$39.95 USA/$55.95 CAN Finance/Investment ‘ The first complete guide to mastering the forecasting techniques essential for short-term trading success PORE Ca Rec Eset Lene eee ROR ee ee eee ray establish future activity, Now, Ed Gately, 2 Leading computerized trading systems developer, creates 2 groundbreaking approach to forecasting that includes setting Pe Cee en cue eee Se eo ee Cer ec ee ee Ree eer ee oem ee es ene eg Se ee ac eet Ce ee ee er ee Tee en ee eee ce es obtain confirmation, thereby strengthening reliability, These key concepts far max- imizing profits over short periods of time incluc Forecasting price movements of securities by using technical analysis SU eR ee Confirming change of trend with moving averages, candlesticks, and other De ae ed Using Fibonacci, Gann’s, Carolan’s, and other number series to target future prices and establish timing of future changes in trend Deere a a cea ee ee ee ee for learning haw to forecast accurately—and successfully ED GATELY, an authority on computerized training, designs and tests neural net- He author of Networks works for financial professionals and traders. He is i Forecasting, published by Wiley. Bese w York © Chichester + Weinheim Besbane » Singapore + Toronto [SBN O-47L-1553 * ] ANN AN I] Sohn Witey & sons, Tne Mh i sore | 5 i | abba SS08 4 5595 fessional, Reference ad Trade Grovp ] N S1T4OUd INL VIFAOS M2229 nee eis a _ z a uv aw = a m goWiley Trader's Exchange Series William Blau, Momentum, Direetion, and Divergence Joba F. Ehlers, MESA and Trading Market Cycles Mark Etzkorn, Trading with Oscillators Robert Fischer, Fibonacci Applications and Strategies for Troders Ed Gately, Neural Networks for Financing Forecasting Ed Gately, Forecasting Profits Using Price & Time Gary Klopfenstein and Jon Stein, Trading Currency Cross Rates Robert Pardo, Design, Testing, and Optimization of Trading Systems Joh Sweeney, Maximum Adverse Excursion FORECASTING PROFITS USING PRICE & TIME ED GATELY ® JOHN WILEY & SONS, INC. NewYork + Chichester + Weinhefm + Bristane + Singapore * Toronto"This text ia printed on aid-roe Copyright © 1998 by Bd Gately Published ty John Wiley & Sins, fae, All sights reserved. Published simltaneously in Canada, er, Reproduction or translation of any gart ofthis work beyond that permitted hy Section {07 or 106 of the 1976 United States Copyright Act without che permission of the copyright mer is unlawful. Request for permission or further Information zhould be addreased to the Peraussions Departament, ohn Wiley & Sons, Ine, 608 Third Avenue, New York, NY 10158-0012, ‘This publication is designed to provide accurate and ‘authoritative information in regard to the subject ‘matter covered. Iti old withthe understand.ag chat the publisher ia not engaged in rendering loge, soeouting, Dcother profesional saraces, Irlogal adviee or other ‘npertaststance is pquited, the services ofa competent Drofescional person should be sought. Library of Congrese Cataloging in Publication Data: (oun 0471-15598 Printed inthe United States of America weetesea2 1 ‘FREE TRADERS CATALOG Fax: To Ted Gately and Joe Norris wha have always been there when I needed them—thank you both,PREFACE ‘Trading securities, futures contracts, indexes, currencies, and agricul- tural products is a risky business at best. The ability to set goals or tar- Sets that might be reached in the future helps to take some of the risk out of the transaction, hecause if the price moves away from the tanget we are able to take corrective action. Risk can be further reduced by not only setting targets, but by confirming the proposed price movement by establishing the probable movement using technical analysis, Although this text does cover a number of technical analysis tools, it is not intended to replace the excellent texts by John Murphy, Mat- tin Pring, and many others. The chapters on technical analysis are in- tended to acquaint newcomers to the basics of technical analysis and to show how oscillators and moving averages can be combined with vari- ous channeis to gain a workable, relatively risk-free technique, ED GaTELy Mahwar, New Jersey December 1997ACKNOWLEDGMENTS I thank the following people who have contributed direetly and indi- rectly to the writing of this book: My wife, Susan, who did the first oditing of the completed copy. Eddie Kwong of Kasanjian Research wrote the manual on which Chapter 8 is based. He also reviewed Chapter 6 and made many excellent suggestions on how it could be improved ‘My step-daughter, Anita Greiter, who compiled zhe bibliography; a major undertaking, I might add Pamela van Giessen, my editor at John Wiley & Sons, who has been very helpfal in a variety of ways, BG ! ieCONTENTS CHAPTER 1 The importance of Targets 1 CHAPTER 2 Fundamental Methods 7 CHAPTER 3 Technical Methods for Setting Time and Price Targets 33 CHAPTER & Oscillators 63, CHAPTER 5. Other Patterns and Events 77CONTENTS, CHAPTER 6 Using Fibonacci and Other Number Series to Establish Price and Time Targets 89 CHAPTER 7 Positioning a Stop-Loss and Profit Objective 107 CHAPTER 8. The Bradley Model: Guidance from the Stars? 