasts from a month ago were in the 4-5% range for the quarter that just ended, and the actual growth was closer to 7.5%. Corporate profits are growing. Stocks are firm. We are seeing evidence that the unemployment rate will start to head down. When you put all of that together, the outlook for stocks is very good.

When I was listening to Terry Odean, I realized that I made a big and stupid mistake with the Opportunity Trust. The mistake was to put shorts on during the summer. That is not to say that it is still a mistake to have them on now, but it was a mistake back then. The mistake was looking back at similar circumstances in history and trying draw too many parallels. My reasoning for this decision, vastly oversimplified, is as follows. The last time I was in this "movie" in the summer of 1987, we were up over 47% in the Value Trust in the first six months of the year. By the end of the year, though, we had lost money because the market crashed. It's pretty bad to have a 40% gain in your pocket and then to end the year with a loss. The Opportunity Trust was up about 40% going into June 2003, the market was "way ahead of itself", and the stocks in the portfolio had done far better than I had expected them to in a short period of time. I thought that the risk-reward outlook in that situation wasn't all that great - a terrorist attack or any number of other things could have brought the market down again. I didn't want to be in the position again of having a 40% gain in the first five and a half months of the year, only to have it drop to 5% at the end of the year if things go awry. I decided to take a lot of volatility out of the portfolio by putting a significant amount of shorts on, which will give me only about 60% net long exposure. After listening to Terry, though, I realized that this decision was stupid because this market is nothing like 1987. The market is a conditional probability exercise. The conditions in 1987 were significantly different - the Fed was tightening in 1987. They wanted the economy to slow down, and they were pulling liquidity out of the system. In 2003, the Fed wanted the economy to accelerate, and they were putting liquidity into the system. The risk-reward profile was actually very different than what I thought it was. It is still the case that a terrorist attack could happen to drive the market down, but the circumstances and context that caused 1987 to be bad for us were nowhere in evidence in 2003, and yet I was using those two situations as though they were analogous. That was a dumb mistake.

Let me summarize. We are in a bull market. The most likely path of the market, given the current evidence, is that we're going to have a really good year this year, we will probably have a pretty good year (above average) next year, and then the risk-return profile may shift in 2005 and 2006. Absent the Fed tightening and absent a geopolitical event, the odds are stacked overwhelmingly in your favor with equities. I hope that this clarifies our overall thinking.

If the outlook is positive for equities, then the question becomes, "What kind of equities?" Again this leads us to think about expectations and what's discounted. Many people have pointed out that the types of stocks that have done the best this year are the worst stocks. These are the stock with no or low earnings, high volatilities, high debt loads, or low absolute prices. When you hear people say that, it makes you think (as Jeremy Rantham put it) that this is the greatest sucker rally in history. Some people believe that this represents a return to speculation and a return of the bubble because all of these low-quality stocks are going up. I think this is wrong. The reason that these so-called low-quality stocks are going up is that they went down too far in the bear market. This is the area that saw the greatest under-valuation. Now those stocks are coming back to levels that make a lot more sense. AES should not have been $.75. Nextel should not have been $2.50. Amazon.com should not have been $6-$7. These prices were radically wrong. If you use those prices as the starting reference, then it appears that the stocks have increased 300-500%, but their prices should never have been so low in the first place.

When you look at the valuation matrix in the market today, things are very different. The market is in what I would call (probably incorrectly) some sort of thermodynamic equilibrium. It is very hard for us to see how we can change the portfolio significantly and improve the returns, adjusted for trading costs and market impact. There are no broad areas of the market, in my judgment, where it is clear that the pricing is wrong. There are many areas of the market in which the pricing looks moderately more attractive than the stocks we currently own, but not so much more attractive that it moves outside the estimation risk that is involved. If you then look at history, then those types of stocks should do better. Those types of stocks tend to be much bigger companies and much higher-quality companies that tend to trade at somewhat lower valuations. We don't own these companies, but this list includes names like GE, Merck or Pfizer or AIG. These are companies with good quality names, good management, good long-term histories, solid dividend yields, that would be included in a portfolio like Chris's. That is the area of the market where I think we have the best probability of seeing concentrated returns over the next year. These are the types of names that we're having our analysts spend a lot of time on. A name that we're working on right now is Mattel, which Carol Siedman-Becker owns. This company fits those characteristics, but we are still figuring out whether or not it is attractive for us.

