m SGeometric Trading Figure 510 shows an example ofa stealth triangle Fist, the most recent pairs of pivet highs and pivet tows are located based on a minimum Stremgdh within a
range specified by Length Then, an imaginary trend line is projected across each pair of pivts to the current bar Ifthe high of the current bar is less than the Value ofthe projected trend line for the high pivot pir, then the fist condition fora triangle paltern is satisfied Similasty, ifthe low ofthe current bari greater ‘than the value ofthe projected trend line for the low pivet pair, then the second ‘condition is satisfied Finally, the slope ofeach trenkl line must be les than the ‘maximum specified slope to avoid acute triangles We scan for obtuse triangles ‘that resemble shims’, as shown in Figure 5.10
TART Day CORA Mov pT ie TEA Rene Tange
Figure 10, Stealth Triangle
To trade a triangle, we wait for the stealth formation and then enter atrade on a ‘break ofthe highest or lowest bar ofthe nesting triangles In each ofthe triangle ‘examples in Figures 511 and 5.12 the slope ofthe moving average is a guide to trade direction Trade triangles inthe direction ofthe trend, checking whether ‘or not the triangle is above or below the moving average Like the rectangle, the triangle is a short-term formation for day trading al- ‘though st does net have as much rovers value asa rectangle, The triangle is _more biased towards the prevailing trend, giving the trader a chance to enter on ‘consolidation pattern,
¬
Trang 2
6 Volatility Trading
There are some things
You learn best in calm, and some in storm
Willa Cather
Most human beings are conditioned from childhood that rfwe can buy some- thing at a cheaper price, then it must be a great deal, and we feel good about the purchase Since stocks trade in prices, we take the mental leap and assume that a cheaper stock is a bargain, expecting those good feelings in return for a higher stock price When the stock continues lower, more shares are purchased because the price is an even better bargain, and the buyer is wondering what these sellers must be thinking This "buy the dip" strategy worked well during the bull mar- ket ofthe 1990's but fell apart in the early 2000's
As with any strategy, the efficacy ofa system depends on where and when it is applied, as shown by the bottom-fishing example The Acme V system 1s self-checking because it takes a long position only within the context of what
it defines as bullish conditions For example, the system requires the stock to be trading above its 50-day moving average-a simple yet effective filter
The Acme V System is the most unorthodox system ofall Even with the moving average filter, it still breaks most of the rules because it is a counter- trend system for volatile stocks It is the only Acme system that does not take short positions because it has a few other tricks under its sleeve, and one of those tricks is the so-called Tuesday Turnaround effect (16
Trang 31 6 Velatiity Trading
“The V System is a strategy especially suited for options because ofthe extreme volailly Buying a stock in the V Zone is dangerous because there may be no apparent reason for the decline, especially ifit is bucking the market and sector trend, This strategy works best during general market declines and is tailored to the sector indices When a sector signal is generated find the best-performing socks in that sector, and buy a basket ofthem
‘This eystem does nct work well for industris! and financial stocks Because oftheir cyelical nature and tendency to tend, these socks tend not to have miid- ‘week reversals, In contrast, the strategy weeks well for both bictechnology and technology stocks
1 Linear Regression
“The basis forthe V system isa statistics! method known 8 linea-regression [20) Linear regression analyzes past data to project fuure values using least squares filing, computing a formula for a ine drawn theough these daa, For a stock ‘hat the regression line can reference any tar price inthe formula’ open, high tow, or close ‘Since We me attempting to pick a bottom, we use the low prices, so the repression line is drawn through the lows ofthe data, as shown in Figure 6.1 ‘The rectangle contains the projection of the previous four lows to Low 5 We select regression length of five bars because the V system works on the weekly
6) Linear Regression rs
