Jasmina Hasanhodzic, a Massachusetts Institute Of Technology grad student, recently interviewed me as part of an MIT project which seeks to introduce new quantitative approaches to technical analysis and highlight the ways in which technical analysis is used by some of its most experienced practitioners. The interview was videotaped, and a transcript follows. The fact that this was an MIT undertaking explains the several references to MIT in the transcript. I also mentioned Mike Epstein several times; Mike, an old friend and extremely insightful trader on Wall Street, is an MIT visiting scholar who helped put the program together.
-- Walter Deemer
An Interview with Walter Deemer
THE EARLY DAYS
Q: When did you first get interested in technical analysis? What first triggered your interest? Did someone or something in particular inspire you?
WD: I first got interested in technical analysis when I was in college. I went to Penn State. I was on the main campus of Penn State in 1961, 1962, and 1963. There was only one brokerage office in town and two out of three people there were very much into technical analysis. One of them was also a part-time teacher at the university. In 1962 the stock market crashed. It was something that was easily predictable using technical analysis, and I was impressed by the fact that they were able to predict it. I would say that they essentially bent the twig for me, and from then on the tree was inclined towards technical analysis. So I first started actually charting indicators in1962 when I was still an undergraduate at Penn State.
Q: Did you have a mentor? What was his or her role in your development as a technical analyst?
WD: My mentor was a guy who worked at the brokerage office named Dick Williams, who is now deceased. At Penn State I was in the honors business administration course, and essentially in your senior year there they let you work on the topic of your own choice. And the topic that I chose was technical indicators vs. leading economic indicators, and Dick helped me chose the technical indicators, use the parameters and research them. We compared a whole bunch of technical indicators, including Lowry's buying power and selling pressure, with a whole bunch of leading economic indicators. So he helped me chose the technical indicators and the parameters to use, and guided me in my research.
Q: Did you learn the craft by studying the literature on your own, or with a teacher?
WD: On my own. I talked with Dick Williams and other people, but essentially it was reading books. I was lucky. They did not teach technical analysis in school, and still to this day very largely do not teach technical analysis in schools. But I was lucky because I sort of made them teach me technical analysis as much as I could. So I learned as much as I could, but essentially it was mostly from the books. As the matter of fact, the first book I read was Joseph Granville's book. I'll never forget, I had just started to get interested in technical analysis, and I took it out of the library at Penn State over the weekend. There was a bunch of us, and we were carpooling home back to Philadelphia. And when we were carpooling home I was reading the book and the trip never went so quickly, because I was just fascinated by the book. And, of course, Joe Granville went into technical indicators. This was the book before On Balance Volume. He was very much into technical indicators, and after reading the book I started plotting things like the advance-decline line, an overbought-oversold oscillator, the odd-lot indexes and so forth. I had a statistical book that I kept back in 1962 on technical indicators.
Q: How much time did you spend learning technical analysis before you felt prepared to use it in your trading?
WD: Unfortunately not enough. I started using it with real money really when I started to work at Merrill Lynch, but unfortunately that's the type of thing that you never stop learning. I've been doing it from 1962, for 42 years, and I am still learning.
Q: Which mistake did you learn the most from?
WD: It's hard to say which particular mistake. Technical analysis is not infallible, and as Stan Berge once said, we are dealing with probabilities rather than certainties, and it's sometimes difficult to realize that no matter how strong you think the probabilities are, in a given situation they are still probabilities, and not certainties.
PERSONAL STYLE
Q: Could you describe your own distinct style of technical analysis?
WD: In a word it would be anticipatory analysis. When I went to work for Putnam in 1970, in one of my very early days, I remember telling one of the big fund managers that a stock had just broken out. He said: "Fine, but I can't buy it. It's already broken out and moving, the price is rallying; I can't really buy it in size. So what I need you to do is tell me before it breaks out." So I went back to my room and spent quite a bit of time working on that, but I soon found out that that, when you are dealing with major institutions, managing large sums of money, you need to tell them, the only time they can really buy in quantity is during a decline, and the only time they can really sell in quantity is in a rally. So, what I need to do is tell them just before the market makes a top, or just before a stock makes a top. Or, when the market is going down, I have to tell them just before the market makes a bottom, so they can buy that last supply and sit there with a basket and have everybody throw stocks in it. Once it starts back up it is much more difficult for people to buy in major size. So where a lot of technicians can say well, as soon as the rally is over, and it starts back down, I will get out, I don't have that luxury. Or they say what you do is you keep tightening up your stop in a long position. Well, you can't sell two million shares on a stop on a floor of the New York Stock Exchange, you just can't handle it that way. So what I need to do is look at things that give me an idea that trend is starting to reverse rather than waiting for it to actually reverse. So I am much different from a lot of my colleagues, because of the fact that I've been dealing only with institutional investors for 34 years.
Q: How much of what you learn from others do you directly apply in your trading?