119 CHAPTER 9 Using Neural Networks and Genetic Algorithms to Establish. Time and Price Targets. 127 CHAPTER 10 Putting It Together: Utilizing a Variety ‘of Techniques 133 APPENDIX Software Used in Writing This Book 141, BIBLIOGRAPHY 1493 INDEX 161 THE IMPORTANCE OF TARGETS TULIPOMANIA OR THE MADNESS OF CROWDS Th 1841 Charles Mackay published his book Extraordinary Popular Delusions and the Madness of Crowds, in which he explores a number of financial disasters created by the actions of crowds of people. Proba- bly the most famous of these disasters was Tulipomania, where people in Holland bid the price of tulip bulbs to astronomical prices, only to have the price plummet back to a rational level. Crowd paychaiogy also Produced the stock market crash of 1929, where stock prices were bid to extremes, only to have them all come tumbling down. History shows that these situations are not unique, and, Mackay in his book covers “The Mississippi Scheme,” “The South Sea Bubble,” and “The Cru- sades.” One thing is certain, the specuiative bubble always bursts, leav- ing many “investors” (read “speculators”) penniless, ‘The actions of crowds have been deseribed as the Greater Fool The- ory, which goes as followa: IfT buy something at an inflated price, then Tam a fool, but will be okay when I want to sell it as long as I can find ‘4 greater fool who will pay me more for the item than I paid. When no more greater fools ean be found, the price stops rising, and everyone thors Note: Exéraordinary Poplar Delusions and the Madness of Crowds haa been ia Print since 1944, ranking it among the cideat continuously availabe hocks 12 ‘THE IMPORTANCE OF TARGETS tries to cash in their profits. Since no one wants to buy, the price comes rashing down. sessand Baruch wrote, in the foreword to the 1982 edition of Extraordinary Popular Delusions and the Maclness of Crowds, “All eco- ‘nomie movements, by their very nature, are motivated hy crowd pay- chology .... Anyone taken as an individual is tolerably sensible and reasonable—asa member of « crowd, he at once becomes a blockhead.” Prices on the stock and commodity markets are ia constant motion and apparently chaotic; they ean run up very quickly, and back down, even faster. It is a rare event when a trading day is so quiet that the daily open, high, low, and close prices are all the same. It is much more likely to see an unaccountable price jump, or an apparently irrational gain following a poor press release. All this seeraingly random motion dows, as we will see, have a sense to it. It is due to the crowd effect, CROWDS AND THE MARKET A business needs money, s0 it sells stock. A retiring couple sell their bigger-than-necessary house and move to an apartment, leaving large amount of money to be invested, so they buy some stock. A widow gets an insurance check and uses it to buy more shares in a mutual fund. A middle-aged man decides to open a business, so he liquidates some of his stock to have available cash. It is these types of transac- tions, multiplied by millions of events, that cause the bulk of price changes in the markets. This group of people, each one acting for per- sonal reasons, constitutes a crowd. Bad news causes a rush to sell, and the price goes down, Ifthe news is bad enough, the price “gape” down. A downward gap on the chart oc curs when the next price is below the previous low, ax in Figure 1 Gaps usually occur on openings. An upward gap is an open above the previous day's high, Usually, bad news causes selling to overshoot, buy orders come in, and the price rises some and stabilizes at some new rice. That is, equilibrium is established, In the case of good news, buy orders flood the market resulting in a price rise. Ifthe rise is sufficient, then profit taking occurs, and the price falls slightly as a new equilit- rium is established. (CROWDS AND THE MARKET 3 TTT rr SSE r TNE ETT TTT sek suns | Figure 1-1 Daily plot of A. G. Edwards stock. During this 2% month period there were Five gaps. Four of the gaps were up (Paints A, 8, C. and €) and one was down (Point D). Figure 1-1 shows four upward gaps (A, B, C, E) and one downward fap at D. Gaps rarely occur on charts longer than daily. They usually oecur on the open, and on a weekly chart the probable day for a gap to oecur would be on Monday's open. Figure 1-2 is a weekly chart of the same stock covering approximately 2 years. Here there are only two aps. Notice that there were