Proceeding to the nest bar, we calculate the Tinear repression for the previous four bars to project a line towards Low 5, highlighted by the rectangle on the chart in Figure 62 Now, compare the slope ofthe repression line in Figure 6.1 ‘with the Tine in Figure 62 The former slope is al a steeper downward angle, ‘hile the latter is more Horizont] This is the basis ofthe V system As soon the slope starts to faten out, we want to consider along entry
igure 62 Linear Regression Line, Pint 2
‘The easiest way to detect the changing slope is to connect the dots foreach lin- ‘earregression projection The resulis the linear regression curve showntin Figure 63 Note how the curve descends and then ticks up al the point where the long entry is taken
“The astute trader will speculate about how the V system went Jong when the slope ofthe regression curve was down on the previous bar, given thatthe value ofthe curve is not known until the end ofthe trading day The answer is that the linear regression value is projected one barinto the future, giving usa statistical jump onthe cdher traders (refer to the discussion on resl-time trade entry versus ‘end-of-day entry in Chapter 5) Instead ofparticipaing injust a follow-through day, the traders able to capitalize onthe first day as well
Trang 416 6 Veli Trading (62 Volatility Trading System (Acme Vì wa
Figure 6.3 Lineur Regression Carve
62 Volatility Trading System (Acme V)
Since V bottoms are tricky, the system has srt requirements for entry The system takes entries only on Tuesdays ot Wednesdays Inthe past several years, ‘Monday has been a relatively bullish day as well [16], so the trader may wish to change the code to accommodate Mondays A stock that has not reversed by late Wednesday or Thursday wil tend to close on the low ofthe week
‘The second condition is thatthe stock must be above its 50-day moving a ‘rage We are trying to simlate bullish conditions and to filter out ell stocks and indices trading below their 50-day moving average During the protonged Dull market, we did not need this filler but learned quickly ence the market ‘tured down in the spine of 200 ‘The third Condition is tha the low ofthe current bar is greater than the pro- jected low ofthe regression curve When a stock is falling sharply, it tends 10 ‘catpace the regression curve, ic, the lows are below the curve As soon as the projected "The other entry conditions are that the high ofthe current bar must be the low is above the curve then this condition issatisfied, lowest high ofthe regression range, the current tars range must be less than & fen percentage ofthe ATR, and the high of the current bar must be greater than yesterday's low (no gap dover
[Calculations
1 Calculate the Average True Range for the past 20 bars (ATRs0 2 Calculate the Moving Average forthe past 50 bars (MAs
3, Caleutate the Linear Regression value of the projected low from the last S bars (LR) 4, Get the Lowest High of the previons 4 bars (LH) 62.1 Long Signal [Eng Rals 1 Today is Monday or Tuesday 2, High < LH 3, Range < RangeFactor* ATRis 4, Low> LR, 5 Low>MAs
6 High > Yesterday's Low
7 Buy the nextbar at or above the Close + (EntryFactor* ATR) Exit Ruder: Profi Target
1 Sell half ofthe position ator above the High + (ProfieFactor* ATR)
2 Sellhalf ofthe position ator above the High of ProfitBars ago + (2° ProfirFactor* ATR) Exit Rules: Stop Loss
1 Sella shares at or below the Lowest Low for StopBars ~ (ExieFuctor* ATRs) 622 Sheat Signal
‘The V System does not have a corresponding short entry The design of this ‘drategy is “an exercise lel to the reader.” Our recommendation is thatthe short ‘entry be symmetric tothe tong entry Use the linear regression ofthe high and a stock below its S0-dsy moving average
Trang 52s 6 Voltitty Trading Example 61 Acme V System #99*99189154X6198084e£errk4retetrterRe9ortofrrTtxrerrietr ‘ene V systen: Anticipate a “V" Botton based on Linear Regxession ngân HC Đạt geoEv41441232901e420944132.aeErerEeooeas] Inputs: {W Parameters) VolatilityFactor(2.0), Regressict®ars(s), actr(1.0), {Position Paraneters) Equity(100000), Risktodel(3), iskPercent(?.0), iskATR(I.0), EntryFactor(6.25), ‘rawtargets(Tave); Variables: No), THR(O.0), th(20), ret, raters), Livalue(0.0);
fa aii, rapt mates
UUhaate = Eueartainie( ton, Regressions, -1)5
{Entry Signal)
1 (OayOFeek(Date) = 1 or DayOFVeek(Date) = 2) Then Hein
{calaulate shares based on risk model)
Na heretetShares(Equity, Riskiodel, RiskPercent, RISKATR);
