Editor’s Note: What follows is part I of Paul Desmond’s interview with TheStreet.com contributor Barry Ritholtz. Tune in tomorrow for part II.
Paul Desmond, president of Lowry’s Reports, is known as a “technician’s technician.” In 2002 he won the Charles H. Dow Award for excellence in the field of technical analysis for his studies on how market bottoms are formed.
More recently, he’s been looking in the other direction, studying how market tops are formed. It has been a long time coming: Many years ago, Desmond’s firm bought microfiche of The Wall Street Journal for 1920-1933. They laboriously converted the printed stock tables into digital form — that’s all market activity for every operating company stock listed in the stock tables, including the opening and closing prices and high and low volumes. From this unique data source, Desmond analyzed the 14 major market tops from 1929 to 2000, trying to identify similarities. His findings are startling and impressive.
Let’s talk about bottoms a little bit because I recall reading a paper that you did that won the 2002 Charles Dow Market Technician Association award. The study on 90% downside days.
Yes. I had been reading a great deal of material about what market bottoms looked like. And one of the people that I happened to be reading was a fellow named of S. Gould, who talked about a classic market bottom in which he assumed it all occurred in a single day. That the volume was very heavy in the morning on the down side, that is stabilized in midday, and then by the afternoon, it was again rallying strongly again on substantially expanding volume. I simply went back through our history — which extends back to 1933 — and I was looking for those classic bottoms as Gould had defined them. I found very few, maybe one or two cases that fit his definition. But it became apparent to me that what he was talking about was an idealized situation and not actual experience.
So I went not only through his work, but through a number of other people’s work where they were talking about what a market bottom looks like. I could not find any of them that really worked or fit preconceived notions. So we started looking for some pattern that would help us to identify a market bottom. We knew that the most important consideration of a market bottom was panic; the final step in a downtrend is that investors panic and throw in the towel. They want to abandon the stock market without any consideration of the value of their portfolios.
The classic expression is, just get me out, I don’t care about the price, I gotta make the pain stop.
That’s exactly right.
Looking at 1987, many people generally think that that was a one-day wonder to the downside — that it was a one-day debacle. But I’ve looked at the month before and saw a big build-up in volume and a pretty hefty decrease in price before that single-day crash. How did you find 1987 to be, compared to other bottoms?
Well, the panic stages of it occurred in three particularly important days. The first one was on the 13th of October, a Wednesday. And then the 14th was a Thursday, and that was a 100-point downside day. Now at that point, a 100-point downside day then was something very spectacular. The 14th was a 100-point downside day on the Dow and it showed intensity, and that is where we had our major sell signal on the 14th for October. Then Friday the 15th, the market also dropped 100 points. So, to have two back-to-back 100-point downside days was pretty spectacular. Then on Monday morning, the 19th, the real crash, the 500-point drop, occurred. Now that was 90% downside day, which showed that there was real panic. (Editor’s note: Lowry’s defines a 90% downside day as a session with 90% downside volume in conjunction with 90% downside price action, meaning 90% (or more) of the price movement of all stocks on a given exchange is lower.)
I’m looking at a chart of October ’87, and the Dow was about 2700 in the beginning of the month. Before we even got to that Wednesday (the 13th), the Dow was down to 2500. The volume really started ticking up on that 13th, 14th, 15th. The 19th and 20th were both the biggest-volume days of the selloff.
That’s right. Actually, the interesting thing about 1987 is that most people incorrectly say ‘it just came out of nowhere.’ That the market was going up one day and suddenly crashed. And yet, we had a whole series of classic warning signs that the market was weakening. For example, the advance/decline line, which is a simple measurement of the number of stocks going up vs. the number of stocks going down, topped out in early April of 1987, showing that that was the point in which the largest bulk of stocks was starting to peak in price.
Our buying power index, which again is the measurement of the amount of buying enthusiasm present in the market, topped out in late March of 1987. From that point on, the market was still going higher but was doing so with less gas in the tank, so to speak. And also we run an index called the selling pressure index that measures the amount of selling activity present in the market in any given time, and that was in a strong uptrend pattern, particularly starting in early August. And during that last rally attempt — the failed rally attempt in early September — the buying power index was dropping at a very rapid pace and the selling pressure index was rising at a very rapid pace, showing that buying enthusiasm had really been lost and that the sellers were trying to dump stocks as quickly as they could. That all occurred almost two months ahead of the actual break in prices.
Not Your Father’s NYSE
A question that has come up about your work is how are your buying power and selling pressure calculated. Is it proprietary, or is it something that anybody could find in the daily market data?
Well, it is proprietary in the sense that Coca-Cola is proprietary (laughs). In other words, you can look at a bottle of Coke and it will tell you exactly what the ingredients are, but they won’t tell you how they cook it.
Without giving away the secret formula, what goes into your buying power and selling pressure indexes?
Well, the buying power index is a measurement of demand, based on the law of supply and demand. So the ingredients that go into it are upside volume and what we refer to as points gained. Points gained are simply a very simple calculation of the amount of price change in every stock that advances for the day.
And you only count operating companies, not closed-end funds? Or bond funds, REITs, things like that?
We do now. Back in the 1980s and so on, there were not the distortions in the listings on the New York Stock Exchange that there are today.
In fact, you recently said we should not be calling it the NY Stock Exchange. More than half of the companies actually aren’t operating companies — now NY ‘Fund’ Exchange is more like it.