five gaps in a 2%-month period using daily plots and only two gaps in a span of nearly 2 yoars using a weekly plot uuauapaene: prereert geen:4 ‘THE IMPORTANCE OF TARGETS Figure 1-2 Weekly plot of A. G. Edwards stock—compare to daily plat shown in Figure 1-1. During the nearly 2-year period covered there are enly two gaps {Points A and 8}, This occurred because the price gapped up on a Monday. Gaps fom the rest of the days of the week do not show on a weekly alat. SETTING PRICE AND TIME TARGETS Tis by setting price targets and time targets that we can anticipate how big the move may be and when it might take place. If we do not set targets, then we are merely followers. For example, if we target that the stock price can rise from $10.00 to $12.00, at which point there will be some profit taking, then we can anticipate this event and be among the profit takers by putting in a sell stop at $12.00. “Most forms of chart analysis and technical analysis are attempts to measure the demand for a stock or commodity; that is, to predict whether the priee will go up or down and, by using certain indicators, how far the price will move. We call this the magnitude of the move. SETTING PRICE AND TIME TARGETS 5 Other technica} analysis, for example, cycle and #ibonacci analysis, is used to establish the time at which a change in trend will occur. We refer to this as the duration of the move. ‘At this time we must stress the importance of establishing not only the probable direction the price of a security will move, but also the probable price and time targets on a continaal basis. Because the tar- ets are based on the initial conditions, and these eonditions contina- ally changs, it is essential that the targets be reevaluated on a regular ot periodie basis. It is possible to invest in the markets without setting price and time targets, as is the case with popular trend-following methods. How- fever, the risk of large equity fluctuations is greater when targets are not used. For example, even a plan with a target that falls short uf ita goal may produce better performance on a risk basis than a program without targets. Option traders know thst it is almost impossible to ‘make money buying or selling options without having very carefully de- fined price and time targets, ‘Targets improve performance by getting you out af the market be- fore a turn, Although you fall short of maximizing return, you ean greatly reduce the equity swings. Also, targets imply being out of the ‘market more, not panicking at the time of exit, aot fighting the crowds. es not unusual to have more than one target for a given security. One target may be a short term, intended tobe reached during the next, mont, while a second target may not he achievable until next year In the following chapters we discuss a variety of methods to set both price and time targetsFUNDAMENTAL METHODS Future price movements can be forecast using fundamental analysis or technical analysis. Fundamental analysis incorporates events that im- pact the market and can be used to estimate future price movement. ‘Technical analysis uses past prices to foretell future prices. This ehap- ter shows ways of using fundamental analysis ta forecast future price movements of securities and for establishing when the move will occur. LINKAGES Many economic events are linked; that is, what happens to one part of the economy, market, or security affects some other elements in the economy. By studying these relationships, we can target when, by how ‘much, and in what direction a security's price may move. It is accepted that interest rate yields and the stock market are linked, and that the bond market leads the stock market. Thus, if it is widely believed that the Federal Reserve will lower interest rates, then bond prices will rise in anticipation of a rate cut. Most often, the pur- pose of lowering rates is to stimulate the economy; therefore, a rise in the stock market will also follow. [we can estimate how much interest ates will fall, and compare what happened in the past under similar circumstances, we can then approximate how much the stock market ?3 FUNDAMENTAL METHODS might rise, Considering that the Federal Reserve (Fed) meets during the third week of the month, we can expect the markets will begin to ‘move in the first two weeks of the month in anticipation of a decision to lower interest rates. Depending on the size of the anticipatod move, the market may or may not make an additional move after the announce- nent, Thus, we are able to determine a time window in which the mar- ket should move. The study of how the bond and stock markets reacted to other rate cuts allows us to estimate how far the markets might move if the interest rates are eut. In other words, we are able to