The next question might be whether these attractive stocks will find their way into our portfolio in the next year. I don't know. We may or may not buy them. It would not be inconsistent for me to tell you that I think this is the best area of the market, and then not put any money into it. Again, it comes down to estimation error and trading costs. A lot of the data that Terry Odean showed us about what individuals do improperly when they trade reflects the very same decisions that institutions make. Unless an institution is run by a computer, it is run by individuals just like all of you. The whole mix of people in the markets is no more immune to these behaviors than the people in the E-Trade commercials.

I would like to focus on how we are thinking strategically. Our turnover this year is only about 4%. This illustrates in real numbers exactly what I've been talking about. We are going to sit there right now, and let the opportunities arise and come to us. What we heard on the Hedge Fund Panel was interesting. I think Lee is right about the catastrophe insurance business. Brian Posner has some interesting thoughts on natural gas - his thinking is very consistent with everything we can see about the energy sector. We have not owned energy since 1986, when Ernie correctly put the Value Trust into energy stocks with the collapse of oil prices down into the $7 range. This is one of the very few times that have looked at energy, because it appears as though there might be a long-term issue when it comes to natural gas. Because the markets are related, there may also be opportunities in oil as well. We have to overcome the long history of commodity prices going down in real terms. All extractive commodities without exception have gone down in real terms for as long as there has been data. This is not to say that there may not be trading opportunities because of OPEC or the 1970's. These commodities have been historically bad investments, but good trades from times to time. Again, that may be changing, though, and we will try to be sensitive to that type of change.

Much of the excess returns that we have been able to generate in the past have been due to just avoiding things that other people do. We avoid economic forecasting or any forecasting where the forecast is critical to how you construct your portfolio. As Lisa made clear, our process is about a central tendency of value under a wide range of scenarios. The market hops from scenario to scenario. I remember giving a talk in Chicago seven or eight years ago. Someone asked me how we know that our forecasts are right. The answer is that we have so many different forecasts that one of them is bound to be right. We just don't know which one it is. We look at what the evidence shows, what is discounted in the stock price, and move from there as we look for opportunities.

That concludes my prepared remarks. I would like to take some time to respond to questions from the audience.

There were so many questions that we will not be able to go through them all. There were four basic groups of questions: Broad questions (market, economy, etc.), Fund construction, Industry, and stock selection.

Question: What are your expectations for inflation and total returns over the next ten years?

Bill: As Brian said in response to another question, the technological change that's underway is generating very high productivity growth. Why is that? Bill Gurley said that it might not be a fundamental change that has kicked us onto a higher productivity growth curve. Instead, the growth we are seeing might be the residual result of all of the IT investments that were made during the boom years. These productivity gains may just be working their way through the system, and as it passes through, we should see productivity drop back down to the more normal path. I would tend to agree with Brian's assessment. Bill was not making a forecast - he was just putting forward a hypothesis. Productivity began to boom in 1995, and there was no major investment proceeding that boom. In fact the capital expenditure cycle had been lower, yet productivity had started to increase. This would tend to support Brian's hypothesis that we were beginning to see the effects of the long-cycle trends of digitalization and widespread use of PC's. If that's the case, then productivity growth could be very high over the next few years.

