by Martin Zweig
In the spring of 1972 I published still another technical indicator in Barron’s, this one on short-selling activity. Again I predicted a down market and again the market obliged by easing lower over several months. That didn’t hurt my new business. I began to advertise The Zweig Forecast in Barron’s in the spring of 1972, and it quickly built into a real business. Soon income from my letter and my money management activities surpassed my college-teaching earnings. However, I really enjoyed teaching and accepted a position as associate professor at Iona College in New Rochelle, New York, where I taught for seven more years, finally taking a leave of absence only because I couldn’t devote enough time to my teaching chores.
When time permits, I’d like to teach again at the college level. What I enjoy most about teaching is the stimulation. I prod the students into asking questions and am not embarrassed if I don’t know the answers. I never try to finesse my way through. If I don’t really know, I tell them so. Often they will force me to think through an issue. In the thinking process I frequently get new ideas, some of which have led to theories I have used in market forecasting.
While still publishing The Zweig Forecast, I launched The Zweig Fund, a closed-end stock mutual fund, in September 1986, with an initial offering of $340,000,000. Two years later came The Zweig Total Return Fund. Both funds are traded on the New York Stock Exchange. Thereafter, we launched several open-end funds.
Ever since I can remember, I have had an almost overwhelming desire to learn all I could about the stock market and to play it successfully. Perhaps my urge was not too different from that attributed to the mountain climber who must assault the mountain just because it’s there. Now, I don’t relate particularly well to mountains, but from an early age I wanted to surmount the summit of the stock market, so to speak. It was a challenge I couldn’t resist.
It hasn’t been all peaches and cream. The stock market at times has driven me up the wall. Then again, it has also earned me a fair sum of money and given me many emotional highs. But, no matter how rough the stock market road, it has never, never bored me. I always find the market fascinating and filled with surprises. Perhaps that’s because no one on this planet will ever know all there is to know about the market—and no one can expect to be right all of the time or even most of the time.
You can, however, be right more often than you are wrong. If you are right 60 percent of the time, ride your profits, and rein in your losses, you’ll find that when you’re right you’re very right, and when you’re wrong you’re only moderately wrong. In the long run, a 60 percent success rate translates into huge gains, a 50 percent rate into solid gains, and even a 40 percent rate can beat the market.
In playing the market, remember you must deal with probabilities, employ sensible strategies to limit risk, and get aggressive only when conditions warrant. I’ve spent my adult life trying to fathom the stock market and, in the following chapters, will try to give you the best information I’ve acquired.
CHAPTER 3
The Market Averages—What They Mean
The world’s best-known stock market index is the Dow Jones Industrial Average. It has been maintained since 1897, when it consisted of 12 large industrial corporations. In 1916 the list was expanded to 20 stocks, and it grew to its present size of 30 industrials in 1928. Many of the biggest manufacturing firms in major U.S. industries are included. They are:
Allied-Signal Eastman Kodak Merck
Aluminum Co. Exxon Minn. M&M
Almer Express General Electric J. P. Morgan
AT&T General Motors Philip Morris
Boeing Goodyear Procter & Gamble
Caterpillar Hewlett-Pachard Sears Roebuck
Chevron IBM Travelers
Coca-Cola Inter Paper Union Carbide
Disney Johnson & Johnson United Tech
DuPont McDonald’s Wal-Mart
Originally, when the Industrial Average included 12 stocks, it was calculated by adding the prices of the 12 issues and dividing by 12. The same rule would apply today for 30 stocks if the average were brand-new (except that the appropriate divisor would be 30). However, over the years stocks frequently split. Moreover, Dow Jones has on occasion substituted one industrial firm for another. The most recent changes came in March 1997, when Hewlett-Packard, Johnson & Johnson, Travelers, and Wal-Mart replaced Bethlehem Steel, Texaco, Westinghouse, and Woolworth.