target the new price. Ivis important to keep a certain perspective. If the corrective aetion of the Fed is to lower interest rates following six previous interest rate rises, the market reaction will be much greater than if it comes after six other rate cuts. The first rate cut, after a series of increases, is recog- nized by the market as a change in policy and often causes much more ofa reaction than if't is just-« continuation of the existing trend. Precious metal prives are also considered as “linked”; for example, the prices of gold, platinum, and silver tend to move together. A 10- percent change in the price of gold usually causes a 5- to 15-pervent change in the price of platinum and silver. Similarly, if platinum moves 10 percent, it usually eauses gold and silver to move 5 to 15 per- cent. A 10percent move in silver has less effect on the price of gold and platinurn, ‘The price of soybeans is linked to the prices of corn and wheat, be- cause each may be used as feed for livestock, based on their protein con- tent (and other factors). In turn, the prices of cattle and hogs are Linked to the prices of soybeans, corn, and wheat. Thus the threat of drought ‘might drive up the price of soybeans and corn, which in turn will drive up the price of wheat, which in turn will drive up the price of livestock. ‘These feedgrain linkages are illustrated in Figures 2-1 to 2-6, charts covering soybeans, soybean meal, soybean oil, eattle, corn, and wheat, Notice how much these charts look alike, Note that wheat is, harvested in May, yet its price rises are entirely driven by corn and beans. Figure 2-4 shows a rise in cattle prices based on the cost of feed, Although the new crop will not be used to feed cattle for Decem- ber delivery, anticipated higher prices effect a rise in currently stored rain The economy has thousands of market linkages. For example, ifthe price of steel goes up, the cost of automobile manufacturing increases. one 2 sgn oy Figure 2-1 Soybean futures. Figure 2-2 Soybean meal futures. 9 GESERSZORSRERGAERITERSGE B8S888E3RERRsez3283 R888 SSURDSRERNU BORN ERYRNZREERRLTTERIA ELSE Razer esses SHABERCEIUE TGS INRUAE RAR 2-6 Winter wheat futures. Figure 2-5 Corn futures, Figure 2 pany ARAMLAEERAALUUAUIDELELEE TRE 3 Soybean oil futures. Figure 2-4 Live cattle futures. Figure 2-2 FUNDAMENTAL METHODS ‘Assuming that the car companies cannot pass on part or all af the cost increase to consumers, or that a price increase meets resistant buyers, ‘automobile companies make less money and the stock goos down. "There is a relationship among most major currencies. Some of these links are loose, such as the dollar/deutache mark (DM) and dollar/ven Tink. When. the news is dollar-related, both the yen and mark may have similar reactions, but when the significant change is reflected in the Japanese economy, the $/DM relationship may behave very differently. ‘The way the U.S. economy is affected by Japanese poliey may be quite different from its effect on Germany. This is particularly true when the policy involves U.S, trade quotas. A tight relationship exists among the continental European cur rencies, The Danish and Swedish kroner track the deutsche mark as though they were all the same currency, because their economies are interdependect in many ways and they share a common heritage, The French franc tracks the mark more loosely than the kroner, because these economies are only slightly linked and their eultures and her- itage are very different. These relationships allow us to forecast to a lesser or greater degree, and with a varying amount of confidence, what will happen. If the deutsche mark goes up, we can expect that the Dan- ish and Swedish kroner wil follow closely behind, but not necessarily the French frane. EARNINGS, NEW PRODUCT, MERGERS, AND OTHER ANNOUNCEMENTS. A number of expected and unexpected events can affect the price of a stock, such as anticipated earnings, new produets, possible mergers, and other nevis. While the entire stock market tends to move with the economy, individual stock prices are closely tied to earnings. The ratio of the stock price to its earnings is a measure of the rate at which the company and its earnings are expected to geaw. High-technology com- panies that are growing rapidly, such as Intel and Microsoft, carry bigh price-tovearnings ratios, Companies with limited growth prospects pos- sess small ratios; those in contracting industries have even smaller ratios, REGRESSION ANALYSIS 3 Zacks and Standard & Poor's estimate the quarterly earnings of ail major companies. A company that does nat meet the “Street's” esti- mate will find its stock down sharply on the announcement of its trae earnings. A company that exceeds its earning’s forecast will find it ing’ forecast will