If productivity is high, then it will be very hard for inflation to return. Then we will hit the Baby Boom trends. The effects of the Baby Boom will be very interesting. Some people at the Santa Fe Institute wrote a paper on demographics and long-term returns. The work indicated that there is an astonishing long-term correlation between demographics and interest rates and returns. Roughly speaking, the older the population, the lower the level of interest rates. The lowest interest rates in the world are where the population is the oldest, namely Japan. High interest rates tend to be in areas with younger populations. In some ways this is just supply and demand. Housing prices took off when the Baby Boomers got to the age of buying a house. Consumption as a percentage of GDP jumped from 61% to 67% when Baby Boomers moved into their high-consuming years. The stock market took off and interest rates fell when the Boomers got into their 40's and started to accumulate capital for long-term growth. So what happens when these Boomers start to liquidate some of their assets? They are not necessarily liquidating in order to consume, but in order to change their risk profile. In the year 2010 or 2020 Boomers' income will get lower and lower because interest rates will be so low. This will have to factor into our thinking. What happens when the cyclical nature of interest rates completely changes? What if interest rates are driven down to 1% or 0.5% or even to zero? That will have implications for lots of different things. The shortage of income hypothesis is one we will keep in mind.

This would change the expected returns of different asset classes going forward. Absent that scenario, expected returns would be 6-8% for stocks and 4%+ for the ten-year bond. We think the expected rate of return for stocks is in the mid-to-high single digits. This represents our cost of capital when we look at companies, and that is the cost of capital that we have to beat. Dividends become much more important in this kind of environment because it's cash in your pocket. I was told ten or twelve years ago that our portfolio looked good, but that our dividend yield was too low. I responded that financial theory tells us that the rate of return on an asset is independent of the way in which it is paid out. Therefore it makes no difference whether the dividend is retained on the company level or paid out, unless you believe that the market systematically undervalues dividends, or if you can systematically reinvest that capital better than your portfolio companies can. Some people believe both of these things, and build portfolios that reflect it.

In either the high productivity growth scenario or the low-income scenario (due to demographics), inflation should not be a problem.

Mark Cuban, the owner of the Dallas Mavericks, once said something very smart (believe it or not). He did manage to sell his company for over a billion dollars right at the peak of the Internet boom. About a year ago, he said that he was not interested in stocks until they start paying lots of dividends. Why would he be interested in dividends? He is worth billions and doesn't need the income. He said that his return on a stock will either be the dividend that he receives or the capital appreciation. The problem that he sees is that the Baby Boom generation is heading into retirement, and they will want to start collecting some income. They can get this income either by selling their stock or by having companies pay it to them through dividends. He went on to say that that's true for any individual, but it's not true for society as a whole. Keynes said that there is no liquidity for the economy as a whole. Let's say that the demand for dividend yield on the existing asset base is 3%. Let's also say that the yield in the market is 1.5% The society as a whole in this scenario will have to see 1.5% of their stocks every year in order to get the necessary income. If that's the case, then that will put greater pressure on the market at the margin than would otherwise be there. Cuban said that until he could see that the economy as a whole could generate enough income, he did not want to be in stocks and make a commitment to an asset class that might be under long-term pressure.

Question: Part of your answer says that you expect to receive 6-8% return on stocks before taxes. If that's the case, then why not just get a 4-5% no-risk, after-tax return on muni's?

Bill: There are multiple answers to that question. I could be radically wrong. If you want to concentrate your risk into a single asset class, then that strategy will work brilliantly if you happen to be right about the risk-adjusted rate of return of that asset class.

If you take a look at muni's right now, they look attractive. I disagree that there is no risk associated with them. If it is a long-dated asset, then there is risk involved. Muni's have credit risk and all kinds of other risk associated with them. Anyone who has owned muni's for a long period of time during different scenarios can attest to this. My wife had a great-aunt who passed away some years ago. She bought bonds in the 1940's at 2%. She asked me in the early 1980's to take a look at her portfolio. She wanted to be conservative and take no risk. She had a fair amount of money, and she had been through the Depression - she had seen what had happened to stocks. So she put her money in bonds at 2% for 35 years, and she lost money every single year. She wiped out almost all of her wealth.

Years ago, people would do bond-swaps every year. At the end of the year, they would look at their portfolio, sell off the ones that had lost value (for the tax write-off) and then repurchase them 31 days later. The operative principle here was that you had losses in your bond portfolio every year because bonds experienced about a 40-year bear market. Then the bond market turned. Bond yields peaked and then began to fall. December rolled around and people continued to call their brokers asking to do the bond swap. They had never experienced profits in their bonds before!