WD: Well, when you say trading, I don't trade, I am only long term. But in my own work and things, I would say, in all honestly, practically everything I use is from others. I build on it. Information comes from various places, and I've been lucky enough to be with some very smart people, like, for example, Dean LeBarron, who is one of the real investment geniuses. He is the fellow that started index funds. He is so contrary that when index funds became popular, he got out of the index fund business. One day Dean and I were talking, and Fidelity had just come out with some of their sector funds. These sector funds are very small, narrowly focused mutual funds that invest only in a single sector of the market, such as energy services, precious metals, networking, electronics, or semiconductors. There are 42 of them, and they price these things hourly. They are designed for very aggressive traders, and they are run by Fidelity analysts who follow a particular industry. So Dean and I were talking about what this might mean, and Dean said I think there is some information in there somewhere. And I went back to my office and thought about it. I realized that what Dean was telling me was that rather than following industry groupings that someone like Standard and Poor's puts out, Fidelity funds give you a chance to follow actively managed portfolios: the best perceived bio-technology stocks vs. the best perceived precious metals stocks vs. the best perceived electronics stocks. You are measuring the best of each group at any particular time. So, back in 1986, we started doing relative strength work on the Fidelity sector funds, and we've done a lot with it, and it's been very helpful. But the original idea was not mine. I picked up the ball and ran with it, if you want to put it that way. The idea came from Dean LeBarron, and I was smart enough to be at the right place at the right time and listen to him.
Q: How do you learn what works for you and what does not, without making big losses?
WD: The only way, I think, to learn is to make mistakes. And what you have to do is try not to make the same mistake twice. And I would think at my age I had made all the mistakes possible, but unfortunately I haven't. It's all supposed to work, it never always does, so you learn from the mistakes.
Q: Is your analysis more effective when you are working by yourself or when you are working with others? Also, in general, is technical analysis better done working individually or in teams?
WD: In my opinion it's done individually, because technical analysis is part contrary, you are leaning against the prevailing wind all the time. So, in coming up with the ideas, you are really standing on your own, it's your own experience. Basically, what I am selling clients is experience. I've been following this for forty years. I've seen it all, I've done it all. I've made most of the mistakes. I've made a lot of mistakes, hopefully learned from them, so when the market does something, I know how to react, because it's something that's happened sometime before. So really it's experience. And they don't want somebody else's experience, they want mine. If they want somebody else's experience, they'll work with that somebody else. And that somebody else and I should not sit down and try to come to a common ground because that's probably not going to be as accurate as the individual opinion.
Q: In what kind of market conditions do you make most mistakes?
WD: All-encompassing. Again, I am not a trader, so I am not trading the markets, and so what I am trying to do is catch trend-reversals. Remember that my job is to anticipate the trend reversing, and the general mistake is anticipating it too early. It was obvious in late 1999 that the technology bubble was going through a once in a lifetime topping process, but it took longer to play out, and more importantly, it carried prices much further than logic would dictate.
Q: How much of what you do are you willing to share with others?
WD: I think probably all of it. I have no secrets. We had a presenter in our brain-storming session in Orlando this past weekend, who had a market timing service for mutual funds. He gave us all the signals, but he did not tell us what was behind them. We learned nothing from him. And if I go to a client, and I say, I have an indicator that says you should sell stocks, unless they know what the indicator is, and the rationale behind it, and the logic behind it, and how it fits in with other indicators, they are not going to act on it. So unless I share with them not only what the indicator is, but why logically it means something, they are not going to pay any attention to it.
Q: If all patterns/indicators/strategies that you use are in the public domain, what is it about the way you use these tools that accounts for your superior returns?
WD: Interpretation.
Q: How do you deal with the problem of tradeoff between early signal detection and sensitivity to random noise?
WD: If I am dealing with the longer term, I don't get as much random noise. Random noise is more generated on a very short term basis, on a day-to-day or an intraday basis. What I am doing essentially is stepping back, I let the trading desk at the institution worry about that, and I worry more about the longer term trends.
Q: Is technical analysis more effective when used on its own, or when combined with fundamental analysis?
WD: Combined with fundamental analysis.
Q: Is that what you do?
WD: No. What I do is supply technical analysis, and then the client integrates it with fundamental analysis, with the interest rate analysis, with political analysis, and with all sorts of other factors, and combines everything to make the best decision. So my position is, from a long term standpoint, that integrating technical analysis correctly with everything else is the most important thing that an institution has to do in order to realize what technical analysis can do, and, more importantly, what it cannot do. Technical analysis does have limitations. For example, no matter what the technical position in the market is right now, if the Federal Reserve comes out and says, `speculative activity is too high, we are going to dampen speculative activity, we really mean it, and our actions are going to cause pain and suffering as we take them,' the market will ultimately respond to it. And that did happen in the midst of the speculative binge of the late 1960's, when William McChesney Martin said there was going to be pain and suffering. At the Manhattan Fund where I worked, we were sort of smirking at the fact that this guy was going to close our game down. Well, he did. He did. So you had all of the charts in uptrends that ultimately became downtrends.
Q: How much of your technical analysis is done on an intuitive and subconscious level?
WD: Quite a bit, but then you try to back it up with facts. Again, it's the experience of having been there so many times, and what you are trying to do is grasp the precedents and then try to relate the current experience to the precedent. So the intuitiveness, perhaps, is trying to figure out what part of the precedent is most applicable in a current situation, and then the analysis is trying to back it up with some numbers.
FAVORITE PATTERNS AND INDICATORS
Q: What do you consider to be the most and the least reliable technical indicators?
WD: Hard question, because what we do, as was once described by John Bennett, who was then my boss at Putnam, is we use exception analysis. You follow a whole bunch of indicators, a whole bunch of factors, and if they are within normal parameters you kind of ignore them, but when they go to one extreme or the other, you start paying attention to them. Usually, when an indicator goes to an extreme, it's trying to tell you something. Having said that, I should also mention that a lot of the indicators that have been reliable in the past are no longer reliable because of changing market conditions and the changing market framework. The best single indicator statistically used to be member firm short sales as the percentage of total short sales, but because of arbitrage activities and other changes that no longer works. Whatever parameters used to work first changed and then became meaningless. Unfortunately, I have a whole bunch of indicators in that camp that used to work real well and don't work any more, and I don't have a whole lot of indicators that work real well as a result of all the changes. We are still working.
