The traditional point-and-figure method calls for the use of a 3-box reversal that is. the price must reverse direction by an amount equal to 3 boxes from the most extreme box of the last column before a new column can begin (it actually must fill the -Ith box since the extreme box is left blank). The importance of keeping the 3-box reversal has long been questioned by experienced point-and-figure traders, it should be noted that the net reversal amount (the box size times the number of boxes in the reversal) is the critical value. For example, a $4 box for gold with a 3-box reversal means that gold must reverse by $12 to indicate a trend change. The opposite combination, a $3 box and a 4-box reversal, would actually signal a reversal of frend at the same time. The proper selection is based on the subsequent sensitivity of the box size following the reversal. Therefore, a smaller box size may interpret the smallest price movements as a continuation of the frend and ultimately capture more of the price move; it is considered the preferable alternative. The choice of box size and reversal boxes will be considered later in more detail.

CHART FORMATIONS

It would be impossible for the average speculator to follow the original method of recording every change in price. When applied to stocks, those charts became so lengthy that inte retations similar to those used for the traditional bar chart were necessarj to select the besl pattems. In 1969, Robert E. Davis published Profit and Profitability, a point-and-figure shidy that detailed eight unique buy and sell signals. The study covered two stocks for the years 1914-1964, and 1, 100 stocks for 1954-1964. The intention was to find specific bull and bear pattems that were more reliable than others. The best buy signal was an ascending friple top, and the best sell signal was the breakout of a friple bottom (Figures 11-3 and 11-4).

Plotted daily, ftitures prices do not offer the variety of formations available in the stock market. The small number of markets and the high correlation of movement between many of the delivery months make the limitations of signal selection impractical. Instead, the most basic approach is used, in which a buy signal occurs when an X in the current column is 1 box above the highest X in the last column of Xs, and the simple sell signal is an 0 plotted below the lowest 0 of the last descending column. The variability in the sjstem lies in the size of the box; the smaller the size, the more sensitive the chart will be to price moves. In 1933, Wyckoff noted that it was advisable to use a chart with a different box size when the price of the stock varied substantially.

Trendlines

The bullish and bearish ttendlines important to bar charting also exist for point-and-figure charts. The top or bottom box that remains blank when a reversal occurs often forms the beginning of a descending or ascending pattem at a 45 angle (diagonally through the boxes, providing the graph paper has square boxes). These 45 lines represent the major anticipated trends of the maiket. Once a top or bottom has been identified, a 45 line can be drawn down from the upper comer of the top boxes of Xs toward the right, or up from the bottom of the lowest box of Os toward the right (Figure 11-5). These trendlines are used to confirm the direction of price movement and often filter the basic point-and-figure trading signals so that only long positions are taken when the 45 frendline is up and only shorts are entered when the trendline is down

Point-and-Figure Studies

In 1970, Charles C. Thiel, Jr., with Robert E. Davis, completed the first purely ftitures market point-and-figure shidj3 that calculated profitability of a reasonably large sanple of markets

TicUardD Wyck..a",S..ckJI«ketTecUuique,Nun,ber:i,eWyck..a"NewY.d . 1933,p ))

SCuThieleJE D™lE.PManrtglmC.i ,odltYTra.lug.AO.mpute Ev lu. hon.rUl &H.« tt WestLafayette.H iWfl

FIGURE 11 -3 Best formations from Daviss shidy.

AscerxingTripldap BrMkouiofaTriptaBotnim XBUr X X

X X OKOXO

FIGURE 11-4 (a) Compound point-and-figure buy signals, (b) Compound point-and-figure sell

ffe-j---

FIGURE 11 -5 Point-and-figure trendlines.

by varjing both the value of a box and the reversal criteria (the number of boxes necessarj to start a new column). With the standard 3-box reversal and only simple buy and sell formations, the tests showed 799 signals, of which were profitable; the average net profit on all trades was $311 realized in approximately 50 dajs. The period studied was I960 through 1969. In the mid- 1970s, Zieg and Kaufinan performed a computerized study using the same rules but limiting the test period to 6 months ending May 1974, an extremely active maricet period. For the 22 commodities tested, 375 signals showed 40°o of the trades were profitable; the net profit over all the trades was $306 and the average duration was 12.4 dajs. it is interesting to note that the most significant difference in the results of the two studies is in the average length of a trade, from 50 to 12.4 days, indicating a change apparently induced by more volatile maikets. Although the two tests varied in many of the details, the results are a strong argument for the consistency of the point-and-figure method as a trading tool.

In its current role, point-and-figure differs from other forms of charting because it provides a rigid set of rules for entering long and short positions. Many of the formations are still subjected to interpretation and are frequently used in the original sense by floor fraders; but as a tool for the speculator without access to infraday prices and the experience to understand the subtleties of chart analjsis, point-and-figure fills an important gap. it will tell the trader exactly what peneUation of a resistance or support level is necessarj to generate a buy or sell signal and exactly where the stop-loss order should be placed to limit rid;. It is this well-defined nature of point-and-figure charting that allows computer testing and evaluation.

The basic study of the point-and-figure method involves rules of charting, buy and sell signals, trendlines.

geometric formations, and price objectives. Since tliese pomts have been covered effectively in at least two books currently available, they will not be repeated here. More advanced point-and-figure topics will be discussed, including its relationship to bar charting, alternate plotting rules, risk-limited tradiqg, and varjing box size.

POINT-AND-FIGURE BOX SIZE

The box size used in a point-and-figure chart determines the sensitivity, or frequency of signals, and allows the identification of frends and trading ranges of various duration. For many years Chartcraft was the only major service that produced a frill set of point-and-figure charts (see Table 11-1).

4KeniutC Ziee, Jr, andPeirv J Tjiifhian. Plut andFieure Ciumodity Tradiue Tecliuiqu ciuplete tai.ul.ni-edreailts vf b-fli point-.Midieure tejt

5 AW C..Uen, .. - teWse the Ibree-P..4ut Reversal Method ..f P.Jut and Figure £t..ck K.mftuan, P.4ut andFigure Tra*ug TecUuicians (1S75.

f ibvest. .-intelligence, Larcliniont.lTV. 1975i ThiEbo"kc< i«k et Tra.liug 0:b.=tcr.ift. L.ncliiuont, ITV. 1S7J., and Zi

TABLE 11-1 a Point-and-Figure Box Sizes*