We have now had a 23 year bull market in bonds. I don't know that we'll necessarily have a bear market in bonds, but that's a reason you wouldn't want to have all of your money in muni's.

Question: You mentioned that we are in a bull market. A question that was submitted indicates that there has not been a leadership change since the last bull market. Please comment.

Bill: There is a general view that each bull market comes with new leadership. That is the general view because it has generally been correct. I would caution you against putting too much emphasis on the past.

I have a fairly strong view on this topic. What we are seeing now is similar to what we have seen at the beginning of each decade and bull market. The beginning of a bull market is characterized by the same leadership as the previous bull market. As the bull market develops, however, the leadership changes. If you go back to 1981-1982, the groups that led off the bottom of the bear market were the same groups that led in the 1970's - the natural resources and capital goods-based stocks and the oil service stocks were big home runs from 1982 to 1983. The reason this happens in psychological. If oil stocks have been up for the last ten years, then when they go down a lot of people jump at the chance to get into them again. They don't want to miss them this time. When the bull market begins, people rush to buy what they wish they had owned in the last bull market. This runs up the prices on these stocks initially, and then they die. This happened in 1981 with the oil stocks. It happened in 1991 with the consumer stocks, since these were the best stocks of the 1980's. It is happening again this time with the technology stocks, although technology is so broad that this is misleading. The technology stocks that I consider to be vulnerable are the great winners of the 1990's - the old, hardware-based technology stocks. They were the great winners because they were the ones who transformed the economy in the 1990's. The beneficiaries will probably be different this time. Jeff Bezos talked about Amazon earlier and emphasize that it is still Day One for the Internet.

Several different things are required to establish the new bull market leaders for a ten-year period. First, these companies need really low valuations and expectations to start with. Then something unexpected, fundamental, and long-lasting must change in the economy. In the 1970s, we saw a ten year change in the global economy because of inflation and OPEC. In the 1980's, we saw deflation and the Baby Boomers hitting their consuming years. In the 1990's we saw the technological boom that no one expected. Dell and Cisco were brand new companies in 1990. If you look at today, what are the candidates for this kind of change? What during the bear market, reached the lowest valuations and has the lowest expectations, but has the potential to be transformative? There are two pretty clear candidates: Internet and Biotech. Lots of different sectors will go up initially, but many of them will start dropping away. If I'm right, then the ones that will drop away will be the old leaders. The ones that keep on going will be the new leaders.

Question: We have a number of related funds with overlapping mandates. Can you talk about the overlap between our funds?

Bill: We had the same question at the Fund Board Meeting. If the funds have overlap, so what? The implication, I guess, is that this overlap effects their results in some way. I'm not sure what it means.

Think about DNA. A mouse shares 96% of its DNA with a human, but I would not be terribly concerned about that overlap. There is a big difference between us and that mouse. The overlap between us and a flatworm is 93%. With a chimp, it's 99.7%! The point is that you can get very different results with extremely high overlap. This is what we see in our funds. Our funds have very different results this year, even though they have moderate overlap.

If you think about it from a market perspective, a number of funds operate in the same parts of the market. We have very different results this year. What is more interesting is that we have very similar results over the last three years, even though we have very different mandates. David did a whole lot better than I did when the market favored conservative portfolios. I am doing better this year because the market is rewarding more volatile portfolios. With respect to our clients, we try to match our clients' particular objectives to the fund with the most appropriate mandate. It is a very bad idea to try to guess which fund will do the best over the next 1-3 years. If you're just doing that, then the better approach will be to figure out what investments will be the best in absolute terms and put your money there.

Question: This question is specific to the Opportunity Trust.2 Some people say that long-only managers don't make good hedged managers. Can you describe your process for picking shorts for the Opportunity Trust portfolio, and how does the process differ from being an upside down version of your standard process?

Bill: I would describe my process for picking shorts as weak and incomplete. This is why we don't do a lot of it.