Q: Have you encountered patterns or indicators that generally work well but that, under certain market conditions, become misleading? Could you give me some examples?
WD: I don't really know.
Q: Is the number of indicators you follow greater when you are more cautious than when you are less cautious?
WD: I follow all the indicators all the time, and for the ones that I don't follow personally, I rely on other technicians to bring them to my attention. Again, I am looking for the exceptions, so when somebody brings out that a certain indicator which I may not necessarily follow has gone to an extreme, I start wondering what's going on. So the number of indicators that I follow really depends on the indicators themselves, on how many of them are signaling something at a given time, rather than on market conditions.
EVOLUTION OF TECHNICAL ANALYSIS
Q: How has the craft evolved since when you first started?
WD: Well, in many ways. One important change, from a practitioner's standpoint, is that the formation of the Market Technicians Association in 1972 fostered sharing of information. In the old days people would come to you with an indicator but they wouldn't tell you what it was -- it was proprietary. Back then you heard the terms `proprietary indicator' or `my work' all the time, which, again, didn't help. If I went to MIT with the perfect indicator, but I didn't tell you what it was and only gave you the signals, it wouldn't help you a whole lot. If I went to MIT and told you what was behind it, then you could take it apart and find out whether it was just random luck or something more. The MTA encouraged the sharing of information. And so technical analysis is a lot more open now that it was when I first came into the business, there is a lot more sharing of information. The other thing, obviously, is the information age. When I first started in business on my own in 1980 my main data source was a VCR hooked to a television set that recorded the stock market channel. I don't need to do that any more, because now, with the Internet, all the information is available for free and instantaneously. The amount of information that's available is just incredible, and that's something else that has changed.
Q: As the field evolves, new indicators and patterns are being introduced. Do you try to stay on top by familiarizing yourself with the new inventions?
WD: I do my best.
Q: Do you study the new inventions just to know what others might be doing, or do you also update your own strategies as the field evolves?
WD: I update my own strategies. I am always trying to get the best feel for the market no matter what tools I can use. Let me go back to what I was saying earlier about Fidelity. At the vary beginning, when they first introduced their sector funds, Fidelity had 30 some sector funds. They charged you a 2 percent fee to come into the funds, as well as a 1 percent redemption fee, which is no longer legal. Someone at Fidelity described their sector fund operation as a big casino. They said, like a big casino, we don't care what game you play as long as you are in the casino. It cost you 2 percent to come into the casino, the management fees that you are paying while you are at the casino are higher than in the non-sector funds, and then, when you leave the casino, there is a 1 percent exit fee. So if somebody wanted to take a break from playing the sector funds, if they wanted to move into cash, they had to move into the sector money market fund or pay a fee to leave the casino and a fee to come back into the casino. That was the only way. So by following the assets in the sector fund money market as a percent of total select assets, you could get a feel of how fully invested these aggressive players were. The sector fund cash ratio was a true contrary indicator -- the lower the cash ratio, the more bearish it was, the higher the cash ratio, the more bullish it was. But Fidelity eventually dropped the redemption fee; they were forced to. And to make a very long story very short, the sector fund cash ratio doesn't work that well any more, because there is no way to measure just how much sideline cash there is; it's spread over many other money market funds as well. In its day, it was a wonderful indicator. Now along comes an outfit called Rydex, and Rydex has a whole bunch of mutual funds, many of which are indexed to the market, and some of which are indexed inversely to the market. So Rydex has an index fund that moves with the S&P500 and an index fund that moves inversely to the S&P500. And Rydex is nice enough to report the assets in each of their funds at about midnight every night. So by comparing the change is assets with the change in the net asset value you can find out how much money came into and went out of those funds every single day. That's very different from the opinion polls -- with Rydex you are looking at what people are actually doing with their money, rather than what people think about the market. And I am convinced that these 6 or 7 billion dollars of assets in Rydex reflect the activity of hedge funds in general, who are the driving force in the market these days. So this is something that we've been doing a lot of work with in the last two years, getting these numbers every single day from Rydex, figuring out what the flows are, and trying to find out how best to measure them. This is one new indicator that's come along to replace the many that have fallen by the wayside. So we are always looking, but usually, it's kind of tough. There is so much arbitrage going on in the market now, it's very difficult to find out where a short sale is just hedging something and where it's a reflection of somebody thinking that the market is about to crash.
Q: To what extent has the introduction of the variety of computer software aided the craft?
WD: It helps you draw charts. We used to draw charts by hand, but now we draw them by computer. It saves us a lot of time calculating, and it saves us a lot of time drawing charts.
Q: To what extent do you rely on computer generated signals? What are the advantages and disadvantages of relying on computer generated signals?
WD: We don't rely on computer generated signals, we just look at indicators. There is no magic-type thing. Computers just provide a much easier way to follow and test the data. Now fortunately, people like you at MIT are doing a lot more on crunching numbers than we are, so if there are truths out there, you will probably find them sooner than we will. Hopefully, you'll share them with us. Again, I don't trade, so that's probably not applicable to me. We don't have a buy signal today and a sell signal tomorrow, or anything like that. It's all subjective analysis, rather than objective analysis. If I say I think that the Nasdaq is vulnerable to a decline, I don't have some red light that's going to go off at the precise moment when it starts declining, or something like that.
Q: Some technicians believe that it is still important to construct your own charts by hand. Do you agree?
