back start next

[start] [1] [2] [ 3 ] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54]


doctor or lawjer accumulates experience through caseworic. Attaining success is the application ofwisdom and the alalitj to match technique with various market conditions.

Most traders have a bias as to the direction of the maiket and position themselves accordingly; however, maiket-neutral strategies are becoming popular for professionals who are tired oftrading on the gerbil wheel of Level II quotes and one-minute charts. By going into every trading d with both long and short opportunities, the trader lets the market pick the direction.

The last point to emphasize is lhatprice leads news. Instead ofreacting to the news or analj-st recommendations, strive to dewlop trading sjstems that detect imusual price movement. Deploy a diversity oftrading sjstems, and watch for combinations of signals in the same direction. When signab, conflict, a\oid the trade.

1.1 Acme Trading Systems

In the following chapters, we present a group of trading sjstems named the Acme TracMng Sysiemr. The Acme sjstems were derived empirically-they are based on historical studies of daHy and intraday price pattems that occur with regularity in the stock market. We use the inductive process preferred by some of the traders profiled in the Market Wizards books [27, 28], who discm-ered price anomahes in diverse instmments such as mutual fund sectors, futures, and options. In contrast, many of the current sjstems are based on deductive, top-down combinations oftechnical analjsis indicators.

The Acme Trading Systems do not rely on traditional technical analysis, mainly because technical indicators derived from price lag the real price action. Moreover, because many traders use these indicators as a foundation for their sjstems, their o-eruse renders them ineffective; instead, the indicators are more useful as trade filters, not as trade signals.

The main strength of the Acme sjstems is that they are mechanical, and nothing is left to chance. They take long and short positions with specific entrj and exit points. Each of these sjstems has been prcgrammed in a trading programming language, EasjLanguages. Consequently, a trader can run stock scans each night and then generate real-time order alerts for the foUowing d.

- till- woril iiciw r-)>rf4(iil! iIk- ); !1 iiiml, llic >iiii).ii ic ol siutcss. in (»iilr;iM, llic l\ctw l<> siipplifil llic lu wile 1 .1\1>1(- wilh (yil<>siv>->- .mil iHlitT kiilic (liililk-r

........<<>11>. il(-sif>.iicil lo t .!]>..... uimtl kiinm I wi- I lio.<- llic iMiiu ..... ix-. .iiim- i .

iv.llo,, I l,.„llll l.vlll,, ..m>..l( ..I.........b.-)..,il, ).. .

im.....t l,,.,-, ii.....„ lull .,1........1, i ,., i.nlll. , l„l,„l..........I.ltl,. i l.ll .„l,-t,.l14

.......•.-.i- 1...... i..,

1.1 Acme Trading Sjstems

For those ofyou watching business tele\iaon during the day, we have one recommendation: Tum it off. Trading is hard enough without having to listen to a money manager pumping his latest highflier down 30° o. Remember that his dual motive is to keep hisjob andto take your money for self-preserv-ation. The so-called business reporters are usually the last to know about breaking news; experienced traders know that media hype is afade, i.e., doing the opposite of the emotional choice. The bottom line is that nobody knows where the market is headed, even though many pretend to know so. Let price be the guide.

The trading sjstems have been designed with one goal in mind: consistent profltahility based on aunique maiket insight. They are aH based on high probability price pattems that do not appear ftequently in a single stock, but can be found often in auniverse ofover ten thousand stocks. The sjstems are shown in Table 1.1.

Tablel.l. Acme Trading Sjstems







Daily. Weekly


Multiple pattems


Intraday, Daily


Narrow range


Intraday. Daily

Pair trading






Intraday, Daily


Acme V

Daily Only


The trading sjstems span the spectrum of complexity. Ifjust starting out, then focus on the Acme N and R sjstems. Both sjstems are based on simple bar formations . The calculations are minimal, so sophisticated trading software is not required, although automationwiU make the sjstems easiertotrade.

The Acme M and V sjstems are designed for the intermediate trader. Each requires knowledge oftechnical analjsis to identify certainbar pattems. As the trader becomes more proflcient at identifying the various market pattems. the M System becomes more powerful in the traders hands. The Acme V Sjstem is a riskier strategy but is based on a single concept. Use this strategy with smaher positions at first to experience the volatility.

The Acme Fand Acme P Systems are the most technical systems for the advanced trader. The F System requires extensive calculations and works best with trading software such us TraileStation or MetaStock*. The P System requires areal time trailing platform with multiple chart windows

Finally, mthe spint of open source, we encourage the trader to make each sj-s-tem his or her own. Experiment with the source code, the rnput parameters, and the trading filters to create or derive new sj-stems. Trading sj-stem development is a laboratory, and each trader has to "own" the system to trade it effectively. Watch the sjstems work in real-time to confirm that trading entries and exits are realistic in terms of slippage andhquiditj.

1.2 system summarj

The Acme F System is based on the technical work of W.D. Gann and a book by Steve Woods called The Precision Profit Float Indicator [38]. The sj-stem uses the fioat of a stock to analyze supply and demand patterns created by custom fioat indicators. The F Sj-stem then pinpoints breakout and turning points by combining fioat turnover points with geometric patterns such as triple bottoms and retracement patterns such as pullbacks.

The Acme M System identifies combinations of bar patterns. For example, a bar that forms a Tail and a Test is a combination of two distinct bar patterns (these patterns are discussed in Chapter 3) The M Sj-stem scans for bars that have two, three, or even more patterns. The success rate of this sj-stem is directly proportional to the mmiber of identified patterns. Associated with each bar pattem is asetofquahfiers. For example, abarmaybe a narrow range bar, orabar may overiap its 50-day mo\Tng a-erage. Since technicians attach significance to these conditions tliey are denoted on the chart.