Of the people on the panel discussion, Brian is actually short on a number of things where we are long. That's why I don't ask him very much about what goes on in his portfolio. It would just confuse me! Brian, Karen, and Lee were all long-only managers. When you go short, there is a different mindset and mentality. There is a different reward structure. My friend Jim Levitas (who is now retired) used to be a partner in a short-only fund. Levitas left to form his own hedge fund, and said that a short-only fund is a really bad idea. The odds are against you because the market goes up two thirds of the time. Worst of all is that when you're wrong in a long-only portfolio, your problem becomes a smaller part of your portfolio, while in a short-only fund, your mistakes become an even larger part of your portfolio when you're wrong. This is a very different psychological experience.

The Opportunity Fund is allowed to do just about anything. We are trying to run this fund without thinking that we have a competitive advantage. The shorting has been fine so far. One short went very very well. They have served their basic mission, which was to take volatility out of the portfolio and reduce the risk. My problem was not in the execution of the short, but in the conceptualization - there was not the risk there to be taken out.

Question: When you have a very large position in the portfolio, like you do with Amazon right now, how do you reduce your position without significantly affecting the price?

Bill: Because we are value investors, we share a common trait with all other value investors. We provide liquidity to the market. Even though we are very large in some cases, our size does not have much market impact because usually we buy when most people want to sell, and we sell when most people want to buy. Normally the price momentum on the upside is far greater than the downward pressure we would exert by selling.

Question: You wrote about the physical limits of size in a recent shareholder letter. Can you talk about the limitations, both physical and fiscal, with regard to Amazon's model (including average order size, gross margin limitations, delivery costs and inventory)?

Bill: Those last five items constitute knowledge about Amazon. I try to operate from the position of maximum ignorance, as Murray Gell-Mann talks about it. I can address the question of physical limitations. As I parse the issues that I have to think about, some of the big issues that I focus on are the shortage of income, the question of size, fiscal limitations imposed on corporations by form and function.

The thing that is interesting about companies like Amazon, Yahoo!, eBay, and Interactive (all of which we own) is that they are free from the practical constraints that operate on real world companies. If WalMart is going to expand its physical footprint by 8-9% a year, it must spend $12B a year to build a store, build distribution centers, make sure it has ways to move goods from one place to another, etc. All of that capital runs into institutional constraints. For example in Germany, you can only be open a certain number of hours. In Tokyo, you can't get land. In California, they're mad at you because you hire illegal workers. Each part of the world has different issues. These are limitations that have significant impact on the appropriate potential value of the business. So far it seems to be the case that these constraints do not exist for Amazon, Yahoo and the others. From a purely theoretical standpoint, their theoretical value could be far higher than any business we have seen before. This could be why the market got so excited about them in 1999 and 2000.

The top five companies in the US by market cap include Exxon, WalMart, GE, Pfizer, and Microsoft. They operate in totally different industries with totally different economics and historical growth rates, yet their total market capitalization is clustered in a fairly narrow band. Because these are the biggest and best-known and most-studied companies in the world, and because the market is forward-looking, this tells me that there is something in common about their future prospects which clusters them together through this economic rendering process. This range is $250B-$300B for total equity market cap.

I can see a scenario in which Amazon might have a trillion dollar market cap. I don't know that it will ever do that - that depends on the future. But that potential exists. It will depend on the actual dynamics of the business and how they relate to the return on capital.

Question: Lisa, you might also address another question. What is not in the price of Amazon at $55?

Lisa Rapuano: If you look at the evolution of Amazon's business model from 1996 to now (the gross margins in the 25% range, inventory turns in the 20X range, etc), none of these components of the business need to change in order for Amazon to become extraordinarily good. They can stay exactly where they are or even shrink their gross margins a bit. Look at the way that the model returns value to the consumer. Look at how the model is maintaining an operating margin that is more steady. Look at everything on the basis of total rate of return of invested capital. We are looking at the mature model right now in the US books, music and video business. It does not have to get better in order for Amazon to continue to create value for shareholders.

Because the model is evident, the company needs to expand sales, grow markets and expand the number of customers using Amazon. We need to expand geographies and expand the available product set. We also need to expand consumers' awareness of that product set. These things will come with more adoption of broadband, with more maturity of the business, with better prices for the consumers, and with all of the other things that the company is currently focused on. I don't feel that I need to get into a lot of detail on the particular items listed in this question because I'm not counting on any improvement in those areas. In fact the company's business model can tolerate some contraction in the margin structure and still create triple-digit returns on invested capital.