TY ágh « Lowst(high, Reressionfors - 3) sn ce Rangeautor + Ak ad tan LRaloe and
ee Ma
High > toe) Then depts "tora tty Targets the ert) TP name Đen Centers = ncnenteargts(‘¥"
loses (entyroctor tA, 6,0, 0; Diy "Ace UL) shares fest O8t on Close Ceotiyractor* A) Stns fess
lạm
ae xs
63 Examples
‘The following chats are examples of trades generated by the Aeme V System Each example uses Equity of8100,000 and the Percent Volailty Model with 2 isk of 2%,
(641 Merosem Cerporabon
Figure 6.4 shows an entry right st the S0-day moving average, Note the differ- cence between entering af the moving average on the same day vereus entering ‘he next day on a treeKout above the high- a difference ofelmest two posts
igure 6 Microsemi: Corpertin Volatility
Given the performance ofthe market from easly 2000 to eaily 2002, with the [Nasdaq declining over 60%, we (ested the performance ofthe V system over tis period since itis @ long-only system Over one thousand stocks from various sectors were buck-tested using daily data from the TradeStatjon histories! data- ‘base, The results are shovenin Table 61L ‘The profit factor for the test period is 175 Although the winning percen!- ‘age is under 50% the average winner was nearly twice the amount ofthe average loser The next step is to test the V system new the 50 day mo
sec if results are improved by using a support Level
Trang 6130 6 Votaility Trading ‘Table 61 acme V System Performance from March 2000 - March 2002 ®%Wimes 47.200 Winners Hồi Avg Win $2,609.83 Losers 1669 ` ProfitFacor 15
‘Table 62 shows the results ofconfining entsies to prices within halfthe ATR of the So-day moving average The profit factor decreased from 175 to 1.20, with ‘2 winning percentage of only 29% Now, confess that you expected the profit factor to be higher because ofsupport at the moving average Inntvely, such a ‘conclusion is logical, but in trading one teams quickly thatthe logics! choice is not the best choice Lets explore the reasons forthe dispasity in results Return to the begiaaing ofthe chapter and read the frst page Assume the trader V signal that is ive points above the moving average and another signal that is has a choice between a ‘one point above the moving average Considering the umber of points above te moving average, describe the key factors that differentiate these two trades CCleady, there are two distinguishing factors, and they are both psychologics) by the traders mind, the second trade is both "cheaper" (comfort factor #1) and also conformist (comfort factor #2 became the literature tells the trader to buy \when a stock approaches the 50-day moving average) The reality is that a stock that has been trading above a key moving average and then proceeds to test that average will strike fear among the long holders and inspite short entries as wel ‘Our medus operandi is: Supper is meart tobe broken “The 62, Acme V Sytem Performance near Shy MA from March 2000 = March 2002 ve Wianers 2% Winners BL Ave Win $2,463.57 Losers 203 Ave Los $1,322.13 Profit Peter 120 63 samples BL 632Vertas Sofwere
‘The V system enters near the low ofthe day, as shown in Figure 6 This is the nly way fo pul the odds in your favor when a stock is ina downtrend Estering ‘on ahigh stop gives too much ofthe profit away In general, the V system is an ‘excellent system forintraday range trading The tader can enter when the stock ‘goes green and either clase He position f Me end ofthe day witha profit or get stopped out case to the low
633 webNethods
Figure 6.6 shows two examples of V entries well above the filty-day moving =v- erage The advantage ofthe V system over traditional ADS/DMI combination systems that the ADS and DMI can filter out trades even ifa stocks trading above its moving average Further, the DMLis deceptive because when astockis in along shallow downtrend, the DMI ratio will flip from postive to negative, even though the long-term trendis up ‘Do nt eliminate stocks priced below $20 per share Both of the entries in Figure 66 occurred in the $15-S16 range, and atthe time, webMethods had an ATR of 13 Most of the industria} and eyclicst stocks trade at much higher prices with lover ATRs We remind you to drink from the foustatn of nqucity
Trang 7m €VelaflityTrading, 634 Seachange “The second Acme V entsy in Figure 6.7 is a losing trade that followed downtrend Entsies after inside days are slighlly more dffienlt but risk limited a choppy’ Vigan 7 Se 63 Examples m (635 Bictecmndogy Indo:
‘Rum the V system on all ofthe indives to get a sense of where the sectors are trading For the entry in Figure 68, we buy either the Biotechnology HOLDR (BBH:Amex) or a basket of biotechnology stocks in the Nasdaq 100 such as Amgen! (AMGN:Nasdaq), Biogen (BGEN: Nasdaq), and Protein Design Labs (PDLENasdaq) The advantage of using the sector indices to trigger trades is that they trend smcether than individual stocks and the average holding period is Tonger The disadvantage of trading a basket ofstocks is that itis a difienlt ‘exmbination of maintaining muliple positions and picking stocks that may not trade synchronously with the index Instead we prefer high-cap stocks that are components ofthe BEH
Figure 68, Biotechnology Index Vain,
636 Computer Ascecates
‘The chart in Figue 6.9 shows a V entry in Computer Associates (CA:NYSE), ‘The problem with this entry is the gap down that occurred two days easter Our reaction to this chart is that the V system code shonld be changed to look for down gaps over the entice inearregression range Iftere are any gaps overthe range,thenthe tradeisnullified ‘Ultimately, the trader's gos isto eliminate mistakes, which means net tak-
Trang 8i 6 Volatility Tradine irecly affect the bottom Tine, and so these observations will become ingrained ‘with practice and experience Figure 6.9 Compater Associates Volatility 7 Range Trading A scaler 5 @ wa Mhoobservesthe future, coud ects before a eceus eruadBarveh
Iu his book The ae ofan, Ted Wiliams describes how he ealeusted that the strike zone was approximately seven bells wide and eleven tls high “The resull was a mats oftaseballs, snd he esiculated his bating average for cach bell in the mats While the ordinary tater dcided between bell and strike, Wiliams refined the tke into seventy-seven separate categories [57] Villans alo determined that cnce a bse stated swinging at pices jut several inches outside ofthe stike zone, the strike zone expanded from 4.2 ưa fet to 58 square fet, an increase of mest 37 pewcent Once 2 picher Jesmed a batter woud swing at bad pitches, then tha’ al the beter woul get, and the batter was destined ‘Now imagine if tie "Splendid Splinter” applied his analyses tothe stock tobe @ 20 hte smasket and tured his altntion to the rang He would som ell ofthe ranges into hee various sizes and then determine li túng average, or profit factor, for each range He would conclude that when the range is arrow Gi the strike
zone his profi factorishigher In const when the rang is wide (out ofthe strike zone), his profit factors lower ‘The average trader analyze a trade as either a winner or aoses- tll or
suite The professional trader analyzes a ade fm it sskvewa rato 13} IF the trader uss range to determine stops, then the isk numerator is the range it- self (Ge higher the age, the higher te sis) and the rewant denominator i the profit tart For example, if long entry is triggered tthe high ofthe tar, ‘nd the range is 15 times the ATR, then the trader is probably swinging eu of the strike zone
“Ta Wotan att I th
Trang 91= ‘TRangeTrading 7.1 Range Ratio
“The Acme N System is based on a simple concept called the Range Ratio We ‘want 2 ratio value less than one becmise a day with a low Range Ratio (RR) is ‘generally followed by a wide range (WR) day under volatile market conditions ‘To ealeulate the RR, divide the current days range by the Average True Range (ATR) over a certain reference range fo estimate today’s volatility For example, ifthe ATR of'uniper Networks forthe past seven days is 2.5, and today’s range is 20, then the Range Ratio is2.0 / 25, or 08 ‘The Range Ratio has two inputs Zeng and Lengyl2 The defanlt values are one and seven, referred to as RR 1:7 The fist range does not have tobe the ‘range of just the current bar; it can span @ number of tars, 80 one can experiment Wilh other ratio values such 2s 2:10 or 3:12 The Aeme N system uses a default ‘threshold of0.7, once the RR fills below this value, the system trades a breakout in the direction ofthe trend “The Range Ratio indicators a separate plct that tracks the ratio ofthe ATR fone range ofbars to the ATR of another range oftars When the ratio is less than a certain pereentage, the chart is in consolidation and is poised to break cout, When the ratio is groater than a certain percentage, then the next bar will probably be a miurow range (NR) bar tu Figure 7.1, each time the ratio is Tess ‘tham.0.7 of 70%, the next day is awide range day