Such switches and splits require a change in the divisor to keep the average continuous. If there were just two stocks, with A at $20 and B at $40, one could calculate the average by adding the prices ($60) and dividing by the divisor (2). The answer, of course, is $30. However, suppose B split 2-for-l and the price went to $20 (no “real” change). If you added the prices of both stocks and divided by 2, your ministock average would fall from $30 to $20. Obviously, that’s not realistic, since there was no real change in value (B may sell for one-half its former price, but it has twice as many shares, so stockholders are equally well off).
A downward adjustment to the divisor is required to maintain the old average at the appropriate $30 level. Now the sum of the two current prices (after the split) is $20 for A plus $20 for B, or $40. By dividing the $40 total by 1.333, you would get $30, the true average. Thus, the new divisor would be only 1.333, even though there are two stocks in the average. Over the years the 30 Dow Industrials have undergone so many splits and changes that the divisor is down to 0.346. In other words, you could add up the prices of the 30 Industrials, divide by 0.346, and wind up with the actual Dow Jones Industrial Average.
Most investors will not need to use the divisors, but it is important to understand what is meant by a Dow Jones Industrial Average of 5,500, or 6,000, or whatever.
There are two other Dow Jones averages, the Transportation Average and the Utility Average, plus a 65-stock average that combines all three. You’ll often see them mentioned in the financial newspapers, but since they cover narrow industry groups, I don’t see much point in discussing them in detail in this book. Occasionally, the Utility Average, because it represents a group of stocks that are very sensitive to interest rates, can be an effective leading indicator for the rest of the stock market. There is such a tendency, but it is not overwhelming.
GRAPH A
Ned Davis Research, P.O. Box 1287, Nokomos, FL 34274-1
AVERAGES AND MARKET ACTIVITY
Let’s see how the most significant averages reflect market activity. The top half of graph A (pp. 26–27) shows the monthly plot of the Dow Jones Industrial Average back to 1962. Each vertical dash on the graph represents the range between the Dow’s high and low for the month. You ought to take a minute or two to look at the graph and familiarize yourself with the market’s history over this span. You can readily see that there were bear markets in 1962, 1966, 1969–70, 1973–74, 1976–78, early 1980, in 1981–82, in 1987, and in 1990. The 1973–74 bear market, when the Dow fell from 1052 to 578, was the worst since the Depression in the 1930s. The 1987 crash bear market was the second worst.
There was a long bull market between 1962 and early 1966, when the Dow nearly doubled. Other bull markets occurred in 1967–68, 1970–73, late 1974–76, 1980–81, 1982–87, and from 1991 to this writing in June, 1996. You’ll also find periods of intermediate-type advances and declines, including the sell-off of 1983 and 1984, which I would classify as an intermediate-type decline.
The lower half of graph A shows the Standard & Poor’s 500 Stock Index. This average concerns institutional investors the most since performance of their own results is usually compared to this benchmark. As its name implies, the S&P consists of 500 stocks, most of which are blue chips. The S&P is weighted by market capitalization. This means that if a stock has 100,000,000 shares outstanding and sells for $20 per share, its market capitalization is 100,000,000 x $20, or $2,000,000,000. The greater the capitalization, the greater the weight given in the S&P Index. As of March 31, 1996, General Electric was the most heavily weighted stock, accounting for 2.70% of the index valu
e. Other heavily weighted stocks included Coca-Cola, Exxon, AT&T, and Merck. Since the S&P 500 is dominated by very large companies, stocks of the smaller firms have correspondingly less weight.
As graph A shows, the Dow Industrials and the S&P Index generally move in the same direction. However, the magnitude of the gains and declines can differ. For example, in April 1981 the Dow peaked at about 1024, slightly under its high of 1052 in early 1973. Meanwhile, the S&P peaked earlier at over 140 in November 1980, considerably above its top of 120 in 1973. While the Dow actually declined almost 3% between the peaks, the S&P climbed nearly 17%.
This can happen when one is dealing with a fully weighted sample of 500 stocks vis-à-vis a much smaller sample of 30 stocks weighted solely for price and not market capitalization. It may seem frustrating at times, but there is no perfect gauge of the stock market. There are merely alternative ways to measure it.