Yeah, that is the thing that most investors are not aware of. That more than 52% of all the issues traded on the NYSE are not domestic common stocks. They are made up of closed-end bond funds. So those are issues that are actually bonds trading on the NYSE. There are real estate partnerships, REITs, a lot of ADRs and foreign stocks that don’t necessarily represent our economy. For example, back in the days when the Japanese market was in an incredibly strong uptrend pattern, the ADRs such as Sony listed on the NYSE were creating quite a distortion, showing strength in our market that was really a reflection of strength of the Japanese market, rather than the U.S. market.
As we saw this tendency of the NYSE to list more and more issues that were really not domestic common stocks, we felt the need to create a series of computations that excluded all of the things that could be considered to be distortions. So what we run our analyses on is what we call our operating companies-only statistics. For that, we create upside volume, downside volume, points gained or lost, advances and declines, new highs and lows, and a whole series of other indicators as well.
It sounds somewhat similar to the Trin or Arms index, in a way, what you are actually looking at.
Well, the Trin Index or the Arms Index is based just on the advances and declines, and that index to my knowledge, the last time I talked to Dick Arms, was based on the traditional advance/decline numbers as are found in The Wall Street Journal that do include all of these potential distortions that I was talking about. So to my knowledge, Arms has not ever gone back and redone their indicators because of these things. In fact, the last time I talked to Dick about it, he said that he had done some studies and felt comfortable that the distortions were not significant enough to worry about.
But what we have seen in some instances — one of the classic instances occurred in August of 2001, a few months before 9/11 — in which stocks were in a relatively dull period, in which most stocks were just moving sideways, but all of a sudden, the advance/decline line began to rise sharply. And a number of analysts pointed to that sharp increase in the advance/decline line, and I think it was reflected as well in the Arms Index. They viewed that improvement in the advance/decline line as a sign of strength, a building up in the market … that would clearly lead to a strong advance. But our operating companies-only advance/decline line at the same time was in a strong downtrend pattern, precisely the opposite direction of the conventional advance/decline line. So it was pretty clear that the difference between the two was primarily bond funds and foreign issues. But when we looked at foreign issues, foreign issues weren’t doing anything particularly strong. So the improvement in the A/D line [at that time] was coming primarily from bonds — and not from stocks.
Greed and Fear
You mentioned Japan: Do you look overseas? I have been bullish on the Japanese market for a couple of years, but I am starting to get a little concerned when I see these up 500-, down 500-type days. Is that a churning top? Is it something to watch? What are your thoughts on what has been going on with the Nikkei?
We don’t watch them with the same degree of intensity that we concentrate on the U.S. markets. I have tried on several occasions in the past, dating back to the mid 1970s, when I first got the idea that it would be fun, if not profitable, to be a world-wide investor. In other words, when our markets were topping out, rather than going into defensive position into Treasury bills, wouldn’t it be better to find some other market somewhere else in the world? At that time, in the mid-70s, there were very few investors who were really interested in foreign markets.
What we found was that the currency conversions created an incredibly complicated system for investors to keep track of. In other words, there were many cases in which you might have bought into a foreign market and made money in the market but when you tried to bring the currency back into our markets you ended up losing money. So we tended to stay away from the foreign markets because of the currency factor. But the new ETFs, exchanged traded funds that are all — or a very large number of them — are dollar-denominated, they make a wonderful way to watch foreign markets.
You are a pure technician, looking at market-derived data. Do you care about things like ‘Are rates particularly low?’ or what Goldman Sachs has called the Brics countries — Brazil, Russia, India, China — and their newfound demand? Did they change the calculus of bottoms or tops? Or is that just background noise?
Well, that’s the interesting thing about it. On a fundamental basis, the fundamental factors are always different in every bull market or every bear market. But the technical factors are based upon something much simpler. They are based on human psychology.
Investors tend to go from periods of extreme depression at market bottoms, to extreme elation at market tops. And there are always a different set of circumstances that help boost that change in psychology. But the range of human psychology remains pretty much the same. And we simply move from panic at market bottoms, fear at market bottoms, and finally we move to greed at market tops. And that is the limit of what technical analysis is really doing — measuring the psychology of investors regardless of events that may have inspired their bullishness or bearishness.
This is a good a point as any to transition away from talking about bottoms in general, and talking about tops. You recently did an analysis of 14 historical tops of the past century, ranging from 1929 until 2000. One of the things that I find pretty fascinating is the Nasdaq, which was really the dominate index of the 2000 crash, dropped about 78%. And the 1929 crash in the Dow was down a comparable amount. Before we specifically talk about identifying tops, I am curious, how would you compare the 2000 crash to the ’29 crash?
Well there are always areas of extreme speculation in any market advance. In the ’29 case, the equity market in the U.S. was much simpler, less complicated than it is today. The NYSE was by far the dominate exchange, the Amex was simply a shadow of the New York. All of the technology stocks at that time were listed on the New York Stock Exchange, stocks like RCA and so one. Now, the markets are more complex and we have several places that we have to look. The Amex, for a period of years, developed into what the Nasdaq is today. In other words, the Amex was a place where companies that couldn’t qualify for listing on the NYSE went to register. And that is the way the Nasdaq really started out, as initial stocks were just getting off the ground. Now it is still the dominate area for micro-cap companies and therefore an area of extreme speculation.
Editor’s note: Tune in tomorrow for part II of the interview, where Desmond discusses his theory that market tops, as well as bottoms, give very, very identifiable signals and offers his thoughts on the current environment.