igre 7.8, Range Ratio
72 Range Paterns bó
7.2 Range Patterns
“The Acme N system integrates the Range Ratio with other narrow range pat- terms developed by Cooper and Crabel [4.6] Further, we have developed other ‘variations, such a8 wo NRs days in a tow and an NR bat All ofthese other [NRpattems are part ofthe Acme Range Patterns, as shown in Table 71 ‘Table7.1 RangePattems
LÊ
1D2 — TwoComeulvelnsdeys
IDNR, Inside Day with the Natowest Range ofthe last n bars NR2, Two Consecutive Narrowest Range bars over bars NR, Narrowest Range ofthe last n bars
RY, ints ofthe Average'True 7.24 Inside Day 2 (102)
‘The Inside Day 2 pattern (11D2) is two consecutive Inside Days (IDY’, as shown
Trang 10ng “7Ranee Trading 122 lnside Day-Narow: Relee 4 (IDNR,)
“The Inside Day-Narrow Range 4 pattem IDNR.) is an inside day with the narrowest range ofthe past four days [3], as shown in Figure 7.3
Figure73.IDNR,Example 723 Narrow Range? (NR2)
‘The Narrow Range 2 pattem (NR2) is two consecutive NR bars over a given singe, Figure 7.4 shows a chatt with two consecutive NRs days (NR)
YieMe 74 ME Eemjle
72 Range Pattems 18
1724 Narrow Range 10 (NRie)
“The Narow Range 10 pattern ONRjo) isthe narrowest range ofthe last ten days,
as shown in Figure 7
igure7.S.NRj¢Example 125 Narrow Range°6(NR°)
Trang 11140 7 Range Trading
7.3 Range Trading System (Acme N)
With all ofthe patterns defined, we can now implement a range trading system known as the Acme N System The Acme N System is based on a combination of Cooper's short-term swing techniques [4]; Crabel's narrow range patterns [6] and the Acme Range Ratio The N system is a traditional momentum system because it uses the ADX and pullbacks - the difference is that trades are entered
7.3 Range Trading System (Acme N) 141
Ifany ofthe above range conditions are true, then the Narrow Range Condition is satisfied Finally, the Range Percentage is applied to the bar to qualify it as a potential Acme N trade entry
7.3.1 Long Signal
Inside day and narrow range bar (IDNR pattern), or
only on breaks of NR bars Calculations
The Acme N system requires one of the following five criteria to establish 1 Calculate the ATR for the past 20 bars (ATR,»)
the existence ofa "Narrow Range Condition": > Multiply the Range Percentage (RP) ofthe current bar by ATRep
1 Two consecutive NR bars (NR2 pattern), or 3 Calculate the ADX for the filter length (ADX,a)
2 Two consecutive ID bars (ID2 pattern), or 4 Calculate the 50-bar moving average so)
3._Narrowest range of the last n bars (NR pattern), or 5 Calculate the historic volatility for the filter length (HV ,)
4
5 The Range Ratio (RR) is less than a certain percentage
Once a low-volatility condition has been established, the range of the current bar must be less than a certain percentage of the ATR, 1.¢., it is an NR% bar of 70% or less (this is the RangePercent parameter)
The trader should choose the option ofusing traditional technical filters for a momentum system Historical testing has shown that the higher each of these values is set, the better the performance of the system The N system uses the following filters:
- Minimum Price - Minimum ADX - Minimum HV
We now define the rules ofthe system, including the filters First, we enumerate the narrow range conditions
Narrow Range Condition
Is the current bar an NR; bar and the previous bar an NR; bar?
Is the current bar an ID bar and the previous bar an ID bar? Is the current bar an NR¡ạ bar?
Is the current bar an ID bar and an NR, bar?
Calculate the Range Ratio (RR) tor the current bar divided by the range of the last 7 bars (RR 1:7) Is the RR 1:7 less than 0.7? th ew nN Entry Rules 1 Narrow Range Condition = True 2 Range <=RP*ATRo59 3 Close > 20 4 ADX {4 >= 1 8 5 HV ,4>=0.5 6 Retracement Bars >= 2
7 Median Price > MAso
8 Buy the next bar at or above the High + (EntryFactor * ATR29)
Exit Rules: Profit Target
1 Sell half ofthe position at or above the High + (ProfitFactor * ATR») 2 Sell half ofthe position at or above the High of ProfitBars ago +
(2 * ProfitFactor * ATR29)
Exit Rules: Stop Loss
Trang 12„g 7 Range Trading 782 Shert Signal
Calculations
Calculate the ATR for the past 20 bars (ATR¿)
‘Muliply the Range Percentage (RP) ofthe current bar by ATR Calculate the ADX for the filter length (ADX.)