At $55, Amazon is discounting about 15.3% revenue growth, in my opinion. This assumes that I have the business model roughly correct over the longer term. The question is whether or not you believe that this is the right growth rate. I believe that there is not a lot of risk to that growth rate. Over the next eight years, I would be very surprised if Amazon did not grow 15%. Of course, back in 1999, people were anticipating growth rates of 25-30%. In 2001, people talked about 10-13% growth. I believe 15% sounds pretty good.

Bill: Jeff Bezos set me up with two guys from a venture capital fund called Cambrian Ventures, and I recently had lunch with them. What they said was so interesting and so complicated that I cannot repeat it - I need to study their presentation to understand it. But I will give you the dumbed-down version. We are going to see a layered approach to technological change. When you search on the Internet, Google is the dominant player. When you punch in your search request on Google, Google uses the number of links to each page to prioritize the results. These links are all the same in Google's view, which doesn't make any sense.

Moreover, when you do a search on Google, Google doesn't actually do any searching. They are not taking your request and searching the Internet for it. All of the results are computed offline and reside in Google's database. They are not showing you results from the Internet - they are showing you results from their database. So, this venture firm funded a small company that only lasted three months because Google bought it. This company did a personalized search. This company did the functional equivalent of what happens when you enter your name into Amazon.com and Amazon remembers what you like and recommends other things the people who are similar to you seemed to like. This company developed a tool to better understand your personal search history, and it builds a profile of your preferences.

This is very cool, but there's a lot more to come. Google today is very primitive. You cannot search Google with a computer program - you have to key in your search requests. You cannot create programs to search Google yet, but someday they will. These guys began to talk about the vast number of ways in which technological innovation is possible. This is like money on the street waiting to be picked up - all they need is someone to build the technology.

This has appeared to be a very tech-centric conference or series of answers. Our portfolios are not terribly tech-centric. I think we are significantly underweight in technology in the traditional sense. We are being tech-centric here, though, because technology is the driving force behind change. We must understand changes in our portfolios. Chris Davis and I had dinner with Warren Buffett. Warren loves long, predictable business, and what could be more predictable than World Book Encyclopedias? This was a company that had been around for 80 years, and now it wasn't worth a quarter. Microsoft annihilated World Book with Encarta. Every industry can be radically transformed by technology, and we cannot afford to be blindsided. It is not that Warren Buffett doesn't understand technology. He understands it very well. He and Bill Gates are good buddies. When we went to his office, he started rattling off all kinds of data about Geico's Internet orders and all sorts of things. He is very technologically savvy as it relates to business, but he does not want to make investments directly into technology.

Question: What is your opinion on the pharmaceutical industry? How will banks act if short-term interest rates go up?

Bill: Banks will act as they should, depending on what is discounted in the stock price. Historically when interest rates go up, banks go down for awhile. Why short-term rates go up will determine how badly their stocks go down. I have made a pledge to at least consider cutting back our bank exposure when interest rates go up. We have never done so in the past, but we will consider it.

The pharmaceutical question is very difficult. The basic pharmaceutical model is being changed by molecular medicine. At Michael Mauboussin's conference in September, two speakers talked about pharmaceuticals, Juan Enriquez from Harvard and Alph Bingham from Eli Lilly. My view is that the pharmaceutical business model has very serious long term structural problems. The question that is important for us is whether that is being reflected in the price. I don't care if the industry will eventually go out of business in 20 years. If the market thinks it's going out of business in four years, then we can still do well investing in it. It is much more complicated that just thinking that their price is cheap compared to historical levels.

Question: You are a big holder of Eastman Kodak, and you expressed concern about their heavy move to digital. What happened at that recent meeting?