WD: I think you get more information because you are forced to look at it. The problem is that it takes so much time to follow. If you do a chart of each of the S&P 500 and each of the Nasdaq 100 stocks by hand, that's 600 daily charts. If you can do one in every 10 seconds, that's 6 a minute, so if you're lucky, it's going to take you a hundred minutes to do that. So you've taken almost 2 hours to do this, before you've even started doing any other work. On the other hand, when the computer draws the charts, you can go through and review them much more quickly. You can also set up parameters of things you are looking for and have the computer flag them, for example.
Q: So, do you still do it sometimes?
WD: I do it a little bit, but I am lazy, I let the computer do it for me.
THE INNOVATIVE PROCESS
Q: What drives your innovative process?
WD: What drives my innovative process? Fear of being wrong and losing all my clients, greed in trying to be more accurate in the market, and a desire to hone my skills. Somebody once asked me why I didn't manage any money, and I said well, that's a whole different set of skills than market forecasting, and when I perfect my skills in market forecasting, I'll try another field, but I am still trying to perfect my skills in market analysis.
Q: Do you and to what extent collaborate with others during the innovative process?
WD: It's mostly on my own. I ask for other people's help when I am looking for data, for example. In some cases when I have tentative conclusions I'll share them with some people. But basically, if it's something I am developing, I know more about what's going on with the particular indicator or whatever it is that I am developing, so it's usually on my own.
Q: Were there moments when you felt that relying on classical patterns and indicators was simply insufficient, and that developing new technical tools was necessary?
WD: All the time, because as somebody once said, whenever you think you've got a key to Wall Street, somebody comes along and changes the lock. So whatever indicators work, however well, however long, something will come along to change them, and you always have to be alert for new things to do.
Q: How soon after you develop a particular technical tool do you make it accessible to public?
WD: Well, when I think it works, I will make it accessible to my clients right away. Again, my job is to give them tools. Most of them are interested in what I think the indicators are saying rather than seeing the indicators themselves. So what we usually do is publish the reasoning behind our market opinion, and we have all the indicators to back it up, but we usually don't publish them, we furnish them on request.
Q: Are there tools that you developed but never shared with the rest of the world?
WD: No.
Q: How often do you use the technical tools you developed?
WD: Constantly, perpetually.
EMOTIONAL ASPECTS OF THE CRAFT
Q: How did you feel when you first lost a lot of money? Has it become easier to lose as you became more experienced?
WD: How did I feel when I first lost a lot of money? MIT really wants to know how people feel when they lose money? I am not a masochist, I did not feel good, I felt that my gods had failed me. And how do I feel when I still lose money? I still feel bad, especially as I get older, because I don't have as many chances to make it back as I did before. Sometimes you get a little more conservative. You show me somebody who is happy about losing money and we'll send them over to Harvard.
Q: Has a big loss ever made you doubt the validity of technical analysis?
WD: Sometimes, but mostly it's the interpretation. The problem with technical analysis is that you can always build a very bullish case and you can always build a very bearish case -- and here I am talking long term, though I suppose it could be the case for trading too. There are some people right now who have a very bullish case and some very good reasons behind it, and there are people who have a bearish case and some very good reasons behind it. There are always some indicators that are frightfully bullish and there are always some indicators that are very bearish. So the problem is, which ones do you listen to? And the big problem is, when you are wrong there are always those indicators that did work, there are always some of them, and that's the humbling part of the business. Because, you see, technicians don't seasonally adjust things; we don't revise data or something, the price is the ultimate arbiter. If you are a weather forecaster and it gets colder, you can sort of fudge away that some factors changed, but in the stock market, if it goes down more than you thought, it went down more than you thought, period. And there are always reasons, beforehand and afterwards, to explain why it did. But you were listening to something else.
Q: To what extent do your emotions interfere with your craft?
WD: Emotions always interfere, that's part of the problem. The reason I don't trade is because I get too emotional. If I start trading something like stock index futures, I end up looking at every trade on them. I get upset with every tick that goes against me, and I get excited with every tick that goes for me. Some people are just not emotionally suited to do it. That's why I've chosen to be a long term market analyst -- I get too emotional over the short term fluctuations, so I ignore them.
Q: Has this changed since when you first started?
WD: No, it's probably the same.
Q: In general, is the ability to separate emotions from technical analysis an inherent trait, or do you feel that it can be learned?
WD: I think it can be learned, but it's very difficult to do it. Mike Epstein has been able to do it over the years, and I admire him for being able to do it. I could not do what he does.
Q: According to Joseph de la Vega ``every speculator seems to have two bodies so that astonished observers see a human being fighting himself.'' To what extent is this statement true in your case?
WD: I think it maybe goes back to the fact that there is always a bullish case and always a bearish case that can be made. Especially when the market is going against the case you are making, sometimes it can awfully tempting to listen to the opposing view. So there is always the conflict between the forces of good and evil. And the evil ones are always tugging at you, because they are always out there. I mean, whenever you play this videotape back you can call me up and I'll give you someone who is very bullish and I'll give you someone who is very bearish, as well as the case for each side. And I don't know whether you are going to be bullish or bearish, but somebody is going to be exactly agreeing with your view, and somebody is going to say that your view is totally wrong. And they'll both be compelling. So there is that tug of war. Always.
THE ROLE OF CREATIVITY
Q: What role does creativity play in technical analysis? Can this creativity be learned?
WD: It plays a great role, and I don't think it can really be learned. I mean, I don't think Rembrandt learned to be creative. I think maybe he learned to help his creativity, but he was really one of those geniuses who are born with it.
Q: Is there such a thing as ``talent for technical analysis''? Could you define it?