‘Caleutate the 50-bar moving average (MA)
(Caleulate the historic volailty for the filter length (HV Entry Rules 1 Nacrow Range Condition = True Range <= RP* ATR» Close > 20 ADK >= 18 HVU=05 Retracement Bars >= 2 Median Price < MAw
8 SellShort the next bar at or below the Low ~ (EntryFactor* ATRa)
pe
aen
vit Roles: Profit Target
1 Cover half fthe position s or below the Low = (ProftFactor * ATR)
2 Coverhalfofthe position at or below the Low of ProfBars azo (2 *ProfitFacter * ATR) Exit Rules: Stop Less
1 Coverall shares at or above the Highest High for StopBars + (ButFactor” ATR) “The BasyLanguage code fr the Acme N System isshown in Example 7.1:
Example7.1 Acme N System
fe Syston: Me th ane atl te find arom Rare Pattee Iepats: TH Paromters} Ratiatemabtt, 173 Range Trading Systan (Acme N) RatioLergtta(7), RangePercent(0.7), Monangeratio(©.75, Retracebars(2), (Filter Paraneters} Filterson( Tie), FilterLength(1) (Position Parateters) Equity (30000), Risteocel (3), RiskFercent(2-0), RiskATR(2.0), Entryfactor(0.10), Drawtargets(Tr1e); TH Ấ ie, es ments, engtrate,” TERR cts AM < VlstlityAtRtength, ÂN eni(ClSe, HA);
{Set Entry and Exit Stops)
BuyStep = High + (Entryfactor * ATR); Shortstop = Low ~ (Entxyfactor * ATR); {here W Setup)
Content = enlarge 4, 4) and
Geraint tv's Ln on High Hg} an Lu] » tna) and
Wot chi, ais IELANH
tim - Aeeisidcaye eng, 6);
nition ~ Armetngetatic(atiotngtaatlctengta) < Lowdolatility ~ Conditions or Condition2’er Conditions of Conditions or MaxRangetatio; Conditions;
Trang 13
th “7Range Trading Traderilter « Close > KintmnPrice and
ADX(Et]kerlengtb) >= RinisunlÐ1 sẻ ‘AceeVclatility(Filtertength) >= KinirueH; IF Lowolatility and ‘ce RangePercent * ATR and
Filter Then Begin
mm Riskiodel, RishPercent, RISKATR); Traásgilter = trues If FiltersOe Then
‘Trader iter = RedianPrice > Mis If Teadefilter and
‘AcnefetraceDoun(RetraceBars) and VradeFilter Then Begin {braw Entry Targets on the Chart}
If Drawtargets Then Cnditiont ~ ActekatsyTargets(°W", BuyStep, 0, 0, 0); Buy(hene LEW") N Shares Next Bar at BuyStop Step; nd TradeFilter = True; TF Filterson Then TradeFTlter Medlarfrice ‹ My là /NG Aerletacelp(fetrsecbars
‘radefilter Then Begin Oraw Entry Targets on the Chart}
Tf Growtargets Then Conditions » AcreéntzyTargets("N", 0, 0, ShortStop, 0); sell(hene SE W°) W shares Net Bar at ShortStop Stop; fre;
‘Afler performing some price calculations athe beginning ofthe code, the N stem alls aloft Acme funetios fr determining nartow range condition “The AemeNarrowange function is designed to locate any nartow rang bat ing the Inde Fer example, it can determine whether the cureat bar is an NR ‘arforthe lat ten bars, crwhether the bar seven days 420 was an NR bar
With the catalog of Acme trading putes defined, the trder can see how such computing power is required for cach bas The TradeStation indicators lines and leters) ze provided so the trader can reoognize all possible patterns that are enceded within a single bar Multiply this horsepower by the number of sod, and one reals the chained captives in Ben-Hur straining their oats at Trmmingspeed Hon bn cH m1.11 How 14 Puenple us 74 Examples
‘The following chats are examples oftrades generated by the Acme N’ System Each example uses Equity of $100,000 and the Percent Volatility Model with risk of %e For stocks, trade filtering tumed offbecause oftheirlower ADX readings is turned on For indices trade filtering is
TAA Nasdaq Composit Index
‘This chart shows the Nasdaq Composite Index with the Range Rat falling be: Tow 07 four times (circled in Figure 7.7) The chart shows two long entries ‘ovo short enisies in a relatively choppy market The DMI would have elini- and nated most ofthese trades The problem is that the DMI is typically based on a star study period By the time it catches up tothe tren, te tend has already changed A volatile steck or maket index is characterized by several sudden tren changes within much shorter periods
lanre 77 Nasdaq, Composite Index
‘The Acme N System simply enterstradeson consolidation days orretracement ays Many traders wat for pullbacks, but the chart in Figure 77 illustrates how A trade ean be entered even in the midst of
Trang 14M6 “Range'Tralng,
applica, the number oftrades is cut in hal, but the profit factor improves fem 2001 to'291, while the Tetal Net Profits reduced by ust 21%
