coordinate to determine tlie price objective. If the half-span is moving up, the price objective uses the minimum price and reflects the distance above the projected crossing. It should be noted that this calculation of distance is simplified because the trend is established by a straight line; for nonlinear fits, the measurement of D will be more complicated.

9. Recalculate the moving averages, the half-span projection (6), the projected crossing (7), and price objective (8) each day until the actual crossing occurs. At that time D is fixed.

10. Follow the trading rules:

a. Enter a new long position when the half-span moving average tums up; cover any existing short positions regardless of the price objective.

b. Enter a new short position when the half-span moving average turns down; close out any long positions.

c. Close out both long and short positions if the price objective is reached. An allowable error factor is considered as 10°o of the height of the fun cycle (lowest to highest point).

This approach to cycles should be studied carefully as an example of a complex problem solved using elementary mathematics. There are many techniques for determining trends and a number of seasonally oriented sjstems, but a cyclic approach is rare. Whereas Hursts explanation is more complete and more sophisticated, the interpretation presented in this section should be considered only a reasonable approximation.

Charting

Nowhere can a picture be more valuable than in price forecastmg. Elaborate theories and complex formulas may ultimately be successful, but the loss of perspective is rarely corrected without a simple chart. We should remember the investor who, anxious after a long technical presentation by a research analjst, could only blurt out, "Bui is it going up or down?" Even the most sophisticated market strategies must cture the obvious trends or countertrends. Before any trading method is used, the past buy and sell signals should be plotted on a chart. Those signals should appear at logical points; otherwise, the basis of the sU-ategj or the testing method should he questioned.

Through the mid-1980s technical analjsis was considered chart inte retation. In the equities industrj thai perception is still sU-ong. Most traders begin as chartists, and many retum to it or use it along with their other methods. William L j iler. a great trader and founder of Commodity Research Bureau, wrote:

One of the most significant and intriguing concepts derived from intensive chart studies by this writer is that of characterization, or habit. Generally speaking, charts of the same commodity tend to have similar pattem sequences which may be different from those of another commodity. In other words, charts of one particular commodity may appear to have an identity or a character peculiar to ttiat commodity. For example, cotton charts display many round tops and bottoms, and even a series of these constractions, which are seldom observed in soybeans and wheat. The examination of soybean charts over the years reveals that ttiangles are especially favored. Head and shoulders formations abound throughout the wheat charts. AU commodities seem to favor certain behavior patterns.

In addition to jilers observation, the catUe market is recognized as also having the unusual occurrence of "V" bottoms. Both the silver and pork belly markets have tendencies to look very similar, with long periods of sidewajs movement and short-lived, violent price shocks, where prices leap rather than ttend to a new level. The financial markets have equally unique personalities. The S&P traditionally makes new highs, then immediately falls bade it has fast short-lived drops and slower, steadj gains. Qirrencies show intermediate ttends bounded by noticeable major stopping levels, while long-term interest rates have long-term direction.

Charting remains a most popular and practical form for evaluating commodity price movement, and numerous works have been written on methods of interpretation. This chapter will summarize some of the accepted approaches to charting, and then consider advanced concepts of both standard charting methods and sjstems designed to take advantage of behavioral pattems found in charts. Some conclusions .N-ill be drawn as to what is most tikely to work and why.

WiUiamL jiler,H..W bartEAreUEedin Cuiuodity.Pricef.Teciliug (CumoditvIese.ncl,llublican on-NrwY.dc 1977.

FIM5ING CONSISTENT PATTERNS

A price chart is often considered a representation of human behavior. The goal of any chart analjst is to find consistent, reliable, and logical patterns that predict price movement. In the classic approaches to charting, there are consolidation forms, ttend channels, top-and-bottom formations, and a multitude of other pattems that can only be created by the repeated action of large groups of people in similar circumstances or with similar objectives. As of this date, quantitative sttidies relating the psjdiologj of behavior to the reliability of chart formations have not been reliable. Traditional trading techniques found in the most popular stocks and commodities literature may themselves be the cause of the repeated pattems. Novice speculators approach the problem with great enthusiasm and often some rigidity in an effort to stick to the rules. They will sell double and ttiple tops, buy breakouts, and generally do everjttiing to propagate the standard formations. In that sense, it is wise to know the most popular and well-read techniques and act accordingly.