Bill: Nothing. We have had extensive meetings and discussions to look at these issues. My conclusion is that it's too early to tell whether or not the new operating strategy will be successful. I will say that I am more optimistic than I was when I first heard the strategy enunciated. I believe that the financial strategy announced at the same time was a mistake. They wanted to cut the dividend and use that capital to fund the new operating strategy. Most people are interested in what I think about the operating strategy - I am agnostic with a positive bias.

Question: Where is Time Warner going?

Bill: It's going up. That's why we own it. We had dinner with Dick Parsons of AOL Time Warner. Coming out of that meeting, I have much greater confidence around the issues that have most people concerned. They generate about $4B per year in free cash flow. Their net debt after disposals that have been announced and other actions taken will be $20B. They will be debt-free in five years if they don't do anything. It is very clear that they're going to do something. What that "something" is will be to extend their footprint in cable (which would be a reasonably expensive, but probably good deal), buy in a lot of stock, or some combination of the two. I also think that the AOL business has a lot of interesting potential. It is by far the cheapest Internet asset on a private market basis that you can find out there.

Question: How are you able to gauge excellent managers, particularly when the business is fairly young?

Bill: We try to look at the decision procedures. Are they thinking probabilistically? Are they thinking on the basis of expected returns? Do they really care about the shareholders? Are they committed and passionate?

If the business is young, the situation is slightly easier. It's like when I was flying with a nervous flyer once. The pilot came to talk to us and said, "You've got the best of all possible worlds - a really young plane and really old pilots!" Old managers (or experienced managers) in new businesses are good.

Question: Given your success, how do you keep your edge and guard against overconfidence and/or over-conservatism (trying to protect your record)?

Bill: If you are properly paranoid, you can guard against overconfidence. You can guard against things that you are aware of as risks better than you can guard against things you are unaware of. I have been blindsided by risks in the portfolio enough times that I know better than to get too confident. Also, if you understand that overconfidence is a big problem, psychologically, then you have to stay vigilant.

Given the size of the assets in our portfolio, there is nothing that can be done to impact our return over the short-term. It has only happened twice, as far as I can recall. In early 2000, we sold a huge amount of technology stocks because we thought that the end had finally come in the early spring. If we had held those stocks rather then selling them, we would have had a disastrous year, rather than just a down year. Those actions had an impact on that same year's performance, but that was a huge amount of activity in a very volatile year.

In 2002, the constant averaging down of Nextel and Amazon really helped us. When the market rebounded in the fourth quarter, it really brought us back. In March of 2002, we looked like Silky Sullivan. At the first turn, we couldn't even see the market it was so far ahead of us. In normal circumstances, you cannot do anything to change the results of your current year unless you are willing to make dramatic changes to the portfolio.

Early in my career, I was doing both sell-side and buy-side stuff. I went with one of our sales people to a big institutional account in Boston. This was one of the first times I'd gone to pitch an institutional client because I had been working outside of the securities business. I had analyzed RJ Reynolds and was very excited about this opportunity. I met with all of their portfolio managers and gave them the pitch. The pitch was that the company was very cheap, and that strategically the company should be broken up. If and when that happened, they would double their money in about three years with very low short-term risk. At the end of the presentation, the Chief Investment Officer said that the presentation was really good, but that they could not buy the stock. He said, "You didn't tell us why this stock would go up in the next 60-90 days." I didn't know that it would go up in the next 60-90 days. He told me, "You're new to this business, but in this business there is a lot of performance pressure. You have to perform every year. Telling me what will happen in three or four years in nice, but in need to know what's going to contribute to my performance this year or I'm not going to buy it." I asked him how he was doing this year, and he said, "Terrible! That's why I need something to work fast." I told him, "If you'd bought stuff like this opportunity three years ago, you wouldn't be feeling these performance pressures today! You'd be doing fine, and you could buy this stock which would give you your performance in two or three years." He thought I was being a wise guy.

This is one of the critical ways we think about our portfolio. We expect the stocks we buy today to contribute to our performance several years hence. While it's nice if they contribute to this year's performance, this year's performance should be driven by decisions we made in previous years. If we keep doing this, we hope that we will provide adequate returns in the future.

I appreciate you all coming out here and spending this time with us. We are working for you, and we are going to keep doing it! Thank you.