“Table7.2- TradeStation Performance AcrneN Stratezy CONB-Drily (271994-31/2002)
——— $804/7650 Open poston PL seo,
‘cos oi, S178 16160 GessLess (887228800)
“Tond#ofUrades “201 Percent potable 801%
uber wining Fades 163 Numberlocng Wades 128
Large winning trade $9976700 Lagestesngtade (6228800 Average wing ade ‘10,780.18 Average losing ace (8881273) Reto ang wining oes +88 Avg tade (wn 8 oss) 904081
Max conve Wines 42 Maxcones loss
Asgfbasinvisers 3 Aug Pbaeinlosers 1
‘Wax today reudow (88072960)
Profit Factor 201 Maxi cortets hod 300,
“Fable 73 TadeStton Performance Acre N Stestegy COMPX-Dally with ier
— 6086263 Open postion PA, son
(ross Pott 6406663266 Gross Lose (688802000)
Tel#efedes Mo Percent poate sso
Nunborwienng aces 981 Numbertoang ¥ades “
Largest waning ade $90797.00 LwgeatiengUaee (962,256.00) ‘Average wining wade S11.71025 /vengelosrgimde (8746980) ati avg wnlovg oss 187 Avgttade (wm loss) s490723
‘Max consee Winners © Maxeorsoc: losers 3
‘Avg barsin wines 4 AngEbaoinlosere 2
Nexinrodaycrawdown (86229600)
Pict Factor 201 Mex compacts 00
14 Examples wa
‘These performance reports have problems, however, because they are based on indices that cannot directly be traded The stock that clsaly tracks the Nasdaq ‘Composite Index is the Nasdag-100 Index Tracking Stock (QQQ-Amex) Al- though the QQQ did not start trading until 199, its profit factor matches the performance ofthe indices, as shown in Table 74
“Table 7.4 TradeStation Performance - Acme N Streey CQC-Dally (1051999312002)
— ‘52453630 OpmpoelenPi sọ
ross Prt s48 (ross Loss (62531025)
‘ola # oftades 5 PHO proetie an
umber viering trades 2 Number wesing trades 2 Largest wining wade $4230.20 Largest sng wage ($2500.00) ‘Average viering ade to avg wing tess $1.Bi645 169 Average esngtrade Augode ain Bess) 65110075) $480.73,
Max exec Wires 3 tx cone sere
Ag bars nvr: 4 pag bes sere 2
He ky đradotn (6162.50)
Prot Factor 7 ax contracts hal 1500
1742 Securties Broker/Deater ndex
“The chart in Figme 78 displays some Acme N tone seals for Che Secu, BrokedDeater Indes (SBD) As with ether sector indies, the SBP doesnt have a direct proxy One possibility isthe Exchange Traded Fund, ex ETP The ETF isjust a tock that is composed ofa gsoup ofstecks ina specific sector The problem withthe ETFs is thal most are net yet liquid enough for short-term trading, andthe spreads perform well on some ofthese stocks (ee Chapter 8) are wide enough such thatthe Acme N sytem doesnot Another altemative isa busker ofstocks Select hc or four representative stocks, and buy or sel them when the Acme N sia fires on the sector index “The advantage ofthis approachis thatthe trader that are much mose Liquid than the ETF.'The disadvantage isthatone crt of can selec a few volatile stocks the socks may not trade in Fine with the index The bes approach is to scoct
Trang 15us ` Range Trading
[igure 78, Securities Broker/Dealer Index
‘The performance report in Table 7.5 shows the unfiltered performance for the Securities Broker/Dealerindes for the past four years “Table75-tradeStationPerformance AcmeNStrategyXBD.X-Daily(3201998-3/ 2002)
"Fa Nel rae 779900 Open poston P| so
‘Gross Prof $7236.00, GrossLoss (488200) “Yotd#cfses 28 Pecenlpofele 029% Number wining voces 4Ð NanbeloangUases e Largestiingtrade ($4,060.00) ‘Averagolosing ade ($1,624.00) "1
‘Average wooing Wace $4021 72
ato avg vniavgloss 220 AsgtadeisnRem) $222204
Maxconeoe Winners 7 Maxeorsec losers
‘Aug bars vines Avg frbas inlets 2
Maxictaday dmedoen— (3957000)
Prot Factor 496
7A Examples 1“
743 Analog Devices
Here is an example ofthree Acme N enties, as shown in Figmre 7.9 Examine the price patteras preceding the occurrence ofthe narrow range bar For the frst ‘ety, the stock consolidated for al leat three days before the signal The sezond entry was an extended pullback and the third entry was a three-bar etracement, Lock: fr rectangles and tables preceding the narrow range bar because this type ofentry has abetterrisk'rewardratio
“The longer the Range Ratio is below the threshold, the more explosive the ‘move, Study the contour ofthe RR curve when the Range Ratio dips below the Ureshotd, Fither the ratio spikes, ort forms along, shallow bottom While he Range Ratio is below the threshold the stock is storing potential energy for 2 protracted move
Trang 16150 Range Trading 744 Tero Pharmaceutics!