find its sock sharply. Knowle of the true and estimated earnings of a company therefore ean lead to large trading profits. Fortunately, all People who would know the exact figures aro corporate insiders who are barred from trading ahead of this information. _ ihe announcements about new products can also move stock Prices. Most announcements are intended to push the stock price up- wards, If the new product represents a breakthrough in technotogy, then the upward move ean be quite dramatic. An announcement about & merger, Purchase of a business, sale of a division, or other asset sale san all affect the stock price. The direction of the movement sill depend. on the offect of the announcement on future earnings = Other events can aifeet a stock's pri fect a stock's price. For example, the govern- ment ban on DDT led to the manufacturers taking @ one-time charge for the discontinued business. REGRESSION ANALYSIS ‘Studying how far a price is from some normal value often gives clues to ‘when to expect changes in trend. Regresaion analysis, sometimes called least-squares analysis, ean be thought of as a way of determining the best straight line through a series of data points. This technique anini, mizes the distances between the prices and the resulting trendline. Figure 2-7 shows a historical plot of a stock price with a regression line plotted through the data, Prices are evenly distributed on both sides of the regression line. Notice that during this trending period, prices fluctuated a similar amount above and below the regression line. In this case, once the price moved approximately 1K points above or below the regression line, it reversed direction. In essence, we could say that the prices oscillate: EERad dhretion i fe could say that the prices oscillated Figure 2-8 is a weekly plot of Boatmans Bank stock over a 4%4- ‘year span. Here two regression lines have been drawn: one from April6 FUNDAMENTAL METHODS Figure 2-7 Chart of A. Edwards with 2 regression tine drawn through the data, 1992 to September 1994, a period of 2% years; the second from No- vember 1994 to January 1996, a period of 1% years. This chart shows, that regression analysis is applicable to weekly as well as daily charts Figure 2-9 is similar to Figure 2-8, but with a shorter-torm viow of the regression analysis, having six regression lines rather than the two shown in Figure 2-8, ‘The fact that the data stay within a given distanee from the re- gression line means that we can use these amounts as targets for buy- ing and selling. That is, buy when the price is at the distance below the regression line; sell when the price is at the distance above the regres: sion bine, Figure 2-8 Chart of Boatmans Bank with regression line different time pavioas, Figure 2-9 Chart of Boatmans Bank different time periods, it | anh ih ‘a ve " with regression lines draum through six 15ts FUNDAMENTAL METHODS SEASONALITY Wo can take advantage of natural phenomena and fundamentals to help identify targets. Seasonality ean be used to identify potential ‘changes in trend. Usually the price of grains drops at harvest time, and ‘the price of fuel oil rises as winter approaches. Therefore, the harvest signals that hedged accounts in crops should exit shorts, while late summer roquires a shifl to the long side of fuel ol. Figure 2-10 is a weekly chart of winter wheat from mid-1990 through most of 1995, Note that the points marked A on the chart are the yearly lows, and that they occur in the middle of the year when the harvest is nearly complete and good assesament of yield can be made. ‘The points marked Bare the high points for the year. Notice that these ‘oceur around the first of the year when it is impossible to estimate the coming harvest, and existing inventory is being drawn down rapidly. ug0gaeeenbsyseneneaeyy ‘ot 13 1953 1955 Figure 2-10 Chart of wheat with the highs and tows marked, CONVENTIONAL cycLES 4 pore 8 S0838 2908 Figure 2-41 Chart of heating oil with the highs and lows marked, ‘The harvest of winter wheat in the southern states ia probably the cate CONVENTIONAL CYCLES Cyeles are another intrinsie component of price movement and ean be (207 helpful in determining when a change ia iver might ake pee Cyeles allow us to accurately predict events in nature: bed mines {he tides, and planetary movements. Cycle analysis can he unt! to pre, Act changes in the financial markets, althoust not normally sith de saTEHVEL: aPr FUNDAMENTAL METHODS i «4, moat commodity price cycles are suracy found in nature. As expected, aoe aero seasonality, which is clearly a subset of eycles, Some stock pri eee tinked to the seasonal cycles. This is most common for companies arate macy produts are ariultoral or whose services are them= Ives seasonal. Other stock prices are