WD: I think so. Some people just seem to have a knack of choosing the right case rather than a wrong case, of knowing what to look at. I think the psychological thing has also something to do with it, the fact that you have to be unemotional especially if you are a trader. Also, a technician is usually a contrarian, because the market usually goes against the crowd. People are most bullish generally at the top, and most bearish at the bottom. So you have to be bullish when people are bearish, and you have to be bearish when people are bullish. I am not just talking about the masses that are walking down the street; I am talking about the money managers too. Because they are people too, and so you have to be a very strong willed person to stand up to these people and say, `precisely because everybody in America says that things look bad, you should be buying stocks.'
Q: Can the absence of favorable personal traits be overcome by hard work and dedication?
WD: I think so, but I think it may be more a psychological training that an educational training.
Q: In your opinion, will the artificial intelligence ever be sophisticated enough to replace a human technical analyst? If not, why not? If yes, please explain.
WD: No. Maybe on a very short term basis, but the problem with the stock market is that it is the result of an infinite number of variables with an infinitely varying number of weightings on them. Someday the market will be excited about something that's going on concerning Washington's foreign policy, another day it might not be interested at all. There are all sorts of variables, and the artificial intelligence model first has to be big enough to have every single variable including let's say, astrology. If any single person ever buys or sells a single share of stock based on something astrological, then astrology is, by definition, a market factor. So you have to put that in your artificial intelligence model. Now you've got this long, infinite list of variables in your model, and you have to figure out what weights they have -- some days astrology may have tremendous weight, other days fear may have tremendous weight. It's just an infinitely varying degree of weights. I think that the human brain can probably cope with this a little better than artificial models. Having said that, I am sure some day somebody at MIT will play this videotape back and laugh at it, because you will have an AI model. When I first came into the business and Jerry Tsai had started the Manhattan Fund, the rumor about Wall Street was `Jerry Tsai is buying, Jerry Tsai is selling, and that's what's making the market go up, and that's what's making the market go down.' And the one thing I realized when Mike was explaining some of the research that you all are doing is that some day it's going to be `MIT is buying, MIT is selling -- they've got the computers that know it all.'
Q: Consider the statement ``technical analysis is what you want it to be.'' If, indeed, technical analysis can be viewed as partly art, then, by definition, it should be capable of many interpretations, in which case this statement would become valid. What do you think?
WD: It's an art, because, if I understand your question, you are going back to my point about the bullish argument and the bearish argument. The science is developing the argument, the art is deciding which argument to embrace, and that's the tough thing.
LUCK, ASTROLOGY, etc.
Q: What is the role of luck in technical analysis?
WD: It shouldn't play much of a role, should it? But it always seems to, doesn't it? And maybe even your artificial intelligence model will have a little luck in it. In other words, sometimes it may get the market right because of luck, rather than thanks to all the variables and weights you put in.
Q: Do you think that inclusion of astrology in technical analysis undermines the credibility of the craft?
WD: Yes, it does, because technical analysis basically is looking at the market internals. What I am doing as a technical analyst is analyzing the price of a stock. My fundamental colleagues are analyzing the value of a stock, which really depends on the value of a company. And the problem is that, as an investor, you are not buying the company, you are buying a little piece of paper that represents part ownership of the company. The value of that little piece of paper goes up and down for all sorts of reasons, not all of which have to do with what the company is doing. Sometimes the value of the little piece of paper will go down because a fund manager has a hangover and decides to do some selling of a bunch of stocks and this is one of them -- and that has nothing to do with the fundamentals. What technical analysts are doing is analyzing the price of that little piece of paper, and, in aggregate, we are analyzing the stock market rather than the economy. There are many things that play a role in that analysis, for example, money supply, because money makes the world go around and it makes the stock prices go up and down. Foreign policy would also play a role because it would influence some people's buy and sell decisions. One thing that as a technical analyst you're doing is trying to measure the impact that a certain stimulus has on the market, and trying to quantify the unquantifiable. In a bear market, no amount of good news will make the market go up, and in a bull market, hardly any amount of bearish news will make the market go down. So if you have something bearish happen and the market ignores it, you know it's a bull market, and so forth. Astrology tries to put a cause and effect on something that cannot really be related as the causal and the effective. Linking astrology to the stock market, saying that by observing something in the heavens you can automatically infer what is going on in the stock market, is, I think, a mistake.
Q: There are technicians who believe that structures such as the Elliott Wave, Gann's natural order postulates, Fibonacci numbers, etc. underlie the market action. Some prominent technicians have an enormous amount of faith in these theories, even though none of them have been scientifically proven. What is your opinion?
WD: I don't use any of the foregoing. I've just never used them, they've never worked for me, and I've never learned enough to try to make them work for me. Call me an old fuddy daddy if you wish but I can't get them to work. I think Elliott Wave is wonderful in hindsight, but I am living today and now and not in hindsight. When you get a 10 page paper, the first 5 pages of which gives you an Elliott wave count, and the last 5 pages of which give you an alternate wave count, that doesn't really help much.
Q: Do you believe that these structures underlie the market action, that they are the governing principles?