Figuce 710 is an example ofthree Acme N long entries over a period of one ‘month This system works best on strongly trending stocks with the following, ‘characteristics, some ofthem taken from Investors Business Daily (IBD):
- IBD Relative Price Strength Rating (RS) * 90 ~ IBD Earnings Per Share Rating (EPS) = 90 = New 52-Week: High
‘At the time, Taro Pharmaceutical (TARO:Nasdag) had RS and EPS rankings ‘of greater than 99, We are certain that a trader could make decent living by
trading just this strategy
Figure 710 Taro Pharmaceutical
‘The number ofretracement Lars varies for each trade shown in Figue 7.10 ‘The first trade pulled back two bars the second trade two bars including ene inside day, and the third trade two bars with ene inside day A parameter to the ‘Acme N System is RetraceBars-iis the minimum number ofretracement bars required to trigger an N entry Ifthe trader chooses net to wait fora retracement ‘and just wants fo enter on a narrow range bar, Hen the RetraceBars parameter canbe set tozero, When net using retracement bars, examine the range ofthe few bars preceding the NR bar Ifthe stock hes appreciated dramatically in this period, then the trade is @ pss The advantage of using no retracement is that the N system picks up consolidations that would normslly be filtered out
74 Bamples 1st
TASMudtimediaGames
‘The chart in Figure 7.11 shows a losing trade Yourjob is to court the number ofproblems with this trade entry before proceeding with this example We find al leas four problems with this trade entry:
a The sock has gapped up
2 The sock has already risen Ite over two days 8 The stock is in the mds oF @retracement a The soxk’s trend is not cleady defined
Trang 178 Market Models di itts simply transferred _fromoneperceptiontoanother GordonGekko Wall Sweet the Motion Picture
‘The market can be handicapped, just as a horseplayer bets on thoroughbreds One might be surprised at just how complicated the betting at the track is-the average bettor is probably not aware ofthe potent speed sires or Diazo's Center of Distribution [8] These data provide the edge to differentiate the professional horseplayer from the amateur As with any game, the player competes for a stz- tistical edge, and this search leads the player to a deeper exploration of diverse subjects such as mathematics, physics, and even philosophy ‘Trading evolves as
a Glass Bead Game’ as the trader attempts to build the ultimate market model
In this chapter, we construct two market models, one using data that are relatively hard to automate First, we apply a set of the Acme trading systems to some market and sector indices Because indices do not have a float, we omit the Acme F system The Acme M N, R, and V systems are combined to form the market model; each of the systems is applied without trade filters to climinate many of the stock-specific requirements This first market model is our Systems Model
‘Second, we develop a special version of the Acme M system using the mar- ket sentiment and breadth indices shown in Table 8.1 For each market index, we specify a rule based on an overbought or oversold reading; the rule interprets the reading based on the index’s correlation with the market For example, the VIX makes a new 20-day high Because the VIX is negatively correlated with the market, the letter “V" is displayed above the current bar As with the Acme M system, a signal is generated when 2 minimum number of pattern criteria in the same direction are met This market model is our sentiment Model
muna