linked to the cycles of new prod- selves seasonal. , “ Suosimple wey of finding cycles is to use the popular Metastock™ software program, which has &eubprogramn to lcato possible els ‘After the program hes been given the mamber of day (week, months to use, it draws the lines on the chart and the mouse ean be used to ‘move the lines to see various fits. By trying different periods and mov- ing the lines around, a best fit can be achieved. Figure 2-12 shows a CONVENTIONAL CYCLES 18 chart of A. G. Edwards with 36-day cycle lines fitted to the chart. The stochastic oscillator is an oscillator that shows when a security is over- sold or overbought. When the oscillator ig a low value, the security is oversold, and when it has a high value, the security is overbought. No- tice that the eycle lines are located where the stochastic oscillator is making major changes. When a eycle haa been found, the probable time of the next change in trend can be identified Figure 2-13 is a 5-year chart of Boatmans Bank. Twelve-week cycle lines have been fitted to the chart, Notice that the eyele lines correlate well with changes of trend. Cycle lines are another technique for tar- geting changes in trend. By themselves they can be erratic, and fre~ quently not as precise as the ones illustrated in Figures 2-12 and 2-13; however, when used with Fourier analysis, their accuracy can be en. hanced. TEESE PPT PTT PTT TT wenteceee eeeuuyewugeseanes Figure 2-12. Chart of A. G. Edwards stock with 36-day cycles plotted. Figure 2-13 Weekly chart af Boatmans Bank stock with 12-week cycle lines plotted.20 FUNDAMENTAL METHODS FOURIER ANALYSIS Cycles can be found using a mathematical evaluation called Fourier ‘analysis, available in the Metastock software package. It changes the price data into a set of sine waves representing major and minor cycles. ‘When all the waves are combined, the result is the original data series. Almost any price serios can be represented as a sine wave plus a series, of harmonics. A harmonic is @ sine wave that is a multiple of the fon damental sine wave, that is, a wave in which the eycles oceur twice as, often, or three times as often. For example, in the United States elec- trical power has a fundamental sine wave of 60 hertz (Hz), or 60 cycles per second icps). Because of the way the power is generated there are also some harmonics included with the power. Mostly these are the third and fifth harmonics. That is to say, along with the 60-Hz compo- nent are components of 180 Hz and 300 Hz, For example, a square wave can be expressed as a fundamental sine wave plus a series of odd harmonics. ‘Take the unlikely case of the square wave shown in Figure 2-14. The formula for a square wave rends as follows: Square wave = Sinelivt} + 1/3 Sinet3ut) + 1/5 Sino(Swt) + U7 Sine(Twei + 1/9 Sine(Qut) + W/11 Sine(1 Lue) +113 Sine(12.0¢) and so forth. ‘This equation should be read: “Square wave is the fundamental sine wave plus one-third of the third harmonic plus one-fifth of the fifth harmonic plus one-seventh of the seventh harmonie plus one- sninth of the ninth harmonic plus one-eleventh of the eleventh harmonic plus one-thirteenth of the thirteenth harmonic... and so forth.” Figure 2-14 shows how this works, Here we have only added the fundamental sine wave with one-third of the third harmonic, which are the first two terms of the previous equation. Notice how, with only this one har- monic added, the wave is closer to a square wave than a sine wave Now imagine adding all the other odd harmonies. Voila, a square wave! Any eyelie data can be separated into fundamental and multiple harmonic sine wave components. When this is done, the resulting equa- tions can be used to target likely times when major and minor changes in trend can be expected. FOURIER ANALYSIS a Square Wave SLND / Fundamanta Sino wt DSDSSSINSININS Third Harmonie: 18 Sine a wt SISA SPASA/S Fundamental Pus Third Harmonic: Sina wes 1/9 Sine 3 wt Figure 2-14 Develogment of a square wave from its fundament owe 2-14 square wave from its fundamental and harmonic Sinee the price movement of a stock tends to be cyclic, Four naa canbe npplied wo the Hatriea data, and orecants ino che he tare can be made. Fourier analysis is a standard mathematical tech- ‘nique for finding cycles, and is sometimes called “spectral analysis.” Figure 2:15 is a daily bar chart of Augat with the results of a Fourier analysis shown below, This chart was made using the Ad- vanced G.E.T. 3.0 software package. This program combines analysis. and forecasting by applying Fourier analysis to determine the sine- ‘wave components of the price data, and then combining these cycles to Show the probable point in time when trend changes might take place2 FUNDAMENTAL METHODS Figure 2-15 Changes in trend