WD: The governing principles? I don't think so. I think there are long term generational cycles -- you go through a boom-bust cycle over a period of a generation. And then you have shorter term cycles. For example, a presidential cycle is well documented as having an effect. I find it fascinating that there was a four year cycle in Great Britain, which does not have presidential elections every four years, and the cycle existed there before the British and the US economies were linked nearly as closely as they are today. So I think there are some long term psychological cycles, but I am not sure that they are quantifiable enough to be useful except in a very broad sense. For example, according to the Kondratieff cycle there is supposed to be a horrible decline every 50 or 60 years or so. Some people say that it is somewhere ahead of us. Some people say that the Kondratieff cycle did indeed occur and that it was in Japan rather than the United States this time, that is, that the Japanese stock market and the economy went through a true Kondratieff bust cycle. So there is some long term evidence, but whether that's an underlying structure that explains everything in the market, is another question. I guess nobody has proven it to my satisfaction yet. There is too much irrationality to make it that rational.
LEVEL OF CONVICTION
Q: Have you always been convinced about the validity of technical analysis?
WD: Yes.
Q: Did you become more or less convinced since when you first started?
WD: I've become more convinced, because it's the basic underlying truth of the stock market. As I said earlier, the only thing that analyzes the stock or the stock market is technical analysis, everything else is really peripheral. When Cisco was at 80, there were some people who thought it was drastically overpriced, and some people who thought it was the greatest growth stock of all time and that it was still cheap. You can come up with the value, but the underlying truth is the law of supply and demand, which is technical analysis. If I were working for Warren Buffet and buying companies, I would have a different perspective on life, but I am working for people who buy these little pieces of paper and so I use technical analysis. If you want to follow and forecast the stock market, you have to use technical analysis. Everything else is just peripheral.
Q: Were you always equally convinced?
WD: More respectful of it over the years.
Q: Which special moments of your career have been critical in determining your level of confidence and conviction in technical analysis?
WD: I can't really think of one.
Q: Did the lack of credit many academics give to technical analysis ever discourage you?
WD: No, because technical analysts are contrarians. We are used to being told that we are wrong about the market. Most people don't agree with us about the market. Back when I worked at Putnam, I used to give weekly presentations and I used to tell the story that if, after I made a very forceful and very insightful presentation to a room filled with money managers saying that the market was going to go up, the fund managers all stood up, applauded, hoisted me upon their shoulders and carried me down the hallway yelling Hosanna, Hosanna, then dumped me as they rushed into the trading room with their buy tickets, I would go back into my office and realize that I had done something wrong. But it was when they did not generally agree with me that I knew I was right. That's the theory of contrary opinion. So, the fact that academics don't agree with technical analysis makes me think that they haven't looked at technical analysis in the right way, even given the understanding that there are some terrific limitations as to what technical analysis can and cannot do. And yet, if the academics are looking to forecast financial markets, then they need to analyze financial markets and the things that impact them, and that's where technical analysis comes in. With all due respect to the academics, the fact that they don't accept it simply means that they don't understand it then.
Q: What, in your opinion, is the best proof of the validity of technical analysis?
WD: That I am not in debtor's row, that I am not standing on the street corner begging for food. I think just the fact that it works. I know some people want track records. The problem is, when your analysis is long term in nature, it gets difficult sometimes to do track records. Let's say the market is coming into a bottom and I think that my clients should be turning more bullish. I can't turn an out and out bear into an out and out bull overnight. Nobody can. What I can do, if I am lucky and if I do my job properly, is turn that person into a little bit less of a bear and then, the next week or the next month, into a little bit less of a bear still; in other words, I am working on the margin. Unfortunately, if I really do my job well, the money manager will think that it's their idea rather than mine. So trying to get a track record is difficult. Remember that the money manager or the institutional investor is getting inputs from a whole variety of sources, fundamental, economic, monetary, and I am just one input. So what I have to do is work against other inputs as well in trying to convince him that my interpretation is correct. And if I do it well, what I am doing is sending that money manager back to his other sources to question their bearishness and ask them, for example, `Do you think that this may be already priced into the market, the market having gone down the economy is now bad?' Well, that would mean that the stock market didn't go down in vain. The stock market went down, the economy is bad, that's the way it's supposed to work. The stock market is the leading economic indicator, so now the stock market should turn back up before the economy. And so you are not going to get the good economic news at the bottom. You won't get it until six months into the uptrend, you'll have to anticipate it, and so you have to let them know that they cannot be bearish because the economy looks bad. The problem is that translating this into a track record gets very difficult.
Q: Did you find that your experience with technical analysis contradicted statements made in books? Did that ever discourage you?
WD: Well, I wouldn't say contradicted. I would say that some indicators that are presented as foolproof are not foolproof. No indicator works all the time, no indicator gives constant information as John Bollinger once said. And the other thing is, there are some things that are presented in books which either I don't understand or at least I can't make work, such as the Elliott Wave. There are a lot of books on the Elliott Wave, all I know is I can't make it work, I can't make it work to forecast. I suppose I could make it work if I wanted to explain something that's happened, but that's not my job. My job is to predict, not explain what's happened.
Q: Does the fact that in technical analysis there are no hard and fast rules and no proven theories ever bother you?
WD: No, it's wonderful. It's part of the wonderful, fascinating field of technical analysis. There are no rules. Nothing works all the time. But there are no rules really in a lot of things where think there are anyway. It goes back to what I was saying earlier -- when you think you've found a key to the market, somebody changes the lock. There are no rules. These are all probabilities, not certainties.
Q: Do you believe that with the help of technical analysis you can make up any loss, regardless of how large?
WD: No. If my wife passes away, technical analysis will not help me make up that loss. Financial loss, I suppose in time, but it depends what made the loss. To answer your question, yes, but it's sort of begging it. You shouldn't have that large a loss if you are an experienced technician in the first place. It's a little bit loaded question, so I know you did not do that question, one of your colleagues did, probably Epstein, because Epstein has these large losses that he tries to make up, so that that's what he does, and he always blames these other traders.