Another major market average is the New York Stock Exchange Composite Index. It is constructed like the S&P 500, except that it gives weight to every common stock on the NYSE, roughly 2,500 stocks. That is, it uses market capitalization to determine the weights, where once again General Electric is the most heavily weighted stock. The movements of the New York Composite Index and the S&P 500 are very similar.
THE UNWEIGHTED PRICE INDEX
Another way to measure the market’s performance is to look at the broad spectrum of stocks not weighted for capitalization. Such an average is called an unweighted price index. For this purpose, I constructed my own measure, called the Zweig Unweighted Price Index, or ZUPI for short, whose base is 100 at the beginning of 1965. I get the raw input for my ZUPI from Quotron, a basic computer service for stock quotations. If the ZUPI rises, say 1%, on a given day, it means that the average stock rose 1% that day regardless of the size of the company.
For example, if you had a two-stock average comprised of General Electric and some very small company, on a weighted basis such as the S&P 500 the performance of the small stock would make virtually no difference and General Electric would account for 90-odd percent of the weight. Your average would almost reflect General Electric’s price movement. But in an unweighted index of two stocks, General Electric’s percentage change would be given just half the weight and the small stock’s price change the other half. When this is done for all stocks on the exchange, General Electric has no more weight than the small company.
GRAPH B
Ned Davis Research
The unweighted average is quite useful for the individual investor, who has the flexibility to buy stock in any company, large or small. But the S&P 500 is probably a better measure for institutional investors, who buy millions of dollars’ worth of General Electric or Coca-Cola but who would have difficulty acquiring very large dollar amounts of extremely small companies.
The performance of the ZUPI can occasionally differ markedly from that of the Dow or the S&P 500. That’s because the ZUPI is heavily biased toward the performance of the smaller stocks, often called secondary stocks, while the major averages are dominated by the blue chips. There are times when blue chips do fairly well, such as from the spring of 1972 to early 1973, while the secondary stocks decline. At other times, such as in 1977, the secondaries can rise while the blue chips come under pressure and fall. But in a major bear market such as 1973–74, most stocks weaken and all the averages drop significantly. It’s a reflection of the old saw, “When the paddy wagons come they take the good girls with the bad.” Similarly, during major bull markets, both the weighted and unweighted averages will tend to advance.
In the studies that follow, I have tried to test various indicators and models against both the ZUPI and the S&P 500 Index. Occasionally, tests will be made against the Dow Industrials or the Value Line Composite Index. The Value Line is seen on the lower half of graph B (pp. 30–31) just under the ZUPI. This index is constructed by Arnold Bernhard & Co., publishers of the Value Line service. This is an unweighted index of approximately 1700 stocks (which is somewhat smaller in size than the NYSE Composite) the bulk of which are on the New York Stock Exchange. It is constructed exactly the same as the ZUPI except that the ZUPI recognizes all common stocks on the NYSE and none from the AMEX and OTC markets. The Value Line ignores a few hundred stocks on the NYSE but includes some from the AMEX and OTC. As seen in the graph, the two unweighted indexes perform almost identically, although they may deviate by small fractions in the short run.
In the spring of 1982, the Kansas City Board of Trade began the first trading ever of stock index futures and based the activity on the Value Line Composite Index. The futures can trade at prices above or below the actual value of the index. But discounts and premiums aside, you can buy or sell the market as a whole as measured by the Value Line unweighted index. Also in 1982, stock index futures began trading in Chicago on the S&P 500 Index. Later that year a third market in stock index futures began in New York with trading on the New York Composite Index.
The activity in the S&P 500 Index is by far the largest, and the combined dollar volume in the three stock index futures now exceeds the dollar volume of all stocks traded on the NYSE. In a recent typical week the dollar volume of all trading on S&P futures was $113.7 billion, Value Line futures traded $169 million, and New York Composite futures traded $2.8 billion, bringing index futures to $116.7 billion, significantly greater than the NYSE volume of $80 billion. Clearly, stock index futures have grown tremendously since their origins a few years ago and are becoming increasingly important to the investment community and to speculators.