determined by Fourier analysis (see text). ‘The heavy black lines represent major changes in trend where multiple cycles or events coincide, and the lighter lines indicate minor changes in trend. This program uses an iterative approach to determining the cycles. In the example shown here, the number of iterations w: 373,000, which took 20 minutes on a Pentium 90 computer. ‘The DOS and the Windows versions of Metastock include a feature that studies price movements by using Fourier analysis and indicates the three most dominant sine-wave components. Figure 2-16 shows the result of that analysis. In the top panel the Fourier analysis has iden- tiffed the three sine-wave components with periods 51 days, 26 days, and 9 days ‘Using the eyele program contained in both the DOS and Windows versions of Metastock, we can see whether these components have value, In the bottorn panel of Figure 2-17 the eycle program plotted the first of these sine wave components, the 51-day cycle lines, and we can Man Ss N : nf penta gael ! t Hy wid he Htc iuidbbdlal J 3 [at ta ld tl | at haf y y ener sulle aaa hs hat dll Buwinva seangt i es oo wo F | Figure 2-17 Conformation of 51-day cycle determined in Figure 2-16. 23Tees, PE: hy sll llth i f = a Wy 2 i 3 ay a) jill Te oa Se IOVN AG Aas USS Ow ES Figure 2-18 Conformation of 25-day cycle determined in Figure 2-16. a li I, shtllfNolo La Tis vt AS a iy x5] xf ay il x ad asl us Puen ge rd a ae il Li |” GeIW VA ATH DASH Figure 2-19 Conformation of 9-day cycle determined in Figure 2-16. 24 FOURIER ANALYSIS. as see that many of the changes in trend occur at the 54-day cycle, while changes in the down trend occur near the 48-day cycle for an average of 81 days, Twenty-five-day eycles are plotted in Figure 2-18. Again, many of the trend changes occur at or near the 25-day cycle. A Suday cycle is plotted in Figure 2-19. Notiee that many of the trend changes occur at or near the 9-day cycle points, Because of the effect of no’ this short day cyele is not as necurate as the 51- ‘cause changes in trend tend to accur at the cycl ‘tended into the future to pinpoint possible Fat ise, and 25-day cycles. Be- le ines, lines can be ex- ‘ure changes in trend. If the individual cycles represent. methods of targeting probable ese components coineida fy points where trend changes and 25-day eyeles have been changes in trend, then the points where th (come together) are more likely to identi ‘might take piace. In Figure 2-20 the 51- Mh | sul lla an ul iit ° Sb ivvs svar eso wo Figure 2-20 Conformation of how 51- and 25-day cycles line up to establish a Dossibie time wien a change of trend can accu,26 FUNDAMENTAL METHODS plotted an the same geaph. Notice that both cycles line up very closely round August [4 and September 18, 1995. When many lines cluster at the samo point, itis a stronger signal that a change will take place at that point than is given by either line by itself, GANN AND HIS ANGLES William D. Gann, born in 1878 in Lufkin, Texas, was known as The Master Trader. His techniques helped to target the time when a change in trend might oceur and tho possible magnitude of that change. His methods have been used successfully for many years, and have in- creased in popularity in recent years, ‘The following appeared in the December 1909 issue of Ticker Mag- azine describing Gann’s ability to forecast time and price targets: ‘One of the most astonishing calculations made by Mr. Gann was during last summer [1909] when he predicted that September Wheat would soll at $1.20. This meant that it must touch that fig- ture before the end of the month of September. At owelve o'dock, Chicago time, on September 30th (the Iast day) the option was selling below $1.08 and it looked as though his prediction would not he fulfilled, Mr. Gann said “IF it does not touch $1.20 by the close of the market, it will prove that there is something wrong ‘with my whole method of calculations. Ido not care what the price ig now, it must go there.” It is common history that September ‘Wheat surprised the whole country by selling at $1.20 and no higher in the last hour of trading, closing at that figure, Gann believed in the existence of a natural order of everything in the universe. He was also very religious and often used biblical pas- sages as a basis for his trading. The following passage from Ecclesi- astes 1:9-10 was often quoted by Gann: ‘What has been, that will be; what has boen done, that will be done. Nothing is new under the sun, Even the thing of which we say, “See, this is new!" has already existed in the ages thet preceded GAN AND MIS ANGLES 2 Gann believed that this universal order of nature existed in the stock and commodity markets, Price movements were not random, but Were predetermined. These predictable movements