Speculators have many habits which, taken as a whole, can be used to interpret charts and help trading. The tjpical screen trader (not on the exchange floor) will place an order at an even number, from 50 to $1.00 per bushel in the grains, 10 to 50 points in other products. This pattem far outweighs the number of orders entering the market to buy at odd numbers, for example ttie S&P at 863.50 rattier ttian 863.35 or bonds at 105 I6 32 instead of 105 19/32. The public is also known to enter into the bull markets alwajs at the wrong time. When the major media, such as television news, sj-ndicated newspapers, and radio, carrj stories of outrageous prices in cattle, sugar, or coffee, the public enters in what WD. Gann calls the grand rush, causing the final runaway move before the collapse; this behavior is easily

identifiable on a diart. Gann also tall of lost motion, the effect of momentum that carries prices slightly past its goal. A common notion of the professional trader who is close to the maiiet is that a large move may carrj I0°o over its objective. A downward swing in the U.S. dollar/Japanese yen fran 1.2000 to a support level of I. I could overshoot the bottom by.Ol 00 without being considered significant.

The behavioral aspects of prices appear rational. In the great bull marlcets, the repeated price pattems and variations fran chance movement are indications of the effects of mass psjchologj. The greatest single source of information on this topic is Maclys Extraordinary Popular Delusions and the Madness of Crowds, originally published in 1841. In the preface to the edition of 1852 the author says:

We find that whole communities suddenly fix their minds on one object, and go mad in its pursuit; that millions of people become simultaneously impressed with one delusion...

In 1975, sugar was being rationed in supermarlcets at the highest price ever loiown. The public was so concemed that there would not be enough at any price that they would buy (and horde) as much as possible. This extreme case of public demand coincidedwith the peak prices, and Portly afterward the public found itself with an abundant supply of high-priced sugar in a rapidly declining market. The world stock markets are often the target of acts of mass psjchologj. Many have taken their turn reaching incredible heights, only to drop suddenly at a time when buyers are most confident, to start the long climb up again, it should not be difficult to understand why contrary thinking has developed.

Chartmg is a broad topic taken to great detail; the chart paper itself and its scaling are sources of controversj. A standard bar chart (or line chart) representing highs and lows can be plotted for daily, weekly, or monthly intervals to smooth out the price movement over time. The use of laige increments representing price levels will reduce the volatile appear

Rerintedm 1995 cv J-b. Wiley & Sons, Inc

ance of price fluauations. Bar chans have been drawn on logarithmic and exponential scales, where the significance of greater volatility at higher price levels is put into proportion with the quieter movement in the low ranges by using percentage changes. Each variation gives the chartist a unique representation of price action. The shape of the box and its ratio of height/width will alter subsequent interpretations based on angles. Standard techniques applied to bar graphs, point-andfigure charts, and other representations use support and resistance trendlines, frequently measured al 45-degree angles (and at various other angles in more complex theories). Selection of the charting paper may have a major effect on the results. This chapter will use daily price charts and square boxes unless otherwise noted.

it may be a concem to todays chartist that the principles and rules that govern chart interpretations were based on the early stock market, using averages instead of individual contracts. This will be discussed in the next section. For now, refer to Edwards and Magee, who said that the similarity of an organized exchange frading, "anything whose maiket value is ddermined solely by the free interplay of supply and demand," will form the same graphic representation. They continued to say that the aims and psjchologj of speculatcrs in either a stock or commodity environment would be essentially the same; the effect of postwar govemment regulations have caused a-moreorderly" market in which these same charting techniques can be used.

INTERPRETING THE BAR CHART

The bar chart also called the line chart, became known through the theories of Charles H. Dow, who expressed them in the editorials of the Wall Sfreet Joumal. Dow first formulated his ideas in 1897 when he created the stock averages for the piMpose of evaluating the movements of stock groups. After Dows death in 1902, William P Hamilton succeeded him and continued the development of his work into the theory that is known today. Those who have used charts extensively and understand their weak and sfrong points might be interested in just how far our acceptance has come. In the 1920s, a New York newspaper was reported to have written:

One leading banker deplores the growing use of chans by professional stock fraders and customers men, who, he sajs, are causing unwarranted market declines by purely mechanical interpretation of a meaningless set of lines. It is impossible, he contends, to figure values by plotting prices actually based on supply and demand; but, he adds, if too many persons play with the same set of charts, they tend to create the very unbalanced supply and demand which upsets market frends. In his opinion, all charts ould be confiscated, piled at the intersection of