The Acme N System is based on a simple concept: identify nan-ow range bars on strongly trending stocks, entering a trade in the direction ofthe trend on a breakout ofthe narrow range bar. The appeal ofthis sjstem is that the risk on the trade is limited to the range ofthe narrow range bar, but the reward is high because the trending stock is in transition from low to high volatilitj.

Acme P System is apair trading strategy that has been gaining popularitj because it is a Iwdged trade, i.e., the trader enters both a long trade and short trade simultaneously. The allure ofpair trading is that it is a strategy with httle risk: however, no stock is immime to the riak of a trading halt or an eamings warning. As with every other sjstem, apecific entrj points, exit points, profit targets, and stop losses are defined.

The Acme R Systeni is based on a simple pattem: the ivclangle [2, 11]. The theorybehind the rectangle is that it represents aperiod of consohdation where traders have already taken positions over several days, but the stock has not moved decidedly in cither direction. Once the stock breaks the rectangle range, the move is usually explosive; further, the narrow range of the rectangle allows the trader to reverse direction if the initial move is a head fake.

The Acme VSystem goes against all the trading truisms such as "the trend is your friend", "dont try to pick bottoms", " catch a falhng knife", etc. In general, these obseri-ations are correct, but at times the trader wants to catch the knife and hold it for a few ds before releasing it. This sjstem is called the V sjstem because the chart formation traces the letter V. The sjstem exploits this pattem withastatistical method known as/ ( 7 / .

The M and N sjstems are swing-trading sjstems. Performance impro\es lineariy with higher values for momentum indicators such as the ADX. The performance ofthe other sjstems does not improve with such filters. Although each sjstem can be improved with proper optimization, none ofthe sjstems has been optimized to avoid overstating results.

1.3 chart indicators

Each Acme sjstem has correaponding chart indicators that aleit the trader to apecific maricet conditions; these indicators are known as lims and letters. Each indicator is presented in the rele-ant chter along with its related sjstem A summary of each indicator is shown in Table 12. Note that the Raii Patterns Indicator is actually a series ofPaintBars™ designed to identifjvarious tj-pes of narrow range bars.

table 1.2. Acme Indicators


Indicator Description

Acme F System

Float Box Channel lines Float Channel Channel lines Float Percent Histogram

Acme M System

Market Patterns Text on Bar High Zone Low Zone PaintBar

Atmc N System

Range Patterns PaintBar Ranic Ratio Separate Rot

Acme 1* Swtcm

Spread Separate Hoi

Ai-im- SyMoiii

liirliiiifif 1 .iiics i>n Chart

Acme V SyslLiii

U.-Kri-ssi<>iiCmve 1 .ii,.-oi, Clwit

Airiii- m.iilu-I SvMi-m

.11 mojtl l(4l till t.ll

1.4 Trading Model

Given a set of trading sjstems, we construct a framework for trading them within the context of an - portfoUo. This Trading Model has three main components:

a PoitfoUo

a Systems

a Trade Manager

The Portfolio is a dynamic set oftrading positions, as shown m Figure 1.1. It specifies the uniform money management criteria, passing them to each ofthe Systems. The Systems enter trades, creating positions based on the equity and position-sizing model. As the Systems run, the Trade Managermomtors profit targets, stop losses, and holding periods, closing any positions that meet the exit criteria closed positions are sent to a trade leg file for spreadsheet analjsis.


pitbn position 2: (Csitions)

Iif.<ni- I.I. Ii.i.lm,-. M.mIiI

1.4 A Trading Model

Although not shown in the diagram, each sjstem is designed with a specific set oftrading filters. The trader has the option of turning the filters on or off to compare filtered performance with unfiHered performance for benchmarking.

1.4.1 Portfolio

Costs are associated with both the Portfoho and the Trade Manager. Portfolio costs are items such as your own salarj, data and exchange fees, and other fixed expenses such as software subscriptions andnews sen-ices. The trading costs encompass commissions, ahppage, andmaigin interest.


Many traders underestimate the initial trading capital and retum required to be a full-time trader. If trading is your profession, then running it as a business is the only way to determine whether or not it wiU be a profitable endeavor. Ifthe trader has no other source ofincome, then cost-of-U\Tng expenses wiU have to be withdrawn from the trading account on a regular basis. A full-time trader starting out should sei aside at least six months ofh\Tng expenses and add these expenses to the fixed costs.

The trader calculates fixed costs on a monthly basis. Achie-ing consistent profitabihty is difficult enough, so every cost must be quantified. For the full-time trader, the added expenses translate into requiring a higher retum on capital. A trader with a $100,000 accountwho must pay several thousand dohare in monthly expenses has significant hurdles to overcome, as shown below

To estimate monthly trading income, start with known quantities: equity, portfoUo costs, trading costs, and tax rate. Then, based on each trading sjstem, estimate the number of trades per month and the amount of capital that wiU be aUocated to each trade. Determine how many positions wUl be maintained si-multaneoualy, and estimate how often the average position wiU be turned over. For example, ifthe average holding period is three dajs, and the portfolio has four positions at any one time, then the estimated number of trades per month is22X (4 / 3)= 29, assuming twenty-two trading days per month.

Table 1.3 shows the expected monthly income for a trading account with $100,000. The fixed costs are $1000 per month, and the commissions and shp-page per round trip total $200. The number oftrades per month is estimated at 20, and 50°o of equity is allocated to each trade, implying a two-day holding period. The dollar amount per trade can either be calculated from actual trading records or extracted from a historical performance report. For example, the av erage Acme trade (win &. loss) based on a hypothetical $50,000 aUocation per hade is about $425 per trade. Using the taUe as a guide, a trader in the 30"o tax bracket could tlieorehcally earn a monthly return of approKimately 4 95"..

[start] [1] [2] [ 3 ] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54]