Q: Do you believe that technical analysis works even when applied to data other than the market action data (e.g. the weather data or the river flow data)? Please explain.
WD: The answer is yes. Now I am going to get in over my head very quickly on this one, so everybody who watches this videotape, do not snicker for the next minute. Technical analysis essentially is measuring trends, it's measuring momentum, whether the trend is gaining or loosing strength. Frank Peluso once told me that a skilled physicist would make an excellent technician because market action really can be explained by following the laws of the swing of a pendulum. What you are measuring in the stock market is the whole sum of human intelligence focused at the corner of Broad and Wall Street. All the supply and all the demand come down to the price of the stock market as measured by the S&P500 average or whatever you want. This price measures all the human irrationality, including the fact that somebody may be buying the stock because they've spent 5 weeks studying all the stocks and this is the one they've come up with, and the fact that somebody may be selling the stock because little Johnny had an appendix out and needs the money. We know what the S&P is right now, so we have a starting point. So what we are trying to do is measure the trend, the direction in which it is going, how forcefully it's going in that direction, whether that direction is changing and how rapidly it's changing, whether that change is increasing or decreasing, that is, the rate of change or the derivative, or whatever it is. The same can be true of weather systems. I happen to track hurricanes the same way that I track the stock market, because, number one, a hurricane is a big physical force, and, number two, a hurricane is going in a particular direction at any particular time, it can either be accelerating or decelerating, and it can either be changing direction or staying in the same direction. If it's changing direction, it can either be changing direction at an increasing or a decreasing rate of change. So you can try to forecast this by charting a hurricane on a map, just as you would chart a stock or a stock market on a piece of paper. And remember, the only reason you are doing this is because, as the Chinese proverb would say, a picture is worth a thousand words. Well, a chart is worth a thousand numbers. It's easier for the human mind to grasp a chart than it is to grasp masses of data. By plotting a hurricane on a map and charting it, you can tell whether it's changing direction or not. Similarly, you can plot various river states and other things related to the flow of a river. As a matter of fact, as we were talking last night, some of the things that you were doing in electrical engineering I think have applicability in the financial markets. Because I think the same things that govern the things you were measuring in the field of electrical engineering also have an application in the field of financial management.
Q: Is it true that these patterns and indicators capture something that's specific to and fundamental of the financial markets so that when they would be applied to data other than the market action data (for example the weather data, the river flow data, etc.) they simply wouldn't work?
WD: I don't think it's true. Let's say MIT has got their artificial intelligence model for the stock market, and now they want to do one on weather systems. That's the same thing. There are models on the internet of weather systems, they've got weather maps 10 days out projected with precipitation, wind speed, and everything like that, and the important thing is, there are changes in trends. Now, the five best models - and they run these weather models every 12 hours - never agree with each other. So to be a good weather forecaster, you should not be forecasting only based on the 10-day or the 240-hour model. You should compare the 240-hour model with the 228-hour one, which will have a slightly different set of parameters, and take the changes. Then you should take the changes every 12 hours from now on as you are coming into the actual event, and you'll come up with a pretty good forecast. You'll come up with a pretty good forecast not by looking at the forecast map which is set in stone, but my monitoring the change in the forecast map from the forecast map 12 hours ago and to the one 12 hours in the future. You'll find the trends. And a trend is a trend, be it in the atmosphere or be it in the financial markets. As long as you are dealing with broadly based forces, and both the atmosphere and the financial markets are broadly based forces, things are not going to change from black to white overnight. So I think there is great deal of application for technical analysis. I think this is one of the things mathematicians have been doing - they have been applying mathematics from other fields to the financial markets with some success, and teaching those of us in the financial markets a thing or two in the process.
Q: Could you apply patterns such as head-and-shoulders, inverse head-and-shoulders, rectangles, triangles, etc. to data other than the market action data to detect trend reversal or continuation?
WD: Sure. I am a radio ham, and the radio propagation depends greatly on the sunspot cycle and the amount of sunspot activity. There are charts of sunspot activity over its 11 year cycle, and in order to find a peak and the projected high in the sunspot activity, they are using, in essence, technical analysis. And I think you can do it with all sorts of things. My friend Dean LeBarron is very heavily involved with the Santa Fe Institute, which is very much mathematically oriented, and they are trying to integrate mathematics from other fields into the stock market. It's something that I don't understand, but I am in awe of what is being done.
LIFESTYLE
Q: Could you describe your working day?