I cannot cover stock index futures in greater detail here because this is a whole subject in itself. However, in later chapters, when you see tests run of an indicator or of a model, bear in mind that you could approximate the results of those tests by trading stock index futures. That way you can avoid most of the transaction costs associated with trading stocks and get the diversification of the market average involved.
INFLATION ADJUSTMENTS
Graph C (pp. 34–35) shows the Dow Jones Industrial Average plotted monthly since 1921. It’s similar to the Dow in graph A except that it goes back much further. Also note that the price scale is in percentages. Thus, a rise in the Dow from 100 to 200 is a gain of 100% and would run the same vertical distance on the graph as a move from 500 to 1000, which is also a gain of 100%. The problem with using nominal prices (prices as they appeared at the time) to construct the market average is that over time they are distorted by the effects of inflation or extreme deflation. On a short-term basis this usually doesn’t matter too much, say over a period of a few days or a few weeks. Or, if the inflation rate is “normal,” say in the 2%-to-4% range, even over a period of one to two years, the effects are not that significant. But if inflation gets up to 10% or so, as it did during a part of the 1970s, or if you have extreme deflation as seen in the early 1930s, it causes a tremendous distortion in the nominal price averages.
GRAPH C
Ned Davis Research
For example, assume that you invested money in the Dow Industrials when the Dow Average stood at 1000. Suppose that over the next five years the cumulative ravages of inflation doubled the consumer price index, a compounded inflation rate of about 15% a year. Consequently, you would need $2000 to purchase what $1000 would have bought five years earlier. Now, suppose that while inflation doubled, the Dow remained at 1000. To conclude that your investment was even at the end of that period is nonsense. Your net worth had actually declined by 50%. The greater the inflation rate, the greater the adjustment necessary in the stock price averages.
Graph D (pp. 38–39) shows the Dow Jones Industrials plotted monthly from 1921, adjusted for the effects of inflation and deflation. It is properly called “Deflated Dow Jones Industrials.” This shows how the Dow performed in “real” dollars, the actual buying power of the dollars that you might have invested in the market. Thus, using the example above, had the Dow in nominal terms remained unchanged over a five-year span while the inflation rat
e doubled, a graph of the deflated Dow would show a gradual decline from 1000 down to 500 during that period, losing one-half in real terms from its beginning point.
To brush up on the performance of the market averages during most of this century, study graph D of the deflated Dow. On a long-term basis there was a tremendous bull market from 1921 to the peak in 1929. This was probably the greatest bull market in our history. Interestingly, prices were quite stable during the decade of the 1920s, with no significant inflation.
WHAT REALLY HAPPENED IN 1929
The market peaked around Labor Day of 1929 and began to sink lower through September. In October the decline picked up steam, and prices literally collapsed on October 23 and October 28. Indeed, at that time, the decline on October 28 was the largest one-day drop in the history of the stock exchange. On that day alone, the Dow plunged from 298.97 to 260.64, a sickening crash of 12.8%! (The October 23 decline had been a whopping 6.3%.) In mid-1996, with the Dow in the 5600s, a percentage drop like that of October 28, 1929, would lop an amazing 700 points off the Dow … and create a splash on the network evening news!
Most people think the market simply crashed on October 29, 1929, and know little about what happened before or after. True, stocks did collapse that day, with the Dow closing at 230.07, down a hefty 11.7%, but the drop was somewhat less severe than the prior day’s loss. Volume hit an all-time-record 16.4 million shares, an amount not seen again on the exchange for another 3½ decades! Actually, on October 29, stocks put on one of their greatest rallies ever during the last hour of trading. The next day, on October 30, the Dow soared 12.3% to 258.47. But the rally was short-lived, and by mid-November the Dow closed at about 199, down by nearly one-half from its September 3 high of above 381. That so-called 1929 crash encompassed far more than just October 29 and represented the greatest decline ever in a fairly short period. But the major damage really occurred between 1930 and 1932.