back start next
[start]                         [ 25 ]                             
directly affea the bottom line, and so these obser%ations wih become ingrained withijractice and exXiience.
CALASHJaily OSf1IW2001 Mov Avg 1 line 29.288 Acme V Stralefly Linear Rl
Figure 6.9. Computer Associates Volatility
7 Range Trading
A iptcidator i% a .- observestlieJutmv, Qiidccts before it occurs.
hi his 1 The Science offlitting, Ted describes how he calculated that the strike zone was approximately seven behs wide and ele-en bahs high. The result was a matrix of baseballs, and he cslculated his battii a-erage for each beh in the matrix. Wlule the ordinarj batter decided between beh and strike, Williams refined the strike into se-entj-seven separate categories .
Williams also determined that once a batter started switit at pitches just se-eral inches outside of the strike zone, the strike zone expanded from 4.2 square feet to 5.8 square feet, an increase of almost 37 percent. Once a pitcher learned a batter would swit at bad pitches, then thats ah the batter would get, and the batter was destined to be a .250 hitter.
Now, imagine if the "Splendid Splinter" apphed his analysis to ftie stock market and tumed his attention to the range. He would sort eh ofthe raises into fheh -arious sizes and then determine his battit a-erage, or profit factor, for each range. He would conclude that when the range is narrow (in the strike zone I his profit factor is higher. In contrast, when the range is wide (out ofthe strike zone), his profit factor is lower.
The a-erage trader analyzes a trade as either a winner or a loser-a beh or a strike. The professional trader analyzes atrade from its risk/reward ratio . If the trader uses range to determine stops, then the risk numerator is the range itself (the higher the range, the higher the risk), and the reward denominator is the profit target. For example, if a long entrjis tiiggered at the high ofthe bar, and the range is 1.5 times the ATR, then the trader is probably swingii out of the strike zone.
ImI W,)),.ii„-. )..m<-.l 1 ) ii. Ml... ii
.iv. i..,.. I.......-..... , ill. I...
»li,-i.-ii 1>.1 1 -M
lilt. /.........)...m..,.> till......... I,.......
.......II III ».....l..» ....I..ill ..I,.
7.1 Range Ratio
The Acme N Sj-stem is based on a simple concept called the Saiige Ratio. We want a ratio value less than one because a day with a low Raie Ratio (RR) is generally followed by a wide range (WR) day imder volatile maiket conditions. To calculate the RR, divide the current days range by the AverageTme Range (ATR) over a certain reference raie to estimate todays volatility. For example, ifthe ATR ofJimiper Networks forthe past se\en days is 2.5, and todays rare is 2.0, then the Range Ratio is 2.0 / 2.5, or 0.8.
The Range Ratio has two irjuts. letigfhl and lengfh2. The default values are one and se\en, referred to as RR 1:7. The first range does not have to be the rare of just the current bar; it can span ammiber ofbars, so one can experiment with other ratio v-alues such as 2:10 or 3:12. The Acme N sj-stem uses a default threshold of 0.7; once the RR Mis below this vralue, the sj-stem trades abreakout in the direction ofthe trend.
The Raiffie Ratio indicator is a sqjarate plot that tradis the ratro ofthe ATR of one range ofbars to the ATR of another raie ofbars. When the ratio is less than a certain percentage, the chart is in consohdation and is poised to break out. When the ratio is greater than a certain percentage, then the next bar will probably be a narrow range (NR) bar. hi Figure 7.1, each time the ratio is less than 0.7 or 70°o, the next day is awide range day.
iXLAST4)aJly 0 2000 MovAvs1lirte1CK>se.50,0) 10B1.9S
Ac f°la="* «00.00
1150.00 1100.00 10SO,00
7.2 Range Patterns
The Acme N sj-stem integrates the Raie Ratio with other narrow range pattems de\eloped by Cooper and Crabel [4. 6]. Further, we have de\eloped other vraiiations, such as two NR5 days in a row and an NRo bar. Ah ofthese other NR pattems are part of the Acme Range Pattems, as shown in Table 7.1.
TableT.l. RangePattems Pattern
ID2 Two Consecufive hiside
IDNRj, hiside Dwith the Nancwest Range ofthe last n bars Two Consecutive Narrowest Range bars over n bars Narrowest Range ofthe last n bars RaiKe is x° ofthe Average True Range
7.2.1 Inside Day 2 (ID2)
The Inside Day 2 pattem (ID2) is two consecutive Inside Days ( ), as shown in Figure 7.2. It is the same as Coopeis Boomer pattem .
rip«-7-2. Il>2 i:)(.imi>l.-
I )mii„,;Jii.. •Iiiiijh-.......Ik Iiim.I, 1 >.<i i. ili. Iii.l.l. \KMif, , ivhi.). In •• i.«l< .1 .>-. lit
7.22 Inside Day-Narrow Range 4 (IDNR4)
The Inside Day-Narrow Raie 4 pattem flDNRj) is an inside day with the narrowest range ofthe past four dajs , as shown inFigure 7.3.
7.23 Narrow Range 2 (NR2)
The Narrow Raie 2 pattem (NR2) is two consecutive NR bars o-er a given range. Figure 7.4 shows a chart with two consecutive NR5 dajs (NR25).
Figure 7.4. .2 E:;ample
7.2 Range Pattems
7.24 Narrow Range 10 (NRio)
The Narrow Raie 10 pattern (NRio) is the narrowest raie ofthe last ten dajs, as shown in Figure 7.5.
The Narrow Range % pattern (NR%J is based upon a percentage ofthe ATR. Dunnigan defines an NR in the context of an Upswit or Downswh, where an NR bar is any bar with a raie less than half of the widest range bar in the swit . Four examples of an NR" ojo bar are shown in Figure 7.6.
[start]                         [ 25 ]