wore the result of points of force found in nature. These forves of nature not only moved Price, but could be predicted, For the most part, he relied on seasonal- ity, which certainly satisfied his eriteria, Gann used a number of different mathods to reach his predictions not all of which were fully disclosed, Two af the disclosed methods, and those most widely used, are Gann cyclic time analysis and Gann lines, InGann cyclic time analysis, a single fixed cycle is projected. In Figure 2-21, using only one pivot date, or anchor point, A, we choose ta project 30-day cycle, point B, Noties that this is the distance from the first. low to the first high. The forecast change in trend is projected 30 days from point B, placing it at point C. Gann had several cycles that he favored for his projections. Begin- ning with 91 days, he used a sequence of eycles $1, 182, 273, and 365 days which, of course, represents the calendar quarters associated with seasonality, Gann considered calendar periods very important und closely associated with the natural orde Projactad Change B Rosuting intend a “ [+ 20 Days» /—30 Days —+| Figure 2-21 Projections of possible Gann cycles.a FUNDAMENTAL METHODS Projected Chango in Tara ON auat 7 Pons Numose of Mts oF Custers Pally C:VG\ORTA FILED aE A, 6. EMANIS (dally) Figure 2-23 Automated multiple projections. 1975 TUE 24 Figure 2-22 Multiple projection of possible Gann changes in trend. Another form of Gann analysis projected several fixed cycles, as i+ \wstrated in Figure 2-22, Here we used cycles of 15, 30, and 46 days, projected from two pivot points, A and B, resulting in target points C, D, and E. In practice Gann used the following cycle periods, 30, 60, 91, 121, 152, 212, 248, 273, 304, 335, and 365 days, Because of the even di- vision of the calendar year, Gann clearly put great store in natural cy- cles and their harmonies (in this eave, subcycles). In Figure 2-23 we have applied the technique demonstrated in Fig- ure 2-22 to an actual security. Here the pivot points are marked with a “P," and the cycle periods were to Gann periods given above. The bars along the bottom of the chart show the number of cycle hits or clusters, Notice thet changes of trend often take place at points where there are a large number of hits. Figure 2-24 illustrates the use of the data from Figure 2-28 to establish possible future turning points if Pats C:WWS\DATAN, FILE» ASE Figure 2-24 Using the data of Figure 2-23 to establish points. 2 a 6. FIWARDS — (datly? possible future turning30 FUNDAMENTAL METHODS. Gann believed that. specific geometric patterns and angles had unique characteristics that could be used to predict price action, All of Gann’s methods specify that equivalent time and price intervals be ‘used on the chart so that a rige-to-run ratio of 1 appears as a 45 de- igreo angle. Gann believed that prices could follow other angles that are related to the 1 x 1 ratio by integers. These ratios are 4x 1,21, 1x 1, 12, 14. A set of these angles, called Gann fans, is shown in Figure 2.25, Notice that the price rise starts by following a2 I line, and after several weeks changes to follow the I x 1 line, and several months later switches to the 1 x2 line. Figure 2-25 An example of Gana lines. GANN AND HIS ANGLES. n ny [ed [i si [= Lt 2} Ie! Figure 2-26 Example of using Gann lines to predict when a change of trend will occur. Gann fans can be drawn from either the lows or highs of price ‘movement, Gann also believed that the points where Gann lines inter- Sect were where trend reversals would occur. In Figure 2-26 we have Placed sets of Gann fans: one from a low point and one from a high point. Notice that trend reversals have taken place at point B. Also note [at at points A and B the price action has been contained by the Gann3 TECHNICAL METHODS FOR SETTING TIME AND PRICE TARGETS ‘This chapter covers a variety of methods for setting time and price tar- gets using supporvresistance, channels of various types, and trend- nes, Also, it discueses a number of moving averages, candlesticks, and other methods of plotting price movement that ean confirm when a change of trend has taken place, SUPPORT ‘Support can be defined as a price level of a commodity or stock at which buyers come into the market, thus forming a base that supports the Price at that level. Support is a combination of psychological factors Working upon the market participants, such as anticipation and gen- eral awareness of the possible development of a support level, plus ex ‘temmal justification, such as economic factors, the cost of production, net worth (liquidated value), and so forth. For example, ifa stock goes up from $16.00 to $17.50, some poten- tial purchasers will have missed getting in on the move. Itis likely that 33