WD: Yes! The first thing I do for my clients in the morning is send them a daily update. It's sort of like a daily weather forecast for them. You don't necessarily run your life on a daily weather forecast, but it's nice to know whether it's going to be warm or cool, and whether it's going to be wet or dry. You are still going to go to work the same way you always go to work but you may decide to take an umbrella or a coat, depending on what the weather forecast is. And, my job is to give them a very quick, if you will, weather forecast. Also, if there may a question on their mind, my job is to answer it before they ask it, so that they don't have to call me at 9 o'clock. In the old days the stock market opened at 10, so I used to issue my daily update about 10 minutes of 10. The stock market now opens earlier, and we are now almost 24/7 in the financial markets, so the daily update goes out now at 20 minutes to 7 in the morning. So I get up at 20 minutes to 6 and go into the office. I get those Rydex numbers that we've talked about a while back. They come out at midnight so I get them first thing in the morning, then I look at the overnight trading to see what's happened, and issue my daily update. So, my day starts about 20 minutes to 6. The really nice part of my life is that I so love what I do that I don't set an alarm clock. I am always up at 20 minutes to 6. I always wake up at 20 minutes to 6 without an alarm clock, and the stock market is such a fun thing to analyze because every morning when I come downstairs and come into the office, I think I know what I am going to see, but it never turns out exactly that way, there is always some wrinkle or something that's happened. So at 20 minutes to 7 I have set that up. Then I collect statistics, I do the regular things of running the business, and then I start collecting the closing statistics, which come out starting with the closing of the bond market at 3 o'clock. We keep a number of broad sector averages which we started at Putnam in the early 70's. We actually had some sector averages at the Manhattan Fund in the 60's, so we've been following the market sectors for a long while. We followed currencies and the bond market at Putnam in the 1970's. So we were doing intermarket analysis before some of the people who popularized it were even born, I think. After collecting the closing statistics, we issue a rough draft of the daily update, sometime around 6 or 6:30 at night because there are some people in Tokyo who want a rough idea of what's going on as their market opens. And then usually I go to bed early. I've had three detached retinas, and my eyes get very tired very quickly, so after looking at television and computer screens all day, my eyes get tired, so I go to bed early and rest my eyes for a while. And then, once every two weeks, we publish a sort of an in-depth report, which takes the better part of three days to prepare. Wednesdays, Thursdays, and Fridays of every other week we are doing that. During the day it's just answering emails and answering any client questions which I didn't anticipate in my daily update. So, I am in the office at 20 to 6 in the morning and at 8 o'clock in the night, but every once in a while I go out and get a breadth of fresh air.
Q: How many hours each day do you spend practicing your craft?
WD: Whenever I am awake, because I am always thinking about it. That's all I do. I don't manage money, I don't do anything else. I am just trying to figure out what the stock market is going to do. And so it's really every waking hour.
Q: Are you sufficiently convinced in the ability of technical analysis to forecast future price moves so that you can live without stress?
WD: There is always stress because, number one, you are dealing with probabilities, not certainties, which means you are never always going to be right; number two, I anticipate moves, rather than react to them, so the market is always going to go against me, hopefully not much, but some of the time it goes against me more than I think it's going to do before it reverses, if indeed it reverses at all. So there is always a certain amount of stress. The other thing is, we don't have a big corporation. I can't blame the committee down the hall for coming up with a bum forecast. It's always me, so there is always stress. And now why I can handle the stress, which I can, and not the emotion of trading the market, I don't know. I have to go talk to a shrink to find out.
Q: In his book, ``Confusion de Confusiones,'' Joseph de la Vega wrote:
"When the speculators talk, they talk shares; when they run an errand, the shares make them do so; when they stand still, the shares act like a rein; when they look at something, it is shares that they see; when they think hard, the shares provide content of their thoughts; if they eat, the shares are their food; if they meditate or study, they think of the shares; in their fever fantasies, they are occupied with shares; and even on the death bed, their last worries are the shares." (De la Vega, Joseph. Confusion de Confusiones. Harvard University Printing Office; Massachusetts: 1959. p. 22).
Would you agree with de la Vega? To what extent does your trading control your life?
WD: I love what I do. You'd have to ask my wife to get the best answer perhaps to that. She once asked me a couple of years ago, `Are you thinking of retiring, shouldn't we be thinking about retiring like normal people and everything?' And I said, why would I want to retire, I am having fun with what I am doing, I enjoy what I am doing. When I go on vacation and I flip on the stock market it's not because I have to, it's because I want to. When I follow these indicators it's not because I have to, it's because I want to. I am curious as to what's going to happen, what's the next act in this never ending play. It's the most fascinating story, why would I want to do anything else? Everything influences the stock market to some degree, so whenever I am looking at anything, doing anything, it really has something to do with the stock market. So why do I need something else? It's fun. And I enjoy it. It's not work. If I thought it was work, I'd do something different, or I'd retire or something, but this is fun.
ADVICE
Q: What kind of formal education is most compatible with the profession of technical analysis?
WD: Experience. Experience is the best teacher. Whatever experience you can get in the academic world, get it. But, as I said, I was very lucky at Penn State. We had in the early 1960's an IBM 7094, which, I was told, was one of the five biggest computers in the country and was doing Department of Defense work and things like that. And because I was in the honors program, I was allowed to program that thing. I actually punched out programs and data on punch cards and everything like that. And I was allowed to do that, so I was getting a little bit into computers in the 1960's and I was getting a little bit into technical analysis, because I wanted to, I created my own curriculum if you will. But you can't teach it, you learn it. Somebody shows you something, but you've got to find out yourself that that's not going to work all the time. Experience is the best teacher, the only way you can learn. I've seen the wheel go down and into the mud, and come back out again I don't know how many times now over the years, but you've got to live through it a couple of times to realize that things at the bottom really, really, really look ugly. You can't get an idea out of a book how ugly it can be at the bottom. And you can't get an idea out of a book how euphoric it can be at the top. And you can't get an idea out of a book how to realize that when things really look ugly and everything is screaming at you logically that this is not the time to buy, you should go in and buy. And when everything is giddy and euphoric, that's the time to sell -- so you sell and go to a party, and people are telling you how much money they've made and everything, and you can't really tell them you just did some selling because tou thought it was probably a good idea. So, experience, experience.
Q: What advice would you give to technical analysis students? What is the key to success?
WD: Experience, open mindedness, realizing that the market is bigger than you are, realizing that no matter how smart you think you are the market is going to make you look like an idiot every once in a while, and realizing that it's all the matter of probabilities. I mean, in baseball, if you've got 4 out of 10 right, you could be a national champion, but in the stock market 4 out of 10 right is not very good.